diff --git "a/checkpoint-52800/trainer_state.json" "b/checkpoint-52800/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-52800/trainer_state.json" @@ -0,0 +1,369633 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.528, + "eval_steps": 500, + "global_step": 52800, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 1e-05, + "grad_norm": 1.7129806011276105, + "learning_rate": 3e-06, + "loss": 10.8348, + "step": 1 + }, + { + "epoch": 2e-05, + "grad_norm": 1.6872753303603527, + "learning_rate": 6e-06, + "loss": 10.8356, + "step": 2 + }, + { + "epoch": 3e-05, + "grad_norm": 1.6970020903903387, + "learning_rate": 9e-06, + "loss": 10.834, + "step": 3 + }, + { + "epoch": 4e-05, + "grad_norm": 1.690199421583159, + "learning_rate": 1.2e-05, + "loss": 10.8334, + "step": 4 + }, + { + "epoch": 5e-05, + "grad_norm": 1.6936208883930068, + "learning_rate": 1.5e-05, + "loss": 10.8294, + "step": 5 + }, + { + "epoch": 6e-05, + "grad_norm": 1.6935154610161474, + "learning_rate": 1.8e-05, + "loss": 10.8281, + "step": 6 + }, + { + "epoch": 7e-05, + "grad_norm": 1.6333694173725648, + "learning_rate": 2.1000000000000002e-05, + "loss": 10.8133, + "step": 7 + }, + { + "epoch": 8e-05, + "grad_norm": 1.4463755011186001, + "learning_rate": 2.4e-05, + "loss": 10.783, + "step": 8 + }, + { + "epoch": 9e-05, + "grad_norm": 1.3815123169360315, + "learning_rate": 2.7e-05, + "loss": 10.7779, + "step": 9 + }, + { + "epoch": 0.0001, + "grad_norm": 1.3507621465484316, + "learning_rate": 3e-05, + "loss": 10.7629, + "step": 10 + }, + { + "epoch": 0.00011, + "grad_norm": 1.257508561634155, + "learning_rate": 3.2999999999999996e-05, + "loss": 10.7454, + "step": 11 + }, + { + "epoch": 0.00012, + "grad_norm": 1.224298046820689, + "learning_rate": 3.6e-05, + "loss": 10.7321, + "step": 12 + }, + { + "epoch": 0.00013, + "grad_norm": 1.1609107458726389, + "learning_rate": 3.9e-05, + "loss": 10.7098, + "step": 13 + }, + { + "epoch": 0.00014, + "grad_norm": 1.1251765756585856, + "learning_rate": 4.2000000000000004e-05, + "loss": 10.6986, + "step": 14 + }, + { + "epoch": 0.00015, + "grad_norm": 1.1021031797679595, + "learning_rate": 4.4999999999999996e-05, + "loss": 10.6882, + "step": 15 + }, + { + "epoch": 0.00016, + "grad_norm": 1.05231707077907, + "learning_rate": 4.8e-05, + "loss": 10.6681, + "step": 16 + }, + { + "epoch": 0.00017, + "grad_norm": 1.0082613504118885, + "learning_rate": 5.1000000000000006e-05, + "loss": 10.6513, + "step": 17 + }, + { + "epoch": 0.00018, + "grad_norm": 0.9840324393168475, + "learning_rate": 5.4e-05, + "loss": 10.6344, + "step": 18 + }, + { + "epoch": 0.00019, + "grad_norm": 0.953923239589578, + "learning_rate": 5.7e-05, + "loss": 10.6196, + "step": 19 + }, + { + "epoch": 0.0002, + "grad_norm": 0.9458057853193742, + "learning_rate": 6e-05, + "loss": 10.6069, + "step": 20 + }, + { + "epoch": 0.00021, + "grad_norm": 0.9177457999897578, + "learning_rate": 6.3e-05, + "loss": 10.5922, + "step": 21 + }, + { + "epoch": 0.00022, + "grad_norm": 0.9134874433162486, + "learning_rate": 6.599999999999999e-05, + "loss": 10.576, + "step": 22 + }, + { + "epoch": 0.00023, + "grad_norm": 0.908696989628468, + "learning_rate": 6.9e-05, + "loss": 10.5635, + "step": 23 + }, + { + "epoch": 0.00024, + "grad_norm": 0.9051143007426985, + "learning_rate": 7.2e-05, + "loss": 10.5499, + "step": 24 + }, + { + "epoch": 0.00025, + "grad_norm": 0.9082451576693834, + "learning_rate": 7.500000000000001e-05, + "loss": 10.5361, + "step": 25 + }, + { + "epoch": 0.00026, + "grad_norm": 0.9099344486243927, + "learning_rate": 7.8e-05, + "loss": 10.521, + "step": 26 + }, + { + "epoch": 0.00027, + "grad_norm": 0.9053293550746107, + "learning_rate": 8.1e-05, + "loss": 10.5103, + "step": 27 + }, + { + "epoch": 0.00028, + "grad_norm": 0.9002471153364864, + "learning_rate": 8.400000000000001e-05, + "loss": 10.4955, + "step": 28 + }, + { + "epoch": 0.00029, + "grad_norm": 0.9068699186733776, + "learning_rate": 8.7e-05, + "loss": 10.4811, + "step": 29 + }, + { + "epoch": 0.0003, + "grad_norm": 0.9095271030063902, + "learning_rate": 8.999999999999999e-05, + "loss": 10.4648, + "step": 30 + }, + { + "epoch": 0.00031, + "grad_norm": 0.9097010936405139, + "learning_rate": 9.3e-05, + "loss": 10.4503, + "step": 31 + }, + { + "epoch": 0.00032, + "grad_norm": 0.9047462396891427, + "learning_rate": 9.6e-05, + "loss": 10.4348, + "step": 32 + }, + { + "epoch": 0.00033, + "grad_norm": 0.9068703333942145, + "learning_rate": 9.900000000000001e-05, + "loss": 10.418, + "step": 33 + }, + { + "epoch": 0.00034, + "grad_norm": 0.9072965837486595, + "learning_rate": 0.00010200000000000001, + "loss": 10.4, + "step": 34 + }, + { + "epoch": 0.00035, + "grad_norm": 0.9110003633890357, + "learning_rate": 0.00010500000000000002, + "loss": 10.3835, + "step": 35 + }, + { + "epoch": 0.00036, + "grad_norm": 0.9049119959927198, + "learning_rate": 0.000108, + "loss": 10.3652, + "step": 36 + }, + { + "epoch": 0.00037, + "grad_norm": 0.8970709544624084, + "learning_rate": 0.000111, + "loss": 10.3479, + "step": 37 + }, + { + "epoch": 0.00038, + "grad_norm": 0.8959068278842482, + "learning_rate": 0.000114, + "loss": 10.3275, + "step": 38 + }, + { + "epoch": 0.00039, + "grad_norm": 0.9005947927478184, + "learning_rate": 0.000117, + "loss": 10.3069, + "step": 39 + }, + { + "epoch": 0.0004, + "grad_norm": 0.9014442598894896, + "learning_rate": 0.00012, + "loss": 10.2842, + "step": 40 + }, + { + "epoch": 0.00041, + "grad_norm": 0.8992939718171602, + "learning_rate": 0.000123, + "loss": 10.2657, + "step": 41 + }, + { + "epoch": 0.00042, + "grad_norm": 0.8994818536906172, + "learning_rate": 0.000126, + "loss": 10.2444, + "step": 42 + }, + { + "epoch": 0.00043, + "grad_norm": 0.9062946670458473, + "learning_rate": 0.000129, + "loss": 10.2208, + "step": 43 + }, + { + "epoch": 0.00044, + "grad_norm": 0.9072550424345267, + "learning_rate": 0.00013199999999999998, + "loss": 10.1985, + "step": 44 + }, + { + "epoch": 0.00045, + "grad_norm": 0.908308760029939, + "learning_rate": 0.000135, + "loss": 10.1758, + "step": 45 + }, + { + "epoch": 0.00046, + "grad_norm": 0.8994605508976834, + "learning_rate": 0.000138, + "loss": 10.1528, + "step": 46 + }, + { + "epoch": 0.00047, + "grad_norm": 0.904955141258698, + "learning_rate": 0.000141, + "loss": 10.1274, + "step": 47 + }, + { + "epoch": 0.00048, + "grad_norm": 0.9044693581157806, + "learning_rate": 0.000144, + "loss": 10.1031, + "step": 48 + }, + { + "epoch": 0.00049, + "grad_norm": 0.8992120995192336, + "learning_rate": 0.000147, + "loss": 10.0777, + "step": 49 + }, + { + "epoch": 0.0005, + "grad_norm": 0.905676588399281, + "learning_rate": 0.00015000000000000001, + "loss": 10.0519, + "step": 50 + }, + { + "epoch": 0.00051, + "grad_norm": 0.9066841497261428, + "learning_rate": 0.000153, + "loss": 10.0251, + "step": 51 + }, + { + "epoch": 0.00052, + "grad_norm": 0.9046656683417261, + "learning_rate": 0.000156, + "loss": 9.9981, + "step": 52 + }, + { + "epoch": 0.00053, + "grad_norm": 0.8943714853313668, + "learning_rate": 0.000159, + "loss": 9.974, + "step": 53 + }, + { + "epoch": 0.00054, + "grad_norm": 0.9141658233846578, + "learning_rate": 0.000162, + "loss": 9.9419, + "step": 54 + }, + { + "epoch": 0.00055, + "grad_norm": 0.9035944774643171, + "learning_rate": 0.000165, + "loss": 9.9169, + "step": 55 + }, + { + "epoch": 0.00056, + "grad_norm": 0.895407870582166, + "learning_rate": 0.00016800000000000002, + "loss": 9.8872, + "step": 56 + }, + { + "epoch": 0.00057, + "grad_norm": 0.9021731997760362, + "learning_rate": 0.000171, + "loss": 9.8601, + "step": 57 + }, + { + "epoch": 0.00058, + "grad_norm": 0.8980871554912008, + "learning_rate": 0.000174, + "loss": 9.8343, + "step": 58 + }, + { + "epoch": 0.00059, + "grad_norm": 0.9073832283363998, + "learning_rate": 0.000177, + "loss": 9.8028, + "step": 59 + }, + { + "epoch": 0.0006, + "grad_norm": 0.8921071838486323, + "learning_rate": 0.00017999999999999998, + "loss": 9.777, + "step": 60 + }, + { + "epoch": 0.00061, + "grad_norm": 0.8918001641348363, + "learning_rate": 0.000183, + "loss": 9.7484, + "step": 61 + }, + { + "epoch": 0.00062, + "grad_norm": 0.897401330332219, + "learning_rate": 0.000186, + "loss": 9.717, + "step": 62 + }, + { + "epoch": 0.00063, + "grad_norm": 0.8914516241190131, + "learning_rate": 0.000189, + "loss": 9.6894, + "step": 63 + }, + { + "epoch": 0.00064, + "grad_norm": 0.8896652156254973, + "learning_rate": 0.000192, + "loss": 9.6587, + "step": 64 + }, + { + "epoch": 0.00065, + "grad_norm": 0.8995447585153489, + "learning_rate": 0.00019500000000000002, + "loss": 9.6261, + "step": 65 + }, + { + "epoch": 0.00066, + "grad_norm": 0.8896939041293862, + "learning_rate": 0.00019800000000000002, + "loss": 9.6026, + "step": 66 + }, + { + "epoch": 0.00067, + "grad_norm": 0.8935314234316469, + "learning_rate": 0.000201, + "loss": 9.5723, + "step": 67 + }, + { + "epoch": 0.00068, + "grad_norm": 0.8971584182008717, + "learning_rate": 0.00020400000000000003, + "loss": 9.5393, + "step": 68 + }, + { + "epoch": 0.00069, + "grad_norm": 0.8806455604370673, + "learning_rate": 0.00020700000000000002, + "loss": 9.5119, + "step": 69 + }, + { + "epoch": 0.0007, + "grad_norm": 0.892956094531968, + "learning_rate": 0.00021000000000000004, + "loss": 9.4751, + "step": 70 + }, + { + "epoch": 0.00071, + "grad_norm": 0.8848452972865632, + "learning_rate": 0.00021299999999999997, + "loss": 9.4495, + "step": 71 + }, + { + "epoch": 0.00072, + "grad_norm": 0.8831148731992822, + "learning_rate": 0.000216, + "loss": 9.4223, + "step": 72 + }, + { + "epoch": 0.00073, + "grad_norm": 0.887150899449638, + "learning_rate": 0.00021899999999999998, + "loss": 9.3879, + "step": 73 + }, + { + "epoch": 0.00074, + "grad_norm": 0.8878619769377328, + "learning_rate": 0.000222, + "loss": 9.3616, + "step": 74 + }, + { + "epoch": 0.00075, + "grad_norm": 0.8808154408936898, + "learning_rate": 0.000225, + "loss": 9.3275, + "step": 75 + }, + { + "epoch": 0.00076, + "grad_norm": 0.8908035269749474, + "learning_rate": 0.000228, + "loss": 9.3009, + "step": 76 + }, + { + "epoch": 0.00077, + "grad_norm": 0.884531047332737, + "learning_rate": 0.000231, + "loss": 9.2727, + "step": 77 + }, + { + "epoch": 0.00078, + "grad_norm": 0.8838664917591654, + "learning_rate": 0.000234, + "loss": 9.2422, + "step": 78 + }, + { + "epoch": 0.00079, + "grad_norm": 0.8858668201182466, + "learning_rate": 0.00023700000000000001, + "loss": 9.2056, + "step": 79 + }, + { + "epoch": 0.0008, + "grad_norm": 0.8856967305037482, + "learning_rate": 0.00024, + "loss": 9.1711, + "step": 80 + }, + { + "epoch": 0.00081, + "grad_norm": 0.8942846826675519, + "learning_rate": 0.00024300000000000002, + "loss": 9.1382, + "step": 81 + }, + { + "epoch": 0.00082, + "grad_norm": 0.897767651472895, + "learning_rate": 0.000246, + "loss": 9.1142, + "step": 82 + }, + { + "epoch": 0.00083, + "grad_norm": 0.8951752702012633, + "learning_rate": 0.00024900000000000004, + "loss": 9.0778, + "step": 83 + }, + { + "epoch": 0.00084, + "grad_norm": 0.8980395289622467, + "learning_rate": 0.000252, + "loss": 9.0469, + "step": 84 + }, + { + "epoch": 0.00085, + "grad_norm": 0.8894006576183595, + "learning_rate": 0.000255, + "loss": 9.0242, + "step": 85 + }, + { + "epoch": 0.00086, + "grad_norm": 0.8907945566480024, + "learning_rate": 0.000258, + "loss": 8.9886, + "step": 86 + }, + { + "epoch": 0.00087, + "grad_norm": 0.8869170795764568, + "learning_rate": 0.000261, + "loss": 8.9664, + "step": 87 + }, + { + "epoch": 0.00088, + "grad_norm": 0.8860541210154026, + "learning_rate": 0.00026399999999999997, + "loss": 8.9293, + "step": 88 + }, + { + "epoch": 0.00089, + "grad_norm": 0.8822605600899943, + "learning_rate": 0.000267, + "loss": 8.9037, + "step": 89 + }, + { + "epoch": 0.0009, + "grad_norm": 0.8817151929172502, + "learning_rate": 0.00027, + "loss": 8.8766, + "step": 90 + }, + { + "epoch": 0.00091, + "grad_norm": 0.877617615465877, + "learning_rate": 0.000273, + "loss": 8.8478, + "step": 91 + }, + { + "epoch": 0.00092, + "grad_norm": 0.8822716293479064, + "learning_rate": 0.000276, + "loss": 8.8156, + "step": 92 + }, + { + "epoch": 0.00093, + "grad_norm": 0.8823661552266111, + "learning_rate": 0.000279, + "loss": 8.7863, + "step": 93 + }, + { + "epoch": 0.00094, + "grad_norm": 0.8830384482321385, + "learning_rate": 0.000282, + "loss": 8.7609, + "step": 94 + }, + { + "epoch": 0.00095, + "grad_norm": 0.8735042737334501, + "learning_rate": 0.000285, + "loss": 8.7321, + "step": 95 + }, + { + "epoch": 0.00096, + "grad_norm": 0.8799214796836804, + "learning_rate": 0.000288, + "loss": 8.7028, + "step": 96 + }, + { + "epoch": 0.00097, + "grad_norm": 0.8704594748643596, + "learning_rate": 0.000291, + "loss": 8.6791, + "step": 97 + }, + { + "epoch": 0.00098, + "grad_norm": 0.8706415983834461, + "learning_rate": 0.000294, + "loss": 8.642, + "step": 98 + }, + { + "epoch": 0.00099, + "grad_norm": 0.8683426041650804, + "learning_rate": 0.000297, + "loss": 8.62, + "step": 99 + }, + { + "epoch": 0.001, + "grad_norm": 0.8690594926543161, + "learning_rate": 0.00030000000000000003, + "loss": 8.5941, + "step": 100 + }, + { + "epoch": 0.00101, + "grad_norm": 0.8744725161423202, + "learning_rate": 0.00030300000000000005, + "loss": 8.5597, + "step": 101 + }, + { + "epoch": 0.00102, + "grad_norm": 0.8626538117869429, + "learning_rate": 0.000306, + "loss": 8.5407, + "step": 102 + }, + { + "epoch": 0.00103, + "grad_norm": 0.8630292491448714, + "learning_rate": 0.000309, + "loss": 8.5165, + "step": 103 + }, + { + "epoch": 0.00104, + "grad_norm": 0.8566580756350954, + "learning_rate": 0.000312, + "loss": 8.4948, + "step": 104 + }, + { + "epoch": 0.00105, + "grad_norm": 0.8588931967033124, + "learning_rate": 0.000315, + "loss": 8.4689, + "step": 105 + }, + { + "epoch": 0.00106, + "grad_norm": 0.8531867230327145, + "learning_rate": 0.000318, + "loss": 8.4397, + "step": 106 + }, + { + "epoch": 0.00107, + "grad_norm": 0.8474808010000593, + "learning_rate": 0.000321, + "loss": 8.4149, + "step": 107 + }, + { + "epoch": 0.00108, + "grad_norm": 0.858890949163445, + "learning_rate": 0.000324, + "loss": 8.3866, + "step": 108 + }, + { + "epoch": 0.00109, + "grad_norm": 0.862504115159085, + "learning_rate": 0.000327, + "loss": 8.3673, + "step": 109 + }, + { + "epoch": 0.0011, + "grad_norm": 0.8797254817902618, + "learning_rate": 0.00033, + "loss": 8.3453, + "step": 110 + }, + { + "epoch": 0.00111, + "grad_norm": 0.8938450574121157, + "learning_rate": 0.000333, + "loss": 8.3162, + "step": 111 + }, + { + "epoch": 0.00112, + "grad_norm": 0.8984693362159062, + "learning_rate": 0.00033600000000000004, + "loss": 8.2961, + "step": 112 + }, + { + "epoch": 0.00113, + "grad_norm": 0.8741969532880793, + "learning_rate": 0.000339, + "loss": 8.2543, + "step": 113 + }, + { + "epoch": 0.00114, + "grad_norm": 0.8263135137442741, + "learning_rate": 0.000342, + "loss": 8.2446, + "step": 114 + }, + { + "epoch": 0.00115, + "grad_norm": 0.8311105019716521, + "learning_rate": 0.00034500000000000004, + "loss": 8.2164, + "step": 115 + }, + { + "epoch": 0.00116, + "grad_norm": 0.8585271561560018, + "learning_rate": 0.000348, + "loss": 8.1938, + "step": 116 + }, + { + "epoch": 0.00117, + "grad_norm": 0.8687047969468357, + "learning_rate": 0.000351, + "loss": 8.1623, + "step": 117 + }, + { + "epoch": 0.00118, + "grad_norm": 0.8460500876754268, + "learning_rate": 0.000354, + "loss": 8.1456, + "step": 118 + }, + { + "epoch": 0.00119, + "grad_norm": 0.80734714043103, + "learning_rate": 0.000357, + "loss": 8.131, + "step": 119 + }, + { + "epoch": 0.0012, + "grad_norm": 0.7912470130916918, + "learning_rate": 0.00035999999999999997, + "loss": 8.103, + "step": 120 + }, + { + "epoch": 0.00121, + "grad_norm": 0.811181199244069, + "learning_rate": 0.000363, + "loss": 8.0751, + "step": 121 + }, + { + "epoch": 0.00122, + "grad_norm": 0.8641427784894331, + "learning_rate": 0.000366, + "loss": 8.0581, + "step": 122 + }, + { + "epoch": 0.00123, + "grad_norm": 1.059706847686213, + "learning_rate": 0.000369, + "loss": 8.038, + "step": 123 + }, + { + "epoch": 0.00124, + "grad_norm": 1.1727027216994725, + "learning_rate": 0.000372, + "loss": 8.021, + "step": 124 + }, + { + "epoch": 0.00125, + "grad_norm": 0.8130002892684417, + "learning_rate": 0.000375, + "loss": 7.9874, + "step": 125 + }, + { + "epoch": 0.00126, + "grad_norm": 0.9195848585380069, + "learning_rate": 0.000378, + "loss": 7.9767, + "step": 126 + }, + { + "epoch": 0.00127, + "grad_norm": 1.0843433185909894, + "learning_rate": 0.000381, + "loss": 7.9572, + "step": 127 + }, + { + "epoch": 0.00128, + "grad_norm": 0.7822919696307823, + "learning_rate": 0.000384, + "loss": 7.9285, + "step": 128 + }, + { + "epoch": 0.00129, + "grad_norm": 0.8822428605457112, + "learning_rate": 0.00038700000000000003, + "loss": 7.9179, + "step": 129 + }, + { + "epoch": 0.0013, + "grad_norm": 0.844355538937723, + "learning_rate": 0.00039000000000000005, + "loss": 7.8895, + "step": 130 + }, + { + "epoch": 0.00131, + "grad_norm": 0.7902535444057679, + "learning_rate": 0.000393, + "loss": 7.8528, + "step": 131 + }, + { + "epoch": 0.00132, + "grad_norm": 0.8777082094723105, + "learning_rate": 0.00039600000000000003, + "loss": 7.8441, + "step": 132 + }, + { + "epoch": 0.00133, + "grad_norm": 0.7973277807473979, + "learning_rate": 0.00039900000000000005, + "loss": 7.8195, + "step": 133 + }, + { + "epoch": 0.00134, + "grad_norm": 0.7889088832890946, + "learning_rate": 0.000402, + "loss": 7.8056, + "step": 134 + }, + { + "epoch": 0.00135, + "grad_norm": 0.7461125825498439, + "learning_rate": 0.00040500000000000003, + "loss": 7.7735, + "step": 135 + }, + { + "epoch": 0.00136, + "grad_norm": 0.7727896835908762, + "learning_rate": 0.00040800000000000005, + "loss": 7.7579, + "step": 136 + }, + { + "epoch": 0.00137, + "grad_norm": 0.6932995987295251, + "learning_rate": 0.000411, + "loss": 7.7341, + "step": 137 + }, + { + "epoch": 0.00138, + "grad_norm": 0.758084762416224, + "learning_rate": 0.00041400000000000003, + "loss": 7.7117, + "step": 138 + }, + { + "epoch": 0.00139, + "grad_norm": 0.7171019453691133, + "learning_rate": 0.00041700000000000005, + "loss": 7.6963, + "step": 139 + }, + { + "epoch": 0.0014, + "grad_norm": 0.6814920611933867, + "learning_rate": 0.00042000000000000007, + "loss": 7.6775, + "step": 140 + }, + { + "epoch": 0.00141, + "grad_norm": 0.7091532995122851, + "learning_rate": 0.000423, + "loss": 7.6638, + "step": 141 + }, + { + "epoch": 0.00142, + "grad_norm": 0.6928279523561562, + "learning_rate": 0.00042599999999999995, + "loss": 7.6501, + "step": 142 + }, + { + "epoch": 0.00143, + "grad_norm": 0.6614572727332786, + "learning_rate": 0.00042899999999999997, + "loss": 7.6195, + "step": 143 + }, + { + "epoch": 0.00144, + "grad_norm": 0.6903462553659518, + "learning_rate": 0.000432, + "loss": 7.6015, + "step": 144 + }, + { + "epoch": 0.00145, + "grad_norm": 0.690019772183536, + "learning_rate": 0.000435, + "loss": 7.5953, + "step": 145 + }, + { + "epoch": 0.00146, + "grad_norm": 0.6908198257220046, + "learning_rate": 0.00043799999999999997, + "loss": 7.5557, + "step": 146 + }, + { + "epoch": 0.00147, + "grad_norm": 0.7009866965495668, + "learning_rate": 0.000441, + "loss": 7.5482, + "step": 147 + }, + { + "epoch": 0.00148, + "grad_norm": 0.6832764187147686, + "learning_rate": 0.000444, + "loss": 7.5366, + "step": 148 + }, + { + "epoch": 0.00149, + "grad_norm": 0.59797192318343, + "learning_rate": 0.00044699999999999997, + "loss": 7.5272, + "step": 149 + }, + { + "epoch": 0.0015, + "grad_norm": 0.6655702435683013, + "learning_rate": 0.00045, + "loss": 7.4963, + "step": 150 + }, + { + "epoch": 0.00151, + "grad_norm": 0.732396941583091, + "learning_rate": 0.000453, + "loss": 7.48, + "step": 151 + }, + { + "epoch": 0.00152, + "grad_norm": 0.5836278900992692, + "learning_rate": 0.000456, + "loss": 7.4694, + "step": 152 + }, + { + "epoch": 0.00153, + "grad_norm": 0.6777912087785298, + "learning_rate": 0.000459, + "loss": 7.4588, + "step": 153 + }, + { + "epoch": 0.00154, + "grad_norm": 0.727978180952039, + "learning_rate": 0.000462, + "loss": 7.442, + "step": 154 + }, + { + "epoch": 0.00155, + "grad_norm": 0.7368922682622268, + "learning_rate": 0.000465, + "loss": 7.4241, + "step": 155 + }, + { + "epoch": 0.00156, + "grad_norm": 0.8391325557731037, + "learning_rate": 0.000468, + "loss": 7.4013, + "step": 156 + }, + { + "epoch": 0.00157, + "grad_norm": 0.8289929528374833, + "learning_rate": 0.000471, + "loss": 7.3995, + "step": 157 + }, + { + "epoch": 0.00158, + "grad_norm": 0.5070337070851558, + "learning_rate": 0.00047400000000000003, + "loss": 7.3713, + "step": 158 + }, + { + "epoch": 0.00159, + "grad_norm": 0.783946493417518, + "learning_rate": 0.000477, + "loss": 7.3668, + "step": 159 + }, + { + "epoch": 0.0016, + "grad_norm": 0.6957053326984224, + "learning_rate": 0.00048, + "loss": 7.3475, + "step": 160 + }, + { + "epoch": 0.00161, + "grad_norm": 0.547833885334286, + "learning_rate": 0.00048300000000000003, + "loss": 7.3204, + "step": 161 + }, + { + "epoch": 0.00162, + "grad_norm": 0.8547649122041628, + "learning_rate": 0.00048600000000000005, + "loss": 7.3325, + "step": 162 + }, + { + "epoch": 0.00163, + "grad_norm": 0.8673949773728752, + "learning_rate": 0.0004890000000000001, + "loss": 7.316, + "step": 163 + }, + { + "epoch": 0.00164, + "grad_norm": 1.0619539506126108, + "learning_rate": 0.000492, + "loss": 7.3191, + "step": 164 + }, + { + "epoch": 0.00165, + "grad_norm": 0.6324744711420325, + "learning_rate": 0.000495, + "loss": 7.2864, + "step": 165 + }, + { + "epoch": 0.00166, + "grad_norm": 0.5662260261966024, + "learning_rate": 0.0004980000000000001, + "loss": 7.26, + "step": 166 + }, + { + "epoch": 0.00167, + "grad_norm": 0.7262900850309921, + "learning_rate": 0.000501, + "loss": 7.2554, + "step": 167 + }, + { + "epoch": 0.00168, + "grad_norm": 0.6121691436587496, + "learning_rate": 0.000504, + "loss": 7.2353, + "step": 168 + }, + { + "epoch": 0.00169, + "grad_norm": 0.5390794603769147, + "learning_rate": 0.0005070000000000001, + "loss": 7.2263, + "step": 169 + }, + { + "epoch": 0.0017, + "grad_norm": 0.5999036994585554, + "learning_rate": 0.00051, + "loss": 7.213, + "step": 170 + }, + { + "epoch": 0.00171, + "grad_norm": 0.4637320512013434, + "learning_rate": 0.000513, + "loss": 7.1933, + "step": 171 + }, + { + "epoch": 0.00172, + "grad_norm": 0.5250975302523401, + "learning_rate": 0.000516, + "loss": 7.1953, + "step": 172 + }, + { + "epoch": 0.00173, + "grad_norm": 0.40559164125903624, + "learning_rate": 0.0005189999999999999, + "loss": 7.1764, + "step": 173 + }, + { + "epoch": 0.00174, + "grad_norm": 0.4505921111310584, + "learning_rate": 0.000522, + "loss": 7.1953, + "step": 174 + }, + { + "epoch": 0.00175, + "grad_norm": 0.4234331150208657, + "learning_rate": 0.000525, + "loss": 7.1572, + "step": 175 + }, + { + "epoch": 0.00176, + "grad_norm": 0.3852967422981744, + "learning_rate": 0.0005279999999999999, + "loss": 7.1322, + "step": 176 + }, + { + "epoch": 0.00177, + "grad_norm": 0.3685443025565043, + "learning_rate": 0.000531, + "loss": 7.1378, + "step": 177 + }, + { + "epoch": 0.00178, + "grad_norm": 0.44280593644992733, + "learning_rate": 0.000534, + "loss": 7.1301, + "step": 178 + }, + { + "epoch": 0.00179, + "grad_norm": 0.3638226120256115, + "learning_rate": 0.000537, + "loss": 7.1191, + "step": 179 + }, + { + "epoch": 0.0018, + "grad_norm": 0.37841703582661185, + "learning_rate": 0.00054, + "loss": 7.0921, + "step": 180 + }, + { + "epoch": 0.00181, + "grad_norm": 0.3275750999054276, + "learning_rate": 0.000543, + "loss": 7.0801, + "step": 181 + }, + { + "epoch": 0.00182, + "grad_norm": 0.3469517461565544, + "learning_rate": 0.000546, + "loss": 7.0774, + "step": 182 + }, + { + "epoch": 0.00183, + "grad_norm": 0.3965623823212328, + "learning_rate": 0.000549, + "loss": 7.0674, + "step": 183 + }, + { + "epoch": 0.00184, + "grad_norm": 0.47244712960577356, + "learning_rate": 0.000552, + "loss": 7.0582, + "step": 184 + }, + { + "epoch": 0.00185, + "grad_norm": 0.7068086356604425, + "learning_rate": 0.000555, + "loss": 7.0369, + "step": 185 + }, + { + "epoch": 0.00186, + "grad_norm": 0.9665650200874053, + "learning_rate": 0.000558, + "loss": 7.0604, + "step": 186 + }, + { + "epoch": 0.00187, + "grad_norm": 1.1379887499575514, + "learning_rate": 0.000561, + "loss": 7.0366, + "step": 187 + }, + { + "epoch": 0.00188, + "grad_norm": 0.5005933831438132, + "learning_rate": 0.000564, + "loss": 7.0008, + "step": 188 + }, + { + "epoch": 0.00189, + "grad_norm": 0.4490325126563288, + "learning_rate": 0.000567, + "loss": 6.997, + "step": 189 + }, + { + "epoch": 0.0019, + "grad_norm": 0.6949112483193859, + "learning_rate": 0.00057, + "loss": 6.9846, + "step": 190 + }, + { + "epoch": 0.00191, + "grad_norm": 0.4887612962658467, + "learning_rate": 0.000573, + "loss": 6.9724, + "step": 191 + }, + { + "epoch": 0.00192, + "grad_norm": 0.5374763602633008, + "learning_rate": 0.000576, + "loss": 6.9655, + "step": 192 + }, + { + "epoch": 0.00193, + "grad_norm": 0.4491815623326969, + "learning_rate": 0.000579, + "loss": 6.9637, + "step": 193 + }, + { + "epoch": 0.00194, + "grad_norm": 0.4044031823800156, + "learning_rate": 0.000582, + "loss": 6.9565, + "step": 194 + }, + { + "epoch": 0.00195, + "grad_norm": 0.5115147380417242, + "learning_rate": 0.000585, + "loss": 6.9386, + "step": 195 + }, + { + "epoch": 0.00196, + "grad_norm": 0.45947827433809557, + "learning_rate": 0.000588, + "loss": 6.9258, + "step": 196 + }, + { + "epoch": 0.00197, + "grad_norm": 0.5289316721591154, + "learning_rate": 0.000591, + "loss": 6.9226, + "step": 197 + }, + { + "epoch": 0.00198, + "grad_norm": 0.4416511613975406, + "learning_rate": 0.000594, + "loss": 6.9132, + "step": 198 + }, + { + "epoch": 0.00199, + "grad_norm": 0.36314866008934127, + "learning_rate": 0.0005970000000000001, + "loss": 6.8916, + "step": 199 + }, + { + "epoch": 0.002, + "grad_norm": 0.4299454881914127, + "learning_rate": 0.0006000000000000001, + "loss": 6.8932, + "step": 200 + }, + { + "epoch": 0.00201, + "grad_norm": 0.2786890859011363, + "learning_rate": 0.000603, + "loss": 6.8645, + "step": 201 + }, + { + "epoch": 0.00202, + "grad_norm": 0.4553990879307791, + "learning_rate": 0.0006060000000000001, + "loss": 6.855, + "step": 202 + }, + { + "epoch": 0.00203, + "grad_norm": 0.49491513980041124, + "learning_rate": 0.0006090000000000001, + "loss": 6.8592, + "step": 203 + }, + { + "epoch": 0.00204, + "grad_norm": 0.5750090076580947, + "learning_rate": 0.000612, + "loss": 6.8457, + "step": 204 + }, + { + "epoch": 0.00205, + "grad_norm": 0.6904749130436038, + "learning_rate": 0.000615, + "loss": 6.8381, + "step": 205 + }, + { + "epoch": 0.00206, + "grad_norm": 0.7582947936777445, + "learning_rate": 0.000618, + "loss": 6.83, + "step": 206 + }, + { + "epoch": 0.00207, + "grad_norm": 0.728748472942146, + "learning_rate": 0.000621, + "loss": 6.8214, + "step": 207 + }, + { + "epoch": 0.00208, + "grad_norm": 0.5163586812963157, + "learning_rate": 0.000624, + "loss": 6.8116, + "step": 208 + }, + { + "epoch": 0.00209, + "grad_norm": 0.5726761174567752, + "learning_rate": 0.000627, + "loss": 6.7933, + "step": 209 + }, + { + "epoch": 0.0021, + "grad_norm": 0.6890311463252623, + "learning_rate": 0.00063, + "loss": 6.7854, + "step": 210 + }, + { + "epoch": 0.00211, + "grad_norm": 0.9174002778722206, + "learning_rate": 0.000633, + "loss": 6.7849, + "step": 211 + }, + { + "epoch": 0.00212, + "grad_norm": 0.8086617968740898, + "learning_rate": 0.000636, + "loss": 6.7808, + "step": 212 + }, + { + "epoch": 0.00213, + "grad_norm": 0.6685717599500662, + "learning_rate": 0.000639, + "loss": 6.7542, + "step": 213 + }, + { + "epoch": 0.00214, + "grad_norm": 0.511917016650173, + "learning_rate": 0.000642, + "loss": 6.7483, + "step": 214 + }, + { + "epoch": 0.00215, + "grad_norm": 0.5132261185164837, + "learning_rate": 0.000645, + "loss": 6.7465, + "step": 215 + }, + { + "epoch": 0.00216, + "grad_norm": 0.3896647006337605, + "learning_rate": 0.000648, + "loss": 6.7354, + "step": 216 + }, + { + "epoch": 0.00217, + "grad_norm": 0.53153444147609, + "learning_rate": 0.000651, + "loss": 6.7114, + "step": 217 + }, + { + "epoch": 0.00218, + "grad_norm": 0.4560253950102483, + "learning_rate": 0.000654, + "loss": 6.7136, + "step": 218 + }, + { + "epoch": 0.00219, + "grad_norm": 0.38246603109839156, + "learning_rate": 0.000657, + "loss": 6.6847, + "step": 219 + }, + { + "epoch": 0.0022, + "grad_norm": 0.502249830770979, + "learning_rate": 0.00066, + "loss": 6.7061, + "step": 220 + }, + { + "epoch": 0.00221, + "grad_norm": 0.555840042257826, + "learning_rate": 0.0006630000000000001, + "loss": 6.6817, + "step": 221 + }, + { + "epoch": 0.00222, + "grad_norm": 0.7008290795132841, + "learning_rate": 0.000666, + "loss": 6.6751, + "step": 222 + }, + { + "epoch": 0.00223, + "grad_norm": 0.9665649898158697, + "learning_rate": 0.000669, + "loss": 6.6759, + "step": 223 + }, + { + "epoch": 0.00224, + "grad_norm": 1.0460190685952617, + "learning_rate": 0.0006720000000000001, + "loss": 6.6821, + "step": 224 + }, + { + "epoch": 0.00225, + "grad_norm": 0.9709238336565439, + "learning_rate": 0.000675, + "loss": 6.6643, + "step": 225 + }, + { + "epoch": 0.00226, + "grad_norm": 0.9675609159629996, + "learning_rate": 0.000678, + "loss": 6.6602, + "step": 226 + }, + { + "epoch": 0.00227, + "grad_norm": 0.6069059963900193, + "learning_rate": 0.0006810000000000001, + "loss": 6.6251, + "step": 227 + }, + { + "epoch": 0.00228, + "grad_norm": 0.6661980684886607, + "learning_rate": 0.000684, + "loss": 6.6314, + "step": 228 + }, + { + "epoch": 0.00229, + "grad_norm": 0.5755819115869941, + "learning_rate": 0.000687, + "loss": 6.6231, + "step": 229 + }, + { + "epoch": 0.0023, + "grad_norm": 0.48612508126201437, + "learning_rate": 0.0006900000000000001, + "loss": 6.6015, + "step": 230 + }, + { + "epoch": 0.00231, + "grad_norm": 0.42902206098374773, + "learning_rate": 0.000693, + "loss": 6.5844, + "step": 231 + }, + { + "epoch": 0.00232, + "grad_norm": 0.4617696239896233, + "learning_rate": 0.000696, + "loss": 6.5964, + "step": 232 + }, + { + "epoch": 0.00233, + "grad_norm": 0.397560103207551, + "learning_rate": 0.0006990000000000001, + "loss": 6.5819, + "step": 233 + }, + { + "epoch": 0.00234, + "grad_norm": 0.39436477469071923, + "learning_rate": 0.000702, + "loss": 6.5732, + "step": 234 + }, + { + "epoch": 0.00235, + "grad_norm": 0.37818254545129, + "learning_rate": 0.000705, + "loss": 6.5584, + "step": 235 + }, + { + "epoch": 0.00236, + "grad_norm": 0.39793300295732814, + "learning_rate": 0.000708, + "loss": 6.539, + "step": 236 + }, + { + "epoch": 0.00237, + "grad_norm": 0.32880148477167265, + "learning_rate": 0.0007109999999999999, + "loss": 6.5486, + "step": 237 + }, + { + "epoch": 0.00238, + "grad_norm": 0.33186649759843, + "learning_rate": 0.000714, + "loss": 6.5374, + "step": 238 + }, + { + "epoch": 0.00239, + "grad_norm": 0.3861082150171924, + "learning_rate": 0.000717, + "loss": 6.5195, + "step": 239 + }, + { + "epoch": 0.0024, + "grad_norm": 0.3818382545466509, + "learning_rate": 0.0007199999999999999, + "loss": 6.5368, + "step": 240 + }, + { + "epoch": 0.00241, + "grad_norm": 0.26517003068907236, + "learning_rate": 0.000723, + "loss": 6.5167, + "step": 241 + }, + { + "epoch": 0.00242, + "grad_norm": 0.30652105574179844, + "learning_rate": 0.000726, + "loss": 6.4934, + "step": 242 + }, + { + "epoch": 0.00243, + "grad_norm": 0.3382942246099826, + "learning_rate": 0.000729, + "loss": 6.4799, + "step": 243 + }, + { + "epoch": 0.00244, + "grad_norm": 0.4164388540502762, + "learning_rate": 0.000732, + "loss": 6.4843, + "step": 244 + }, + { + "epoch": 0.00245, + "grad_norm": 0.4035007765909141, + "learning_rate": 0.000735, + "loss": 6.4741, + "step": 245 + }, + { + "epoch": 0.00246, + "grad_norm": 0.4484502106991885, + "learning_rate": 0.000738, + "loss": 6.458, + "step": 246 + }, + { + "epoch": 0.00247, + "grad_norm": 0.6057401118193197, + "learning_rate": 0.000741, + "loss": 6.4543, + "step": 247 + }, + { + "epoch": 0.00248, + "grad_norm": 1.1775332556723501, + "learning_rate": 0.000744, + "loss": 6.4781, + "step": 248 + }, + { + "epoch": 0.00249, + "grad_norm": 1.0888595785598245, + "learning_rate": 0.000747, + "loss": 6.4635, + "step": 249 + }, + { + "epoch": 0.0025, + "grad_norm": 0.536872031808477, + "learning_rate": 0.00075, + "loss": 6.4479, + "step": 250 + }, + { + "epoch": 0.00251, + "grad_norm": 0.7926645181932281, + "learning_rate": 0.000753, + "loss": 6.4169, + "step": 251 + }, + { + "epoch": 0.00252, + "grad_norm": 0.6863348407685264, + "learning_rate": 0.000756, + "loss": 6.4273, + "step": 252 + }, + { + "epoch": 0.00253, + "grad_norm": 0.7123800606299509, + "learning_rate": 0.000759, + "loss": 6.4195, + "step": 253 + }, + { + "epoch": 0.00254, + "grad_norm": 0.839405849029746, + "learning_rate": 0.000762, + "loss": 6.4177, + "step": 254 + }, + { + "epoch": 0.00255, + "grad_norm": 0.76501143998226, + "learning_rate": 0.0007650000000000001, + "loss": 6.4159, + "step": 255 + }, + { + "epoch": 0.00256, + "grad_norm": 0.597608858095952, + "learning_rate": 0.000768, + "loss": 6.3889, + "step": 256 + }, + { + "epoch": 0.00257, + "grad_norm": 0.6526176951631347, + "learning_rate": 0.000771, + "loss": 6.3981, + "step": 257 + }, + { + "epoch": 0.00258, + "grad_norm": 0.651228257980475, + "learning_rate": 0.0007740000000000001, + "loss": 6.3725, + "step": 258 + }, + { + "epoch": 0.00259, + "grad_norm": 0.5603901273931662, + "learning_rate": 0.000777, + "loss": 6.3719, + "step": 259 + }, + { + "epoch": 0.0026, + "grad_norm": 0.41845020316479425, + "learning_rate": 0.0007800000000000001, + "loss": 6.3536, + "step": 260 + }, + { + "epoch": 0.00261, + "grad_norm": 0.5144884019867095, + "learning_rate": 0.0007830000000000001, + "loss": 6.3665, + "step": 261 + }, + { + "epoch": 0.00262, + "grad_norm": 0.5548811083770797, + "learning_rate": 0.000786, + "loss": 6.3412, + "step": 262 + }, + { + "epoch": 0.00263, + "grad_norm": 0.702068573310266, + "learning_rate": 0.0007890000000000001, + "loss": 6.353, + "step": 263 + }, + { + "epoch": 0.00264, + "grad_norm": 0.9481897048028406, + "learning_rate": 0.0007920000000000001, + "loss": 6.3404, + "step": 264 + }, + { + "epoch": 0.00265, + "grad_norm": 1.2297805755386195, + "learning_rate": 0.000795, + "loss": 6.3478, + "step": 265 + }, + { + "epoch": 0.00266, + "grad_norm": 0.5034998722006886, + "learning_rate": 0.0007980000000000001, + "loss": 6.3233, + "step": 266 + }, + { + "epoch": 0.00267, + "grad_norm": 0.8457797300321087, + "learning_rate": 0.0008010000000000001, + "loss": 6.3268, + "step": 267 + }, + { + "epoch": 0.00268, + "grad_norm": 0.7624901894608749, + "learning_rate": 0.000804, + "loss": 6.3226, + "step": 268 + }, + { + "epoch": 0.00269, + "grad_norm": 0.6803898428780553, + "learning_rate": 0.0008070000000000001, + "loss": 6.3045, + "step": 269 + }, + { + "epoch": 0.0027, + "grad_norm": 0.5891673657315365, + "learning_rate": 0.0008100000000000001, + "loss": 6.3, + "step": 270 + }, + { + "epoch": 0.00271, + "grad_norm": 0.6321969571669588, + "learning_rate": 0.000813, + "loss": 6.3081, + "step": 271 + }, + { + "epoch": 0.00272, + "grad_norm": 0.6080230974854919, + "learning_rate": 0.0008160000000000001, + "loss": 6.2911, + "step": 272 + }, + { + "epoch": 0.00273, + "grad_norm": 0.577176950863229, + "learning_rate": 0.0008190000000000001, + "loss": 6.2786, + "step": 273 + }, + { + "epoch": 0.00274, + "grad_norm": 0.46970800663022055, + "learning_rate": 0.000822, + "loss": 6.2573, + "step": 274 + }, + { + "epoch": 0.00275, + "grad_norm": 0.5095773122618286, + "learning_rate": 0.0008250000000000001, + "loss": 6.2676, + "step": 275 + }, + { + "epoch": 0.00276, + "grad_norm": 0.421367493059458, + "learning_rate": 0.0008280000000000001, + "loss": 6.2547, + "step": 276 + }, + { + "epoch": 0.00277, + "grad_norm": 0.4229723742956301, + "learning_rate": 0.0008310000000000001, + "loss": 6.2503, + "step": 277 + }, + { + "epoch": 0.00278, + "grad_norm": 0.4631977178825306, + "learning_rate": 0.0008340000000000001, + "loss": 6.2346, + "step": 278 + }, + { + "epoch": 0.00279, + "grad_norm": 0.41870110969580987, + "learning_rate": 0.0008370000000000001, + "loss": 6.2332, + "step": 279 + }, + { + "epoch": 0.0028, + "grad_norm": 0.4083314739680453, + "learning_rate": 0.0008400000000000001, + "loss": 6.2161, + "step": 280 + }, + { + "epoch": 0.00281, + "grad_norm": 0.42451645247510567, + "learning_rate": 0.0008430000000000001, + "loss": 6.2058, + "step": 281 + }, + { + "epoch": 0.00282, + "grad_norm": 0.4811013283391871, + "learning_rate": 0.000846, + "loss": 6.206, + "step": 282 + }, + { + "epoch": 0.00283, + "grad_norm": 0.6798083705841664, + "learning_rate": 0.0008489999999999999, + "loss": 6.2015, + "step": 283 + }, + { + "epoch": 0.00284, + "grad_norm": 1.0382201143248402, + "learning_rate": 0.0008519999999999999, + "loss": 6.2055, + "step": 284 + }, + { + "epoch": 0.00285, + "grad_norm": 1.115942818342409, + "learning_rate": 0.000855, + "loss": 6.2129, + "step": 285 + }, + { + "epoch": 0.00286, + "grad_norm": 0.8889955339821247, + "learning_rate": 0.0008579999999999999, + "loss": 6.187, + "step": 286 + }, + { + "epoch": 0.00287, + "grad_norm": 1.2422801265585652, + "learning_rate": 0.000861, + "loss": 6.209, + "step": 287 + }, + { + "epoch": 0.00288, + "grad_norm": 0.8315932342234975, + "learning_rate": 0.000864, + "loss": 6.174, + "step": 288 + }, + { + "epoch": 0.00289, + "grad_norm": 1.2914759013339998, + "learning_rate": 0.0008669999999999999, + "loss": 6.2078, + "step": 289 + }, + { + "epoch": 0.0029, + "grad_norm": 0.8376507056381004, + "learning_rate": 0.00087, + "loss": 6.1757, + "step": 290 + }, + { + "epoch": 0.00291, + "grad_norm": 0.8412780961911104, + "learning_rate": 0.000873, + "loss": 6.1658, + "step": 291 + }, + { + "epoch": 0.00292, + "grad_norm": 1.047021583757866, + "learning_rate": 0.0008759999999999999, + "loss": 6.1758, + "step": 292 + }, + { + "epoch": 0.00293, + "grad_norm": 0.8919470282886952, + "learning_rate": 0.000879, + "loss": 6.151, + "step": 293 + }, + { + "epoch": 0.00294, + "grad_norm": 0.665529628519212, + "learning_rate": 0.000882, + "loss": 6.159, + "step": 294 + }, + { + "epoch": 0.00295, + "grad_norm": 0.5169660787787601, + "learning_rate": 0.0008849999999999999, + "loss": 6.1239, + "step": 295 + }, + { + "epoch": 0.00296, + "grad_norm": 0.5611538425989948, + "learning_rate": 0.000888, + "loss": 6.1363, + "step": 296 + }, + { + "epoch": 0.00297, + "grad_norm": 0.46398604023920087, + "learning_rate": 0.000891, + "loss": 6.1045, + "step": 297 + }, + { + "epoch": 0.00298, + "grad_norm": 0.4361556326298739, + "learning_rate": 0.0008939999999999999, + "loss": 6.1198, + "step": 298 + }, + { + "epoch": 0.00299, + "grad_norm": 0.4319584905904094, + "learning_rate": 0.000897, + "loss": 6.0941, + "step": 299 + }, + { + "epoch": 0.003, + "grad_norm": 0.4255386299160817, + "learning_rate": 0.0009, + "loss": 6.0936, + "step": 300 + }, + { + "epoch": 0.00301, + "grad_norm": 0.3316584659066082, + "learning_rate": 0.0009029999999999999, + "loss": 6.0857, + "step": 301 + }, + { + "epoch": 0.00302, + "grad_norm": 0.37299869635167304, + "learning_rate": 0.000906, + "loss": 6.0685, + "step": 302 + }, + { + "epoch": 0.00303, + "grad_norm": 0.40148217950038195, + "learning_rate": 0.000909, + "loss": 6.0805, + "step": 303 + }, + { + "epoch": 0.00304, + "grad_norm": 0.420191340163935, + "learning_rate": 0.000912, + "loss": 6.0758, + "step": 304 + }, + { + "epoch": 0.00305, + "grad_norm": 0.45307668264044143, + "learning_rate": 0.000915, + "loss": 6.0736, + "step": 305 + }, + { + "epoch": 0.00306, + "grad_norm": 0.6122731249830943, + "learning_rate": 0.000918, + "loss": 6.0651, + "step": 306 + }, + { + "epoch": 0.00307, + "grad_norm": 0.851197326038436, + "learning_rate": 0.000921, + "loss": 6.0633, + "step": 307 + }, + { + "epoch": 0.00308, + "grad_norm": 1.1284591769922636, + "learning_rate": 0.000924, + "loss": 6.0582, + "step": 308 + }, + { + "epoch": 0.00309, + "grad_norm": 0.9596545216263644, + "learning_rate": 0.000927, + "loss": 6.0709, + "step": 309 + }, + { + "epoch": 0.0031, + "grad_norm": 1.039007932956353, + "learning_rate": 0.00093, + "loss": 6.0624, + "step": 310 + }, + { + "epoch": 0.00311, + "grad_norm": 0.9855401820369791, + "learning_rate": 0.000933, + "loss": 6.0524, + "step": 311 + }, + { + "epoch": 0.00312, + "grad_norm": 1.0163701418335827, + "learning_rate": 0.000936, + "loss": 6.041, + "step": 312 + }, + { + "epoch": 0.00313, + "grad_norm": 1.0223663613867633, + "learning_rate": 0.0009390000000000001, + "loss": 6.0491, + "step": 313 + }, + { + "epoch": 0.00314, + "grad_norm": 1.0997292958340223, + "learning_rate": 0.000942, + "loss": 6.0641, + "step": 314 + }, + { + "epoch": 0.00315, + "grad_norm": 0.932276773939602, + "learning_rate": 0.000945, + "loss": 6.0354, + "step": 315 + }, + { + "epoch": 0.00316, + "grad_norm": 0.8624268848533463, + "learning_rate": 0.0009480000000000001, + "loss": 6.0096, + "step": 316 + }, + { + "epoch": 0.00317, + "grad_norm": 0.6867359398602113, + "learning_rate": 0.000951, + "loss": 6.0237, + "step": 317 + }, + { + "epoch": 0.00318, + "grad_norm": 0.5777711812516898, + "learning_rate": 0.000954, + "loss": 6.014, + "step": 318 + }, + { + "epoch": 0.00319, + "grad_norm": 0.5907201170091796, + "learning_rate": 0.0009570000000000001, + "loss": 6.0042, + "step": 319 + }, + { + "epoch": 0.0032, + "grad_norm": 0.5929910498481646, + "learning_rate": 0.00096, + "loss": 6.0021, + "step": 320 + }, + { + "epoch": 0.00321, + "grad_norm": 0.6559636432249029, + "learning_rate": 0.000963, + "loss": 5.9891, + "step": 321 + }, + { + "epoch": 0.00322, + "grad_norm": 0.5844133161497509, + "learning_rate": 0.0009660000000000001, + "loss": 5.9766, + "step": 322 + }, + { + "epoch": 0.00323, + "grad_norm": 0.5466606066369618, + "learning_rate": 0.000969, + "loss": 5.9736, + "step": 323 + }, + { + "epoch": 0.00324, + "grad_norm": 0.563270781105711, + "learning_rate": 0.0009720000000000001, + "loss": 5.9778, + "step": 324 + }, + { + "epoch": 0.00325, + "grad_norm": 0.5312990845923178, + "learning_rate": 0.0009750000000000001, + "loss": 5.9405, + "step": 325 + }, + { + "epoch": 0.00326, + "grad_norm": 0.5118566622058196, + "learning_rate": 0.0009780000000000001, + "loss": 5.9566, + "step": 326 + }, + { + "epoch": 0.00327, + "grad_norm": 0.5259315695578027, + "learning_rate": 0.000981, + "loss": 5.9568, + "step": 327 + }, + { + "epoch": 0.00328, + "grad_norm": 0.5410551164101933, + "learning_rate": 0.000984, + "loss": 5.9324, + "step": 328 + }, + { + "epoch": 0.00329, + "grad_norm": 0.48301414107035934, + "learning_rate": 0.000987, + "loss": 5.931, + "step": 329 + }, + { + "epoch": 0.0033, + "grad_norm": 0.5975532498257996, + "learning_rate": 0.00099, + "loss": 5.9265, + "step": 330 + }, + { + "epoch": 0.00331, + "grad_norm": 0.9746373555768076, + "learning_rate": 0.0009930000000000002, + "loss": 5.9381, + "step": 331 + }, + { + "epoch": 0.00332, + "grad_norm": 1.1994973147641799, + "learning_rate": 0.0009960000000000001, + "loss": 5.9385, + "step": 332 + }, + { + "epoch": 0.00333, + "grad_norm": 0.5144992648158865, + "learning_rate": 0.000999, + "loss": 5.8989, + "step": 333 + }, + { + "epoch": 0.00334, + "grad_norm": 0.9073672400240862, + "learning_rate": 0.001002, + "loss": 5.9262, + "step": 334 + }, + { + "epoch": 0.00335, + "grad_norm": 0.7561451103694465, + "learning_rate": 0.001005, + "loss": 5.9186, + "step": 335 + }, + { + "epoch": 0.00336, + "grad_norm": 0.7872757919528415, + "learning_rate": 0.001008, + "loss": 5.9134, + "step": 336 + }, + { + "epoch": 0.00337, + "grad_norm": 0.7536020827923614, + "learning_rate": 0.0010110000000000002, + "loss": 5.8884, + "step": 337 + }, + { + "epoch": 0.00338, + "grad_norm": 1.0523353883962923, + "learning_rate": 0.0010140000000000001, + "loss": 5.9132, + "step": 338 + }, + { + "epoch": 0.00339, + "grad_norm": 1.2857238182949966, + "learning_rate": 0.0010170000000000001, + "loss": 5.9334, + "step": 339 + }, + { + "epoch": 0.0034, + "grad_norm": 0.7361708327689132, + "learning_rate": 0.00102, + "loss": 5.885, + "step": 340 + }, + { + "epoch": 0.00341, + "grad_norm": 0.6901997441262301, + "learning_rate": 0.001023, + "loss": 5.869, + "step": 341 + }, + { + "epoch": 0.00342, + "grad_norm": 0.6381033550571615, + "learning_rate": 0.001026, + "loss": 5.8702, + "step": 342 + }, + { + "epoch": 0.00343, + "grad_norm": 0.6071718978500397, + "learning_rate": 0.0010290000000000002, + "loss": 5.8743, + "step": 343 + }, + { + "epoch": 0.00344, + "grad_norm": 0.5857796625429044, + "learning_rate": 0.001032, + "loss": 5.861, + "step": 344 + }, + { + "epoch": 0.00345, + "grad_norm": 0.626640702848716, + "learning_rate": 0.001035, + "loss": 5.8537, + "step": 345 + }, + { + "epoch": 0.00346, + "grad_norm": 0.6755670022907736, + "learning_rate": 0.0010379999999999999, + "loss": 5.8603, + "step": 346 + }, + { + "epoch": 0.00347, + "grad_norm": 0.9144508249400731, + "learning_rate": 0.001041, + "loss": 5.8338, + "step": 347 + }, + { + "epoch": 0.00348, + "grad_norm": 1.2125876856754099, + "learning_rate": 0.001044, + "loss": 5.8634, + "step": 348 + }, + { + "epoch": 0.00349, + "grad_norm": 0.6928695941460523, + "learning_rate": 0.001047, + "loss": 5.8236, + "step": 349 + }, + { + "epoch": 0.0035, + "grad_norm": 0.7654262923967496, + "learning_rate": 0.00105, + "loss": 5.8502, + "step": 350 + }, + { + "epoch": 0.00351, + "grad_norm": 0.8300223804260752, + "learning_rate": 0.001053, + "loss": 5.8507, + "step": 351 + }, + { + "epoch": 0.00352, + "grad_norm": 1.1393832643973667, + "learning_rate": 0.0010559999999999999, + "loss": 5.841, + "step": 352 + }, + { + "epoch": 0.00353, + "grad_norm": 0.7670875434573843, + "learning_rate": 0.001059, + "loss": 5.8302, + "step": 353 + }, + { + "epoch": 0.00354, + "grad_norm": 0.8617169684849714, + "learning_rate": 0.001062, + "loss": 5.8072, + "step": 354 + }, + { + "epoch": 0.00355, + "grad_norm": 0.8787230305505044, + "learning_rate": 0.001065, + "loss": 5.8221, + "step": 355 + }, + { + "epoch": 0.00356, + "grad_norm": 0.9037602020080988, + "learning_rate": 0.001068, + "loss": 5.8164, + "step": 356 + }, + { + "epoch": 0.00357, + "grad_norm": 0.785887699185926, + "learning_rate": 0.001071, + "loss": 5.8055, + "step": 357 + }, + { + "epoch": 0.00358, + "grad_norm": 0.6152742029666318, + "learning_rate": 0.001074, + "loss": 5.7887, + "step": 358 + }, + { + "epoch": 0.00359, + "grad_norm": 0.5372063086433791, + "learning_rate": 0.001077, + "loss": 5.78, + "step": 359 + }, + { + "epoch": 0.0036, + "grad_norm": 0.5078257678271803, + "learning_rate": 0.00108, + "loss": 5.7825, + "step": 360 + }, + { + "epoch": 0.00361, + "grad_norm": 0.4885651334266738, + "learning_rate": 0.001083, + "loss": 5.7748, + "step": 361 + }, + { + "epoch": 0.00362, + "grad_norm": 0.5495429650143561, + "learning_rate": 0.001086, + "loss": 5.7596, + "step": 362 + }, + { + "epoch": 0.00363, + "grad_norm": 0.5626950540152672, + "learning_rate": 0.001089, + "loss": 5.7515, + "step": 363 + }, + { + "epoch": 0.00364, + "grad_norm": 0.6199658617744055, + "learning_rate": 0.001092, + "loss": 5.766, + "step": 364 + }, + { + "epoch": 0.00365, + "grad_norm": 0.7392438146286566, + "learning_rate": 0.001095, + "loss": 5.7655, + "step": 365 + }, + { + "epoch": 0.00366, + "grad_norm": 0.9864875055616179, + "learning_rate": 0.001098, + "loss": 5.7524, + "step": 366 + }, + { + "epoch": 0.00367, + "grad_norm": 1.1449768044417052, + "learning_rate": 0.001101, + "loss": 5.7648, + "step": 367 + }, + { + "epoch": 0.00368, + "grad_norm": 0.9013400729864322, + "learning_rate": 0.001104, + "loss": 5.755, + "step": 368 + }, + { + "epoch": 0.00369, + "grad_norm": 0.9093377711089596, + "learning_rate": 0.001107, + "loss": 5.7659, + "step": 369 + }, + { + "epoch": 0.0037, + "grad_norm": 0.575923493278037, + "learning_rate": 0.00111, + "loss": 5.7328, + "step": 370 + }, + { + "epoch": 0.00371, + "grad_norm": 0.6737016310188764, + "learning_rate": 0.001113, + "loss": 5.7102, + "step": 371 + }, + { + "epoch": 0.00372, + "grad_norm": 0.4833347808689347, + "learning_rate": 0.001116, + "loss": 5.7236, + "step": 372 + }, + { + "epoch": 0.00373, + "grad_norm": 0.6361357392920576, + "learning_rate": 0.001119, + "loss": 5.7181, + "step": 373 + }, + { + "epoch": 0.00374, + "grad_norm": 0.6286357986456886, + "learning_rate": 0.001122, + "loss": 5.7192, + "step": 374 + }, + { + "epoch": 0.00375, + "grad_norm": 0.7140127832546589, + "learning_rate": 0.0011250000000000001, + "loss": 5.7248, + "step": 375 + }, + { + "epoch": 0.00376, + "grad_norm": 0.728891228424708, + "learning_rate": 0.001128, + "loss": 5.7207, + "step": 376 + }, + { + "epoch": 0.00377, + "grad_norm": 0.7251122752592066, + "learning_rate": 0.001131, + "loss": 5.7141, + "step": 377 + }, + { + "epoch": 0.00378, + "grad_norm": 0.8109517942362439, + "learning_rate": 0.001134, + "loss": 5.7121, + "step": 378 + }, + { + "epoch": 0.00379, + "grad_norm": 0.7548765882892476, + "learning_rate": 0.001137, + "loss": 5.6981, + "step": 379 + }, + { + "epoch": 0.0038, + "grad_norm": 0.5982490555873449, + "learning_rate": 0.00114, + "loss": 5.7128, + "step": 380 + }, + { + "epoch": 0.00381, + "grad_norm": 0.5479723067602768, + "learning_rate": 0.0011430000000000001, + "loss": 5.6793, + "step": 381 + }, + { + "epoch": 0.00382, + "grad_norm": 0.5400365110175976, + "learning_rate": 0.001146, + "loss": 5.6631, + "step": 382 + }, + { + "epoch": 0.00383, + "grad_norm": 0.4406698702316126, + "learning_rate": 0.001149, + "loss": 5.673, + "step": 383 + }, + { + "epoch": 0.00384, + "grad_norm": 0.5495584711003424, + "learning_rate": 0.001152, + "loss": 5.6782, + "step": 384 + }, + { + "epoch": 0.00385, + "grad_norm": 0.7047837665038742, + "learning_rate": 0.001155, + "loss": 5.6686, + "step": 385 + }, + { + "epoch": 0.00386, + "grad_norm": 1.0039450355838517, + "learning_rate": 0.001158, + "loss": 5.6846, + "step": 386 + }, + { + "epoch": 0.00387, + "grad_norm": 1.2552299335364856, + "learning_rate": 0.0011610000000000001, + "loss": 5.6713, + "step": 387 + }, + { + "epoch": 0.00388, + "grad_norm": 1.0630057286422998, + "learning_rate": 0.001164, + "loss": 5.6615, + "step": 388 + }, + { + "epoch": 0.00389, + "grad_norm": 1.4085777330550793, + "learning_rate": 0.001167, + "loss": 5.6963, + "step": 389 + }, + { + "epoch": 0.0039, + "grad_norm": 0.7893066659624004, + "learning_rate": 0.00117, + "loss": 5.651, + "step": 390 + }, + { + "epoch": 0.00391, + "grad_norm": 0.8891270576556106, + "learning_rate": 0.001173, + "loss": 5.666, + "step": 391 + }, + { + "epoch": 0.00392, + "grad_norm": 1.1058013192110903, + "learning_rate": 0.001176, + "loss": 5.6796, + "step": 392 + }, + { + "epoch": 0.00393, + "grad_norm": 1.170614508410806, + "learning_rate": 0.0011790000000000001, + "loss": 5.6646, + "step": 393 + }, + { + "epoch": 0.00394, + "grad_norm": 0.8391276502601887, + "learning_rate": 0.001182, + "loss": 5.6402, + "step": 394 + }, + { + "epoch": 0.00395, + "grad_norm": 0.9435882620236007, + "learning_rate": 0.001185, + "loss": 5.6277, + "step": 395 + }, + { + "epoch": 0.00396, + "grad_norm": 0.7925001626557522, + "learning_rate": 0.001188, + "loss": 5.6404, + "step": 396 + }, + { + "epoch": 0.00397, + "grad_norm": 0.8633162203152536, + "learning_rate": 0.001191, + "loss": 5.6366, + "step": 397 + }, + { + "epoch": 0.00398, + "grad_norm": 0.9359127674730449, + "learning_rate": 0.0011940000000000002, + "loss": 5.6437, + "step": 398 + }, + { + "epoch": 0.00399, + "grad_norm": 1.0926478209626875, + "learning_rate": 0.0011970000000000001, + "loss": 5.6494, + "step": 399 + }, + { + "epoch": 0.004, + "grad_norm": 0.8943926064407558, + "learning_rate": 0.0012000000000000001, + "loss": 5.6306, + "step": 400 + }, + { + "epoch": 0.00401, + "grad_norm": 1.211025202532141, + "learning_rate": 0.001203, + "loss": 5.6241, + "step": 401 + }, + { + "epoch": 0.00402, + "grad_norm": 0.8585006093020132, + "learning_rate": 0.001206, + "loss": 5.6196, + "step": 402 + }, + { + "epoch": 0.00403, + "grad_norm": 0.8708424012246267, + "learning_rate": 0.001209, + "loss": 5.618, + "step": 403 + }, + { + "epoch": 0.00404, + "grad_norm": 0.6771235232466043, + "learning_rate": 0.0012120000000000002, + "loss": 5.6057, + "step": 404 + }, + { + "epoch": 0.00405, + "grad_norm": 0.5089480196948696, + "learning_rate": 0.0012150000000000002, + "loss": 5.5986, + "step": 405 + }, + { + "epoch": 0.00406, + "grad_norm": 0.5324784457955185, + "learning_rate": 0.0012180000000000001, + "loss": 5.583, + "step": 406 + }, + { + "epoch": 0.00407, + "grad_norm": 0.4806328504890235, + "learning_rate": 0.0012209999999999999, + "loss": 5.575, + "step": 407 + }, + { + "epoch": 0.00408, + "grad_norm": 0.5340674298116082, + "learning_rate": 0.001224, + "loss": 5.5941, + "step": 408 + }, + { + "epoch": 0.00409, + "grad_norm": 0.6817510563164704, + "learning_rate": 0.001227, + "loss": 5.5739, + "step": 409 + }, + { + "epoch": 0.0041, + "grad_norm": 0.8230982603577015, + "learning_rate": 0.00123, + "loss": 5.5739, + "step": 410 + }, + { + "epoch": 0.00411, + "grad_norm": 0.7982833366881755, + "learning_rate": 0.001233, + "loss": 5.5886, + "step": 411 + }, + { + "epoch": 0.00412, + "grad_norm": 0.7882120301866252, + "learning_rate": 0.001236, + "loss": 5.5767, + "step": 412 + }, + { + "epoch": 0.00413, + "grad_norm": 0.9078160072473371, + "learning_rate": 0.0012389999999999999, + "loss": 5.5798, + "step": 413 + }, + { + "epoch": 0.00414, + "grad_norm": 0.8046291024914881, + "learning_rate": 0.001242, + "loss": 5.5582, + "step": 414 + }, + { + "epoch": 0.00415, + "grad_norm": 0.9449024826812693, + "learning_rate": 0.001245, + "loss": 5.5633, + "step": 415 + }, + { + "epoch": 0.00416, + "grad_norm": 0.9578358959807691, + "learning_rate": 0.001248, + "loss": 5.5489, + "step": 416 + }, + { + "epoch": 0.00417, + "grad_norm": 0.7364680005190741, + "learning_rate": 0.001251, + "loss": 5.575, + "step": 417 + }, + { + "epoch": 0.00418, + "grad_norm": 0.5600093653837771, + "learning_rate": 0.001254, + "loss": 5.5419, + "step": 418 + }, + { + "epoch": 0.00419, + "grad_norm": 0.7369458002937045, + "learning_rate": 0.0012569999999999999, + "loss": 5.535, + "step": 419 + }, + { + "epoch": 0.0042, + "grad_norm": 0.7566412883958042, + "learning_rate": 0.00126, + "loss": 5.5478, + "step": 420 + }, + { + "epoch": 0.00421, + "grad_norm": 0.9341471688658377, + "learning_rate": 0.001263, + "loss": 5.5468, + "step": 421 + }, + { + "epoch": 0.00422, + "grad_norm": 0.9387048270351058, + "learning_rate": 0.001266, + "loss": 5.5395, + "step": 422 + }, + { + "epoch": 0.00423, + "grad_norm": 0.738543672170714, + "learning_rate": 0.001269, + "loss": 5.5309, + "step": 423 + }, + { + "epoch": 0.00424, + "grad_norm": 0.879163854006119, + "learning_rate": 0.001272, + "loss": 5.5379, + "step": 424 + }, + { + "epoch": 0.00425, + "grad_norm": 0.878245832078137, + "learning_rate": 0.001275, + "loss": 5.5393, + "step": 425 + }, + { + "epoch": 0.00426, + "grad_norm": 0.8393572375675296, + "learning_rate": 0.001278, + "loss": 5.5388, + "step": 426 + }, + { + "epoch": 0.00427, + "grad_norm": 0.8175993205302655, + "learning_rate": 0.001281, + "loss": 5.5188, + "step": 427 + }, + { + "epoch": 0.00428, + "grad_norm": 0.8492227718152501, + "learning_rate": 0.001284, + "loss": 5.5011, + "step": 428 + }, + { + "epoch": 0.00429, + "grad_norm": 0.8455500448937461, + "learning_rate": 0.001287, + "loss": 5.5167, + "step": 429 + }, + { + "epoch": 0.0043, + "grad_norm": 0.9588196540360735, + "learning_rate": 0.00129, + "loss": 5.5126, + "step": 430 + }, + { + "epoch": 0.00431, + "grad_norm": 1.0358439149859766, + "learning_rate": 0.001293, + "loss": 5.5121, + "step": 431 + }, + { + "epoch": 0.00432, + "grad_norm": 0.853137595287236, + "learning_rate": 0.001296, + "loss": 5.5152, + "step": 432 + }, + { + "epoch": 0.00433, + "grad_norm": 0.9144896540159448, + "learning_rate": 0.001299, + "loss": 5.5075, + "step": 433 + }, + { + "epoch": 0.00434, + "grad_norm": 1.0340397416077374, + "learning_rate": 0.001302, + "loss": 5.5131, + "step": 434 + }, + { + "epoch": 0.00435, + "grad_norm": 1.1136200661191735, + "learning_rate": 0.001305, + "loss": 5.5153, + "step": 435 + }, + { + "epoch": 0.00436, + "grad_norm": 0.7998503424321469, + "learning_rate": 0.001308, + "loss": 5.4814, + "step": 436 + }, + { + "epoch": 0.00437, + "grad_norm": 0.8862208467810537, + "learning_rate": 0.001311, + "loss": 5.5052, + "step": 437 + }, + { + "epoch": 0.00438, + "grad_norm": 0.85557749799579, + "learning_rate": 0.001314, + "loss": 5.4855, + "step": 438 + }, + { + "epoch": 0.00439, + "grad_norm": 0.6596001138977952, + "learning_rate": 0.001317, + "loss": 5.5056, + "step": 439 + }, + { + "epoch": 0.0044, + "grad_norm": 0.5461926920380444, + "learning_rate": 0.00132, + "loss": 5.4734, + "step": 440 + }, + { + "epoch": 0.00441, + "grad_norm": 0.5325344576484976, + "learning_rate": 0.001323, + "loss": 5.4692, + "step": 441 + }, + { + "epoch": 0.00442, + "grad_norm": 0.46029396349038315, + "learning_rate": 0.0013260000000000001, + "loss": 5.4603, + "step": 442 + }, + { + "epoch": 0.00443, + "grad_norm": 0.5200620875251907, + "learning_rate": 0.001329, + "loss": 5.4641, + "step": 443 + }, + { + "epoch": 0.00444, + "grad_norm": 0.511034817927936, + "learning_rate": 0.001332, + "loss": 5.4632, + "step": 444 + }, + { + "epoch": 0.00445, + "grad_norm": 0.61375364791033, + "learning_rate": 0.001335, + "loss": 5.4483, + "step": 445 + }, + { + "epoch": 0.00446, + "grad_norm": 0.7540282970336214, + "learning_rate": 0.001338, + "loss": 5.4549, + "step": 446 + }, + { + "epoch": 0.00447, + "grad_norm": 0.7743861790351634, + "learning_rate": 0.001341, + "loss": 5.456, + "step": 447 + }, + { + "epoch": 0.00448, + "grad_norm": 0.6949785247448689, + "learning_rate": 0.0013440000000000001, + "loss": 5.4375, + "step": 448 + }, + { + "epoch": 0.00449, + "grad_norm": 0.8972954522362333, + "learning_rate": 0.001347, + "loss": 5.4453, + "step": 449 + }, + { + "epoch": 0.0045, + "grad_norm": 1.0136292885280909, + "learning_rate": 0.00135, + "loss": 5.4524, + "step": 450 + }, + { + "epoch": 0.00451, + "grad_norm": 0.7959348815359711, + "learning_rate": 0.001353, + "loss": 5.4372, + "step": 451 + }, + { + "epoch": 0.00452, + "grad_norm": 0.750530581913797, + "learning_rate": 0.001356, + "loss": 5.4212, + "step": 452 + }, + { + "epoch": 0.00453, + "grad_norm": 0.718332283553841, + "learning_rate": 0.001359, + "loss": 5.4094, + "step": 453 + }, + { + "epoch": 0.00454, + "grad_norm": 0.8243339574967999, + "learning_rate": 0.0013620000000000001, + "loss": 5.4327, + "step": 454 + }, + { + "epoch": 0.00455, + "grad_norm": 0.8060545663764288, + "learning_rate": 0.0013650000000000001, + "loss": 5.4278, + "step": 455 + }, + { + "epoch": 0.00456, + "grad_norm": 0.9387057405661987, + "learning_rate": 0.001368, + "loss": 5.4287, + "step": 456 + }, + { + "epoch": 0.00457, + "grad_norm": 1.110172512819111, + "learning_rate": 0.001371, + "loss": 5.4304, + "step": 457 + }, + { + "epoch": 0.00458, + "grad_norm": 0.7485802071411273, + "learning_rate": 0.001374, + "loss": 5.4279, + "step": 458 + }, + { + "epoch": 0.00459, + "grad_norm": 0.846395295484429, + "learning_rate": 0.0013770000000000002, + "loss": 5.4177, + "step": 459 + }, + { + "epoch": 0.0046, + "grad_norm": 1.2095188964594632, + "learning_rate": 0.0013800000000000002, + "loss": 5.4166, + "step": 460 + }, + { + "epoch": 0.00461, + "grad_norm": 1.1548058188436976, + "learning_rate": 0.0013830000000000001, + "loss": 5.417, + "step": 461 + }, + { + "epoch": 0.00462, + "grad_norm": 0.9626057997692408, + "learning_rate": 0.001386, + "loss": 5.4177, + "step": 462 + }, + { + "epoch": 0.00463, + "grad_norm": 1.1365427244526745, + "learning_rate": 0.001389, + "loss": 5.4083, + "step": 463 + }, + { + "epoch": 0.00464, + "grad_norm": 0.7154214826672701, + "learning_rate": 0.001392, + "loss": 5.4147, + "step": 464 + }, + { + "epoch": 0.00465, + "grad_norm": 0.5933778225768791, + "learning_rate": 0.0013950000000000002, + "loss": 5.3806, + "step": 465 + }, + { + "epoch": 0.00466, + "grad_norm": 0.6213055581786315, + "learning_rate": 0.0013980000000000002, + "loss": 5.3973, + "step": 466 + }, + { + "epoch": 0.00467, + "grad_norm": 0.5608640811659587, + "learning_rate": 0.0014010000000000001, + "loss": 5.3871, + "step": 467 + }, + { + "epoch": 0.00468, + "grad_norm": 0.4459725756410885, + "learning_rate": 0.001404, + "loss": 5.3713, + "step": 468 + }, + { + "epoch": 0.00469, + "grad_norm": 0.46857319789964524, + "learning_rate": 0.001407, + "loss": 5.3733, + "step": 469 + }, + { + "epoch": 0.0047, + "grad_norm": 0.4864537455422831, + "learning_rate": 0.00141, + "loss": 5.3823, + "step": 470 + }, + { + "epoch": 0.00471, + "grad_norm": 0.5233417273033707, + "learning_rate": 0.001413, + "loss": 5.3595, + "step": 471 + }, + { + "epoch": 0.00472, + "grad_norm": 0.7276814872840428, + "learning_rate": 0.001416, + "loss": 5.376, + "step": 472 + }, + { + "epoch": 0.00473, + "grad_norm": 0.9313958457119089, + "learning_rate": 0.001419, + "loss": 5.3908, + "step": 473 + }, + { + "epoch": 0.00474, + "grad_norm": 0.9969581851520253, + "learning_rate": 0.0014219999999999999, + "loss": 5.3782, + "step": 474 + }, + { + "epoch": 0.00475, + "grad_norm": 0.7172709684261298, + "learning_rate": 0.001425, + "loss": 5.3626, + "step": 475 + }, + { + "epoch": 0.00476, + "grad_norm": 0.749267541315669, + "learning_rate": 0.001428, + "loss": 5.395, + "step": 476 + }, + { + "epoch": 0.00477, + "grad_norm": 0.7586220730764037, + "learning_rate": 0.001431, + "loss": 5.35, + "step": 477 + }, + { + "epoch": 0.00478, + "grad_norm": 0.8456343691773762, + "learning_rate": 0.001434, + "loss": 5.378, + "step": 478 + }, + { + "epoch": 0.00479, + "grad_norm": 1.0643326937248148, + "learning_rate": 0.001437, + "loss": 5.3628, + "step": 479 + }, + { + "epoch": 0.0048, + "grad_norm": 0.9414664669888669, + "learning_rate": 0.0014399999999999999, + "loss": 5.3552, + "step": 480 + }, + { + "epoch": 0.00481, + "grad_norm": 0.9003358375425434, + "learning_rate": 0.001443, + "loss": 5.3593, + "step": 481 + }, + { + "epoch": 0.00482, + "grad_norm": 0.8978331557974801, + "learning_rate": 0.001446, + "loss": 5.3398, + "step": 482 + }, + { + "epoch": 0.00483, + "grad_norm": 1.1289938883763697, + "learning_rate": 0.001449, + "loss": 5.3578, + "step": 483 + }, + { + "epoch": 0.00484, + "grad_norm": 1.1013965257300222, + "learning_rate": 0.001452, + "loss": 5.3611, + "step": 484 + }, + { + "epoch": 0.00485, + "grad_norm": 0.9590590729365723, + "learning_rate": 0.001455, + "loss": 5.3653, + "step": 485 + }, + { + "epoch": 0.00486, + "grad_norm": 0.8987758447866343, + "learning_rate": 0.001458, + "loss": 5.3363, + "step": 486 + }, + { + "epoch": 0.00487, + "grad_norm": 1.1924974582473045, + "learning_rate": 0.001461, + "loss": 5.3676, + "step": 487 + }, + { + "epoch": 0.00488, + "grad_norm": 0.945268328759551, + "learning_rate": 0.001464, + "loss": 5.3393, + "step": 488 + }, + { + "epoch": 0.00489, + "grad_norm": 0.9295239913179795, + "learning_rate": 0.001467, + "loss": 5.3304, + "step": 489 + }, + { + "epoch": 0.0049, + "grad_norm": 0.8440135575742712, + "learning_rate": 0.00147, + "loss": 5.3401, + "step": 490 + }, + { + "epoch": 0.00491, + "grad_norm": 0.8518595255895461, + "learning_rate": 0.001473, + "loss": 5.3387, + "step": 491 + }, + { + "epoch": 0.00492, + "grad_norm": 0.8521564820940866, + "learning_rate": 0.001476, + "loss": 5.3306, + "step": 492 + }, + { + "epoch": 0.00493, + "grad_norm": 1.037786125041147, + "learning_rate": 0.001479, + "loss": 5.337, + "step": 493 + }, + { + "epoch": 0.00494, + "grad_norm": 0.8602402216970677, + "learning_rate": 0.001482, + "loss": 5.3365, + "step": 494 + }, + { + "epoch": 0.00495, + "grad_norm": 0.833325441881814, + "learning_rate": 0.001485, + "loss": 5.3402, + "step": 495 + }, + { + "epoch": 0.00496, + "grad_norm": 0.7919536178225737, + "learning_rate": 0.001488, + "loss": 5.324, + "step": 496 + }, + { + "epoch": 0.00497, + "grad_norm": 0.873050532284563, + "learning_rate": 0.001491, + "loss": 5.3076, + "step": 497 + }, + { + "epoch": 0.00498, + "grad_norm": 0.8401174504098841, + "learning_rate": 0.001494, + "loss": 5.3331, + "step": 498 + }, + { + "epoch": 0.00499, + "grad_norm": 0.7077904347820982, + "learning_rate": 0.001497, + "loss": 5.3023, + "step": 499 + }, + { + "epoch": 0.005, + "grad_norm": 0.722193970534152, + "learning_rate": 0.0015, + "loss": 5.3084, + "step": 500 + }, + { + "epoch": 0.00501, + "grad_norm": 0.6788524699363854, + "learning_rate": 0.001503, + "loss": 5.3068, + "step": 501 + }, + { + "epoch": 0.00502, + "grad_norm": 0.4688637134987383, + "learning_rate": 0.001506, + "loss": 5.3027, + "step": 502 + }, + { + "epoch": 0.00503, + "grad_norm": 0.5347885266184398, + "learning_rate": 0.0015090000000000001, + "loss": 5.3107, + "step": 503 + }, + { + "epoch": 0.00504, + "grad_norm": 0.4568589587316421, + "learning_rate": 0.001512, + "loss": 5.2883, + "step": 504 + }, + { + "epoch": 0.00505, + "grad_norm": 0.6059364079243326, + "learning_rate": 0.001515, + "loss": 5.2843, + "step": 505 + }, + { + "epoch": 0.00506, + "grad_norm": 0.5690871559871206, + "learning_rate": 0.001518, + "loss": 5.2813, + "step": 506 + }, + { + "epoch": 0.00507, + "grad_norm": 0.6408278920334604, + "learning_rate": 0.001521, + "loss": 5.279, + "step": 507 + }, + { + "epoch": 0.00508, + "grad_norm": 0.6374798421453328, + "learning_rate": 0.001524, + "loss": 5.2759, + "step": 508 + }, + { + "epoch": 0.00509, + "grad_norm": 0.5508180980448808, + "learning_rate": 0.0015270000000000001, + "loss": 5.2603, + "step": 509 + }, + { + "epoch": 0.0051, + "grad_norm": 0.4678264081626914, + "learning_rate": 0.0015300000000000001, + "loss": 5.2734, + "step": 510 + }, + { + "epoch": 0.00511, + "grad_norm": 0.4947375080326457, + "learning_rate": 0.001533, + "loss": 5.252, + "step": 511 + }, + { + "epoch": 0.00512, + "grad_norm": 0.4646141343190718, + "learning_rate": 0.001536, + "loss": 5.2454, + "step": 512 + }, + { + "epoch": 0.00513, + "grad_norm": 0.4614608859817223, + "learning_rate": 0.001539, + "loss": 5.2616, + "step": 513 + }, + { + "epoch": 0.00514, + "grad_norm": 0.47514560040822135, + "learning_rate": 0.001542, + "loss": 5.2462, + "step": 514 + }, + { + "epoch": 0.00515, + "grad_norm": 0.5516172292908152, + "learning_rate": 0.0015450000000000001, + "loss": 5.2495, + "step": 515 + }, + { + "epoch": 0.00516, + "grad_norm": 0.6762883843976203, + "learning_rate": 0.0015480000000000001, + "loss": 5.2514, + "step": 516 + }, + { + "epoch": 0.00517, + "grad_norm": 0.8220059659316783, + "learning_rate": 0.001551, + "loss": 5.2491, + "step": 517 + }, + { + "epoch": 0.00518, + "grad_norm": 0.8472303407674783, + "learning_rate": 0.001554, + "loss": 5.2384, + "step": 518 + }, + { + "epoch": 0.00519, + "grad_norm": 0.8352554081583162, + "learning_rate": 0.001557, + "loss": 5.2403, + "step": 519 + }, + { + "epoch": 0.0052, + "grad_norm": 0.9289527635802941, + "learning_rate": 0.0015600000000000002, + "loss": 5.2527, + "step": 520 + }, + { + "epoch": 0.00521, + "grad_norm": 1.112613217213008, + "learning_rate": 0.0015630000000000002, + "loss": 5.242, + "step": 521 + }, + { + "epoch": 0.00522, + "grad_norm": 0.9402831205824588, + "learning_rate": 0.0015660000000000001, + "loss": 5.2388, + "step": 522 + }, + { + "epoch": 0.00523, + "grad_norm": 1.103453237567065, + "learning_rate": 0.001569, + "loss": 5.2627, + "step": 523 + }, + { + "epoch": 0.00524, + "grad_norm": 1.0118007660209627, + "learning_rate": 0.001572, + "loss": 5.268, + "step": 524 + }, + { + "epoch": 0.00525, + "grad_norm": 0.9435955195760587, + "learning_rate": 0.001575, + "loss": 5.2606, + "step": 525 + }, + { + "epoch": 0.00526, + "grad_norm": 0.8731456366832249, + "learning_rate": 0.0015780000000000002, + "loss": 5.2355, + "step": 526 + }, + { + "epoch": 0.00527, + "grad_norm": 0.7556141048030679, + "learning_rate": 0.0015810000000000002, + "loss": 5.244, + "step": 527 + }, + { + "epoch": 0.00528, + "grad_norm": 0.8959181971092488, + "learning_rate": 0.0015840000000000001, + "loss": 5.2287, + "step": 528 + }, + { + "epoch": 0.00529, + "grad_norm": 1.10956087267792, + "learning_rate": 0.001587, + "loss": 5.2485, + "step": 529 + }, + { + "epoch": 0.0053, + "grad_norm": 0.7373688170406403, + "learning_rate": 0.00159, + "loss": 5.2473, + "step": 530 + }, + { + "epoch": 0.00531, + "grad_norm": 0.9440929198745536, + "learning_rate": 0.001593, + "loss": 5.2364, + "step": 531 + }, + { + "epoch": 0.00532, + "grad_norm": 1.2761807884581948, + "learning_rate": 0.0015960000000000002, + "loss": 5.2627, + "step": 532 + }, + { + "epoch": 0.00533, + "grad_norm": 0.7254918561906474, + "learning_rate": 0.0015990000000000002, + "loss": 5.2411, + "step": 533 + }, + { + "epoch": 0.00534, + "grad_norm": 0.9458095799105631, + "learning_rate": 0.0016020000000000001, + "loss": 5.2364, + "step": 534 + }, + { + "epoch": 0.00535, + "grad_norm": 1.1518877149230409, + "learning_rate": 0.001605, + "loss": 5.2538, + "step": 535 + }, + { + "epoch": 0.00536, + "grad_norm": 1.0003795894462209, + "learning_rate": 0.001608, + "loss": 5.2516, + "step": 536 + }, + { + "epoch": 0.00537, + "grad_norm": 1.0837750602504435, + "learning_rate": 0.0016110000000000002, + "loss": 5.245, + "step": 537 + }, + { + "epoch": 0.00538, + "grad_norm": 0.8341741073056014, + "learning_rate": 0.0016140000000000002, + "loss": 5.2215, + "step": 538 + }, + { + "epoch": 0.00539, + "grad_norm": 0.8405248729905289, + "learning_rate": 0.0016170000000000002, + "loss": 5.218, + "step": 539 + }, + { + "epoch": 0.0054, + "grad_norm": 1.0652634270364731, + "learning_rate": 0.0016200000000000001, + "loss": 5.2348, + "step": 540 + }, + { + "epoch": 0.00541, + "grad_norm": 1.02504417487003, + "learning_rate": 0.001623, + "loss": 5.2275, + "step": 541 + }, + { + "epoch": 0.00542, + "grad_norm": 1.0410836109463524, + "learning_rate": 0.001626, + "loss": 5.2461, + "step": 542 + }, + { + "epoch": 0.00543, + "grad_norm": 1.1820143116937272, + "learning_rate": 0.0016290000000000002, + "loss": 5.2251, + "step": 543 + }, + { + "epoch": 0.00544, + "grad_norm": 0.7766425096978492, + "learning_rate": 0.0016320000000000002, + "loss": 5.227, + "step": 544 + }, + { + "epoch": 0.00545, + "grad_norm": 0.7162190734488068, + "learning_rate": 0.0016350000000000002, + "loss": 5.2344, + "step": 545 + }, + { + "epoch": 0.00546, + "grad_norm": 0.6138609629035112, + "learning_rate": 0.0016380000000000001, + "loss": 5.1997, + "step": 546 + }, + { + "epoch": 0.00547, + "grad_norm": 0.5845964576550613, + "learning_rate": 0.001641, + "loss": 5.199, + "step": 547 + }, + { + "epoch": 0.00548, + "grad_norm": 0.5000045746748227, + "learning_rate": 0.001644, + "loss": 5.1903, + "step": 548 + }, + { + "epoch": 0.00549, + "grad_norm": 0.45001097530426093, + "learning_rate": 0.0016470000000000002, + "loss": 5.1848, + "step": 549 + }, + { + "epoch": 0.0055, + "grad_norm": 0.48852089918970865, + "learning_rate": 0.0016500000000000002, + "loss": 5.1966, + "step": 550 + }, + { + "epoch": 0.00551, + "grad_norm": 0.4504277526474317, + "learning_rate": 0.0016530000000000002, + "loss": 5.1901, + "step": 551 + }, + { + "epoch": 0.00552, + "grad_norm": 0.3849177997836439, + "learning_rate": 0.0016560000000000001, + "loss": 5.1829, + "step": 552 + }, + { + "epoch": 0.00553, + "grad_norm": 0.3648459807382568, + "learning_rate": 0.001659, + "loss": 5.1708, + "step": 553 + }, + { + "epoch": 0.00554, + "grad_norm": 0.36806425815097327, + "learning_rate": 0.0016620000000000003, + "loss": 5.1683, + "step": 554 + }, + { + "epoch": 0.00555, + "grad_norm": 0.399857849170916, + "learning_rate": 0.0016650000000000002, + "loss": 5.174, + "step": 555 + }, + { + "epoch": 0.00556, + "grad_norm": 0.5317181922092629, + "learning_rate": 0.0016680000000000002, + "loss": 5.1608, + "step": 556 + }, + { + "epoch": 0.00557, + "grad_norm": 0.7831777956168835, + "learning_rate": 0.0016710000000000002, + "loss": 5.1496, + "step": 557 + }, + { + "epoch": 0.00558, + "grad_norm": 1.0341253030675983, + "learning_rate": 0.0016740000000000001, + "loss": 5.1915, + "step": 558 + }, + { + "epoch": 0.00559, + "grad_norm": 0.9310990839840096, + "learning_rate": 0.001677, + "loss": 5.1724, + "step": 559 + }, + { + "epoch": 0.0056, + "grad_norm": 1.1320536331440925, + "learning_rate": 0.0016800000000000003, + "loss": 5.1553, + "step": 560 + }, + { + "epoch": 0.00561, + "grad_norm": 1.0798635762834115, + "learning_rate": 0.0016830000000000003, + "loss": 5.1888, + "step": 561 + }, + { + "epoch": 0.00562, + "grad_norm": 0.9955409946448898, + "learning_rate": 0.0016860000000000002, + "loss": 5.1785, + "step": 562 + }, + { + "epoch": 0.00563, + "grad_norm": 0.8046398619546209, + "learning_rate": 0.001689, + "loss": 5.1727, + "step": 563 + }, + { + "epoch": 0.00564, + "grad_norm": 0.8640806614647399, + "learning_rate": 0.001692, + "loss": 5.1564, + "step": 564 + }, + { + "epoch": 0.00565, + "grad_norm": 0.9094590673653183, + "learning_rate": 0.001695, + "loss": 5.1591, + "step": 565 + }, + { + "epoch": 0.00566, + "grad_norm": 1.0763832222921514, + "learning_rate": 0.0016979999999999999, + "loss": 5.1686, + "step": 566 + }, + { + "epoch": 0.00567, + "grad_norm": 0.9106969464088174, + "learning_rate": 0.0017009999999999998, + "loss": 5.1536, + "step": 567 + }, + { + "epoch": 0.00568, + "grad_norm": 1.0612496613421336, + "learning_rate": 0.0017039999999999998, + "loss": 5.175, + "step": 568 + }, + { + "epoch": 0.00569, + "grad_norm": 1.030428797640198, + "learning_rate": 0.001707, + "loss": 5.1588, + "step": 569 + }, + { + "epoch": 0.0057, + "grad_norm": 1.0168109110540526, + "learning_rate": 0.00171, + "loss": 5.1716, + "step": 570 + }, + { + "epoch": 0.00571, + "grad_norm": 1.1348044931505399, + "learning_rate": 0.001713, + "loss": 5.1856, + "step": 571 + }, + { + "epoch": 0.00572, + "grad_norm": 0.761146802779985, + "learning_rate": 0.0017159999999999999, + "loss": 5.1497, + "step": 572 + }, + { + "epoch": 0.00573, + "grad_norm": 0.7662650183514703, + "learning_rate": 0.0017189999999999998, + "loss": 5.1536, + "step": 573 + }, + { + "epoch": 0.00574, + "grad_norm": 0.8733185334950944, + "learning_rate": 0.001722, + "loss": 5.1641, + "step": 574 + }, + { + "epoch": 0.00575, + "grad_norm": 0.8629871927891676, + "learning_rate": 0.001725, + "loss": 5.142, + "step": 575 + }, + { + "epoch": 0.00576, + "grad_norm": 0.8467410686880783, + "learning_rate": 0.001728, + "loss": 5.137, + "step": 576 + }, + { + "epoch": 0.00577, + "grad_norm": 0.8007452144545169, + "learning_rate": 0.001731, + "loss": 5.1615, + "step": 577 + }, + { + "epoch": 0.00578, + "grad_norm": 0.6405300314411674, + "learning_rate": 0.0017339999999999999, + "loss": 5.1325, + "step": 578 + }, + { + "epoch": 0.00579, + "grad_norm": 0.70587507977218, + "learning_rate": 0.0017369999999999998, + "loss": 5.1208, + "step": 579 + }, + { + "epoch": 0.0058, + "grad_norm": 0.6985854652221088, + "learning_rate": 0.00174, + "loss": 5.1378, + "step": 580 + }, + { + "epoch": 0.00581, + "grad_norm": 0.6137529113435678, + "learning_rate": 0.001743, + "loss": 5.1348, + "step": 581 + }, + { + "epoch": 0.00582, + "grad_norm": 0.48264514730128444, + "learning_rate": 0.001746, + "loss": 5.1045, + "step": 582 + }, + { + "epoch": 0.00583, + "grad_norm": 0.5771367675923135, + "learning_rate": 0.001749, + "loss": 5.1293, + "step": 583 + }, + { + "epoch": 0.00584, + "grad_norm": 0.6936037332792736, + "learning_rate": 0.0017519999999999999, + "loss": 5.1229, + "step": 584 + }, + { + "epoch": 0.00585, + "grad_norm": 0.7595444862659105, + "learning_rate": 0.0017549999999999998, + "loss": 5.1095, + "step": 585 + }, + { + "epoch": 0.00586, + "grad_norm": 0.8605796861767604, + "learning_rate": 0.001758, + "loss": 5.1202, + "step": 586 + }, + { + "epoch": 0.00587, + "grad_norm": 0.907920304070859, + "learning_rate": 0.001761, + "loss": 5.1079, + "step": 587 + }, + { + "epoch": 0.00588, + "grad_norm": 0.9246177820411534, + "learning_rate": 0.001764, + "loss": 5.1163, + "step": 588 + }, + { + "epoch": 0.00589, + "grad_norm": 0.8704696950275989, + "learning_rate": 0.001767, + "loss": 5.1211, + "step": 589 + }, + { + "epoch": 0.0059, + "grad_norm": 0.7193563853719346, + "learning_rate": 0.0017699999999999999, + "loss": 5.101, + "step": 590 + }, + { + "epoch": 0.00591, + "grad_norm": 0.6949293332158505, + "learning_rate": 0.001773, + "loss": 5.0928, + "step": 591 + }, + { + "epoch": 0.00592, + "grad_norm": 0.654055845037553, + "learning_rate": 0.001776, + "loss": 5.1147, + "step": 592 + }, + { + "epoch": 0.00593, + "grad_norm": 0.7140132542710956, + "learning_rate": 0.001779, + "loss": 5.1161, + "step": 593 + }, + { + "epoch": 0.00594, + "grad_norm": 0.6270820258277058, + "learning_rate": 0.001782, + "loss": 5.107, + "step": 594 + }, + { + "epoch": 0.00595, + "grad_norm": 0.5627466921921568, + "learning_rate": 0.001785, + "loss": 5.0962, + "step": 595 + }, + { + "epoch": 0.00596, + "grad_norm": 0.6408060930916085, + "learning_rate": 0.0017879999999999999, + "loss": 5.0966, + "step": 596 + }, + { + "epoch": 0.00597, + "grad_norm": 0.7315483645065906, + "learning_rate": 0.001791, + "loss": 5.077, + "step": 597 + }, + { + "epoch": 0.00598, + "grad_norm": 0.7860917802127037, + "learning_rate": 0.001794, + "loss": 5.0917, + "step": 598 + }, + { + "epoch": 0.00599, + "grad_norm": 0.7396694082174577, + "learning_rate": 0.001797, + "loss": 5.0901, + "step": 599 + }, + { + "epoch": 0.006, + "grad_norm": 0.678011019966431, + "learning_rate": 0.0018, + "loss": 5.0859, + "step": 600 + }, + { + "epoch": 0.00601, + "grad_norm": 0.8039567544190375, + "learning_rate": 0.001803, + "loss": 5.0817, + "step": 601 + }, + { + "epoch": 0.00602, + "grad_norm": 0.8362811130198314, + "learning_rate": 0.0018059999999999999, + "loss": 5.0736, + "step": 602 + }, + { + "epoch": 0.00603, + "grad_norm": 0.7437701841167941, + "learning_rate": 0.001809, + "loss": 5.0897, + "step": 603 + }, + { + "epoch": 0.00604, + "grad_norm": 0.7865788629191941, + "learning_rate": 0.001812, + "loss": 5.0711, + "step": 604 + }, + { + "epoch": 0.00605, + "grad_norm": 0.7065527487675252, + "learning_rate": 0.001815, + "loss": 5.0798, + "step": 605 + }, + { + "epoch": 0.00606, + "grad_norm": 0.6576721805135337, + "learning_rate": 0.001818, + "loss": 5.0782, + "step": 606 + }, + { + "epoch": 0.00607, + "grad_norm": 0.6528064751937639, + "learning_rate": 0.001821, + "loss": 5.0831, + "step": 607 + }, + { + "epoch": 0.00608, + "grad_norm": 0.6182798778507802, + "learning_rate": 0.001824, + "loss": 5.0615, + "step": 608 + }, + { + "epoch": 0.00609, + "grad_norm": 0.8163285355074863, + "learning_rate": 0.001827, + "loss": 5.079, + "step": 609 + }, + { + "epoch": 0.0061, + "grad_norm": 0.795800426894062, + "learning_rate": 0.00183, + "loss": 5.053, + "step": 610 + }, + { + "epoch": 0.00611, + "grad_norm": 0.7510077277929101, + "learning_rate": 0.001833, + "loss": 5.0545, + "step": 611 + }, + { + "epoch": 0.00612, + "grad_norm": 0.7863529213037952, + "learning_rate": 0.001836, + "loss": 5.0625, + "step": 612 + }, + { + "epoch": 0.00613, + "grad_norm": 0.718541559578873, + "learning_rate": 0.001839, + "loss": 5.0526, + "step": 613 + }, + { + "epoch": 0.00614, + "grad_norm": 0.648240484004962, + "learning_rate": 0.001842, + "loss": 5.0729, + "step": 614 + }, + { + "epoch": 0.00615, + "grad_norm": 0.6572100797643513, + "learning_rate": 0.001845, + "loss": 5.0607, + "step": 615 + }, + { + "epoch": 0.00616, + "grad_norm": 0.598653088729871, + "learning_rate": 0.001848, + "loss": 5.0431, + "step": 616 + }, + { + "epoch": 0.00617, + "grad_norm": 0.6867319038647199, + "learning_rate": 0.001851, + "loss": 5.058, + "step": 617 + }, + { + "epoch": 0.00618, + "grad_norm": 0.8065427766951909, + "learning_rate": 0.001854, + "loss": 5.0471, + "step": 618 + }, + { + "epoch": 0.00619, + "grad_norm": 0.952336607614539, + "learning_rate": 0.001857, + "loss": 5.0614, + "step": 619 + }, + { + "epoch": 0.0062, + "grad_norm": 0.937144329881144, + "learning_rate": 0.00186, + "loss": 5.0503, + "step": 620 + }, + { + "epoch": 0.00621, + "grad_norm": 0.861992409216581, + "learning_rate": 0.001863, + "loss": 5.0519, + "step": 621 + }, + { + "epoch": 0.00622, + "grad_norm": 0.7937303039626423, + "learning_rate": 0.001866, + "loss": 5.046, + "step": 622 + }, + { + "epoch": 0.00623, + "grad_norm": 0.7729108287648654, + "learning_rate": 0.001869, + "loss": 5.0315, + "step": 623 + }, + { + "epoch": 0.00624, + "grad_norm": 0.7644499866680247, + "learning_rate": 0.001872, + "loss": 5.0449, + "step": 624 + }, + { + "epoch": 0.00625, + "grad_norm": 0.833886296843224, + "learning_rate": 0.001875, + "loss": 5.0598, + "step": 625 + }, + { + "epoch": 0.00626, + "grad_norm": 0.8372671439007264, + "learning_rate": 0.0018780000000000001, + "loss": 5.0351, + "step": 626 + }, + { + "epoch": 0.00627, + "grad_norm": 0.8603722936304884, + "learning_rate": 0.001881, + "loss": 5.0306, + "step": 627 + }, + { + "epoch": 0.00628, + "grad_norm": 0.7718433014027412, + "learning_rate": 0.001884, + "loss": 5.0548, + "step": 628 + }, + { + "epoch": 0.00629, + "grad_norm": 0.6787638881219394, + "learning_rate": 0.001887, + "loss": 5.0263, + "step": 629 + }, + { + "epoch": 0.0063, + "grad_norm": 0.6841857077673691, + "learning_rate": 0.00189, + "loss": 5.0201, + "step": 630 + }, + { + "epoch": 0.00631, + "grad_norm": 0.7297958952205305, + "learning_rate": 0.0018930000000000002, + "loss": 5.0163, + "step": 631 + }, + { + "epoch": 0.00632, + "grad_norm": 0.8611511166687711, + "learning_rate": 0.0018960000000000001, + "loss": 5.0388, + "step": 632 + }, + { + "epoch": 0.00633, + "grad_norm": 0.9514927337678543, + "learning_rate": 0.001899, + "loss": 5.0438, + "step": 633 + }, + { + "epoch": 0.00634, + "grad_norm": 0.8956137982105657, + "learning_rate": 0.001902, + "loss": 5.031, + "step": 634 + }, + { + "epoch": 0.00635, + "grad_norm": 0.952460918530188, + "learning_rate": 0.001905, + "loss": 5.0218, + "step": 635 + }, + { + "epoch": 0.00636, + "grad_norm": 1.0747774477703274, + "learning_rate": 0.001908, + "loss": 5.0402, + "step": 636 + }, + { + "epoch": 0.00637, + "grad_norm": 0.8336825219104492, + "learning_rate": 0.0019110000000000002, + "loss": 5.0466, + "step": 637 + }, + { + "epoch": 0.00638, + "grad_norm": 0.8986808700713379, + "learning_rate": 0.0019140000000000001, + "loss": 5.0353, + "step": 638 + }, + { + "epoch": 0.00639, + "grad_norm": 1.0369332378334974, + "learning_rate": 0.001917, + "loss": 5.0213, + "step": 639 + }, + { + "epoch": 0.0064, + "grad_norm": 1.0511699002679498, + "learning_rate": 0.00192, + "loss": 5.0287, + "step": 640 + }, + { + "epoch": 0.00641, + "grad_norm": 1.0967175650217074, + "learning_rate": 0.001923, + "loss": 5.0426, + "step": 641 + }, + { + "epoch": 0.00642, + "grad_norm": 0.9933027226251017, + "learning_rate": 0.001926, + "loss": 5.0352, + "step": 642 + }, + { + "epoch": 0.00643, + "grad_norm": 0.9768777735178511, + "learning_rate": 0.0019290000000000002, + "loss": 5.0523, + "step": 643 + }, + { + "epoch": 0.00644, + "grad_norm": 1.001060805060802, + "learning_rate": 0.0019320000000000001, + "loss": 5.0363, + "step": 644 + }, + { + "epoch": 0.00645, + "grad_norm": 1.0724455418770003, + "learning_rate": 0.001935, + "loss": 5.0522, + "step": 645 + }, + { + "epoch": 0.00646, + "grad_norm": 1.1074452374796593, + "learning_rate": 0.001938, + "loss": 5.0611, + "step": 646 + }, + { + "epoch": 0.00647, + "grad_norm": 0.9675974316517354, + "learning_rate": 0.001941, + "loss": 5.062, + "step": 647 + }, + { + "epoch": 0.00648, + "grad_norm": 0.7269750089664031, + "learning_rate": 0.0019440000000000002, + "loss": 5.0382, + "step": 648 + }, + { + "epoch": 0.00649, + "grad_norm": 0.6773093371497811, + "learning_rate": 0.0019470000000000002, + "loss": 5.0209, + "step": 649 + }, + { + "epoch": 0.0065, + "grad_norm": 0.6721153823123059, + "learning_rate": 0.0019500000000000001, + "loss": 5.0458, + "step": 650 + }, + { + "epoch": 0.00651, + "grad_norm": 0.6257734701697928, + "learning_rate": 0.001953, + "loss": 5.022, + "step": 651 + }, + { + "epoch": 0.00652, + "grad_norm": 0.6412471052264174, + "learning_rate": 0.0019560000000000003, + "loss": 5.0157, + "step": 652 + }, + { + "epoch": 0.00653, + "grad_norm": 0.7030358108381808, + "learning_rate": 0.0019590000000000002, + "loss": 5.0108, + "step": 653 + }, + { + "epoch": 0.00654, + "grad_norm": 0.6417272703555769, + "learning_rate": 0.001962, + "loss": 4.994, + "step": 654 + }, + { + "epoch": 0.00655, + "grad_norm": 0.6123939930365004, + "learning_rate": 0.001965, + "loss": 5.0066, + "step": 655 + }, + { + "epoch": 0.00656, + "grad_norm": 0.6761646659152818, + "learning_rate": 0.001968, + "loss": 4.9975, + "step": 656 + }, + { + "epoch": 0.00657, + "grad_norm": 0.9015512681716132, + "learning_rate": 0.001971, + "loss": 4.9965, + "step": 657 + }, + { + "epoch": 0.00658, + "grad_norm": 1.1070883554163469, + "learning_rate": 0.001974, + "loss": 5.032, + "step": 658 + }, + { + "epoch": 0.00659, + "grad_norm": 0.8166390486888542, + "learning_rate": 0.001977, + "loss": 5.0022, + "step": 659 + }, + { + "epoch": 0.0066, + "grad_norm": 0.7886404692657704, + "learning_rate": 0.00198, + "loss": 4.9952, + "step": 660 + }, + { + "epoch": 0.00661, + "grad_norm": 0.6930157053874706, + "learning_rate": 0.001983, + "loss": 4.9751, + "step": 661 + }, + { + "epoch": 0.00662, + "grad_norm": 0.6021137246671374, + "learning_rate": 0.0019860000000000004, + "loss": 4.9738, + "step": 662 + }, + { + "epoch": 0.00663, + "grad_norm": 0.5342549595827694, + "learning_rate": 0.0019890000000000003, + "loss": 4.9858, + "step": 663 + }, + { + "epoch": 0.00664, + "grad_norm": 0.5549032292488858, + "learning_rate": 0.0019920000000000003, + "loss": 4.9668, + "step": 664 + }, + { + "epoch": 0.00665, + "grad_norm": 0.5630527450969944, + "learning_rate": 0.0019950000000000002, + "loss": 4.9437, + "step": 665 + }, + { + "epoch": 0.00666, + "grad_norm": 0.5436432724284425, + "learning_rate": 0.001998, + "loss": 4.9653, + "step": 666 + }, + { + "epoch": 0.00667, + "grad_norm": 0.5455501294317395, + "learning_rate": 0.002001, + "loss": 4.9611, + "step": 667 + }, + { + "epoch": 0.00668, + "grad_norm": 0.5771340900782089, + "learning_rate": 0.002004, + "loss": 4.9815, + "step": 668 + }, + { + "epoch": 0.00669, + "grad_norm": 0.5741027689717983, + "learning_rate": 0.002007, + "loss": 4.9757, + "step": 669 + }, + { + "epoch": 0.0067, + "grad_norm": 0.6058348461614704, + "learning_rate": 0.00201, + "loss": 4.969, + "step": 670 + }, + { + "epoch": 0.00671, + "grad_norm": 0.566759025441364, + "learning_rate": 0.002013, + "loss": 4.9434, + "step": 671 + }, + { + "epoch": 0.00672, + "grad_norm": 0.5204858890697072, + "learning_rate": 0.002016, + "loss": 4.9491, + "step": 672 + }, + { + "epoch": 0.00673, + "grad_norm": 0.5776654990466389, + "learning_rate": 0.002019, + "loss": 4.9473, + "step": 673 + }, + { + "epoch": 0.00674, + "grad_norm": 0.5899189939484214, + "learning_rate": 0.0020220000000000004, + "loss": 4.9512, + "step": 674 + }, + { + "epoch": 0.00675, + "grad_norm": 0.73017970913878, + "learning_rate": 0.0020250000000000003, + "loss": 4.954, + "step": 675 + }, + { + "epoch": 0.00676, + "grad_norm": 0.8714534217285366, + "learning_rate": 0.0020280000000000003, + "loss": 4.9527, + "step": 676 + }, + { + "epoch": 0.00677, + "grad_norm": 0.9455529231086871, + "learning_rate": 0.0020310000000000003, + "loss": 4.966, + "step": 677 + }, + { + "epoch": 0.00678, + "grad_norm": 0.7654589233287328, + "learning_rate": 0.0020340000000000002, + "loss": 4.9721, + "step": 678 + }, + { + "epoch": 0.00679, + "grad_norm": 0.7453542715563302, + "learning_rate": 0.002037, + "loss": 4.9514, + "step": 679 + }, + { + "epoch": 0.0068, + "grad_norm": 0.8198542095591255, + "learning_rate": 0.00204, + "loss": 4.9599, + "step": 680 + }, + { + "epoch": 0.00681, + "grad_norm": 1.0035327070670854, + "learning_rate": 0.002043, + "loss": 4.9608, + "step": 681 + }, + { + "epoch": 0.00682, + "grad_norm": 1.1272493939475496, + "learning_rate": 0.002046, + "loss": 4.9829, + "step": 682 + }, + { + "epoch": 0.00683, + "grad_norm": 0.9375943776902779, + "learning_rate": 0.002049, + "loss": 4.9573, + "step": 683 + }, + { + "epoch": 0.00684, + "grad_norm": 0.988255846210588, + "learning_rate": 0.002052, + "loss": 4.9739, + "step": 684 + }, + { + "epoch": 0.00685, + "grad_norm": 1.2166901935378913, + "learning_rate": 0.0020550000000000004, + "loss": 4.9877, + "step": 685 + }, + { + "epoch": 0.00686, + "grad_norm": 0.8834239063706651, + "learning_rate": 0.0020580000000000004, + "loss": 4.9625, + "step": 686 + }, + { + "epoch": 0.00687, + "grad_norm": 0.9980240769781185, + "learning_rate": 0.0020610000000000003, + "loss": 4.9587, + "step": 687 + }, + { + "epoch": 0.00688, + "grad_norm": 1.1254414916617237, + "learning_rate": 0.002064, + "loss": 4.9804, + "step": 688 + }, + { + "epoch": 0.00689, + "grad_norm": 1.0306142304378458, + "learning_rate": 0.002067, + "loss": 4.9703, + "step": 689 + }, + { + "epoch": 0.0069, + "grad_norm": 1.1841729949788287, + "learning_rate": 0.00207, + "loss": 4.9832, + "step": 690 + }, + { + "epoch": 0.00691, + "grad_norm": 0.94703321985356, + "learning_rate": 0.0020729999999999998, + "loss": 4.946, + "step": 691 + }, + { + "epoch": 0.00692, + "grad_norm": 1.0500733994835558, + "learning_rate": 0.0020759999999999997, + "loss": 4.9769, + "step": 692 + }, + { + "epoch": 0.00693, + "grad_norm": 0.8703789610160195, + "learning_rate": 0.0020789999999999997, + "loss": 4.9707, + "step": 693 + }, + { + "epoch": 0.00694, + "grad_norm": 0.8616077913099822, + "learning_rate": 0.002082, + "loss": 4.9464, + "step": 694 + }, + { + "epoch": 0.00695, + "grad_norm": 1.1391889561194677, + "learning_rate": 0.002085, + "loss": 4.9626, + "step": 695 + }, + { + "epoch": 0.00696, + "grad_norm": 0.9584412207038412, + "learning_rate": 0.002088, + "loss": 4.9606, + "step": 696 + }, + { + "epoch": 0.00697, + "grad_norm": 1.0011041706204022, + "learning_rate": 0.002091, + "loss": 4.9624, + "step": 697 + }, + { + "epoch": 0.00698, + "grad_norm": 1.3127609957868622, + "learning_rate": 0.002094, + "loss": 4.9728, + "step": 698 + }, + { + "epoch": 0.00699, + "grad_norm": 0.8537920758308664, + "learning_rate": 0.002097, + "loss": 4.9491, + "step": 699 + }, + { + "epoch": 0.007, + "grad_norm": 1.0018321540367086, + "learning_rate": 0.0021, + "loss": 4.9652, + "step": 700 + }, + { + "epoch": 0.00701, + "grad_norm": 0.9481035749179906, + "learning_rate": 0.002103, + "loss": 4.9422, + "step": 701 + }, + { + "epoch": 0.00702, + "grad_norm": 0.8258463386850557, + "learning_rate": 0.002106, + "loss": 4.945, + "step": 702 + }, + { + "epoch": 0.00703, + "grad_norm": 0.6305963851301584, + "learning_rate": 0.0021089999999999998, + "loss": 4.9167, + "step": 703 + }, + { + "epoch": 0.00704, + "grad_norm": 0.6428932601365419, + "learning_rate": 0.0021119999999999997, + "loss": 4.9232, + "step": 704 + }, + { + "epoch": 0.00705, + "grad_norm": 0.6965384868722897, + "learning_rate": 0.002115, + "loss": 4.9266, + "step": 705 + }, + { + "epoch": 0.00706, + "grad_norm": 0.7410784964260257, + "learning_rate": 0.002118, + "loss": 4.9344, + "step": 706 + }, + { + "epoch": 0.00707, + "grad_norm": 0.721339398040246, + "learning_rate": 0.002121, + "loss": 4.9083, + "step": 707 + }, + { + "epoch": 0.00708, + "grad_norm": 0.675102725576719, + "learning_rate": 0.002124, + "loss": 4.8875, + "step": 708 + }, + { + "epoch": 0.00709, + "grad_norm": 0.6428565761724286, + "learning_rate": 0.002127, + "loss": 4.9151, + "step": 709 + }, + { + "epoch": 0.0071, + "grad_norm": 0.5795546215215135, + "learning_rate": 0.00213, + "loss": 4.8959, + "step": 710 + }, + { + "epoch": 0.00711, + "grad_norm": 0.6667417841414063, + "learning_rate": 0.002133, + "loss": 4.8739, + "step": 711 + }, + { + "epoch": 0.00712, + "grad_norm": 0.7094343142489271, + "learning_rate": 0.002136, + "loss": 4.8945, + "step": 712 + }, + { + "epoch": 0.00713, + "grad_norm": 0.6503826772007358, + "learning_rate": 0.002139, + "loss": 4.9061, + "step": 713 + }, + { + "epoch": 0.00714, + "grad_norm": 0.5916220730293257, + "learning_rate": 0.002142, + "loss": 4.8928, + "step": 714 + }, + { + "epoch": 0.00715, + "grad_norm": 0.6248268778602033, + "learning_rate": 0.0021449999999999998, + "loss": 4.8651, + "step": 715 + }, + { + "epoch": 0.00716, + "grad_norm": 0.7056512941727522, + "learning_rate": 0.002148, + "loss": 4.8619, + "step": 716 + }, + { + "epoch": 0.00717, + "grad_norm": 0.7609530945312164, + "learning_rate": 0.002151, + "loss": 4.8823, + "step": 717 + }, + { + "epoch": 0.00718, + "grad_norm": 0.7494152535556156, + "learning_rate": 0.002154, + "loss": 4.8579, + "step": 718 + }, + { + "epoch": 0.00719, + "grad_norm": 0.7503709346262388, + "learning_rate": 0.002157, + "loss": 4.8753, + "step": 719 + }, + { + "epoch": 0.0072, + "grad_norm": 0.7326012456404516, + "learning_rate": 0.00216, + "loss": 4.8631, + "step": 720 + }, + { + "epoch": 0.00721, + "grad_norm": 0.7261591024591523, + "learning_rate": 0.002163, + "loss": 4.867, + "step": 721 + }, + { + "epoch": 0.00722, + "grad_norm": 0.782783835861699, + "learning_rate": 0.002166, + "loss": 4.843, + "step": 722 + }, + { + "epoch": 0.00723, + "grad_norm": 0.9405574246546834, + "learning_rate": 0.002169, + "loss": 4.859, + "step": 723 + }, + { + "epoch": 0.00724, + "grad_norm": 0.9635066655828881, + "learning_rate": 0.002172, + "loss": 4.8537, + "step": 724 + }, + { + "epoch": 0.00725, + "grad_norm": 1.040794527733107, + "learning_rate": 0.002175, + "loss": 4.8644, + "step": 725 + }, + { + "epoch": 0.00726, + "grad_norm": 0.8259351691267115, + "learning_rate": 0.002178, + "loss": 4.8549, + "step": 726 + }, + { + "epoch": 0.00727, + "grad_norm": 0.9246523356382073, + "learning_rate": 0.0021809999999999998, + "loss": 4.8481, + "step": 727 + }, + { + "epoch": 0.00728, + "grad_norm": 0.8492007448518646, + "learning_rate": 0.002184, + "loss": 4.8551, + "step": 728 + }, + { + "epoch": 0.00729, + "grad_norm": 0.8802732367664233, + "learning_rate": 0.002187, + "loss": 4.8519, + "step": 729 + }, + { + "epoch": 0.0073, + "grad_norm": 0.8143690086226398, + "learning_rate": 0.00219, + "loss": 4.8583, + "step": 730 + }, + { + "epoch": 0.00731, + "grad_norm": 0.7978098404882568, + "learning_rate": 0.002193, + "loss": 4.846, + "step": 731 + }, + { + "epoch": 0.00732, + "grad_norm": 0.8726405483808616, + "learning_rate": 0.002196, + "loss": 4.8538, + "step": 732 + }, + { + "epoch": 0.00733, + "grad_norm": 0.852624212596407, + "learning_rate": 0.002199, + "loss": 4.8348, + "step": 733 + }, + { + "epoch": 0.00734, + "grad_norm": 0.8516024625713395, + "learning_rate": 0.002202, + "loss": 4.8503, + "step": 734 + }, + { + "epoch": 0.00735, + "grad_norm": 1.0296852244953179, + "learning_rate": 0.002205, + "loss": 4.8752, + "step": 735 + }, + { + "epoch": 0.00736, + "grad_norm": 0.8907291733705317, + "learning_rate": 0.002208, + "loss": 4.8478, + "step": 736 + }, + { + "epoch": 0.00737, + "grad_norm": 0.847923173569621, + "learning_rate": 0.002211, + "loss": 4.8138, + "step": 737 + }, + { + "epoch": 0.00738, + "grad_norm": 1.0201562870165168, + "learning_rate": 0.002214, + "loss": 4.8304, + "step": 738 + }, + { + "epoch": 0.00739, + "grad_norm": 0.8864771113941665, + "learning_rate": 0.0022170000000000002, + "loss": 4.8243, + "step": 739 + }, + { + "epoch": 0.0074, + "grad_norm": 0.8107829600567034, + "learning_rate": 0.00222, + "loss": 4.8172, + "step": 740 + }, + { + "epoch": 0.00741, + "grad_norm": 0.9926562607182313, + "learning_rate": 0.002223, + "loss": 4.7981, + "step": 741 + }, + { + "epoch": 0.00742, + "grad_norm": 1.1169095501625559, + "learning_rate": 0.002226, + "loss": 4.8325, + "step": 742 + }, + { + "epoch": 0.00743, + "grad_norm": 1.0154903575297356, + "learning_rate": 0.002229, + "loss": 4.8033, + "step": 743 + }, + { + "epoch": 0.00744, + "grad_norm": 0.8796310645612629, + "learning_rate": 0.002232, + "loss": 4.8021, + "step": 744 + }, + { + "epoch": 0.00745, + "grad_norm": 0.8347673276949072, + "learning_rate": 0.002235, + "loss": 4.8034, + "step": 745 + }, + { + "epoch": 0.00746, + "grad_norm": 1.0163090629032976, + "learning_rate": 0.002238, + "loss": 4.8122, + "step": 746 + }, + { + "epoch": 0.00747, + "grad_norm": 0.9716623005872416, + "learning_rate": 0.002241, + "loss": 4.7984, + "step": 747 + }, + { + "epoch": 0.00748, + "grad_norm": 1.0278655828642824, + "learning_rate": 0.002244, + "loss": 4.8089, + "step": 748 + }, + { + "epoch": 0.00749, + "grad_norm": 1.0480541445842007, + "learning_rate": 0.002247, + "loss": 4.797, + "step": 749 + }, + { + "epoch": 0.0075, + "grad_norm": 0.8547105592298554, + "learning_rate": 0.0022500000000000003, + "loss": 4.8161, + "step": 750 + }, + { + "epoch": 0.00751, + "grad_norm": 1.124867997489978, + "learning_rate": 0.0022530000000000002, + "loss": 4.8181, + "step": 751 + }, + { + "epoch": 0.00752, + "grad_norm": 0.9791690369641056, + "learning_rate": 0.002256, + "loss": 4.777, + "step": 752 + }, + { + "epoch": 0.00753, + "grad_norm": 1.076442658365417, + "learning_rate": 0.002259, + "loss": 4.8205, + "step": 753 + }, + { + "epoch": 0.00754, + "grad_norm": 1.2763721488952993, + "learning_rate": 0.002262, + "loss": 4.8438, + "step": 754 + }, + { + "epoch": 0.00755, + "grad_norm": 0.956893666746658, + "learning_rate": 0.002265, + "loss": 4.7969, + "step": 755 + }, + { + "epoch": 0.00756, + "grad_norm": 0.9171753465968743, + "learning_rate": 0.002268, + "loss": 4.7762, + "step": 756 + }, + { + "epoch": 0.00757, + "grad_norm": 1.0908406750664337, + "learning_rate": 0.002271, + "loss": 4.7959, + "step": 757 + }, + { + "epoch": 0.00758, + "grad_norm": 0.9176733613061218, + "learning_rate": 0.002274, + "loss": 4.7709, + "step": 758 + }, + { + "epoch": 0.00759, + "grad_norm": 0.8908245026099332, + "learning_rate": 0.002277, + "loss": 4.7862, + "step": 759 + }, + { + "epoch": 0.0076, + "grad_norm": 0.9024826649233609, + "learning_rate": 0.00228, + "loss": 4.7716, + "step": 760 + }, + { + "epoch": 0.00761, + "grad_norm": 0.8163892030302853, + "learning_rate": 0.002283, + "loss": 4.7408, + "step": 761 + }, + { + "epoch": 0.00762, + "grad_norm": 0.8370737395441165, + "learning_rate": 0.0022860000000000003, + "loss": 4.7247, + "step": 762 + }, + { + "epoch": 0.00763, + "grad_norm": 0.8480960736158745, + "learning_rate": 0.0022890000000000002, + "loss": 4.7641, + "step": 763 + }, + { + "epoch": 0.00764, + "grad_norm": 0.7323677069229925, + "learning_rate": 0.002292, + "loss": 4.7635, + "step": 764 + }, + { + "epoch": 0.00765, + "grad_norm": 0.517428476031845, + "learning_rate": 0.002295, + "loss": 4.7402, + "step": 765 + }, + { + "epoch": 0.00766, + "grad_norm": 0.5449748434883583, + "learning_rate": 0.002298, + "loss": 4.7349, + "step": 766 + }, + { + "epoch": 0.00767, + "grad_norm": 0.519905380789997, + "learning_rate": 0.002301, + "loss": 4.7297, + "step": 767 + }, + { + "epoch": 0.00768, + "grad_norm": 0.5353014095205901, + "learning_rate": 0.002304, + "loss": 4.7326, + "step": 768 + }, + { + "epoch": 0.00769, + "grad_norm": 0.6021476092106454, + "learning_rate": 0.002307, + "loss": 4.7369, + "step": 769 + }, + { + "epoch": 0.0077, + "grad_norm": 0.6291707943467315, + "learning_rate": 0.00231, + "loss": 4.6951, + "step": 770 + }, + { + "epoch": 0.00771, + "grad_norm": 0.6259748544734598, + "learning_rate": 0.002313, + "loss": 4.726, + "step": 771 + }, + { + "epoch": 0.00772, + "grad_norm": 0.6411937087975925, + "learning_rate": 0.002316, + "loss": 4.71, + "step": 772 + }, + { + "epoch": 0.00773, + "grad_norm": 0.6087209583882061, + "learning_rate": 0.0023190000000000003, + "loss": 4.7089, + "step": 773 + }, + { + "epoch": 0.00774, + "grad_norm": 0.6187878154543281, + "learning_rate": 0.0023220000000000003, + "loss": 4.7216, + "step": 774 + }, + { + "epoch": 0.00775, + "grad_norm": 0.6698244521960122, + "learning_rate": 0.0023250000000000002, + "loss": 4.7125, + "step": 775 + }, + { + "epoch": 0.00776, + "grad_norm": 0.681313335099601, + "learning_rate": 0.002328, + "loss": 4.716, + "step": 776 + }, + { + "epoch": 0.00777, + "grad_norm": 0.564623702064127, + "learning_rate": 0.002331, + "loss": 4.6866, + "step": 777 + }, + { + "epoch": 0.00778, + "grad_norm": 0.5709713255684672, + "learning_rate": 0.002334, + "loss": 4.6954, + "step": 778 + }, + { + "epoch": 0.00779, + "grad_norm": 0.6149821587836622, + "learning_rate": 0.002337, + "loss": 4.6565, + "step": 779 + }, + { + "epoch": 0.0078, + "grad_norm": 0.632842020953678, + "learning_rate": 0.00234, + "loss": 4.6725, + "step": 780 + }, + { + "epoch": 0.00781, + "grad_norm": 0.7467970568022195, + "learning_rate": 0.002343, + "loss": 4.678, + "step": 781 + }, + { + "epoch": 0.00782, + "grad_norm": 0.9047857384333676, + "learning_rate": 0.002346, + "loss": 4.6801, + "step": 782 + }, + { + "epoch": 0.00783, + "grad_norm": 1.0576276659724262, + "learning_rate": 0.002349, + "loss": 4.6678, + "step": 783 + }, + { + "epoch": 0.00784, + "grad_norm": 1.1751809929780095, + "learning_rate": 0.002352, + "loss": 4.6921, + "step": 784 + }, + { + "epoch": 0.00785, + "grad_norm": 0.8865561583369859, + "learning_rate": 0.0023550000000000003, + "loss": 4.6774, + "step": 785 + }, + { + "epoch": 0.00786, + "grad_norm": 1.1372694124786462, + "learning_rate": 0.0023580000000000003, + "loss": 4.7016, + "step": 786 + }, + { + "epoch": 0.00787, + "grad_norm": 1.0092158733190273, + "learning_rate": 0.0023610000000000003, + "loss": 4.6918, + "step": 787 + }, + { + "epoch": 0.00788, + "grad_norm": 1.1115636173510615, + "learning_rate": 0.002364, + "loss": 4.7036, + "step": 788 + }, + { + "epoch": 0.00789, + "grad_norm": 0.9627507497820444, + "learning_rate": 0.002367, + "loss": 4.7196, + "step": 789 + }, + { + "epoch": 0.0079, + "grad_norm": 1.0391564883926698, + "learning_rate": 0.00237, + "loss": 4.706, + "step": 790 + }, + { + "epoch": 0.00791, + "grad_norm": 0.9857967273862618, + "learning_rate": 0.002373, + "loss": 4.6613, + "step": 791 + }, + { + "epoch": 0.00792, + "grad_norm": 1.07807376567954, + "learning_rate": 0.002376, + "loss": 4.6969, + "step": 792 + }, + { + "epoch": 0.00793, + "grad_norm": 1.1762161150555115, + "learning_rate": 0.002379, + "loss": 4.7002, + "step": 793 + }, + { + "epoch": 0.00794, + "grad_norm": 0.8247319634032979, + "learning_rate": 0.002382, + "loss": 4.6948, + "step": 794 + }, + { + "epoch": 0.00795, + "grad_norm": 0.8498192334201926, + "learning_rate": 0.002385, + "loss": 4.676, + "step": 795 + }, + { + "epoch": 0.00796, + "grad_norm": 0.7871292688998256, + "learning_rate": 0.0023880000000000004, + "loss": 4.6606, + "step": 796 + }, + { + "epoch": 0.00797, + "grad_norm": 0.7600244620970749, + "learning_rate": 0.0023910000000000003, + "loss": 4.7021, + "step": 797 + }, + { + "epoch": 0.00798, + "grad_norm": 0.7960964641914068, + "learning_rate": 0.0023940000000000003, + "loss": 4.6698, + "step": 798 + }, + { + "epoch": 0.00799, + "grad_norm": 0.740030757660218, + "learning_rate": 0.0023970000000000003, + "loss": 4.6941, + "step": 799 + }, + { + "epoch": 0.008, + "grad_norm": 0.7620304314170098, + "learning_rate": 0.0024000000000000002, + "loss": 4.6524, + "step": 800 + }, + { + "epoch": 0.00801, + "grad_norm": 0.7109641363313225, + "learning_rate": 0.002403, + "loss": 4.6746, + "step": 801 + }, + { + "epoch": 0.00802, + "grad_norm": 0.6915734125547375, + "learning_rate": 0.002406, + "loss": 4.6614, + "step": 802 + }, + { + "epoch": 0.00803, + "grad_norm": 0.6786740334805601, + "learning_rate": 0.002409, + "loss": 4.6408, + "step": 803 + }, + { + "epoch": 0.00804, + "grad_norm": 0.6631767318006158, + "learning_rate": 0.002412, + "loss": 4.6361, + "step": 804 + }, + { + "epoch": 0.00805, + "grad_norm": 0.6460763714070807, + "learning_rate": 0.002415, + "loss": 4.631, + "step": 805 + }, + { + "epoch": 0.00806, + "grad_norm": 0.5689446567265103, + "learning_rate": 0.002418, + "loss": 4.6413, + "step": 806 + }, + { + "epoch": 0.00807, + "grad_norm": 0.7294155919129667, + "learning_rate": 0.0024210000000000004, + "loss": 4.6407, + "step": 807 + }, + { + "epoch": 0.00808, + "grad_norm": 0.9713010260235329, + "learning_rate": 0.0024240000000000004, + "loss": 4.6372, + "step": 808 + }, + { + "epoch": 0.00809, + "grad_norm": 1.039651350431828, + "learning_rate": 0.0024270000000000003, + "loss": 4.6452, + "step": 809 + }, + { + "epoch": 0.0081, + "grad_norm": 0.7987971948237619, + "learning_rate": 0.0024300000000000003, + "loss": 4.6558, + "step": 810 + }, + { + "epoch": 0.00811, + "grad_norm": 0.866506786468452, + "learning_rate": 0.0024330000000000003, + "loss": 4.6332, + "step": 811 + }, + { + "epoch": 0.00812, + "grad_norm": 0.8940070034144846, + "learning_rate": 0.0024360000000000002, + "loss": 4.6162, + "step": 812 + }, + { + "epoch": 0.00813, + "grad_norm": 0.7601015395596301, + "learning_rate": 0.0024389999999999998, + "loss": 4.6418, + "step": 813 + }, + { + "epoch": 0.00814, + "grad_norm": 0.959371802874838, + "learning_rate": 0.0024419999999999997, + "loss": 4.6561, + "step": 814 + }, + { + "epoch": 0.00815, + "grad_norm": 1.0501943727451555, + "learning_rate": 0.0024449999999999997, + "loss": 4.6758, + "step": 815 + }, + { + "epoch": 0.00816, + "grad_norm": 1.0657571457398118, + "learning_rate": 0.002448, + "loss": 4.6479, + "step": 816 + }, + { + "epoch": 0.00817, + "grad_norm": 0.9466339757774411, + "learning_rate": 0.002451, + "loss": 4.6581, + "step": 817 + }, + { + "epoch": 0.00818, + "grad_norm": 1.0180120345579544, + "learning_rate": 0.002454, + "loss": 4.6439, + "step": 818 + }, + { + "epoch": 0.00819, + "grad_norm": 0.9541848583096257, + "learning_rate": 0.002457, + "loss": 4.6558, + "step": 819 + }, + { + "epoch": 0.0082, + "grad_norm": 0.8369907334148766, + "learning_rate": 0.00246, + "loss": 4.6334, + "step": 820 + }, + { + "epoch": 0.00821, + "grad_norm": 0.9636511535948165, + "learning_rate": 0.002463, + "loss": 4.6201, + "step": 821 + }, + { + "epoch": 0.00822, + "grad_norm": 0.9781201110522014, + "learning_rate": 0.002466, + "loss": 4.6234, + "step": 822 + }, + { + "epoch": 0.00823, + "grad_norm": 0.7544698826164155, + "learning_rate": 0.002469, + "loss": 4.6177, + "step": 823 + }, + { + "epoch": 0.00824, + "grad_norm": 0.8351046210351873, + "learning_rate": 0.002472, + "loss": 4.6395, + "step": 824 + }, + { + "epoch": 0.00825, + "grad_norm": 0.9579203449788568, + "learning_rate": 0.0024749999999999998, + "loss": 4.6552, + "step": 825 + }, + { + "epoch": 0.00826, + "grad_norm": 0.6855903524790949, + "learning_rate": 0.0024779999999999997, + "loss": 4.6015, + "step": 826 + }, + { + "epoch": 0.00827, + "grad_norm": 0.5831382107833026, + "learning_rate": 0.002481, + "loss": 4.6271, + "step": 827 + }, + { + "epoch": 0.00828, + "grad_norm": 0.6353397555552095, + "learning_rate": 0.002484, + "loss": 4.6166, + "step": 828 + }, + { + "epoch": 0.00829, + "grad_norm": 0.7710602439494972, + "learning_rate": 0.002487, + "loss": 4.619, + "step": 829 + }, + { + "epoch": 0.0083, + "grad_norm": 0.8708544431374072, + "learning_rate": 0.00249, + "loss": 4.6284, + "step": 830 + }, + { + "epoch": 0.00831, + "grad_norm": 1.0453965047267408, + "learning_rate": 0.002493, + "loss": 4.6196, + "step": 831 + }, + { + "epoch": 0.00832, + "grad_norm": 1.1447270695034462, + "learning_rate": 0.002496, + "loss": 4.6327, + "step": 832 + }, + { + "epoch": 0.00833, + "grad_norm": 1.0246810397019857, + "learning_rate": 0.002499, + "loss": 4.6176, + "step": 833 + }, + { + "epoch": 0.00834, + "grad_norm": 0.9810039542594958, + "learning_rate": 0.002502, + "loss": 4.6274, + "step": 834 + }, + { + "epoch": 0.00835, + "grad_norm": 0.8876541499902995, + "learning_rate": 0.002505, + "loss": 4.6197, + "step": 835 + }, + { + "epoch": 0.00836, + "grad_norm": 0.903016402602156, + "learning_rate": 0.002508, + "loss": 4.6457, + "step": 836 + }, + { + "epoch": 0.00837, + "grad_norm": 0.9089070943446178, + "learning_rate": 0.0025109999999999998, + "loss": 4.6001, + "step": 837 + }, + { + "epoch": 0.00838, + "grad_norm": 1.0056159571167094, + "learning_rate": 0.0025139999999999997, + "loss": 4.6276, + "step": 838 + }, + { + "epoch": 0.00839, + "grad_norm": 0.9338401740679543, + "learning_rate": 0.002517, + "loss": 4.6258, + "step": 839 + }, + { + "epoch": 0.0084, + "grad_norm": 0.8595710991386702, + "learning_rate": 0.00252, + "loss": 4.6076, + "step": 840 + }, + { + "epoch": 0.00841, + "grad_norm": 0.8408169361023179, + "learning_rate": 0.002523, + "loss": 4.5981, + "step": 841 + }, + { + "epoch": 0.00842, + "grad_norm": 0.9223461051136856, + "learning_rate": 0.002526, + "loss": 4.6018, + "step": 842 + }, + { + "epoch": 0.00843, + "grad_norm": 0.8967298269444864, + "learning_rate": 0.002529, + "loss": 4.5992, + "step": 843 + }, + { + "epoch": 0.00844, + "grad_norm": 1.0054950179651305, + "learning_rate": 0.002532, + "loss": 4.6384, + "step": 844 + }, + { + "epoch": 0.00845, + "grad_norm": 0.9421873102957943, + "learning_rate": 0.002535, + "loss": 4.597, + "step": 845 + }, + { + "epoch": 0.00846, + "grad_norm": 0.7813468730903476, + "learning_rate": 0.002538, + "loss": 4.612, + "step": 846 + }, + { + "epoch": 0.00847, + "grad_norm": 0.7336538825841323, + "learning_rate": 0.002541, + "loss": 4.5922, + "step": 847 + }, + { + "epoch": 0.00848, + "grad_norm": 0.7490910216536115, + "learning_rate": 0.002544, + "loss": 4.5881, + "step": 848 + }, + { + "epoch": 0.00849, + "grad_norm": 0.7281767237417216, + "learning_rate": 0.002547, + "loss": 4.5791, + "step": 849 + }, + { + "epoch": 0.0085, + "grad_norm": 0.6108432594850526, + "learning_rate": 0.00255, + "loss": 4.5847, + "step": 850 + }, + { + "epoch": 0.00851, + "grad_norm": 0.6024151823798927, + "learning_rate": 0.002553, + "loss": 4.5988, + "step": 851 + }, + { + "epoch": 0.00852, + "grad_norm": 0.4780185378158163, + "learning_rate": 0.002556, + "loss": 4.5828, + "step": 852 + }, + { + "epoch": 0.00853, + "grad_norm": 0.5435830639563773, + "learning_rate": 0.002559, + "loss": 4.5614, + "step": 853 + }, + { + "epoch": 0.00854, + "grad_norm": 0.6028384776402124, + "learning_rate": 0.002562, + "loss": 4.5542, + "step": 854 + }, + { + "epoch": 0.00855, + "grad_norm": 0.5885480983638159, + "learning_rate": 0.002565, + "loss": 4.5306, + "step": 855 + }, + { + "epoch": 0.00856, + "grad_norm": 0.6060410896547319, + "learning_rate": 0.002568, + "loss": 4.5634, + "step": 856 + }, + { + "epoch": 0.00857, + "grad_norm": 0.5609885323605243, + "learning_rate": 0.002571, + "loss": 4.5495, + "step": 857 + }, + { + "epoch": 0.00858, + "grad_norm": 0.5199675893089152, + "learning_rate": 0.002574, + "loss": 4.541, + "step": 858 + }, + { + "epoch": 0.00859, + "grad_norm": 0.5200210947429836, + "learning_rate": 0.002577, + "loss": 4.5318, + "step": 859 + }, + { + "epoch": 0.0086, + "grad_norm": 0.5623426902483657, + "learning_rate": 0.00258, + "loss": 4.5597, + "step": 860 + }, + { + "epoch": 0.00861, + "grad_norm": 0.6385070114593622, + "learning_rate": 0.0025830000000000002, + "loss": 4.5763, + "step": 861 + }, + { + "epoch": 0.00862, + "grad_norm": 0.7215967762933786, + "learning_rate": 0.002586, + "loss": 4.5292, + "step": 862 + }, + { + "epoch": 0.00863, + "grad_norm": 0.8778870896923412, + "learning_rate": 0.002589, + "loss": 4.559, + "step": 863 + }, + { + "epoch": 0.00864, + "grad_norm": 1.1440074056398795, + "learning_rate": 0.002592, + "loss": 4.547, + "step": 864 + }, + { + "epoch": 0.00865, + "grad_norm": 0.9219223071258862, + "learning_rate": 0.002595, + "loss": 4.5491, + "step": 865 + }, + { + "epoch": 0.00866, + "grad_norm": 1.065047382603863, + "learning_rate": 0.002598, + "loss": 4.5845, + "step": 866 + }, + { + "epoch": 0.00867, + "grad_norm": 0.9527212983309642, + "learning_rate": 0.002601, + "loss": 4.5736, + "step": 867 + }, + { + "epoch": 0.00868, + "grad_norm": 0.7943624191664302, + "learning_rate": 0.002604, + "loss": 4.5474, + "step": 868 + }, + { + "epoch": 0.00869, + "grad_norm": 0.671885521142916, + "learning_rate": 0.002607, + "loss": 4.5536, + "step": 869 + }, + { + "epoch": 0.0087, + "grad_norm": 0.6527559883208905, + "learning_rate": 0.00261, + "loss": 4.54, + "step": 870 + }, + { + "epoch": 0.00871, + "grad_norm": 0.6266922867904314, + "learning_rate": 0.002613, + "loss": 4.5522, + "step": 871 + }, + { + "epoch": 0.00872, + "grad_norm": 0.5879044949928981, + "learning_rate": 0.002616, + "loss": 4.5345, + "step": 872 + }, + { + "epoch": 0.00873, + "grad_norm": 0.5844885657438569, + "learning_rate": 0.0026190000000000002, + "loss": 4.5441, + "step": 873 + }, + { + "epoch": 0.00874, + "grad_norm": 0.5286751889514939, + "learning_rate": 0.002622, + "loss": 4.5573, + "step": 874 + }, + { + "epoch": 0.00875, + "grad_norm": 0.5494502975686567, + "learning_rate": 0.002625, + "loss": 4.5051, + "step": 875 + }, + { + "epoch": 0.00876, + "grad_norm": 0.6269645893334619, + "learning_rate": 0.002628, + "loss": 4.5356, + "step": 876 + }, + { + "epoch": 0.00877, + "grad_norm": 0.6901692969476347, + "learning_rate": 0.002631, + "loss": 4.5537, + "step": 877 + }, + { + "epoch": 0.00878, + "grad_norm": 0.7433571199864201, + "learning_rate": 0.002634, + "loss": 4.4894, + "step": 878 + }, + { + "epoch": 0.00879, + "grad_norm": 0.7806263902694818, + "learning_rate": 0.002637, + "loss": 4.5303, + "step": 879 + }, + { + "epoch": 0.0088, + "grad_norm": 0.7694352427135146, + "learning_rate": 0.00264, + "loss": 4.5353, + "step": 880 + }, + { + "epoch": 0.00881, + "grad_norm": 0.6938035372615263, + "learning_rate": 0.002643, + "loss": 4.5351, + "step": 881 + }, + { + "epoch": 0.00882, + "grad_norm": 0.672108824408367, + "learning_rate": 0.002646, + "loss": 4.5331, + "step": 882 + }, + { + "epoch": 0.00883, + "grad_norm": 0.6847246415873477, + "learning_rate": 0.002649, + "loss": 4.524, + "step": 883 + }, + { + "epoch": 0.00884, + "grad_norm": 0.8034466162851082, + "learning_rate": 0.0026520000000000003, + "loss": 4.5123, + "step": 884 + }, + { + "epoch": 0.00885, + "grad_norm": 0.8504740488359851, + "learning_rate": 0.0026550000000000002, + "loss": 4.5174, + "step": 885 + }, + { + "epoch": 0.00886, + "grad_norm": 0.6814828631263485, + "learning_rate": 0.002658, + "loss": 4.526, + "step": 886 + }, + { + "epoch": 0.00887, + "grad_norm": 0.7764171535679266, + "learning_rate": 0.002661, + "loss": 4.5254, + "step": 887 + }, + { + "epoch": 0.00888, + "grad_norm": 0.864178556132455, + "learning_rate": 0.002664, + "loss": 4.5323, + "step": 888 + }, + { + "epoch": 0.00889, + "grad_norm": 0.9963366133988169, + "learning_rate": 0.002667, + "loss": 4.4915, + "step": 889 + }, + { + "epoch": 0.0089, + "grad_norm": 0.9684646116139818, + "learning_rate": 0.00267, + "loss": 4.5454, + "step": 890 + }, + { + "epoch": 0.00891, + "grad_norm": 0.8019607803647243, + "learning_rate": 0.002673, + "loss": 4.523, + "step": 891 + }, + { + "epoch": 0.00892, + "grad_norm": 0.9082109544296421, + "learning_rate": 0.002676, + "loss": 4.5431, + "step": 892 + }, + { + "epoch": 0.00893, + "grad_norm": 0.9728985136087724, + "learning_rate": 0.002679, + "loss": 4.5212, + "step": 893 + }, + { + "epoch": 0.00894, + "grad_norm": 1.0580229080074564, + "learning_rate": 0.002682, + "loss": 4.5156, + "step": 894 + }, + { + "epoch": 0.00895, + "grad_norm": 0.8166896686472195, + "learning_rate": 0.0026850000000000003, + "loss": 4.5188, + "step": 895 + }, + { + "epoch": 0.00896, + "grad_norm": 0.8942454442625324, + "learning_rate": 0.0026880000000000003, + "loss": 4.5121, + "step": 896 + }, + { + "epoch": 0.00897, + "grad_norm": 0.9627689980406442, + "learning_rate": 0.0026910000000000002, + "loss": 4.5406, + "step": 897 + }, + { + "epoch": 0.00898, + "grad_norm": 0.915209672072645, + "learning_rate": 0.002694, + "loss": 4.5291, + "step": 898 + }, + { + "epoch": 0.00899, + "grad_norm": 0.8008505248098504, + "learning_rate": 0.002697, + "loss": 4.5313, + "step": 899 + }, + { + "epoch": 0.009, + "grad_norm": 0.974662986907947, + "learning_rate": 0.0027, + "loss": 4.5627, + "step": 900 + }, + { + "epoch": 0.00901, + "grad_norm": 1.1878109077704273, + "learning_rate": 0.002703, + "loss": 4.5336, + "step": 901 + }, + { + "epoch": 0.00902, + "grad_norm": 0.7391945290133592, + "learning_rate": 0.002706, + "loss": 4.5248, + "step": 902 + }, + { + "epoch": 0.00903, + "grad_norm": 0.7310904041760188, + "learning_rate": 0.002709, + "loss": 4.5195, + "step": 903 + }, + { + "epoch": 0.00904, + "grad_norm": 0.763540341773903, + "learning_rate": 0.002712, + "loss": 4.4985, + "step": 904 + }, + { + "epoch": 0.00905, + "grad_norm": 0.922476362328922, + "learning_rate": 0.002715, + "loss": 4.5209, + "step": 905 + }, + { + "epoch": 0.00906, + "grad_norm": 0.9574009618515481, + "learning_rate": 0.002718, + "loss": 4.5379, + "step": 906 + }, + { + "epoch": 0.00907, + "grad_norm": 1.035774941473643, + "learning_rate": 0.0027210000000000003, + "loss": 4.5538, + "step": 907 + }, + { + "epoch": 0.00908, + "grad_norm": 1.0191319400722418, + "learning_rate": 0.0027240000000000003, + "loss": 4.5279, + "step": 908 + }, + { + "epoch": 0.00909, + "grad_norm": 1.1103950353838743, + "learning_rate": 0.0027270000000000003, + "loss": 4.554, + "step": 909 + }, + { + "epoch": 0.0091, + "grad_norm": 0.8691030580510016, + "learning_rate": 0.0027300000000000002, + "loss": 4.5543, + "step": 910 + }, + { + "epoch": 0.00911, + "grad_norm": 0.7723878752960539, + "learning_rate": 0.002733, + "loss": 4.5035, + "step": 911 + }, + { + "epoch": 0.00912, + "grad_norm": 0.8180906874107354, + "learning_rate": 0.002736, + "loss": 4.5117, + "step": 912 + }, + { + "epoch": 0.00913, + "grad_norm": 0.8148069350423363, + "learning_rate": 0.002739, + "loss": 4.505, + "step": 913 + }, + { + "epoch": 0.00914, + "grad_norm": 0.7118677869163097, + "learning_rate": 0.002742, + "loss": 4.5326, + "step": 914 + }, + { + "epoch": 0.00915, + "grad_norm": 0.67221021637466, + "learning_rate": 0.002745, + "loss": 4.4896, + "step": 915 + }, + { + "epoch": 0.00916, + "grad_norm": 0.7521734182214744, + "learning_rate": 0.002748, + "loss": 4.4742, + "step": 916 + }, + { + "epoch": 0.00917, + "grad_norm": 0.6660072639776646, + "learning_rate": 0.002751, + "loss": 4.4809, + "step": 917 + }, + { + "epoch": 0.00918, + "grad_norm": 0.5671697873930601, + "learning_rate": 0.0027540000000000004, + "loss": 4.4963, + "step": 918 + }, + { + "epoch": 0.00919, + "grad_norm": 0.5594822914263968, + "learning_rate": 0.0027570000000000003, + "loss": 4.4754, + "step": 919 + }, + { + "epoch": 0.0092, + "grad_norm": 0.6080421327199805, + "learning_rate": 0.0027600000000000003, + "loss": 4.4903, + "step": 920 + }, + { + "epoch": 0.00921, + "grad_norm": 0.728784609109995, + "learning_rate": 0.0027630000000000003, + "loss": 4.4759, + "step": 921 + }, + { + "epoch": 0.00922, + "grad_norm": 0.8477951683682543, + "learning_rate": 0.0027660000000000002, + "loss": 4.5076, + "step": 922 + }, + { + "epoch": 0.00923, + "grad_norm": 0.6016579194588629, + "learning_rate": 0.002769, + "loss": 4.4637, + "step": 923 + }, + { + "epoch": 0.00924, + "grad_norm": 0.6726777284073512, + "learning_rate": 0.002772, + "loss": 4.4919, + "step": 924 + }, + { + "epoch": 0.00925, + "grad_norm": 0.7061399279754877, + "learning_rate": 0.002775, + "loss": 4.4751, + "step": 925 + }, + { + "epoch": 0.00926, + "grad_norm": 0.6525850527724611, + "learning_rate": 0.002778, + "loss": 4.4844, + "step": 926 + }, + { + "epoch": 0.00927, + "grad_norm": 0.6775355167555962, + "learning_rate": 0.002781, + "loss": 4.4493, + "step": 927 + }, + { + "epoch": 0.00928, + "grad_norm": 0.660488843824027, + "learning_rate": 0.002784, + "loss": 4.493, + "step": 928 + }, + { + "epoch": 0.00929, + "grad_norm": 0.6375906920183253, + "learning_rate": 0.0027870000000000004, + "loss": 4.4714, + "step": 929 + }, + { + "epoch": 0.0093, + "grad_norm": 0.7170906127689122, + "learning_rate": 0.0027900000000000004, + "loss": 4.4557, + "step": 930 + }, + { + "epoch": 0.00931, + "grad_norm": 0.7727209257871167, + "learning_rate": 0.0027930000000000003, + "loss": 4.4706, + "step": 931 + }, + { + "epoch": 0.00932, + "grad_norm": 0.759902395005458, + "learning_rate": 0.0027960000000000003, + "loss": 4.4795, + "step": 932 + }, + { + "epoch": 0.00933, + "grad_norm": 0.9089287828754257, + "learning_rate": 0.0027990000000000003, + "loss": 4.4536, + "step": 933 + }, + { + "epoch": 0.00934, + "grad_norm": 0.9561562356933814, + "learning_rate": 0.0028020000000000002, + "loss": 4.4859, + "step": 934 + }, + { + "epoch": 0.00935, + "grad_norm": 1.010681147518413, + "learning_rate": 0.002805, + "loss": 4.5044, + "step": 935 + }, + { + "epoch": 0.00936, + "grad_norm": 0.960270788501537, + "learning_rate": 0.002808, + "loss": 4.5076, + "step": 936 + }, + { + "epoch": 0.00937, + "grad_norm": 0.8170703531029062, + "learning_rate": 0.002811, + "loss": 4.5151, + "step": 937 + }, + { + "epoch": 0.00938, + "grad_norm": 0.6779931914379209, + "learning_rate": 0.002814, + "loss": 4.4854, + "step": 938 + }, + { + "epoch": 0.00939, + "grad_norm": 0.6509733323350428, + "learning_rate": 0.002817, + "loss": 4.4771, + "step": 939 + }, + { + "epoch": 0.0094, + "grad_norm": 0.6633333111873998, + "learning_rate": 0.00282, + "loss": 4.4796, + "step": 940 + }, + { + "epoch": 0.00941, + "grad_norm": 0.7449615790040228, + "learning_rate": 0.002823, + "loss": 4.4576, + "step": 941 + }, + { + "epoch": 0.00942, + "grad_norm": 0.7696922222349031, + "learning_rate": 0.002826, + "loss": 4.4823, + "step": 942 + }, + { + "epoch": 0.00943, + "grad_norm": 0.6087986992050594, + "learning_rate": 0.002829, + "loss": 4.4688, + "step": 943 + }, + { + "epoch": 0.00944, + "grad_norm": 0.5228736659630974, + "learning_rate": 0.002832, + "loss": 4.4678, + "step": 944 + }, + { + "epoch": 0.00945, + "grad_norm": 0.5754533269053475, + "learning_rate": 0.002835, + "loss": 4.4713, + "step": 945 + }, + { + "epoch": 0.00946, + "grad_norm": 0.6164648444392735, + "learning_rate": 0.002838, + "loss": 4.4832, + "step": 946 + }, + { + "epoch": 0.00947, + "grad_norm": 0.6419505093155794, + "learning_rate": 0.0028409999999999998, + "loss": 4.4556, + "step": 947 + }, + { + "epoch": 0.00948, + "grad_norm": 0.556707209906786, + "learning_rate": 0.0028439999999999997, + "loss": 4.4351, + "step": 948 + }, + { + "epoch": 0.00949, + "grad_norm": 0.5031407625923785, + "learning_rate": 0.002847, + "loss": 4.4737, + "step": 949 + }, + { + "epoch": 0.0095, + "grad_norm": 0.5881138889925357, + "learning_rate": 0.00285, + "loss": 4.4655, + "step": 950 + }, + { + "epoch": 0.00951, + "grad_norm": 0.6194945571715194, + "learning_rate": 0.002853, + "loss": 4.4418, + "step": 951 + }, + { + "epoch": 0.00952, + "grad_norm": 0.6965416314433459, + "learning_rate": 0.002856, + "loss": 4.4704, + "step": 952 + }, + { + "epoch": 0.00953, + "grad_norm": 0.7506696864658969, + "learning_rate": 0.002859, + "loss": 4.4161, + "step": 953 + }, + { + "epoch": 0.00954, + "grad_norm": 0.6440899731036704, + "learning_rate": 0.002862, + "loss": 4.4703, + "step": 954 + }, + { + "epoch": 0.00955, + "grad_norm": 0.7716821398038454, + "learning_rate": 0.002865, + "loss": 4.4601, + "step": 955 + }, + { + "epoch": 0.00956, + "grad_norm": 0.9111507025485583, + "learning_rate": 0.002868, + "loss": 4.4578, + "step": 956 + }, + { + "epoch": 0.00957, + "grad_norm": 0.8443539201487298, + "learning_rate": 0.002871, + "loss": 4.4727, + "step": 957 + }, + { + "epoch": 0.00958, + "grad_norm": 0.7989694313181581, + "learning_rate": 0.002874, + "loss": 4.4582, + "step": 958 + }, + { + "epoch": 0.00959, + "grad_norm": 0.7547037856292312, + "learning_rate": 0.002877, + "loss": 4.4482, + "step": 959 + }, + { + "epoch": 0.0096, + "grad_norm": 0.7121103430025651, + "learning_rate": 0.0028799999999999997, + "loss": 4.4466, + "step": 960 + }, + { + "epoch": 0.00961, + "grad_norm": 0.6758285086831927, + "learning_rate": 0.002883, + "loss": 4.4209, + "step": 961 + }, + { + "epoch": 0.00962, + "grad_norm": 0.5336350485211147, + "learning_rate": 0.002886, + "loss": 4.4436, + "step": 962 + }, + { + "epoch": 0.00963, + "grad_norm": 0.5703472269796636, + "learning_rate": 0.002889, + "loss": 4.4467, + "step": 963 + }, + { + "epoch": 0.00964, + "grad_norm": 0.6524118600131552, + "learning_rate": 0.002892, + "loss": 4.4355, + "step": 964 + }, + { + "epoch": 0.00965, + "grad_norm": 0.7342343756321769, + "learning_rate": 0.002895, + "loss": 4.4144, + "step": 965 + }, + { + "epoch": 0.00966, + "grad_norm": 0.9544017629329996, + "learning_rate": 0.002898, + "loss": 4.4408, + "step": 966 + }, + { + "epoch": 0.00967, + "grad_norm": 1.1065817167809169, + "learning_rate": 0.002901, + "loss": 4.4927, + "step": 967 + }, + { + "epoch": 0.00968, + "grad_norm": 0.9829826223932284, + "learning_rate": 0.002904, + "loss": 4.4568, + "step": 968 + }, + { + "epoch": 0.00969, + "grad_norm": 0.8112557218556605, + "learning_rate": 0.002907, + "loss": 4.4805, + "step": 969 + }, + { + "epoch": 0.0097, + "grad_norm": 0.983264710703698, + "learning_rate": 0.00291, + "loss": 4.4488, + "step": 970 + }, + { + "epoch": 0.00971, + "grad_norm": 1.010591995610007, + "learning_rate": 0.002913, + "loss": 4.4796, + "step": 971 + }, + { + "epoch": 0.00972, + "grad_norm": 0.9930743768968635, + "learning_rate": 0.002916, + "loss": 4.4798, + "step": 972 + }, + { + "epoch": 0.00973, + "grad_norm": 1.1817662624154759, + "learning_rate": 0.002919, + "loss": 4.4931, + "step": 973 + }, + { + "epoch": 0.00974, + "grad_norm": 1.1915796929226945, + "learning_rate": 0.002922, + "loss": 4.4839, + "step": 974 + }, + { + "epoch": 0.00975, + "grad_norm": 0.9880254500237188, + "learning_rate": 0.002925, + "loss": 4.4969, + "step": 975 + }, + { + "epoch": 0.00976, + "grad_norm": 1.0394868289507098, + "learning_rate": 0.002928, + "loss": 4.4906, + "step": 976 + }, + { + "epoch": 0.00977, + "grad_norm": 1.1242485803004214, + "learning_rate": 0.002931, + "loss": 4.5346, + "step": 977 + }, + { + "epoch": 0.00978, + "grad_norm": 1.1287291198302327, + "learning_rate": 0.002934, + "loss": 4.5134, + "step": 978 + }, + { + "epoch": 0.00979, + "grad_norm": 0.937365777135907, + "learning_rate": 0.002937, + "loss": 4.4945, + "step": 979 + }, + { + "epoch": 0.0098, + "grad_norm": 0.9999316178288561, + "learning_rate": 0.00294, + "loss": 4.4906, + "step": 980 + }, + { + "epoch": 0.00981, + "grad_norm": 1.0692695747943242, + "learning_rate": 0.002943, + "loss": 4.5166, + "step": 981 + }, + { + "epoch": 0.00982, + "grad_norm": 0.855671043658802, + "learning_rate": 0.002946, + "loss": 4.4983, + "step": 982 + }, + { + "epoch": 0.00983, + "grad_norm": 0.8819131773225922, + "learning_rate": 0.0029490000000000002, + "loss": 4.4932, + "step": 983 + }, + { + "epoch": 0.00984, + "grad_norm": 0.8474323035176099, + "learning_rate": 0.002952, + "loss": 4.4794, + "step": 984 + }, + { + "epoch": 0.00985, + "grad_norm": 0.8010021292921615, + "learning_rate": 0.002955, + "loss": 4.4955, + "step": 985 + }, + { + "epoch": 0.00986, + "grad_norm": 0.6778548937329826, + "learning_rate": 0.002958, + "loss": 4.4543, + "step": 986 + }, + { + "epoch": 0.00987, + "grad_norm": 0.6335901643582099, + "learning_rate": 0.002961, + "loss": 4.4618, + "step": 987 + }, + { + "epoch": 0.00988, + "grad_norm": 0.6898288520048222, + "learning_rate": 0.002964, + "loss": 4.4934, + "step": 988 + }, + { + "epoch": 0.00989, + "grad_norm": 0.6968481266470886, + "learning_rate": 0.002967, + "loss": 4.4338, + "step": 989 + }, + { + "epoch": 0.0099, + "grad_norm": 0.7462093807424477, + "learning_rate": 0.00297, + "loss": 4.4706, + "step": 990 + }, + { + "epoch": 0.00991, + "grad_norm": 0.7437971839370325, + "learning_rate": 0.002973, + "loss": 4.4697, + "step": 991 + }, + { + "epoch": 0.00992, + "grad_norm": 0.6722774287362101, + "learning_rate": 0.002976, + "loss": 4.4576, + "step": 992 + }, + { + "epoch": 0.00993, + "grad_norm": 0.6583210228081078, + "learning_rate": 0.002979, + "loss": 4.4522, + "step": 993 + }, + { + "epoch": 0.00994, + "grad_norm": 0.6827391668775007, + "learning_rate": 0.002982, + "loss": 4.4553, + "step": 994 + }, + { + "epoch": 0.00995, + "grad_norm": 0.7054299500075728, + "learning_rate": 0.0029850000000000002, + "loss": 4.4538, + "step": 995 + }, + { + "epoch": 0.00996, + "grad_norm": 0.7382372248626952, + "learning_rate": 0.002988, + "loss": 4.4592, + "step": 996 + }, + { + "epoch": 0.00997, + "grad_norm": 0.6621181102222602, + "learning_rate": 0.002991, + "loss": 4.4437, + "step": 997 + }, + { + "epoch": 0.00998, + "grad_norm": 0.6532026522918166, + "learning_rate": 0.002994, + "loss": 4.4467, + "step": 998 + }, + { + "epoch": 0.00999, + "grad_norm": 0.6863970198472243, + "learning_rate": 0.002997, + "loss": 4.4317, + "step": 999 + }, + { + "epoch": 0.01, + "grad_norm": 0.6643394709769835, + "learning_rate": 0.003, + "loss": 4.4529, + "step": 1000 + }, + { + "epoch": 0.01001, + "grad_norm": 0.5664327532419113, + "learning_rate": 0.003, + "loss": 4.4443, + "step": 1001 + }, + { + "epoch": 0.01002, + "grad_norm": 0.5614282124245983, + "learning_rate": 0.003, + "loss": 4.4107, + "step": 1002 + }, + { + "epoch": 0.01003, + "grad_norm": 0.5395347922472447, + "learning_rate": 0.003, + "loss": 4.4319, + "step": 1003 + }, + { + "epoch": 0.01004, + "grad_norm": 0.5638853817459814, + "learning_rate": 0.003, + "loss": 4.4194, + "step": 1004 + }, + { + "epoch": 0.01005, + "grad_norm": 0.5916080142589283, + "learning_rate": 0.003, + "loss": 4.3932, + "step": 1005 + }, + { + "epoch": 0.01006, + "grad_norm": 0.5700688347090367, + "learning_rate": 0.003, + "loss": 4.3913, + "step": 1006 + }, + { + "epoch": 0.01007, + "grad_norm": 0.5618557938751672, + "learning_rate": 0.003, + "loss": 4.4343, + "step": 1007 + }, + { + "epoch": 0.01008, + "grad_norm": 0.6176294799951538, + "learning_rate": 0.003, + "loss": 4.4449, + "step": 1008 + }, + { + "epoch": 0.01009, + "grad_norm": 0.6917308148051521, + "learning_rate": 0.003, + "loss": 4.4152, + "step": 1009 + }, + { + "epoch": 0.0101, + "grad_norm": 0.7355219940735465, + "learning_rate": 0.003, + "loss": 4.4242, + "step": 1010 + }, + { + "epoch": 0.01011, + "grad_norm": 0.8258393023556594, + "learning_rate": 0.003, + "loss": 4.437, + "step": 1011 + }, + { + "epoch": 0.01012, + "grad_norm": 0.7883368033913232, + "learning_rate": 0.003, + "loss": 4.4273, + "step": 1012 + }, + { + "epoch": 0.01013, + "grad_norm": 0.8421670780561094, + "learning_rate": 0.003, + "loss": 4.453, + "step": 1013 + }, + { + "epoch": 0.01014, + "grad_norm": 0.7611364620814401, + "learning_rate": 0.003, + "loss": 4.3906, + "step": 1014 + }, + { + "epoch": 0.01015, + "grad_norm": 0.8948224171004916, + "learning_rate": 0.003, + "loss": 4.4422, + "step": 1015 + }, + { + "epoch": 0.01016, + "grad_norm": 0.9330968648547454, + "learning_rate": 0.003, + "loss": 4.4537, + "step": 1016 + }, + { + "epoch": 0.01017, + "grad_norm": 0.9071988294298332, + "learning_rate": 0.003, + "loss": 4.4046, + "step": 1017 + }, + { + "epoch": 0.01018, + "grad_norm": 0.9028445315923417, + "learning_rate": 0.003, + "loss": 4.4536, + "step": 1018 + }, + { + "epoch": 0.01019, + "grad_norm": 0.701398424995911, + "learning_rate": 0.003, + "loss": 4.4297, + "step": 1019 + }, + { + "epoch": 0.0102, + "grad_norm": 0.7422831527736846, + "learning_rate": 0.003, + "loss": 4.428, + "step": 1020 + }, + { + "epoch": 0.01021, + "grad_norm": 0.7468650100074169, + "learning_rate": 0.003, + "loss": 4.4136, + "step": 1021 + }, + { + "epoch": 0.01022, + "grad_norm": 0.9325291879496833, + "learning_rate": 0.003, + "loss": 4.4294, + "step": 1022 + }, + { + "epoch": 0.01023, + "grad_norm": 0.8401676161943546, + "learning_rate": 0.003, + "loss": 4.4261, + "step": 1023 + }, + { + "epoch": 0.01024, + "grad_norm": 0.8270343729935282, + "learning_rate": 0.003, + "loss": 4.3898, + "step": 1024 + }, + { + "epoch": 0.01025, + "grad_norm": 0.7180913593704098, + "learning_rate": 0.003, + "loss": 4.4147, + "step": 1025 + }, + { + "epoch": 0.01026, + "grad_norm": 0.6277034329789435, + "learning_rate": 0.003, + "loss": 4.4039, + "step": 1026 + }, + { + "epoch": 0.01027, + "grad_norm": 0.5318737503801085, + "learning_rate": 0.003, + "loss": 4.4219, + "step": 1027 + }, + { + "epoch": 0.01028, + "grad_norm": 0.4999785226446009, + "learning_rate": 0.003, + "loss": 4.4015, + "step": 1028 + }, + { + "epoch": 0.01029, + "grad_norm": 0.48973961084120876, + "learning_rate": 0.003, + "loss": 4.3769, + "step": 1029 + }, + { + "epoch": 0.0103, + "grad_norm": 0.5942733082980836, + "learning_rate": 0.003, + "loss": 4.4125, + "step": 1030 + }, + { + "epoch": 0.01031, + "grad_norm": 0.7229752342980397, + "learning_rate": 0.003, + "loss": 4.4274, + "step": 1031 + }, + { + "epoch": 0.01032, + "grad_norm": 0.7373514084562369, + "learning_rate": 0.003, + "loss": 4.4145, + "step": 1032 + }, + { + "epoch": 0.01033, + "grad_norm": 0.6482800157489234, + "learning_rate": 0.003, + "loss": 4.4346, + "step": 1033 + }, + { + "epoch": 0.01034, + "grad_norm": 0.6524184922218951, + "learning_rate": 0.003, + "loss": 4.4051, + "step": 1034 + }, + { + "epoch": 0.01035, + "grad_norm": 0.705389079509784, + "learning_rate": 0.003, + "loss": 4.4073, + "step": 1035 + }, + { + "epoch": 0.01036, + "grad_norm": 0.7235541204605349, + "learning_rate": 0.003, + "loss": 4.3939, + "step": 1036 + }, + { + "epoch": 0.01037, + "grad_norm": 0.6796343657160102, + "learning_rate": 0.003, + "loss": 4.4161, + "step": 1037 + }, + { + "epoch": 0.01038, + "grad_norm": 0.5773141410598708, + "learning_rate": 0.003, + "loss": 4.4121, + "step": 1038 + }, + { + "epoch": 0.01039, + "grad_norm": 0.5565285106963757, + "learning_rate": 0.003, + "loss": 4.4014, + "step": 1039 + }, + { + "epoch": 0.0104, + "grad_norm": 0.526665138631767, + "learning_rate": 0.003, + "loss": 4.4124, + "step": 1040 + }, + { + "epoch": 0.01041, + "grad_norm": 0.5117793067919421, + "learning_rate": 0.003, + "loss": 4.3909, + "step": 1041 + }, + { + "epoch": 0.01042, + "grad_norm": 0.5092553549098504, + "learning_rate": 0.003, + "loss": 4.4167, + "step": 1042 + }, + { + "epoch": 0.01043, + "grad_norm": 0.5543198083705748, + "learning_rate": 0.003, + "loss": 4.3593, + "step": 1043 + }, + { + "epoch": 0.01044, + "grad_norm": 0.600481111898673, + "learning_rate": 0.003, + "loss": 4.3987, + "step": 1044 + }, + { + "epoch": 0.01045, + "grad_norm": 0.5965980004717368, + "learning_rate": 0.003, + "loss": 4.4162, + "step": 1045 + }, + { + "epoch": 0.01046, + "grad_norm": 0.6740252530686676, + "learning_rate": 0.003, + "loss": 4.3928, + "step": 1046 + }, + { + "epoch": 0.01047, + "grad_norm": 0.7361053030528937, + "learning_rate": 0.003, + "loss": 4.3997, + "step": 1047 + }, + { + "epoch": 0.01048, + "grad_norm": 0.767513361829787, + "learning_rate": 0.003, + "loss": 4.3936, + "step": 1048 + }, + { + "epoch": 0.01049, + "grad_norm": 0.6855190682446736, + "learning_rate": 0.003, + "loss": 4.3699, + "step": 1049 + }, + { + "epoch": 0.0105, + "grad_norm": 0.5741691502464937, + "learning_rate": 0.003, + "loss": 4.4062, + "step": 1050 + }, + { + "epoch": 0.01051, + "grad_norm": 0.6230464159511063, + "learning_rate": 0.003, + "loss": 4.3827, + "step": 1051 + }, + { + "epoch": 0.01052, + "grad_norm": 0.6366737529257143, + "learning_rate": 0.003, + "loss": 4.3967, + "step": 1052 + }, + { + "epoch": 0.01053, + "grad_norm": 0.6755427193415695, + "learning_rate": 0.003, + "loss": 4.4056, + "step": 1053 + }, + { + "epoch": 0.01054, + "grad_norm": 0.6231015147472246, + "learning_rate": 0.003, + "loss": 4.3965, + "step": 1054 + }, + { + "epoch": 0.01055, + "grad_norm": 0.5394645878748071, + "learning_rate": 0.003, + "loss": 4.3715, + "step": 1055 + }, + { + "epoch": 0.01056, + "grad_norm": 0.5755128064540977, + "learning_rate": 0.003, + "loss": 4.3719, + "step": 1056 + }, + { + "epoch": 0.01057, + "grad_norm": 0.5491694969265289, + "learning_rate": 0.003, + "loss": 4.3754, + "step": 1057 + }, + { + "epoch": 0.01058, + "grad_norm": 0.7343919109861163, + "learning_rate": 0.003, + "loss": 4.375, + "step": 1058 + }, + { + "epoch": 0.01059, + "grad_norm": 0.9361910412332192, + "learning_rate": 0.003, + "loss": 4.3892, + "step": 1059 + }, + { + "epoch": 0.0106, + "grad_norm": 0.9065515100292879, + "learning_rate": 0.003, + "loss": 4.3901, + "step": 1060 + }, + { + "epoch": 0.01061, + "grad_norm": 1.1211758916188983, + "learning_rate": 0.003, + "loss": 4.3972, + "step": 1061 + }, + { + "epoch": 0.01062, + "grad_norm": 0.7834596971762835, + "learning_rate": 0.003, + "loss": 4.3907, + "step": 1062 + }, + { + "epoch": 0.01063, + "grad_norm": 0.6963669031971059, + "learning_rate": 0.003, + "loss": 4.3832, + "step": 1063 + }, + { + "epoch": 0.01064, + "grad_norm": 0.6238185439704477, + "learning_rate": 0.003, + "loss": 4.3945, + "step": 1064 + }, + { + "epoch": 0.01065, + "grad_norm": 0.6958753605701185, + "learning_rate": 0.003, + "loss": 4.3993, + "step": 1065 + }, + { + "epoch": 0.01066, + "grad_norm": 0.7129893176363267, + "learning_rate": 0.003, + "loss": 4.4025, + "step": 1066 + }, + { + "epoch": 0.01067, + "grad_norm": 0.7862199002009355, + "learning_rate": 0.003, + "loss": 4.4178, + "step": 1067 + }, + { + "epoch": 0.01068, + "grad_norm": 0.7422673244500091, + "learning_rate": 0.003, + "loss": 4.3984, + "step": 1068 + }, + { + "epoch": 0.01069, + "grad_norm": 0.7997129592129516, + "learning_rate": 0.003, + "loss": 4.4072, + "step": 1069 + }, + { + "epoch": 0.0107, + "grad_norm": 0.945099613944759, + "learning_rate": 0.003, + "loss": 4.4674, + "step": 1070 + }, + { + "epoch": 0.01071, + "grad_norm": 0.8590095455876905, + "learning_rate": 0.003, + "loss": 4.387, + "step": 1071 + }, + { + "epoch": 0.01072, + "grad_norm": 0.8651521698594221, + "learning_rate": 0.003, + "loss": 4.4227, + "step": 1072 + }, + { + "epoch": 0.01073, + "grad_norm": 0.80387311045371, + "learning_rate": 0.003, + "loss": 4.4158, + "step": 1073 + }, + { + "epoch": 0.01074, + "grad_norm": 0.8137862583554676, + "learning_rate": 0.003, + "loss": 4.3801, + "step": 1074 + }, + { + "epoch": 0.01075, + "grad_norm": 0.8055768924194764, + "learning_rate": 0.003, + "loss": 4.4185, + "step": 1075 + }, + { + "epoch": 0.01076, + "grad_norm": 0.8990845414650448, + "learning_rate": 0.003, + "loss": 4.4196, + "step": 1076 + }, + { + "epoch": 0.01077, + "grad_norm": 0.9178768409478193, + "learning_rate": 0.003, + "loss": 4.4207, + "step": 1077 + }, + { + "epoch": 0.01078, + "grad_norm": 0.9083931280693812, + "learning_rate": 0.003, + "loss": 4.3748, + "step": 1078 + }, + { + "epoch": 0.01079, + "grad_norm": 0.861335694801563, + "learning_rate": 0.003, + "loss": 4.4104, + "step": 1079 + }, + { + "epoch": 0.0108, + "grad_norm": 0.6640978224934152, + "learning_rate": 0.003, + "loss": 4.3777, + "step": 1080 + }, + { + "epoch": 0.01081, + "grad_norm": 0.6684756830200824, + "learning_rate": 0.003, + "loss": 4.3836, + "step": 1081 + }, + { + "epoch": 0.01082, + "grad_norm": 0.7577090594132565, + "learning_rate": 0.003, + "loss": 4.4054, + "step": 1082 + }, + { + "epoch": 0.01083, + "grad_norm": 0.8160609336744046, + "learning_rate": 0.003, + "loss": 4.3924, + "step": 1083 + }, + { + "epoch": 0.01084, + "grad_norm": 0.7897303142858512, + "learning_rate": 0.003, + "loss": 4.4011, + "step": 1084 + }, + { + "epoch": 0.01085, + "grad_norm": 0.7064792552725174, + "learning_rate": 0.003, + "loss": 4.3862, + "step": 1085 + }, + { + "epoch": 0.01086, + "grad_norm": 0.6262724475441191, + "learning_rate": 0.003, + "loss": 4.3571, + "step": 1086 + }, + { + "epoch": 0.01087, + "grad_norm": 0.6391781173229293, + "learning_rate": 0.003, + "loss": 4.3729, + "step": 1087 + }, + { + "epoch": 0.01088, + "grad_norm": 0.6632808145329404, + "learning_rate": 0.003, + "loss": 4.4092, + "step": 1088 + }, + { + "epoch": 0.01089, + "grad_norm": 0.6686792714887265, + "learning_rate": 0.003, + "loss": 4.3783, + "step": 1089 + }, + { + "epoch": 0.0109, + "grad_norm": 0.5882408044346112, + "learning_rate": 0.003, + "loss": 4.3992, + "step": 1090 + }, + { + "epoch": 0.01091, + "grad_norm": 0.5565191786424966, + "learning_rate": 0.003, + "loss": 4.3662, + "step": 1091 + }, + { + "epoch": 0.01092, + "grad_norm": 0.571165259700944, + "learning_rate": 0.003, + "loss": 4.3627, + "step": 1092 + }, + { + "epoch": 0.01093, + "grad_norm": 0.5483152737763334, + "learning_rate": 0.003, + "loss": 4.379, + "step": 1093 + }, + { + "epoch": 0.01094, + "grad_norm": 0.4816093735859256, + "learning_rate": 0.003, + "loss": 4.3664, + "step": 1094 + }, + { + "epoch": 0.01095, + "grad_norm": 0.4607091969369753, + "learning_rate": 0.003, + "loss": 4.3913, + "step": 1095 + }, + { + "epoch": 0.01096, + "grad_norm": 0.5948450571982326, + "learning_rate": 0.003, + "loss": 4.3662, + "step": 1096 + }, + { + "epoch": 0.01097, + "grad_norm": 0.8991911041937081, + "learning_rate": 0.003, + "loss": 4.3648, + "step": 1097 + }, + { + "epoch": 0.01098, + "grad_norm": 1.00898451325596, + "learning_rate": 0.003, + "loss": 4.4173, + "step": 1098 + }, + { + "epoch": 0.01099, + "grad_norm": 0.837940726063793, + "learning_rate": 0.003, + "loss": 4.3702, + "step": 1099 + }, + { + "epoch": 0.011, + "grad_norm": 1.0135265311826804, + "learning_rate": 0.003, + "loss": 4.3911, + "step": 1100 + }, + { + "epoch": 0.01101, + "grad_norm": 1.0865706640064363, + "learning_rate": 0.003, + "loss": 4.4129, + "step": 1101 + }, + { + "epoch": 0.01102, + "grad_norm": 0.9407674508321733, + "learning_rate": 0.003, + "loss": 4.4141, + "step": 1102 + }, + { + "epoch": 0.01103, + "grad_norm": 0.7826637713653176, + "learning_rate": 0.003, + "loss": 4.4004, + "step": 1103 + }, + { + "epoch": 0.01104, + "grad_norm": 0.6792138170427816, + "learning_rate": 0.003, + "loss": 4.3864, + "step": 1104 + }, + { + "epoch": 0.01105, + "grad_norm": 0.7494779633947588, + "learning_rate": 0.003, + "loss": 4.4098, + "step": 1105 + }, + { + "epoch": 0.01106, + "grad_norm": 0.7074202272250578, + "learning_rate": 0.003, + "loss": 4.3837, + "step": 1106 + }, + { + "epoch": 0.01107, + "grad_norm": 0.6111631455372859, + "learning_rate": 0.003, + "loss": 4.3903, + "step": 1107 + }, + { + "epoch": 0.01108, + "grad_norm": 0.5958647482211449, + "learning_rate": 0.003, + "loss": 4.3565, + "step": 1108 + }, + { + "epoch": 0.01109, + "grad_norm": 0.615214127583215, + "learning_rate": 0.003, + "loss": 4.3726, + "step": 1109 + }, + { + "epoch": 0.0111, + "grad_norm": 0.5327988413049348, + "learning_rate": 0.003, + "loss": 4.3878, + "step": 1110 + }, + { + "epoch": 0.01111, + "grad_norm": 0.47023927407071675, + "learning_rate": 0.003, + "loss": 4.3622, + "step": 1111 + }, + { + "epoch": 0.01112, + "grad_norm": 0.3776559700056012, + "learning_rate": 0.003, + "loss": 4.3759, + "step": 1112 + }, + { + "epoch": 0.01113, + "grad_norm": 0.3763967476164199, + "learning_rate": 0.003, + "loss": 4.3414, + "step": 1113 + }, + { + "epoch": 0.01114, + "grad_norm": 0.3564238802994098, + "learning_rate": 0.003, + "loss": 4.354, + "step": 1114 + }, + { + "epoch": 0.01115, + "grad_norm": 0.30679188146183123, + "learning_rate": 0.003, + "loss": 4.3753, + "step": 1115 + }, + { + "epoch": 0.01116, + "grad_norm": 0.33853412768388663, + "learning_rate": 0.003, + "loss": 4.3633, + "step": 1116 + }, + { + "epoch": 0.01117, + "grad_norm": 0.3107462307737106, + "learning_rate": 0.003, + "loss": 4.3492, + "step": 1117 + }, + { + "epoch": 0.01118, + "grad_norm": 0.32797709961339483, + "learning_rate": 0.003, + "loss": 4.352, + "step": 1118 + }, + { + "epoch": 0.01119, + "grad_norm": 0.3013823503547338, + "learning_rate": 0.003, + "loss": 4.3543, + "step": 1119 + }, + { + "epoch": 0.0112, + "grad_norm": 0.3198307315980734, + "learning_rate": 0.003, + "loss": 4.3378, + "step": 1120 + }, + { + "epoch": 0.01121, + "grad_norm": 0.33595869744145906, + "learning_rate": 0.003, + "loss": 4.3478, + "step": 1121 + }, + { + "epoch": 0.01122, + "grad_norm": 0.4236485736167927, + "learning_rate": 0.003, + "loss": 4.3259, + "step": 1122 + }, + { + "epoch": 0.01123, + "grad_norm": 0.5207438928810102, + "learning_rate": 0.003, + "loss": 4.3165, + "step": 1123 + }, + { + "epoch": 0.01124, + "grad_norm": 0.5897496309635056, + "learning_rate": 0.003, + "loss": 4.3639, + "step": 1124 + }, + { + "epoch": 0.01125, + "grad_norm": 0.6251226930864733, + "learning_rate": 0.003, + "loss": 4.3351, + "step": 1125 + }, + { + "epoch": 0.01126, + "grad_norm": 0.7904920259399445, + "learning_rate": 0.003, + "loss": 4.3485, + "step": 1126 + }, + { + "epoch": 0.01127, + "grad_norm": 0.8706338403024567, + "learning_rate": 0.003, + "loss": 4.3878, + "step": 1127 + }, + { + "epoch": 0.01128, + "grad_norm": 0.7520394017234492, + "learning_rate": 0.003, + "loss": 4.3618, + "step": 1128 + }, + { + "epoch": 0.01129, + "grad_norm": 0.9434772506575799, + "learning_rate": 0.003, + "loss": 4.3617, + "step": 1129 + }, + { + "epoch": 0.0113, + "grad_norm": 0.9515853773144796, + "learning_rate": 0.003, + "loss": 4.3566, + "step": 1130 + }, + { + "epoch": 0.01131, + "grad_norm": 1.1049329865545139, + "learning_rate": 0.003, + "loss": 4.3775, + "step": 1131 + }, + { + "epoch": 0.01132, + "grad_norm": 0.869968096155467, + "learning_rate": 0.003, + "loss": 4.3991, + "step": 1132 + }, + { + "epoch": 0.01133, + "grad_norm": 0.7539769321914868, + "learning_rate": 0.003, + "loss": 4.3532, + "step": 1133 + }, + { + "epoch": 0.01134, + "grad_norm": 0.8275263694532653, + "learning_rate": 0.003, + "loss": 4.3686, + "step": 1134 + }, + { + "epoch": 0.01135, + "grad_norm": 0.8617681055772174, + "learning_rate": 0.003, + "loss": 4.402, + "step": 1135 + }, + { + "epoch": 0.01136, + "grad_norm": 0.7602421394000872, + "learning_rate": 0.003, + "loss": 4.3721, + "step": 1136 + }, + { + "epoch": 0.01137, + "grad_norm": 0.7778258812220352, + "learning_rate": 0.003, + "loss": 4.3963, + "step": 1137 + }, + { + "epoch": 0.01138, + "grad_norm": 0.7594339489071814, + "learning_rate": 0.003, + "loss": 4.3534, + "step": 1138 + }, + { + "epoch": 0.01139, + "grad_norm": 0.7671596268779747, + "learning_rate": 0.003, + "loss": 4.3643, + "step": 1139 + }, + { + "epoch": 0.0114, + "grad_norm": 0.8353688690035959, + "learning_rate": 0.003, + "loss": 4.3455, + "step": 1140 + }, + { + "epoch": 0.01141, + "grad_norm": 0.9712894348927099, + "learning_rate": 0.003, + "loss": 4.3818, + "step": 1141 + }, + { + "epoch": 0.01142, + "grad_norm": 0.8605012156156574, + "learning_rate": 0.003, + "loss": 4.355, + "step": 1142 + }, + { + "epoch": 0.01143, + "grad_norm": 0.7447870609491655, + "learning_rate": 0.003, + "loss": 4.3674, + "step": 1143 + }, + { + "epoch": 0.01144, + "grad_norm": 0.8620443864386994, + "learning_rate": 0.003, + "loss": 4.339, + "step": 1144 + }, + { + "epoch": 0.01145, + "grad_norm": 0.9495119843583, + "learning_rate": 0.003, + "loss": 4.3837, + "step": 1145 + }, + { + "epoch": 0.01146, + "grad_norm": 0.8719913828537735, + "learning_rate": 0.003, + "loss": 4.3682, + "step": 1146 + }, + { + "epoch": 0.01147, + "grad_norm": 0.8406491643071615, + "learning_rate": 0.003, + "loss": 4.3599, + "step": 1147 + }, + { + "epoch": 0.01148, + "grad_norm": 0.7299655375485881, + "learning_rate": 0.003, + "loss": 4.3549, + "step": 1148 + }, + { + "epoch": 0.01149, + "grad_norm": 0.7470410614101604, + "learning_rate": 0.003, + "loss": 4.3727, + "step": 1149 + }, + { + "epoch": 0.0115, + "grad_norm": 0.7232429634523785, + "learning_rate": 0.003, + "loss": 4.364, + "step": 1150 + }, + { + "epoch": 0.01151, + "grad_norm": 0.6295150712699281, + "learning_rate": 0.003, + "loss": 4.3586, + "step": 1151 + }, + { + "epoch": 0.01152, + "grad_norm": 0.652544707149974, + "learning_rate": 0.003, + "loss": 4.3661, + "step": 1152 + }, + { + "epoch": 0.01153, + "grad_norm": 0.5683271703897544, + "learning_rate": 0.003, + "loss": 4.3775, + "step": 1153 + }, + { + "epoch": 0.01154, + "grad_norm": 0.535300431807245, + "learning_rate": 0.003, + "loss": 4.3515, + "step": 1154 + }, + { + "epoch": 0.01155, + "grad_norm": 0.4766589820931375, + "learning_rate": 0.003, + "loss": 4.3377, + "step": 1155 + }, + { + "epoch": 0.01156, + "grad_norm": 0.5530856703722951, + "learning_rate": 0.003, + "loss": 4.342, + "step": 1156 + }, + { + "epoch": 0.01157, + "grad_norm": 0.6154599631908627, + "learning_rate": 0.003, + "loss": 4.374, + "step": 1157 + }, + { + "epoch": 0.01158, + "grad_norm": 0.693370036269422, + "learning_rate": 0.003, + "loss": 4.3417, + "step": 1158 + }, + { + "epoch": 0.01159, + "grad_norm": 0.7413452595288457, + "learning_rate": 0.003, + "loss": 4.3455, + "step": 1159 + }, + { + "epoch": 0.0116, + "grad_norm": 0.7301771780442605, + "learning_rate": 0.003, + "loss": 4.3545, + "step": 1160 + }, + { + "epoch": 0.01161, + "grad_norm": 0.8159054162369931, + "learning_rate": 0.003, + "loss": 4.3334, + "step": 1161 + }, + { + "epoch": 0.01162, + "grad_norm": 0.7760523716092679, + "learning_rate": 0.003, + "loss": 4.3486, + "step": 1162 + }, + { + "epoch": 0.01163, + "grad_norm": 0.6309863142169387, + "learning_rate": 0.003, + "loss": 4.36, + "step": 1163 + }, + { + "epoch": 0.01164, + "grad_norm": 0.559610446270928, + "learning_rate": 0.003, + "loss": 4.3468, + "step": 1164 + }, + { + "epoch": 0.01165, + "grad_norm": 0.5932728337340286, + "learning_rate": 0.003, + "loss": 4.3443, + "step": 1165 + }, + { + "epoch": 0.01166, + "grad_norm": 0.47536439541847164, + "learning_rate": 0.003, + "loss": 4.3373, + "step": 1166 + }, + { + "epoch": 0.01167, + "grad_norm": 0.4935599339579411, + "learning_rate": 0.003, + "loss": 4.3054, + "step": 1167 + }, + { + "epoch": 0.01168, + "grad_norm": 0.530157986351205, + "learning_rate": 0.003, + "loss": 4.3136, + "step": 1168 + }, + { + "epoch": 0.01169, + "grad_norm": 0.4039160553718301, + "learning_rate": 0.003, + "loss": 4.3497, + "step": 1169 + }, + { + "epoch": 0.0117, + "grad_norm": 0.3887795730112425, + "learning_rate": 0.003, + "loss": 4.3331, + "step": 1170 + }, + { + "epoch": 0.01171, + "grad_norm": 0.44146129518485994, + "learning_rate": 0.003, + "loss": 4.3186, + "step": 1171 + }, + { + "epoch": 0.01172, + "grad_norm": 0.3842074037767514, + "learning_rate": 0.003, + "loss": 4.3032, + "step": 1172 + }, + { + "epoch": 0.01173, + "grad_norm": 0.3741626855737148, + "learning_rate": 0.003, + "loss": 4.3368, + "step": 1173 + }, + { + "epoch": 0.01174, + "grad_norm": 0.41728859284602393, + "learning_rate": 0.003, + "loss": 4.3131, + "step": 1174 + }, + { + "epoch": 0.01175, + "grad_norm": 0.41981830032719913, + "learning_rate": 0.003, + "loss": 4.2965, + "step": 1175 + }, + { + "epoch": 0.01176, + "grad_norm": 0.37957919503111737, + "learning_rate": 0.003, + "loss": 4.3296, + "step": 1176 + }, + { + "epoch": 0.01177, + "grad_norm": 0.3992995566049025, + "learning_rate": 0.003, + "loss": 4.2992, + "step": 1177 + }, + { + "epoch": 0.01178, + "grad_norm": 0.4902741144335086, + "learning_rate": 0.003, + "loss": 4.3211, + "step": 1178 + }, + { + "epoch": 0.01179, + "grad_norm": 0.6458507017424667, + "learning_rate": 0.003, + "loss": 4.3023, + "step": 1179 + }, + { + "epoch": 0.0118, + "grad_norm": 0.7925849610797445, + "learning_rate": 0.003, + "loss": 4.3757, + "step": 1180 + }, + { + "epoch": 0.01181, + "grad_norm": 0.8606498974404746, + "learning_rate": 0.003, + "loss": 4.3669, + "step": 1181 + }, + { + "epoch": 0.01182, + "grad_norm": 0.7553862661379784, + "learning_rate": 0.003, + "loss": 4.3259, + "step": 1182 + }, + { + "epoch": 0.01183, + "grad_norm": 0.7241455767877882, + "learning_rate": 0.003, + "loss": 4.3159, + "step": 1183 + }, + { + "epoch": 0.01184, + "grad_norm": 0.6945084653960217, + "learning_rate": 0.003, + "loss": 4.3593, + "step": 1184 + }, + { + "epoch": 0.01185, + "grad_norm": 0.5809073742888989, + "learning_rate": 0.003, + "loss": 4.3238, + "step": 1185 + }, + { + "epoch": 0.01186, + "grad_norm": 0.556060401057652, + "learning_rate": 0.003, + "loss": 4.3461, + "step": 1186 + }, + { + "epoch": 0.01187, + "grad_norm": 0.6174052114657497, + "learning_rate": 0.003, + "loss": 4.3108, + "step": 1187 + }, + { + "epoch": 0.01188, + "grad_norm": 0.721410101268809, + "learning_rate": 0.003, + "loss": 4.3288, + "step": 1188 + }, + { + "epoch": 0.01189, + "grad_norm": 0.7947776232413513, + "learning_rate": 0.003, + "loss": 4.3281, + "step": 1189 + }, + { + "epoch": 0.0119, + "grad_norm": 0.882572116743252, + "learning_rate": 0.003, + "loss": 4.3193, + "step": 1190 + }, + { + "epoch": 0.01191, + "grad_norm": 1.0146227461981676, + "learning_rate": 0.003, + "loss": 4.307, + "step": 1191 + }, + { + "epoch": 0.01192, + "grad_norm": 1.112972484482218, + "learning_rate": 0.003, + "loss": 4.3512, + "step": 1192 + }, + { + "epoch": 0.01193, + "grad_norm": 0.9615449743933697, + "learning_rate": 0.003, + "loss": 4.3541, + "step": 1193 + }, + { + "epoch": 0.01194, + "grad_norm": 0.8644326468862364, + "learning_rate": 0.003, + "loss": 4.3507, + "step": 1194 + }, + { + "epoch": 0.01195, + "grad_norm": 0.8480336292760144, + "learning_rate": 0.003, + "loss": 4.3317, + "step": 1195 + }, + { + "epoch": 0.01196, + "grad_norm": 0.8784381650624298, + "learning_rate": 0.003, + "loss": 4.3413, + "step": 1196 + }, + { + "epoch": 0.01197, + "grad_norm": 0.9498997917047696, + "learning_rate": 0.003, + "loss": 4.3186, + "step": 1197 + }, + { + "epoch": 0.01198, + "grad_norm": 0.8920952388458802, + "learning_rate": 0.003, + "loss": 4.388, + "step": 1198 + }, + { + "epoch": 0.01199, + "grad_norm": 1.0871047324767358, + "learning_rate": 0.003, + "loss": 4.3742, + "step": 1199 + }, + { + "epoch": 0.012, + "grad_norm": 0.9019122174295907, + "learning_rate": 0.003, + "loss": 4.3579, + "step": 1200 + }, + { + "epoch": 0.01201, + "grad_norm": 0.85321834247841, + "learning_rate": 0.003, + "loss": 4.3858, + "step": 1201 + }, + { + "epoch": 0.01202, + "grad_norm": 0.7349368268324239, + "learning_rate": 0.003, + "loss": 4.3505, + "step": 1202 + }, + { + "epoch": 0.01203, + "grad_norm": 0.6774471285416843, + "learning_rate": 0.003, + "loss": 4.357, + "step": 1203 + }, + { + "epoch": 0.01204, + "grad_norm": 0.6937988439444178, + "learning_rate": 0.003, + "loss": 4.3513, + "step": 1204 + }, + { + "epoch": 0.01205, + "grad_norm": 0.6364064699328508, + "learning_rate": 0.003, + "loss": 4.3406, + "step": 1205 + }, + { + "epoch": 0.01206, + "grad_norm": 0.522276421270424, + "learning_rate": 0.003, + "loss": 4.3798, + "step": 1206 + }, + { + "epoch": 0.01207, + "grad_norm": 0.5076942128127889, + "learning_rate": 0.003, + "loss": 4.2994, + "step": 1207 + }, + { + "epoch": 0.01208, + "grad_norm": 0.47287636588909837, + "learning_rate": 0.003, + "loss": 4.3565, + "step": 1208 + }, + { + "epoch": 0.01209, + "grad_norm": 0.5041956058683869, + "learning_rate": 0.003, + "loss": 4.3379, + "step": 1209 + }, + { + "epoch": 0.0121, + "grad_norm": 0.6167670848439076, + "learning_rate": 0.003, + "loss": 4.341, + "step": 1210 + }, + { + "epoch": 0.01211, + "grad_norm": 0.7242996454774496, + "learning_rate": 0.003, + "loss": 4.3379, + "step": 1211 + }, + { + "epoch": 0.01212, + "grad_norm": 0.8213908511974465, + "learning_rate": 0.003, + "loss": 4.3624, + "step": 1212 + }, + { + "epoch": 0.01213, + "grad_norm": 0.6880698765717139, + "learning_rate": 0.003, + "loss": 4.3342, + "step": 1213 + }, + { + "epoch": 0.01214, + "grad_norm": 0.5407616002767914, + "learning_rate": 0.003, + "loss": 4.311, + "step": 1214 + }, + { + "epoch": 0.01215, + "grad_norm": 0.6707389680302347, + "learning_rate": 0.003, + "loss": 4.3237, + "step": 1215 + }, + { + "epoch": 0.01216, + "grad_norm": 0.715155480014794, + "learning_rate": 0.003, + "loss": 4.3206, + "step": 1216 + }, + { + "epoch": 0.01217, + "grad_norm": 0.6669611805526361, + "learning_rate": 0.003, + "loss": 4.3242, + "step": 1217 + }, + { + "epoch": 0.01218, + "grad_norm": 0.6194536332440173, + "learning_rate": 0.003, + "loss": 4.3183, + "step": 1218 + }, + { + "epoch": 0.01219, + "grad_norm": 0.5624325800012369, + "learning_rate": 0.003, + "loss": 4.3151, + "step": 1219 + }, + { + "epoch": 0.0122, + "grad_norm": 0.5994752597130036, + "learning_rate": 0.003, + "loss": 4.3082, + "step": 1220 + }, + { + "epoch": 0.01221, + "grad_norm": 0.5929964237485116, + "learning_rate": 0.003, + "loss": 4.3217, + "step": 1221 + }, + { + "epoch": 0.01222, + "grad_norm": 0.5364789869662485, + "learning_rate": 0.003, + "loss": 4.3034, + "step": 1222 + }, + { + "epoch": 0.01223, + "grad_norm": 0.4883935817063237, + "learning_rate": 0.003, + "loss": 4.2989, + "step": 1223 + }, + { + "epoch": 0.01224, + "grad_norm": 0.5365327424367556, + "learning_rate": 0.003, + "loss": 4.2844, + "step": 1224 + }, + { + "epoch": 0.01225, + "grad_norm": 0.5060580384449018, + "learning_rate": 0.003, + "loss": 4.3078, + "step": 1225 + }, + { + "epoch": 0.01226, + "grad_norm": 0.4316062800492973, + "learning_rate": 0.003, + "loss": 4.297, + "step": 1226 + }, + { + "epoch": 0.01227, + "grad_norm": 0.41324487305946883, + "learning_rate": 0.003, + "loss": 4.3243, + "step": 1227 + }, + { + "epoch": 0.01228, + "grad_norm": 0.4686302520601058, + "learning_rate": 0.003, + "loss": 4.3085, + "step": 1228 + }, + { + "epoch": 0.01229, + "grad_norm": 0.5830158408620507, + "learning_rate": 0.003, + "loss": 4.2989, + "step": 1229 + }, + { + "epoch": 0.0123, + "grad_norm": 0.7143816453247004, + "learning_rate": 0.003, + "loss": 4.3393, + "step": 1230 + }, + { + "epoch": 0.01231, + "grad_norm": 0.8518087843200909, + "learning_rate": 0.003, + "loss": 4.3134, + "step": 1231 + }, + { + "epoch": 0.01232, + "grad_norm": 0.8467846529950723, + "learning_rate": 0.003, + "loss": 4.3028, + "step": 1232 + }, + { + "epoch": 0.01233, + "grad_norm": 0.7843462879273329, + "learning_rate": 0.003, + "loss": 4.3052, + "step": 1233 + }, + { + "epoch": 0.01234, + "grad_norm": 0.8956533482473612, + "learning_rate": 0.003, + "loss": 4.3298, + "step": 1234 + }, + { + "epoch": 0.01235, + "grad_norm": 0.9162978759093423, + "learning_rate": 0.003, + "loss": 4.3266, + "step": 1235 + }, + { + "epoch": 0.01236, + "grad_norm": 0.860682215697324, + "learning_rate": 0.003, + "loss": 4.3299, + "step": 1236 + }, + { + "epoch": 0.01237, + "grad_norm": 0.851315880486755, + "learning_rate": 0.003, + "loss": 4.3393, + "step": 1237 + }, + { + "epoch": 0.01238, + "grad_norm": 0.8294971616591855, + "learning_rate": 0.003, + "loss": 4.3358, + "step": 1238 + }, + { + "epoch": 0.01239, + "grad_norm": 0.9437199934028814, + "learning_rate": 0.003, + "loss": 4.3282, + "step": 1239 + }, + { + "epoch": 0.0124, + "grad_norm": 0.9106084651998833, + "learning_rate": 0.003, + "loss": 4.3137, + "step": 1240 + }, + { + "epoch": 0.01241, + "grad_norm": 0.8351496076806352, + "learning_rate": 0.003, + "loss": 4.3273, + "step": 1241 + }, + { + "epoch": 0.01242, + "grad_norm": 0.6918670619381783, + "learning_rate": 0.003, + "loss": 4.3257, + "step": 1242 + }, + { + "epoch": 0.01243, + "grad_norm": 0.732924073441867, + "learning_rate": 0.003, + "loss": 4.3077, + "step": 1243 + }, + { + "epoch": 0.01244, + "grad_norm": 0.7515551346375118, + "learning_rate": 0.003, + "loss": 4.3219, + "step": 1244 + }, + { + "epoch": 0.01245, + "grad_norm": 0.635919850996669, + "learning_rate": 0.003, + "loss": 4.3108, + "step": 1245 + }, + { + "epoch": 0.01246, + "grad_norm": 0.5245627080366145, + "learning_rate": 0.003, + "loss": 4.2806, + "step": 1246 + }, + { + "epoch": 0.01247, + "grad_norm": 0.5964323822303592, + "learning_rate": 0.003, + "loss": 4.3378, + "step": 1247 + }, + { + "epoch": 0.01248, + "grad_norm": 0.7016619320460757, + "learning_rate": 0.003, + "loss": 4.3217, + "step": 1248 + }, + { + "epoch": 0.01249, + "grad_norm": 0.7231849999092693, + "learning_rate": 0.003, + "loss": 4.3154, + "step": 1249 + }, + { + "epoch": 0.0125, + "grad_norm": 0.6036388669727554, + "learning_rate": 0.003, + "loss": 4.2869, + "step": 1250 + }, + { + "epoch": 0.01251, + "grad_norm": 0.5253366712272899, + "learning_rate": 0.003, + "loss": 4.316, + "step": 1251 + }, + { + "epoch": 0.01252, + "grad_norm": 0.46320485886810614, + "learning_rate": 0.003, + "loss": 4.2912, + "step": 1252 + }, + { + "epoch": 0.01253, + "grad_norm": 0.4963582371471178, + "learning_rate": 0.003, + "loss": 4.2694, + "step": 1253 + }, + { + "epoch": 0.01254, + "grad_norm": 0.5081054207596929, + "learning_rate": 0.003, + "loss": 4.3157, + "step": 1254 + }, + { + "epoch": 0.01255, + "grad_norm": 0.4761071055102706, + "learning_rate": 0.003, + "loss": 4.2641, + "step": 1255 + }, + { + "epoch": 0.01256, + "grad_norm": 0.5232678892632563, + "learning_rate": 0.003, + "loss": 4.2916, + "step": 1256 + }, + { + "epoch": 0.01257, + "grad_norm": 0.5344373546996498, + "learning_rate": 0.003, + "loss": 4.2768, + "step": 1257 + }, + { + "epoch": 0.01258, + "grad_norm": 0.5773429526145795, + "learning_rate": 0.003, + "loss": 4.2867, + "step": 1258 + }, + { + "epoch": 0.01259, + "grad_norm": 0.6584062985843472, + "learning_rate": 0.003, + "loss": 4.2918, + "step": 1259 + }, + { + "epoch": 0.0126, + "grad_norm": 0.6804145752695209, + "learning_rate": 0.003, + "loss": 4.3214, + "step": 1260 + }, + { + "epoch": 0.01261, + "grad_norm": 0.6981998021120679, + "learning_rate": 0.003, + "loss": 4.3161, + "step": 1261 + }, + { + "epoch": 0.01262, + "grad_norm": 0.6112143338587285, + "learning_rate": 0.003, + "loss": 4.304, + "step": 1262 + }, + { + "epoch": 0.01263, + "grad_norm": 0.5927786151181257, + "learning_rate": 0.003, + "loss": 4.298, + "step": 1263 + }, + { + "epoch": 0.01264, + "grad_norm": 0.6346760589695591, + "learning_rate": 0.003, + "loss": 4.2851, + "step": 1264 + }, + { + "epoch": 0.01265, + "grad_norm": 0.6927824976704505, + "learning_rate": 0.003, + "loss": 4.3011, + "step": 1265 + }, + { + "epoch": 0.01266, + "grad_norm": 0.7328319392289776, + "learning_rate": 0.003, + "loss": 4.3026, + "step": 1266 + }, + { + "epoch": 0.01267, + "grad_norm": 0.8489062737553393, + "learning_rate": 0.003, + "loss": 4.325, + "step": 1267 + }, + { + "epoch": 0.01268, + "grad_norm": 0.8961579939511539, + "learning_rate": 0.003, + "loss": 4.3257, + "step": 1268 + }, + { + "epoch": 0.01269, + "grad_norm": 0.821810765966473, + "learning_rate": 0.003, + "loss": 4.2839, + "step": 1269 + }, + { + "epoch": 0.0127, + "grad_norm": 0.8357784457532415, + "learning_rate": 0.003, + "loss": 4.3228, + "step": 1270 + }, + { + "epoch": 0.01271, + "grad_norm": 0.7922720998753914, + "learning_rate": 0.003, + "loss": 4.3037, + "step": 1271 + }, + { + "epoch": 0.01272, + "grad_norm": 0.7964991232532974, + "learning_rate": 0.003, + "loss": 4.3039, + "step": 1272 + }, + { + "epoch": 0.01273, + "grad_norm": 0.7778280400168709, + "learning_rate": 0.003, + "loss": 4.3248, + "step": 1273 + }, + { + "epoch": 0.01274, + "grad_norm": 0.9153345412853071, + "learning_rate": 0.003, + "loss": 4.3241, + "step": 1274 + }, + { + "epoch": 0.01275, + "grad_norm": 0.8587735334056731, + "learning_rate": 0.003, + "loss": 4.294, + "step": 1275 + }, + { + "epoch": 0.01276, + "grad_norm": 0.7198180114679126, + "learning_rate": 0.003, + "loss": 4.2973, + "step": 1276 + }, + { + "epoch": 0.01277, + "grad_norm": 0.7510923187630398, + "learning_rate": 0.003, + "loss": 4.3066, + "step": 1277 + }, + { + "epoch": 0.01278, + "grad_norm": 0.8054501129326099, + "learning_rate": 0.003, + "loss": 4.3179, + "step": 1278 + }, + { + "epoch": 0.01279, + "grad_norm": 0.7824154610465377, + "learning_rate": 0.003, + "loss": 4.2997, + "step": 1279 + }, + { + "epoch": 0.0128, + "grad_norm": 0.7320824999573285, + "learning_rate": 0.003, + "loss": 4.2956, + "step": 1280 + }, + { + "epoch": 0.01281, + "grad_norm": 0.6960943615110375, + "learning_rate": 0.003, + "loss": 4.2806, + "step": 1281 + }, + { + "epoch": 0.01282, + "grad_norm": 0.5992318655887974, + "learning_rate": 0.003, + "loss": 4.3205, + "step": 1282 + }, + { + "epoch": 0.01283, + "grad_norm": 0.6115671413316082, + "learning_rate": 0.003, + "loss": 4.2898, + "step": 1283 + }, + { + "epoch": 0.01284, + "grad_norm": 0.5685168810152473, + "learning_rate": 0.003, + "loss": 4.2893, + "step": 1284 + }, + { + "epoch": 0.01285, + "grad_norm": 0.6350368869509793, + "learning_rate": 0.003, + "loss": 4.2923, + "step": 1285 + }, + { + "epoch": 0.01286, + "grad_norm": 0.6769385854863653, + "learning_rate": 0.003, + "loss": 4.3189, + "step": 1286 + }, + { + "epoch": 0.01287, + "grad_norm": 0.5997038647975563, + "learning_rate": 0.003, + "loss": 4.2792, + "step": 1287 + }, + { + "epoch": 0.01288, + "grad_norm": 0.6318526861304625, + "learning_rate": 0.003, + "loss": 4.2923, + "step": 1288 + }, + { + "epoch": 0.01289, + "grad_norm": 0.6619003980657734, + "learning_rate": 0.003, + "loss": 4.2877, + "step": 1289 + }, + { + "epoch": 0.0129, + "grad_norm": 0.7141266768182228, + "learning_rate": 0.003, + "loss": 4.3104, + "step": 1290 + }, + { + "epoch": 0.01291, + "grad_norm": 0.7063834987249501, + "learning_rate": 0.003, + "loss": 4.2921, + "step": 1291 + }, + { + "epoch": 0.01292, + "grad_norm": 0.6423787800334868, + "learning_rate": 0.003, + "loss": 4.3242, + "step": 1292 + }, + { + "epoch": 0.01293, + "grad_norm": 0.5344471091426154, + "learning_rate": 0.003, + "loss": 4.3062, + "step": 1293 + }, + { + "epoch": 0.01294, + "grad_norm": 0.4523910877254715, + "learning_rate": 0.003, + "loss": 4.2707, + "step": 1294 + }, + { + "epoch": 0.01295, + "grad_norm": 0.5040604805545252, + "learning_rate": 0.003, + "loss": 4.2993, + "step": 1295 + }, + { + "epoch": 0.01296, + "grad_norm": 0.5193740138388458, + "learning_rate": 0.003, + "loss": 4.2656, + "step": 1296 + }, + { + "epoch": 0.01297, + "grad_norm": 0.5346893098464323, + "learning_rate": 0.003, + "loss": 4.2788, + "step": 1297 + }, + { + "epoch": 0.01298, + "grad_norm": 0.5373039682823991, + "learning_rate": 0.003, + "loss": 4.2592, + "step": 1298 + }, + { + "epoch": 0.01299, + "grad_norm": 0.5661548725913075, + "learning_rate": 0.003, + "loss": 4.2709, + "step": 1299 + }, + { + "epoch": 0.013, + "grad_norm": 0.7446298448290246, + "learning_rate": 0.003, + "loss": 4.2678, + "step": 1300 + }, + { + "epoch": 0.01301, + "grad_norm": 0.8230150822772082, + "learning_rate": 0.003, + "loss": 4.2955, + "step": 1301 + }, + { + "epoch": 0.01302, + "grad_norm": 0.7744841619477817, + "learning_rate": 0.003, + "loss": 4.3361, + "step": 1302 + }, + { + "epoch": 0.01303, + "grad_norm": 0.6454504186361478, + "learning_rate": 0.003, + "loss": 4.2765, + "step": 1303 + }, + { + "epoch": 0.01304, + "grad_norm": 0.7188710255689805, + "learning_rate": 0.003, + "loss": 4.3063, + "step": 1304 + }, + { + "epoch": 0.01305, + "grad_norm": 0.757402061055745, + "learning_rate": 0.003, + "loss": 4.2938, + "step": 1305 + }, + { + "epoch": 0.01306, + "grad_norm": 0.691272304301314, + "learning_rate": 0.003, + "loss": 4.2864, + "step": 1306 + }, + { + "epoch": 0.01307, + "grad_norm": 0.706053416610055, + "learning_rate": 0.003, + "loss": 4.3074, + "step": 1307 + }, + { + "epoch": 0.01308, + "grad_norm": 0.6221880747565551, + "learning_rate": 0.003, + "loss": 4.284, + "step": 1308 + }, + { + "epoch": 0.01309, + "grad_norm": 0.5791770340268083, + "learning_rate": 0.003, + "loss": 4.3174, + "step": 1309 + }, + { + "epoch": 0.0131, + "grad_norm": 0.6695730992498613, + "learning_rate": 0.003, + "loss": 4.2672, + "step": 1310 + }, + { + "epoch": 0.01311, + "grad_norm": 0.6922028781625468, + "learning_rate": 0.003, + "loss": 4.3073, + "step": 1311 + }, + { + "epoch": 0.01312, + "grad_norm": 0.8287735129364245, + "learning_rate": 0.003, + "loss": 4.2832, + "step": 1312 + }, + { + "epoch": 0.01313, + "grad_norm": 0.7976431888385418, + "learning_rate": 0.003, + "loss": 4.2757, + "step": 1313 + }, + { + "epoch": 0.01314, + "grad_norm": 0.7220518823683293, + "learning_rate": 0.003, + "loss": 4.2709, + "step": 1314 + }, + { + "epoch": 0.01315, + "grad_norm": 0.6095265673664224, + "learning_rate": 0.003, + "loss": 4.2914, + "step": 1315 + }, + { + "epoch": 0.01316, + "grad_norm": 0.7674466841738673, + "learning_rate": 0.003, + "loss": 4.2855, + "step": 1316 + }, + { + "epoch": 0.01317, + "grad_norm": 0.9059745922778419, + "learning_rate": 0.003, + "loss": 4.3151, + "step": 1317 + }, + { + "epoch": 0.01318, + "grad_norm": 0.9427302613153309, + "learning_rate": 0.003, + "loss": 4.3056, + "step": 1318 + }, + { + "epoch": 0.01319, + "grad_norm": 0.73054171411267, + "learning_rate": 0.003, + "loss": 4.3015, + "step": 1319 + }, + { + "epoch": 0.0132, + "grad_norm": 0.7111720819817223, + "learning_rate": 0.003, + "loss": 4.3007, + "step": 1320 + }, + { + "epoch": 0.01321, + "grad_norm": 0.7456349593949091, + "learning_rate": 0.003, + "loss": 4.2768, + "step": 1321 + }, + { + "epoch": 0.01322, + "grad_norm": 0.7807986643070355, + "learning_rate": 0.003, + "loss": 4.3113, + "step": 1322 + }, + { + "epoch": 0.01323, + "grad_norm": 0.8224389080614981, + "learning_rate": 0.003, + "loss": 4.3093, + "step": 1323 + }, + { + "epoch": 0.01324, + "grad_norm": 1.035353941682105, + "learning_rate": 0.003, + "loss": 4.3086, + "step": 1324 + }, + { + "epoch": 0.01325, + "grad_norm": 0.8946051092509474, + "learning_rate": 0.003, + "loss": 4.3063, + "step": 1325 + }, + { + "epoch": 0.01326, + "grad_norm": 1.0201089618117705, + "learning_rate": 0.003, + "loss": 4.2973, + "step": 1326 + }, + { + "epoch": 0.01327, + "grad_norm": 0.864867336973391, + "learning_rate": 0.003, + "loss": 4.3115, + "step": 1327 + }, + { + "epoch": 0.01328, + "grad_norm": 0.8025779319028616, + "learning_rate": 0.003, + "loss": 4.3034, + "step": 1328 + }, + { + "epoch": 0.01329, + "grad_norm": 0.7807799111244325, + "learning_rate": 0.003, + "loss": 4.2681, + "step": 1329 + }, + { + "epoch": 0.0133, + "grad_norm": 0.8681894612953762, + "learning_rate": 0.003, + "loss": 4.3154, + "step": 1330 + }, + { + "epoch": 0.01331, + "grad_norm": 1.0494856392086225, + "learning_rate": 0.003, + "loss": 4.3161, + "step": 1331 + }, + { + "epoch": 0.01332, + "grad_norm": 0.9504785674947578, + "learning_rate": 0.003, + "loss": 4.3364, + "step": 1332 + }, + { + "epoch": 0.01333, + "grad_norm": 0.9898834831271559, + "learning_rate": 0.003, + "loss": 4.3084, + "step": 1333 + }, + { + "epoch": 0.01334, + "grad_norm": 1.0019143105022317, + "learning_rate": 0.003, + "loss": 4.3012, + "step": 1334 + }, + { + "epoch": 0.01335, + "grad_norm": 0.8650334140122408, + "learning_rate": 0.003, + "loss": 4.3214, + "step": 1335 + }, + { + "epoch": 0.01336, + "grad_norm": 0.8920064299561147, + "learning_rate": 0.003, + "loss": 4.317, + "step": 1336 + }, + { + "epoch": 0.01337, + "grad_norm": 0.7707840908187834, + "learning_rate": 0.003, + "loss": 4.3286, + "step": 1337 + }, + { + "epoch": 0.01338, + "grad_norm": 0.7632598332017204, + "learning_rate": 0.003, + "loss": 4.2919, + "step": 1338 + }, + { + "epoch": 0.01339, + "grad_norm": 0.7360297385886754, + "learning_rate": 0.003, + "loss": 4.3155, + "step": 1339 + }, + { + "epoch": 0.0134, + "grad_norm": 0.6003622832803963, + "learning_rate": 0.003, + "loss": 4.2916, + "step": 1340 + }, + { + "epoch": 0.01341, + "grad_norm": 0.6955896552188154, + "learning_rate": 0.003, + "loss": 4.3105, + "step": 1341 + }, + { + "epoch": 0.01342, + "grad_norm": 0.6883046873976916, + "learning_rate": 0.003, + "loss": 4.2818, + "step": 1342 + }, + { + "epoch": 0.01343, + "grad_norm": 0.7284558823650665, + "learning_rate": 0.003, + "loss": 4.2883, + "step": 1343 + }, + { + "epoch": 0.01344, + "grad_norm": 0.6360126374822531, + "learning_rate": 0.003, + "loss": 4.2837, + "step": 1344 + }, + { + "epoch": 0.01345, + "grad_norm": 0.5334484786655853, + "learning_rate": 0.003, + "loss": 4.3176, + "step": 1345 + }, + { + "epoch": 0.01346, + "grad_norm": 0.5681507381379014, + "learning_rate": 0.003, + "loss": 4.2885, + "step": 1346 + }, + { + "epoch": 0.01347, + "grad_norm": 0.5788404528903132, + "learning_rate": 0.003, + "loss": 4.2816, + "step": 1347 + }, + { + "epoch": 0.01348, + "grad_norm": 0.5418128304616573, + "learning_rate": 0.003, + "loss": 4.2755, + "step": 1348 + }, + { + "epoch": 0.01349, + "grad_norm": 0.525198484250559, + "learning_rate": 0.003, + "loss": 4.2954, + "step": 1349 + }, + { + "epoch": 0.0135, + "grad_norm": 0.46326210720927236, + "learning_rate": 0.003, + "loss": 4.2853, + "step": 1350 + }, + { + "epoch": 0.01351, + "grad_norm": 0.4585121858352072, + "learning_rate": 0.003, + "loss": 4.313, + "step": 1351 + }, + { + "epoch": 0.01352, + "grad_norm": 0.4283890963121618, + "learning_rate": 0.003, + "loss": 4.286, + "step": 1352 + }, + { + "epoch": 0.01353, + "grad_norm": 0.42199420882026434, + "learning_rate": 0.003, + "loss": 4.2702, + "step": 1353 + }, + { + "epoch": 0.01354, + "grad_norm": 0.4262744973588792, + "learning_rate": 0.003, + "loss": 4.2568, + "step": 1354 + }, + { + "epoch": 0.01355, + "grad_norm": 0.40850743293980507, + "learning_rate": 0.003, + "loss": 4.268, + "step": 1355 + }, + { + "epoch": 0.01356, + "grad_norm": 0.41818294724190413, + "learning_rate": 0.003, + "loss": 4.2623, + "step": 1356 + }, + { + "epoch": 0.01357, + "grad_norm": 0.43655941768902196, + "learning_rate": 0.003, + "loss": 4.2503, + "step": 1357 + }, + { + "epoch": 0.01358, + "grad_norm": 0.49395890148457533, + "learning_rate": 0.003, + "loss": 4.2463, + "step": 1358 + }, + { + "epoch": 0.01359, + "grad_norm": 0.616978373290081, + "learning_rate": 0.003, + "loss": 4.2894, + "step": 1359 + }, + { + "epoch": 0.0136, + "grad_norm": 0.735374833430507, + "learning_rate": 0.003, + "loss": 4.2639, + "step": 1360 + }, + { + "epoch": 0.01361, + "grad_norm": 0.7389266502027623, + "learning_rate": 0.003, + "loss": 4.2792, + "step": 1361 + }, + { + "epoch": 0.01362, + "grad_norm": 0.65177426874564, + "learning_rate": 0.003, + "loss": 4.2865, + "step": 1362 + }, + { + "epoch": 0.01363, + "grad_norm": 0.6753488054632895, + "learning_rate": 0.003, + "loss": 4.2893, + "step": 1363 + }, + { + "epoch": 0.01364, + "grad_norm": 0.8240837082437891, + "learning_rate": 0.003, + "loss": 4.2868, + "step": 1364 + }, + { + "epoch": 0.01365, + "grad_norm": 0.7993265568553571, + "learning_rate": 0.003, + "loss": 4.306, + "step": 1365 + }, + { + "epoch": 0.01366, + "grad_norm": 0.6564545276995603, + "learning_rate": 0.003, + "loss": 4.2558, + "step": 1366 + }, + { + "epoch": 0.01367, + "grad_norm": 0.5493737530398193, + "learning_rate": 0.003, + "loss": 4.3003, + "step": 1367 + }, + { + "epoch": 0.01368, + "grad_norm": 0.5650653386363494, + "learning_rate": 0.003, + "loss": 4.2739, + "step": 1368 + }, + { + "epoch": 0.01369, + "grad_norm": 0.46328750272215335, + "learning_rate": 0.003, + "loss": 4.262, + "step": 1369 + }, + { + "epoch": 0.0137, + "grad_norm": 0.5045609042462222, + "learning_rate": 0.003, + "loss": 4.2894, + "step": 1370 + }, + { + "epoch": 0.01371, + "grad_norm": 0.5910891998788811, + "learning_rate": 0.003, + "loss": 4.2822, + "step": 1371 + }, + { + "epoch": 0.01372, + "grad_norm": 0.6815189005820225, + "learning_rate": 0.003, + "loss": 4.2381, + "step": 1372 + }, + { + "epoch": 0.01373, + "grad_norm": 0.842580155535958, + "learning_rate": 0.003, + "loss": 4.2808, + "step": 1373 + }, + { + "epoch": 0.01374, + "grad_norm": 0.907805736475784, + "learning_rate": 0.003, + "loss": 4.2961, + "step": 1374 + }, + { + "epoch": 0.01375, + "grad_norm": 0.8312986426917952, + "learning_rate": 0.003, + "loss": 4.2534, + "step": 1375 + }, + { + "epoch": 0.01376, + "grad_norm": 0.8401650089095936, + "learning_rate": 0.003, + "loss": 4.2829, + "step": 1376 + }, + { + "epoch": 0.01377, + "grad_norm": 0.7706178388137035, + "learning_rate": 0.003, + "loss": 4.2816, + "step": 1377 + }, + { + "epoch": 0.01378, + "grad_norm": 0.6957781157082459, + "learning_rate": 0.003, + "loss": 4.3087, + "step": 1378 + }, + { + "epoch": 0.01379, + "grad_norm": 0.6831406156786277, + "learning_rate": 0.003, + "loss": 4.2792, + "step": 1379 + }, + { + "epoch": 0.0138, + "grad_norm": 0.6334610013963772, + "learning_rate": 0.003, + "loss": 4.2739, + "step": 1380 + }, + { + "epoch": 0.01381, + "grad_norm": 0.5316939383075773, + "learning_rate": 0.003, + "loss": 4.2776, + "step": 1381 + }, + { + "epoch": 0.01382, + "grad_norm": 0.516573615018121, + "learning_rate": 0.003, + "loss": 4.2735, + "step": 1382 + }, + { + "epoch": 0.01383, + "grad_norm": 0.5599212453399529, + "learning_rate": 0.003, + "loss": 4.2692, + "step": 1383 + }, + { + "epoch": 0.01384, + "grad_norm": 0.6188520650392525, + "learning_rate": 0.003, + "loss": 4.3, + "step": 1384 + }, + { + "epoch": 0.01385, + "grad_norm": 0.6353409140192112, + "learning_rate": 0.003, + "loss": 4.2877, + "step": 1385 + }, + { + "epoch": 0.01386, + "grad_norm": 0.6356880916095538, + "learning_rate": 0.003, + "loss": 4.2638, + "step": 1386 + }, + { + "epoch": 0.01387, + "grad_norm": 0.7257450023455765, + "learning_rate": 0.003, + "loss": 4.2537, + "step": 1387 + }, + { + "epoch": 0.01388, + "grad_norm": 0.7014765592615619, + "learning_rate": 0.003, + "loss": 4.2914, + "step": 1388 + }, + { + "epoch": 0.01389, + "grad_norm": 0.6616359972565842, + "learning_rate": 0.003, + "loss": 4.2702, + "step": 1389 + }, + { + "epoch": 0.0139, + "grad_norm": 0.6527699070789852, + "learning_rate": 0.003, + "loss": 4.2735, + "step": 1390 + }, + { + "epoch": 0.01391, + "grad_norm": 0.7095964540484884, + "learning_rate": 0.003, + "loss": 4.2741, + "step": 1391 + }, + { + "epoch": 0.01392, + "grad_norm": 0.7501817269895354, + "learning_rate": 0.003, + "loss": 4.2866, + "step": 1392 + }, + { + "epoch": 0.01393, + "grad_norm": 0.6714529168989051, + "learning_rate": 0.003, + "loss": 4.2555, + "step": 1393 + }, + { + "epoch": 0.01394, + "grad_norm": 0.5871737080135582, + "learning_rate": 0.003, + "loss": 4.2392, + "step": 1394 + }, + { + "epoch": 0.01395, + "grad_norm": 0.5621623507740485, + "learning_rate": 0.003, + "loss": 4.2747, + "step": 1395 + }, + { + "epoch": 0.01396, + "grad_norm": 0.664320619088776, + "learning_rate": 0.003, + "loss": 4.2655, + "step": 1396 + }, + { + "epoch": 0.01397, + "grad_norm": 0.7274104246733604, + "learning_rate": 0.003, + "loss": 4.2817, + "step": 1397 + }, + { + "epoch": 0.01398, + "grad_norm": 0.6424657996436192, + "learning_rate": 0.003, + "loss": 4.2594, + "step": 1398 + }, + { + "epoch": 0.01399, + "grad_norm": 0.6842824738024023, + "learning_rate": 0.003, + "loss": 4.2829, + "step": 1399 + }, + { + "epoch": 0.014, + "grad_norm": 0.8261095933984154, + "learning_rate": 0.003, + "loss": 4.2493, + "step": 1400 + }, + { + "epoch": 0.01401, + "grad_norm": 0.7965575457737929, + "learning_rate": 0.003, + "loss": 4.2708, + "step": 1401 + }, + { + "epoch": 0.01402, + "grad_norm": 0.8026102237656736, + "learning_rate": 0.003, + "loss": 4.2842, + "step": 1402 + }, + { + "epoch": 0.01403, + "grad_norm": 0.816192334350639, + "learning_rate": 0.003, + "loss": 4.2628, + "step": 1403 + }, + { + "epoch": 0.01404, + "grad_norm": 0.9312442614599608, + "learning_rate": 0.003, + "loss": 4.2897, + "step": 1404 + }, + { + "epoch": 0.01405, + "grad_norm": 1.0392313893683864, + "learning_rate": 0.003, + "loss": 4.319, + "step": 1405 + }, + { + "epoch": 0.01406, + "grad_norm": 0.9393883719208338, + "learning_rate": 0.003, + "loss": 4.3, + "step": 1406 + }, + { + "epoch": 0.01407, + "grad_norm": 0.9978258942739143, + "learning_rate": 0.003, + "loss": 4.3193, + "step": 1407 + }, + { + "epoch": 0.01408, + "grad_norm": 1.1500652805383735, + "learning_rate": 0.003, + "loss": 4.2914, + "step": 1408 + }, + { + "epoch": 0.01409, + "grad_norm": 1.0900231626852614, + "learning_rate": 0.003, + "loss": 4.3253, + "step": 1409 + }, + { + "epoch": 0.0141, + "grad_norm": 1.0202054584780536, + "learning_rate": 0.003, + "loss": 4.3009, + "step": 1410 + }, + { + "epoch": 0.01411, + "grad_norm": 0.898714028166491, + "learning_rate": 0.003, + "loss": 4.2744, + "step": 1411 + }, + { + "epoch": 0.01412, + "grad_norm": 0.7435025676246113, + "learning_rate": 0.003, + "loss": 4.3061, + "step": 1412 + }, + { + "epoch": 0.01413, + "grad_norm": 0.5873930770772416, + "learning_rate": 0.003, + "loss": 4.2941, + "step": 1413 + }, + { + "epoch": 0.01414, + "grad_norm": 0.6016255593464158, + "learning_rate": 0.003, + "loss": 4.2747, + "step": 1414 + }, + { + "epoch": 0.01415, + "grad_norm": 0.6099426868013017, + "learning_rate": 0.003, + "loss": 4.2659, + "step": 1415 + }, + { + "epoch": 0.01416, + "grad_norm": 0.5575813775095378, + "learning_rate": 0.003, + "loss": 4.2911, + "step": 1416 + }, + { + "epoch": 0.01417, + "grad_norm": 0.5917623365800833, + "learning_rate": 0.003, + "loss": 4.2803, + "step": 1417 + }, + { + "epoch": 0.01418, + "grad_norm": 0.6068864838347083, + "learning_rate": 0.003, + "loss": 4.2958, + "step": 1418 + }, + { + "epoch": 0.01419, + "grad_norm": 0.6824165624145411, + "learning_rate": 0.003, + "loss": 4.2858, + "step": 1419 + }, + { + "epoch": 0.0142, + "grad_norm": 0.6199413471532862, + "learning_rate": 0.003, + "loss": 4.291, + "step": 1420 + }, + { + "epoch": 0.01421, + "grad_norm": 0.5408211469998263, + "learning_rate": 0.003, + "loss": 4.2544, + "step": 1421 + }, + { + "epoch": 0.01422, + "grad_norm": 0.5533737134579111, + "learning_rate": 0.003, + "loss": 4.2598, + "step": 1422 + }, + { + "epoch": 0.01423, + "grad_norm": 0.5685842699001438, + "learning_rate": 0.003, + "loss": 4.2645, + "step": 1423 + }, + { + "epoch": 0.01424, + "grad_norm": 0.6304522666735822, + "learning_rate": 0.003, + "loss": 4.2833, + "step": 1424 + }, + { + "epoch": 0.01425, + "grad_norm": 0.6515865244102063, + "learning_rate": 0.003, + "loss": 4.2865, + "step": 1425 + }, + { + "epoch": 0.01426, + "grad_norm": 0.7444754270308325, + "learning_rate": 0.003, + "loss": 4.2655, + "step": 1426 + }, + { + "epoch": 0.01427, + "grad_norm": 0.7497304563870452, + "learning_rate": 0.003, + "loss": 4.263, + "step": 1427 + }, + { + "epoch": 0.01428, + "grad_norm": 0.6556010515324793, + "learning_rate": 0.003, + "loss": 4.2594, + "step": 1428 + }, + { + "epoch": 0.01429, + "grad_norm": 0.6636248907749417, + "learning_rate": 0.003, + "loss": 4.265, + "step": 1429 + }, + { + "epoch": 0.0143, + "grad_norm": 0.7211105035251077, + "learning_rate": 0.003, + "loss": 4.28, + "step": 1430 + }, + { + "epoch": 0.01431, + "grad_norm": 0.7061472665079888, + "learning_rate": 0.003, + "loss": 4.2684, + "step": 1431 + }, + { + "epoch": 0.01432, + "grad_norm": 0.6456431394575474, + "learning_rate": 0.003, + "loss": 4.2556, + "step": 1432 + }, + { + "epoch": 0.01433, + "grad_norm": 0.7570512378286692, + "learning_rate": 0.003, + "loss": 4.2568, + "step": 1433 + }, + { + "epoch": 0.01434, + "grad_norm": 0.7648500642290317, + "learning_rate": 0.003, + "loss": 4.2443, + "step": 1434 + }, + { + "epoch": 0.01435, + "grad_norm": 0.7689156199684593, + "learning_rate": 0.003, + "loss": 4.2531, + "step": 1435 + }, + { + "epoch": 0.01436, + "grad_norm": 0.7360234649948437, + "learning_rate": 0.003, + "loss": 4.274, + "step": 1436 + }, + { + "epoch": 0.01437, + "grad_norm": 0.7697472519317102, + "learning_rate": 0.003, + "loss": 4.2879, + "step": 1437 + }, + { + "epoch": 0.01438, + "grad_norm": 0.7889422694170278, + "learning_rate": 0.003, + "loss": 4.2805, + "step": 1438 + }, + { + "epoch": 0.01439, + "grad_norm": 0.6891331546216416, + "learning_rate": 0.003, + "loss": 4.2417, + "step": 1439 + }, + { + "epoch": 0.0144, + "grad_norm": 0.560460778999846, + "learning_rate": 0.003, + "loss": 4.2508, + "step": 1440 + }, + { + "epoch": 0.01441, + "grad_norm": 0.5400270352929513, + "learning_rate": 0.003, + "loss": 4.2578, + "step": 1441 + }, + { + "epoch": 0.01442, + "grad_norm": 0.4911140953080599, + "learning_rate": 0.003, + "loss": 4.2391, + "step": 1442 + }, + { + "epoch": 0.01443, + "grad_norm": 0.45294215274078276, + "learning_rate": 0.003, + "loss": 4.2993, + "step": 1443 + }, + { + "epoch": 0.01444, + "grad_norm": 0.4224101633170518, + "learning_rate": 0.003, + "loss": 4.2618, + "step": 1444 + }, + { + "epoch": 0.01445, + "grad_norm": 0.3961185894381269, + "learning_rate": 0.003, + "loss": 4.2206, + "step": 1445 + }, + { + "epoch": 0.01446, + "grad_norm": 0.33118667900370186, + "learning_rate": 0.003, + "loss": 4.2343, + "step": 1446 + }, + { + "epoch": 0.01447, + "grad_norm": 0.3556733761768107, + "learning_rate": 0.003, + "loss": 4.2433, + "step": 1447 + }, + { + "epoch": 0.01448, + "grad_norm": 0.36561455130036685, + "learning_rate": 0.003, + "loss": 4.2519, + "step": 1448 + }, + { + "epoch": 0.01449, + "grad_norm": 0.46874354720560246, + "learning_rate": 0.003, + "loss": 4.2461, + "step": 1449 + }, + { + "epoch": 0.0145, + "grad_norm": 0.60149549500153, + "learning_rate": 0.003, + "loss": 4.266, + "step": 1450 + }, + { + "epoch": 0.01451, + "grad_norm": 0.6696836890558183, + "learning_rate": 0.003, + "loss": 4.2469, + "step": 1451 + }, + { + "epoch": 0.01452, + "grad_norm": 0.6515350232296636, + "learning_rate": 0.003, + "loss": 4.2476, + "step": 1452 + }, + { + "epoch": 0.01453, + "grad_norm": 0.662277322649583, + "learning_rate": 0.003, + "loss": 4.2608, + "step": 1453 + }, + { + "epoch": 0.01454, + "grad_norm": 0.8972969416580211, + "learning_rate": 0.003, + "loss": 4.2374, + "step": 1454 + }, + { + "epoch": 0.01455, + "grad_norm": 1.030023068993243, + "learning_rate": 0.003, + "loss": 4.2974, + "step": 1455 + }, + { + "epoch": 0.01456, + "grad_norm": 1.0186167234807815, + "learning_rate": 0.003, + "loss": 4.2685, + "step": 1456 + }, + { + "epoch": 0.01457, + "grad_norm": 0.9067819891416041, + "learning_rate": 0.003, + "loss": 4.2556, + "step": 1457 + }, + { + "epoch": 0.01458, + "grad_norm": 0.8145250303169672, + "learning_rate": 0.003, + "loss": 4.2507, + "step": 1458 + }, + { + "epoch": 0.01459, + "grad_norm": 0.8089455723824179, + "learning_rate": 0.003, + "loss": 4.2807, + "step": 1459 + }, + { + "epoch": 0.0146, + "grad_norm": 0.7818105608289873, + "learning_rate": 0.003, + "loss": 4.2825, + "step": 1460 + }, + { + "epoch": 0.01461, + "grad_norm": 0.8037592863895617, + "learning_rate": 0.003, + "loss": 4.2628, + "step": 1461 + }, + { + "epoch": 0.01462, + "grad_norm": 0.7476671417509403, + "learning_rate": 0.003, + "loss": 4.2637, + "step": 1462 + }, + { + "epoch": 0.01463, + "grad_norm": 0.7531868665794315, + "learning_rate": 0.003, + "loss": 4.2649, + "step": 1463 + }, + { + "epoch": 0.01464, + "grad_norm": 0.7893765528220285, + "learning_rate": 0.003, + "loss": 4.2417, + "step": 1464 + }, + { + "epoch": 0.01465, + "grad_norm": 0.6572421930862311, + "learning_rate": 0.003, + "loss": 4.2823, + "step": 1465 + }, + { + "epoch": 0.01466, + "grad_norm": 0.5506048327632588, + "learning_rate": 0.003, + "loss": 4.2484, + "step": 1466 + }, + { + "epoch": 0.01467, + "grad_norm": 0.611053596724818, + "learning_rate": 0.003, + "loss": 4.2828, + "step": 1467 + }, + { + "epoch": 0.01468, + "grad_norm": 0.68776369659849, + "learning_rate": 0.003, + "loss": 4.2597, + "step": 1468 + }, + { + "epoch": 0.01469, + "grad_norm": 0.6702695663961474, + "learning_rate": 0.003, + "loss": 4.2631, + "step": 1469 + }, + { + "epoch": 0.0147, + "grad_norm": 0.6961169309212991, + "learning_rate": 0.003, + "loss": 4.2544, + "step": 1470 + }, + { + "epoch": 0.01471, + "grad_norm": 0.760155282734985, + "learning_rate": 0.003, + "loss": 4.2691, + "step": 1471 + }, + { + "epoch": 0.01472, + "grad_norm": 0.7721176095708484, + "learning_rate": 0.003, + "loss": 4.2542, + "step": 1472 + }, + { + "epoch": 0.01473, + "grad_norm": 0.7877033148700755, + "learning_rate": 0.003, + "loss": 4.2676, + "step": 1473 + }, + { + "epoch": 0.01474, + "grad_norm": 0.7464730946193943, + "learning_rate": 0.003, + "loss": 4.2539, + "step": 1474 + }, + { + "epoch": 0.01475, + "grad_norm": 0.6862863699060794, + "learning_rate": 0.003, + "loss": 4.2711, + "step": 1475 + }, + { + "epoch": 0.01476, + "grad_norm": 0.6949444506676095, + "learning_rate": 0.003, + "loss": 4.2569, + "step": 1476 + }, + { + "epoch": 0.01477, + "grad_norm": 0.6485108315252704, + "learning_rate": 0.003, + "loss": 4.2593, + "step": 1477 + }, + { + "epoch": 0.01478, + "grad_norm": 0.7249395862424638, + "learning_rate": 0.003, + "loss": 4.2474, + "step": 1478 + }, + { + "epoch": 0.01479, + "grad_norm": 0.6746802016444003, + "learning_rate": 0.003, + "loss": 4.28, + "step": 1479 + }, + { + "epoch": 0.0148, + "grad_norm": 0.6561363565990774, + "learning_rate": 0.003, + "loss": 4.2473, + "step": 1480 + }, + { + "epoch": 0.01481, + "grad_norm": 0.7984072714276804, + "learning_rate": 0.003, + "loss": 4.2564, + "step": 1481 + }, + { + "epoch": 0.01482, + "grad_norm": 0.8535873443126762, + "learning_rate": 0.003, + "loss": 4.2436, + "step": 1482 + }, + { + "epoch": 0.01483, + "grad_norm": 0.8309335851279978, + "learning_rate": 0.003, + "loss": 4.2469, + "step": 1483 + }, + { + "epoch": 0.01484, + "grad_norm": 0.7153917621202344, + "learning_rate": 0.003, + "loss": 4.2896, + "step": 1484 + }, + { + "epoch": 0.01485, + "grad_norm": 0.7644465422516641, + "learning_rate": 0.003, + "loss": 4.2586, + "step": 1485 + }, + { + "epoch": 0.01486, + "grad_norm": 0.6575547181769861, + "learning_rate": 0.003, + "loss": 4.2505, + "step": 1486 + }, + { + "epoch": 0.01487, + "grad_norm": 0.616480892995353, + "learning_rate": 0.003, + "loss": 4.2525, + "step": 1487 + }, + { + "epoch": 0.01488, + "grad_norm": 0.5676164891655556, + "learning_rate": 0.003, + "loss": 4.259, + "step": 1488 + }, + { + "epoch": 0.01489, + "grad_norm": 0.5087336805317257, + "learning_rate": 0.003, + "loss": 4.252, + "step": 1489 + }, + { + "epoch": 0.0149, + "grad_norm": 0.5711204937411694, + "learning_rate": 0.003, + "loss": 4.2449, + "step": 1490 + }, + { + "epoch": 0.01491, + "grad_norm": 0.5555383511180694, + "learning_rate": 0.003, + "loss": 4.2594, + "step": 1491 + }, + { + "epoch": 0.01492, + "grad_norm": 0.5910867309514242, + "learning_rate": 0.003, + "loss": 4.2249, + "step": 1492 + }, + { + "epoch": 0.01493, + "grad_norm": 0.5552788367316016, + "learning_rate": 0.003, + "loss": 4.2566, + "step": 1493 + }, + { + "epoch": 0.01494, + "grad_norm": 0.597504731138664, + "learning_rate": 0.003, + "loss": 4.236, + "step": 1494 + }, + { + "epoch": 0.01495, + "grad_norm": 0.7040905961754581, + "learning_rate": 0.003, + "loss": 4.2494, + "step": 1495 + }, + { + "epoch": 0.01496, + "grad_norm": 0.6504979761973873, + "learning_rate": 0.003, + "loss": 4.2253, + "step": 1496 + }, + { + "epoch": 0.01497, + "grad_norm": 0.7479904857429174, + "learning_rate": 0.003, + "loss": 4.2013, + "step": 1497 + }, + { + "epoch": 0.01498, + "grad_norm": 0.8354013055859276, + "learning_rate": 0.003, + "loss": 4.2649, + "step": 1498 + }, + { + "epoch": 0.01499, + "grad_norm": 0.9228916776476623, + "learning_rate": 0.003, + "loss": 4.2868, + "step": 1499 + }, + { + "epoch": 0.015, + "grad_norm": 0.9718172707744857, + "learning_rate": 0.003, + "loss": 4.25, + "step": 1500 + }, + { + "epoch": 0.01501, + "grad_norm": 1.0136419160882046, + "learning_rate": 0.003, + "loss": 4.2532, + "step": 1501 + }, + { + "epoch": 0.01502, + "grad_norm": 0.9325478476141171, + "learning_rate": 0.003, + "loss": 4.2999, + "step": 1502 + }, + { + "epoch": 0.01503, + "grad_norm": 1.045999048375591, + "learning_rate": 0.003, + "loss": 4.2898, + "step": 1503 + }, + { + "epoch": 0.01504, + "grad_norm": 0.9914485824003936, + "learning_rate": 0.003, + "loss": 4.2936, + "step": 1504 + }, + { + "epoch": 0.01505, + "grad_norm": 0.915373742082242, + "learning_rate": 0.003, + "loss": 4.2776, + "step": 1505 + }, + { + "epoch": 0.01506, + "grad_norm": 0.8272684821900249, + "learning_rate": 0.003, + "loss": 4.287, + "step": 1506 + }, + { + "epoch": 0.01507, + "grad_norm": 0.8618015345967726, + "learning_rate": 0.003, + "loss": 4.305, + "step": 1507 + }, + { + "epoch": 0.01508, + "grad_norm": 0.9316744896757736, + "learning_rate": 0.003, + "loss": 4.2861, + "step": 1508 + }, + { + "epoch": 0.01509, + "grad_norm": 0.8440466174421261, + "learning_rate": 0.003, + "loss": 4.2981, + "step": 1509 + }, + { + "epoch": 0.0151, + "grad_norm": 0.7063003162706939, + "learning_rate": 0.003, + "loss": 4.2679, + "step": 1510 + }, + { + "epoch": 0.01511, + "grad_norm": 0.6428638550974116, + "learning_rate": 0.003, + "loss": 4.2518, + "step": 1511 + }, + { + "epoch": 0.01512, + "grad_norm": 0.6112249954297037, + "learning_rate": 0.003, + "loss": 4.2613, + "step": 1512 + }, + { + "epoch": 0.01513, + "grad_norm": 0.5280577640396528, + "learning_rate": 0.003, + "loss": 4.2622, + "step": 1513 + }, + { + "epoch": 0.01514, + "grad_norm": 0.5269385985194992, + "learning_rate": 0.003, + "loss": 4.2672, + "step": 1514 + }, + { + "epoch": 0.01515, + "grad_norm": 0.5487987330042349, + "learning_rate": 0.003, + "loss": 4.2562, + "step": 1515 + }, + { + "epoch": 0.01516, + "grad_norm": 0.5680391134156543, + "learning_rate": 0.003, + "loss": 4.2622, + "step": 1516 + }, + { + "epoch": 0.01517, + "grad_norm": 0.6062107860159427, + "learning_rate": 0.003, + "loss": 4.2129, + "step": 1517 + }, + { + "epoch": 0.01518, + "grad_norm": 0.6323154755891632, + "learning_rate": 0.003, + "loss": 4.2497, + "step": 1518 + }, + { + "epoch": 0.01519, + "grad_norm": 0.5798448883069626, + "learning_rate": 0.003, + "loss": 4.2808, + "step": 1519 + }, + { + "epoch": 0.0152, + "grad_norm": 0.5997731039690386, + "learning_rate": 0.003, + "loss": 4.2326, + "step": 1520 + }, + { + "epoch": 0.01521, + "grad_norm": 0.6222254423875409, + "learning_rate": 0.003, + "loss": 4.248, + "step": 1521 + }, + { + "epoch": 0.01522, + "grad_norm": 0.5290334858261236, + "learning_rate": 0.003, + "loss": 4.2398, + "step": 1522 + }, + { + "epoch": 0.01523, + "grad_norm": 0.5493394171973806, + "learning_rate": 0.003, + "loss": 4.2321, + "step": 1523 + }, + { + "epoch": 0.01524, + "grad_norm": 0.6415152277388406, + "learning_rate": 0.003, + "loss": 4.269, + "step": 1524 + }, + { + "epoch": 0.01525, + "grad_norm": 0.7027827268185114, + "learning_rate": 0.003, + "loss": 4.2694, + "step": 1525 + }, + { + "epoch": 0.01526, + "grad_norm": 0.6962686558421027, + "learning_rate": 0.003, + "loss": 4.259, + "step": 1526 + }, + { + "epoch": 0.01527, + "grad_norm": 0.6968664170247966, + "learning_rate": 0.003, + "loss": 4.2293, + "step": 1527 + }, + { + "epoch": 0.01528, + "grad_norm": 0.6258396233188714, + "learning_rate": 0.003, + "loss": 4.2456, + "step": 1528 + }, + { + "epoch": 0.01529, + "grad_norm": 0.6831052588682119, + "learning_rate": 0.003, + "loss": 4.2426, + "step": 1529 + }, + { + "epoch": 0.0153, + "grad_norm": 0.5524942784192245, + "learning_rate": 0.003, + "loss": 4.2472, + "step": 1530 + }, + { + "epoch": 0.01531, + "grad_norm": 0.5976116227060053, + "learning_rate": 0.003, + "loss": 4.241, + "step": 1531 + }, + { + "epoch": 0.01532, + "grad_norm": 0.5212214454213059, + "learning_rate": 0.003, + "loss": 4.2166, + "step": 1532 + }, + { + "epoch": 0.01533, + "grad_norm": 0.5878087168177306, + "learning_rate": 0.003, + "loss": 4.2439, + "step": 1533 + }, + { + "epoch": 0.01534, + "grad_norm": 0.8062347592746759, + "learning_rate": 0.003, + "loss": 4.2507, + "step": 1534 + }, + { + "epoch": 0.01535, + "grad_norm": 1.1825381040517529, + "learning_rate": 0.003, + "loss": 4.2824, + "step": 1535 + }, + { + "epoch": 0.01536, + "grad_norm": 0.7819176506012738, + "learning_rate": 0.003, + "loss": 4.2554, + "step": 1536 + }, + { + "epoch": 0.01537, + "grad_norm": 0.7933291638944598, + "learning_rate": 0.003, + "loss": 4.2465, + "step": 1537 + }, + { + "epoch": 0.01538, + "grad_norm": 0.7224294373390021, + "learning_rate": 0.003, + "loss": 4.2237, + "step": 1538 + }, + { + "epoch": 0.01539, + "grad_norm": 0.7725279995753968, + "learning_rate": 0.003, + "loss": 4.2589, + "step": 1539 + }, + { + "epoch": 0.0154, + "grad_norm": 0.6639445396789071, + "learning_rate": 0.003, + "loss": 4.2578, + "step": 1540 + }, + { + "epoch": 0.01541, + "grad_norm": 0.5655107354833333, + "learning_rate": 0.003, + "loss": 4.2359, + "step": 1541 + }, + { + "epoch": 0.01542, + "grad_norm": 0.4939945119276493, + "learning_rate": 0.003, + "loss": 4.2606, + "step": 1542 + }, + { + "epoch": 0.01543, + "grad_norm": 0.6098938664838743, + "learning_rate": 0.003, + "loss": 4.2469, + "step": 1543 + }, + { + "epoch": 0.01544, + "grad_norm": 0.58488864006147, + "learning_rate": 0.003, + "loss": 4.2344, + "step": 1544 + }, + { + "epoch": 0.01545, + "grad_norm": 0.5313644010442865, + "learning_rate": 0.003, + "loss": 4.2981, + "step": 1545 + }, + { + "epoch": 0.01546, + "grad_norm": 0.45959833989927285, + "learning_rate": 0.003, + "loss": 4.263, + "step": 1546 + }, + { + "epoch": 0.01547, + "grad_norm": 0.41872682800093836, + "learning_rate": 0.003, + "loss": 4.2626, + "step": 1547 + }, + { + "epoch": 0.01548, + "grad_norm": 0.4070005827014792, + "learning_rate": 0.003, + "loss": 4.2237, + "step": 1548 + }, + { + "epoch": 0.01549, + "grad_norm": 0.40332574830782875, + "learning_rate": 0.003, + "loss": 4.2125, + "step": 1549 + }, + { + "epoch": 0.0155, + "grad_norm": 0.4273425707216713, + "learning_rate": 0.003, + "loss": 4.2323, + "step": 1550 + }, + { + "epoch": 0.01551, + "grad_norm": 0.44340108675114914, + "learning_rate": 0.003, + "loss": 4.2095, + "step": 1551 + }, + { + "epoch": 0.01552, + "grad_norm": 0.4283402844721516, + "learning_rate": 0.003, + "loss": 4.2297, + "step": 1552 + }, + { + "epoch": 0.01553, + "grad_norm": 0.39909368113710497, + "learning_rate": 0.003, + "loss": 4.2548, + "step": 1553 + }, + { + "epoch": 0.01554, + "grad_norm": 0.4303761621068326, + "learning_rate": 0.003, + "loss": 4.2356, + "step": 1554 + }, + { + "epoch": 0.01555, + "grad_norm": 0.5073116437720543, + "learning_rate": 0.003, + "loss": 4.2089, + "step": 1555 + }, + { + "epoch": 0.01556, + "grad_norm": 0.6171177000269511, + "learning_rate": 0.003, + "loss": 4.2523, + "step": 1556 + }, + { + "epoch": 0.01557, + "grad_norm": 0.678464738142453, + "learning_rate": 0.003, + "loss": 4.229, + "step": 1557 + }, + { + "epoch": 0.01558, + "grad_norm": 0.6779041786729552, + "learning_rate": 0.003, + "loss": 4.22, + "step": 1558 + }, + { + "epoch": 0.01559, + "grad_norm": 0.7160170404134826, + "learning_rate": 0.003, + "loss": 4.2601, + "step": 1559 + }, + { + "epoch": 0.0156, + "grad_norm": 0.646776348851303, + "learning_rate": 0.003, + "loss": 4.2484, + "step": 1560 + }, + { + "epoch": 0.01561, + "grad_norm": 0.7362963973677504, + "learning_rate": 0.003, + "loss": 4.232, + "step": 1561 + }, + { + "epoch": 0.01562, + "grad_norm": 0.8173738140090338, + "learning_rate": 0.003, + "loss": 4.2318, + "step": 1562 + }, + { + "epoch": 0.01563, + "grad_norm": 0.6949580272753917, + "learning_rate": 0.003, + "loss": 4.2334, + "step": 1563 + }, + { + "epoch": 0.01564, + "grad_norm": 0.8704354124122435, + "learning_rate": 0.003, + "loss": 4.2705, + "step": 1564 + }, + { + "epoch": 0.01565, + "grad_norm": 1.0516715274291202, + "learning_rate": 0.003, + "loss": 4.2561, + "step": 1565 + }, + { + "epoch": 0.01566, + "grad_norm": 0.9138795915356578, + "learning_rate": 0.003, + "loss": 4.2729, + "step": 1566 + }, + { + "epoch": 0.01567, + "grad_norm": 0.7933215729025597, + "learning_rate": 0.003, + "loss": 4.2643, + "step": 1567 + }, + { + "epoch": 0.01568, + "grad_norm": 0.7524451299585192, + "learning_rate": 0.003, + "loss": 4.2455, + "step": 1568 + }, + { + "epoch": 0.01569, + "grad_norm": 0.8688441972266158, + "learning_rate": 0.003, + "loss": 4.2557, + "step": 1569 + }, + { + "epoch": 0.0157, + "grad_norm": 1.0213964087576262, + "learning_rate": 0.003, + "loss": 4.2839, + "step": 1570 + }, + { + "epoch": 0.01571, + "grad_norm": 0.9658714210834742, + "learning_rate": 0.003, + "loss": 4.2787, + "step": 1571 + }, + { + "epoch": 0.01572, + "grad_norm": 0.8962942943425795, + "learning_rate": 0.003, + "loss": 4.28, + "step": 1572 + }, + { + "epoch": 0.01573, + "grad_norm": 0.936508363757706, + "learning_rate": 0.003, + "loss": 4.2727, + "step": 1573 + }, + { + "epoch": 0.01574, + "grad_norm": 0.8583102869770243, + "learning_rate": 0.003, + "loss": 4.279, + "step": 1574 + }, + { + "epoch": 0.01575, + "grad_norm": 0.8543483448302793, + "learning_rate": 0.003, + "loss": 4.2483, + "step": 1575 + }, + { + "epoch": 0.01576, + "grad_norm": 0.6740643949431316, + "learning_rate": 0.003, + "loss": 4.2592, + "step": 1576 + }, + { + "epoch": 0.01577, + "grad_norm": 0.7068574162099283, + "learning_rate": 0.003, + "loss": 4.2234, + "step": 1577 + }, + { + "epoch": 0.01578, + "grad_norm": 0.679751091133263, + "learning_rate": 0.003, + "loss": 4.2501, + "step": 1578 + }, + { + "epoch": 0.01579, + "grad_norm": 0.6207998089836335, + "learning_rate": 0.003, + "loss": 4.2428, + "step": 1579 + }, + { + "epoch": 0.0158, + "grad_norm": 0.546999417238052, + "learning_rate": 0.003, + "loss": 4.2499, + "step": 1580 + }, + { + "epoch": 0.01581, + "grad_norm": 0.5890586432360596, + "learning_rate": 0.003, + "loss": 4.2581, + "step": 1581 + }, + { + "epoch": 0.01582, + "grad_norm": 0.6738822960276017, + "learning_rate": 0.003, + "loss": 4.2412, + "step": 1582 + }, + { + "epoch": 0.01583, + "grad_norm": 0.8449645723590827, + "learning_rate": 0.003, + "loss": 4.235, + "step": 1583 + }, + { + "epoch": 0.01584, + "grad_norm": 0.8760719049873085, + "learning_rate": 0.003, + "loss": 4.2744, + "step": 1584 + }, + { + "epoch": 0.01585, + "grad_norm": 0.8749576730978259, + "learning_rate": 0.003, + "loss": 4.2653, + "step": 1585 + }, + { + "epoch": 0.01586, + "grad_norm": 0.9817764014825603, + "learning_rate": 0.003, + "loss": 4.2423, + "step": 1586 + }, + { + "epoch": 0.01587, + "grad_norm": 0.916565337571196, + "learning_rate": 0.003, + "loss": 4.2657, + "step": 1587 + }, + { + "epoch": 0.01588, + "grad_norm": 0.5958323489289675, + "learning_rate": 0.003, + "loss": 4.2761, + "step": 1588 + }, + { + "epoch": 0.01589, + "grad_norm": 0.6040948221660047, + "learning_rate": 0.003, + "loss": 4.2694, + "step": 1589 + }, + { + "epoch": 0.0159, + "grad_norm": 0.5541532228259608, + "learning_rate": 0.003, + "loss": 4.2601, + "step": 1590 + }, + { + "epoch": 0.01591, + "grad_norm": 0.5732941807574177, + "learning_rate": 0.003, + "loss": 4.2648, + "step": 1591 + }, + { + "epoch": 0.01592, + "grad_norm": 0.6004965962049119, + "learning_rate": 0.003, + "loss": 4.2552, + "step": 1592 + }, + { + "epoch": 0.01593, + "grad_norm": 0.6750400712704605, + "learning_rate": 0.003, + "loss": 4.2608, + "step": 1593 + }, + { + "epoch": 0.01594, + "grad_norm": 0.6685590290117844, + "learning_rate": 0.003, + "loss": 4.2444, + "step": 1594 + }, + { + "epoch": 0.01595, + "grad_norm": 0.6394603041102755, + "learning_rate": 0.003, + "loss": 4.27, + "step": 1595 + }, + { + "epoch": 0.01596, + "grad_norm": 0.5525501078218852, + "learning_rate": 0.003, + "loss": 4.2582, + "step": 1596 + }, + { + "epoch": 0.01597, + "grad_norm": 0.5429531891975062, + "learning_rate": 0.003, + "loss": 4.2192, + "step": 1597 + }, + { + "epoch": 0.01598, + "grad_norm": 0.5494281969618091, + "learning_rate": 0.003, + "loss": 4.2123, + "step": 1598 + }, + { + "epoch": 0.01599, + "grad_norm": 0.48563975518620667, + "learning_rate": 0.003, + "loss": 4.2331, + "step": 1599 + }, + { + "epoch": 0.016, + "grad_norm": 0.4293130690644179, + "learning_rate": 0.003, + "loss": 4.1975, + "step": 1600 + }, + { + "epoch": 0.01601, + "grad_norm": 0.4385393368882689, + "learning_rate": 0.003, + "loss": 4.2457, + "step": 1601 + }, + { + "epoch": 0.01602, + "grad_norm": 0.6149074026063737, + "learning_rate": 0.003, + "loss": 4.2243, + "step": 1602 + }, + { + "epoch": 0.01603, + "grad_norm": 0.7618943354871598, + "learning_rate": 0.003, + "loss": 4.2459, + "step": 1603 + }, + { + "epoch": 0.01604, + "grad_norm": 0.8231087133002348, + "learning_rate": 0.003, + "loss": 4.2421, + "step": 1604 + }, + { + "epoch": 0.01605, + "grad_norm": 0.792939816240345, + "learning_rate": 0.003, + "loss": 4.2225, + "step": 1605 + }, + { + "epoch": 0.01606, + "grad_norm": 0.8283386222980521, + "learning_rate": 0.003, + "loss": 4.2264, + "step": 1606 + }, + { + "epoch": 0.01607, + "grad_norm": 0.7929167281458711, + "learning_rate": 0.003, + "loss": 4.2525, + "step": 1607 + }, + { + "epoch": 0.01608, + "grad_norm": 0.8238114241929788, + "learning_rate": 0.003, + "loss": 4.2293, + "step": 1608 + }, + { + "epoch": 0.01609, + "grad_norm": 0.8482072969488592, + "learning_rate": 0.003, + "loss": 4.2448, + "step": 1609 + }, + { + "epoch": 0.0161, + "grad_norm": 0.9731734226169589, + "learning_rate": 0.003, + "loss": 4.2157, + "step": 1610 + }, + { + "epoch": 0.01611, + "grad_norm": 0.8477820620248173, + "learning_rate": 0.003, + "loss": 4.2328, + "step": 1611 + }, + { + "epoch": 0.01612, + "grad_norm": 0.8003056368656658, + "learning_rate": 0.003, + "loss": 4.2224, + "step": 1612 + }, + { + "epoch": 0.01613, + "grad_norm": 0.9384869181655229, + "learning_rate": 0.003, + "loss": 4.2759, + "step": 1613 + }, + { + "epoch": 0.01614, + "grad_norm": 1.094095683053054, + "learning_rate": 0.003, + "loss": 4.305, + "step": 1614 + }, + { + "epoch": 0.01615, + "grad_norm": 1.0454054633439653, + "learning_rate": 0.003, + "loss": 4.284, + "step": 1615 + }, + { + "epoch": 0.01616, + "grad_norm": 0.728394920780058, + "learning_rate": 0.003, + "loss": 4.2459, + "step": 1616 + }, + { + "epoch": 0.01617, + "grad_norm": 0.7455143593181587, + "learning_rate": 0.003, + "loss": 4.2373, + "step": 1617 + }, + { + "epoch": 0.01618, + "grad_norm": 0.6872936552392757, + "learning_rate": 0.003, + "loss": 4.2451, + "step": 1618 + }, + { + "epoch": 0.01619, + "grad_norm": 0.5634358260711833, + "learning_rate": 0.003, + "loss": 4.2274, + "step": 1619 + }, + { + "epoch": 0.0162, + "grad_norm": 0.5499920921999998, + "learning_rate": 0.003, + "loss": 4.231, + "step": 1620 + }, + { + "epoch": 0.01621, + "grad_norm": 0.4948415173461953, + "learning_rate": 0.003, + "loss": 4.226, + "step": 1621 + }, + { + "epoch": 0.01622, + "grad_norm": 0.45485020618430855, + "learning_rate": 0.003, + "loss": 4.2573, + "step": 1622 + }, + { + "epoch": 0.01623, + "grad_norm": 0.4128235375978694, + "learning_rate": 0.003, + "loss": 4.2306, + "step": 1623 + }, + { + "epoch": 0.01624, + "grad_norm": 0.35723113909525034, + "learning_rate": 0.003, + "loss": 4.2422, + "step": 1624 + }, + { + "epoch": 0.01625, + "grad_norm": 0.3930391853636471, + "learning_rate": 0.003, + "loss": 4.2175, + "step": 1625 + }, + { + "epoch": 0.01626, + "grad_norm": 0.3980902109371374, + "learning_rate": 0.003, + "loss": 4.2372, + "step": 1626 + }, + { + "epoch": 0.01627, + "grad_norm": 0.4192958733549867, + "learning_rate": 0.003, + "loss": 4.2057, + "step": 1627 + }, + { + "epoch": 0.01628, + "grad_norm": 0.4456736368427327, + "learning_rate": 0.003, + "loss": 4.2132, + "step": 1628 + }, + { + "epoch": 0.01629, + "grad_norm": 0.46084356058160403, + "learning_rate": 0.003, + "loss": 4.2138, + "step": 1629 + }, + { + "epoch": 0.0163, + "grad_norm": 0.4494988261781255, + "learning_rate": 0.003, + "loss": 4.2025, + "step": 1630 + }, + { + "epoch": 0.01631, + "grad_norm": 0.49049706860110326, + "learning_rate": 0.003, + "loss": 4.2248, + "step": 1631 + }, + { + "epoch": 0.01632, + "grad_norm": 0.4941434115670244, + "learning_rate": 0.003, + "loss": 4.234, + "step": 1632 + }, + { + "epoch": 0.01633, + "grad_norm": 0.5015101022129131, + "learning_rate": 0.003, + "loss": 4.2072, + "step": 1633 + }, + { + "epoch": 0.01634, + "grad_norm": 0.5355988850166129, + "learning_rate": 0.003, + "loss": 4.2264, + "step": 1634 + }, + { + "epoch": 0.01635, + "grad_norm": 0.6114378514764486, + "learning_rate": 0.003, + "loss": 4.2077, + "step": 1635 + }, + { + "epoch": 0.01636, + "grad_norm": 0.659185840591732, + "learning_rate": 0.003, + "loss": 4.2261, + "step": 1636 + }, + { + "epoch": 0.01637, + "grad_norm": 0.6890891308278232, + "learning_rate": 0.003, + "loss": 4.2122, + "step": 1637 + }, + { + "epoch": 0.01638, + "grad_norm": 0.7917384547475219, + "learning_rate": 0.003, + "loss": 4.2254, + "step": 1638 + }, + { + "epoch": 0.01639, + "grad_norm": 0.8310494258877913, + "learning_rate": 0.003, + "loss": 4.2496, + "step": 1639 + }, + { + "epoch": 0.0164, + "grad_norm": 0.6945218446546085, + "learning_rate": 0.003, + "loss": 4.2286, + "step": 1640 + }, + { + "epoch": 0.01641, + "grad_norm": 0.643774136427719, + "learning_rate": 0.003, + "loss": 4.2028, + "step": 1641 + }, + { + "epoch": 0.01642, + "grad_norm": 0.5630395282689907, + "learning_rate": 0.003, + "loss": 4.2418, + "step": 1642 + }, + { + "epoch": 0.01643, + "grad_norm": 0.6319019547985867, + "learning_rate": 0.003, + "loss": 4.23, + "step": 1643 + }, + { + "epoch": 0.01644, + "grad_norm": 0.6923812193124619, + "learning_rate": 0.003, + "loss": 4.2497, + "step": 1644 + }, + { + "epoch": 0.01645, + "grad_norm": 0.7835455696779042, + "learning_rate": 0.003, + "loss": 4.2006, + "step": 1645 + }, + { + "epoch": 0.01646, + "grad_norm": 0.9362858528690088, + "learning_rate": 0.003, + "loss": 4.2431, + "step": 1646 + }, + { + "epoch": 0.01647, + "grad_norm": 1.2289381301321516, + "learning_rate": 0.003, + "loss": 4.255, + "step": 1647 + }, + { + "epoch": 0.01648, + "grad_norm": 0.8093728261853784, + "learning_rate": 0.003, + "loss": 4.2206, + "step": 1648 + }, + { + "epoch": 0.01649, + "grad_norm": 0.7780860637025212, + "learning_rate": 0.003, + "loss": 4.2535, + "step": 1649 + }, + { + "epoch": 0.0165, + "grad_norm": 0.8631062415473619, + "learning_rate": 0.003, + "loss": 4.2375, + "step": 1650 + }, + { + "epoch": 0.01651, + "grad_norm": 0.9142840285195195, + "learning_rate": 0.003, + "loss": 4.2523, + "step": 1651 + }, + { + "epoch": 0.01652, + "grad_norm": 0.8336324309893678, + "learning_rate": 0.003, + "loss": 4.2094, + "step": 1652 + }, + { + "epoch": 0.01653, + "grad_norm": 0.856140708339795, + "learning_rate": 0.003, + "loss": 4.2552, + "step": 1653 + }, + { + "epoch": 0.01654, + "grad_norm": 0.9035014747718915, + "learning_rate": 0.003, + "loss": 4.2428, + "step": 1654 + }, + { + "epoch": 0.01655, + "grad_norm": 0.8007220508865799, + "learning_rate": 0.003, + "loss": 4.2521, + "step": 1655 + }, + { + "epoch": 0.01656, + "grad_norm": 0.8140605513593369, + "learning_rate": 0.003, + "loss": 4.2202, + "step": 1656 + }, + { + "epoch": 0.01657, + "grad_norm": 0.752017726600531, + "learning_rate": 0.003, + "loss": 4.2402, + "step": 1657 + }, + { + "epoch": 0.01658, + "grad_norm": 0.8846697411940587, + "learning_rate": 0.003, + "loss": 4.2338, + "step": 1658 + }, + { + "epoch": 0.01659, + "grad_norm": 0.8979228683366066, + "learning_rate": 0.003, + "loss": 4.2417, + "step": 1659 + }, + { + "epoch": 0.0166, + "grad_norm": 0.73714542235099, + "learning_rate": 0.003, + "loss": 4.2224, + "step": 1660 + }, + { + "epoch": 0.01661, + "grad_norm": 0.6566854277166256, + "learning_rate": 0.003, + "loss": 4.2212, + "step": 1661 + }, + { + "epoch": 0.01662, + "grad_norm": 0.5217494005958339, + "learning_rate": 0.003, + "loss": 4.2208, + "step": 1662 + }, + { + "epoch": 0.01663, + "grad_norm": 0.4794160667989573, + "learning_rate": 0.003, + "loss": 4.2395, + "step": 1663 + }, + { + "epoch": 0.01664, + "grad_norm": 0.5082285006914213, + "learning_rate": 0.003, + "loss": 4.2563, + "step": 1664 + }, + { + "epoch": 0.01665, + "grad_norm": 0.5164271930929455, + "learning_rate": 0.003, + "loss": 4.2188, + "step": 1665 + }, + { + "epoch": 0.01666, + "grad_norm": 0.6299180885641164, + "learning_rate": 0.003, + "loss": 4.2234, + "step": 1666 + }, + { + "epoch": 0.01667, + "grad_norm": 0.6409430167727818, + "learning_rate": 0.003, + "loss": 4.2382, + "step": 1667 + }, + { + "epoch": 0.01668, + "grad_norm": 0.6129313286956495, + "learning_rate": 0.003, + "loss": 4.2415, + "step": 1668 + }, + { + "epoch": 0.01669, + "grad_norm": 0.5586023730002122, + "learning_rate": 0.003, + "loss": 4.223, + "step": 1669 + }, + { + "epoch": 0.0167, + "grad_norm": 0.6736111064245857, + "learning_rate": 0.003, + "loss": 4.2169, + "step": 1670 + }, + { + "epoch": 0.01671, + "grad_norm": 0.7441939131840011, + "learning_rate": 0.003, + "loss": 4.2115, + "step": 1671 + }, + { + "epoch": 0.01672, + "grad_norm": 0.8159394752701435, + "learning_rate": 0.003, + "loss": 4.2214, + "step": 1672 + }, + { + "epoch": 0.01673, + "grad_norm": 0.7952390142878619, + "learning_rate": 0.003, + "loss": 4.2239, + "step": 1673 + }, + { + "epoch": 0.01674, + "grad_norm": 0.7199816431285693, + "learning_rate": 0.003, + "loss": 4.2398, + "step": 1674 + }, + { + "epoch": 0.01675, + "grad_norm": 0.7812011305713198, + "learning_rate": 0.003, + "loss": 4.2244, + "step": 1675 + }, + { + "epoch": 0.01676, + "grad_norm": 0.7469280078216791, + "learning_rate": 0.003, + "loss": 4.2309, + "step": 1676 + }, + { + "epoch": 0.01677, + "grad_norm": 0.7271162906259458, + "learning_rate": 0.003, + "loss": 4.2196, + "step": 1677 + }, + { + "epoch": 0.01678, + "grad_norm": 0.8881037129933885, + "learning_rate": 0.003, + "loss": 4.2406, + "step": 1678 + }, + { + "epoch": 0.01679, + "grad_norm": 0.7948545784278963, + "learning_rate": 0.003, + "loss": 4.2229, + "step": 1679 + }, + { + "epoch": 0.0168, + "grad_norm": 0.7189446111250655, + "learning_rate": 0.003, + "loss": 4.2374, + "step": 1680 + }, + { + "epoch": 0.01681, + "grad_norm": 0.7033815727690789, + "learning_rate": 0.003, + "loss": 4.2384, + "step": 1681 + }, + { + "epoch": 0.01682, + "grad_norm": 0.6685587436768649, + "learning_rate": 0.003, + "loss": 4.231, + "step": 1682 + }, + { + "epoch": 0.01683, + "grad_norm": 0.637929998688511, + "learning_rate": 0.003, + "loss": 4.241, + "step": 1683 + }, + { + "epoch": 0.01684, + "grad_norm": 0.628959376324859, + "learning_rate": 0.003, + "loss": 4.2154, + "step": 1684 + }, + { + "epoch": 0.01685, + "grad_norm": 0.6231765249837203, + "learning_rate": 0.003, + "loss": 4.231, + "step": 1685 + }, + { + "epoch": 0.01686, + "grad_norm": 0.5820543845215673, + "learning_rate": 0.003, + "loss": 4.2226, + "step": 1686 + }, + { + "epoch": 0.01687, + "grad_norm": 0.5908668431711505, + "learning_rate": 0.003, + "loss": 4.2167, + "step": 1687 + }, + { + "epoch": 0.01688, + "grad_norm": 0.6244492735925097, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 1688 + }, + { + "epoch": 0.01689, + "grad_norm": 0.601372798813749, + "learning_rate": 0.003, + "loss": 4.2192, + "step": 1689 + }, + { + "epoch": 0.0169, + "grad_norm": 0.5539912290934804, + "learning_rate": 0.003, + "loss": 4.2069, + "step": 1690 + }, + { + "epoch": 0.01691, + "grad_norm": 0.6406257238232621, + "learning_rate": 0.003, + "loss": 4.2196, + "step": 1691 + }, + { + "epoch": 0.01692, + "grad_norm": 0.6715764673442781, + "learning_rate": 0.003, + "loss": 4.2214, + "step": 1692 + }, + { + "epoch": 0.01693, + "grad_norm": 0.594965855834695, + "learning_rate": 0.003, + "loss": 4.2461, + "step": 1693 + }, + { + "epoch": 0.01694, + "grad_norm": 0.5534146564551462, + "learning_rate": 0.003, + "loss": 4.1943, + "step": 1694 + }, + { + "epoch": 0.01695, + "grad_norm": 0.6691590053804509, + "learning_rate": 0.003, + "loss": 4.2273, + "step": 1695 + }, + { + "epoch": 0.01696, + "grad_norm": 0.7988479106506707, + "learning_rate": 0.003, + "loss": 4.2265, + "step": 1696 + }, + { + "epoch": 0.01697, + "grad_norm": 0.9378853835820298, + "learning_rate": 0.003, + "loss": 4.2129, + "step": 1697 + }, + { + "epoch": 0.01698, + "grad_norm": 0.8351693624536569, + "learning_rate": 0.003, + "loss": 4.2098, + "step": 1698 + }, + { + "epoch": 0.01699, + "grad_norm": 0.6689256544140322, + "learning_rate": 0.003, + "loss": 4.2116, + "step": 1699 + }, + { + "epoch": 0.017, + "grad_norm": 0.718191815671613, + "learning_rate": 0.003, + "loss": 4.2331, + "step": 1700 + }, + { + "epoch": 0.01701, + "grad_norm": 0.6236805335228212, + "learning_rate": 0.003, + "loss": 4.2114, + "step": 1701 + }, + { + "epoch": 0.01702, + "grad_norm": 0.6198051313671201, + "learning_rate": 0.003, + "loss": 4.235, + "step": 1702 + }, + { + "epoch": 0.01703, + "grad_norm": 0.6124735544668184, + "learning_rate": 0.003, + "loss": 4.2192, + "step": 1703 + }, + { + "epoch": 0.01704, + "grad_norm": 0.5788411907923544, + "learning_rate": 0.003, + "loss": 4.2046, + "step": 1704 + }, + { + "epoch": 0.01705, + "grad_norm": 0.528954361543718, + "learning_rate": 0.003, + "loss": 4.2048, + "step": 1705 + }, + { + "epoch": 0.01706, + "grad_norm": 0.5207913048275148, + "learning_rate": 0.003, + "loss": 4.203, + "step": 1706 + }, + { + "epoch": 0.01707, + "grad_norm": 0.49956087831547324, + "learning_rate": 0.003, + "loss": 4.2234, + "step": 1707 + }, + { + "epoch": 0.01708, + "grad_norm": 0.5265055046037553, + "learning_rate": 0.003, + "loss": 4.2005, + "step": 1708 + }, + { + "epoch": 0.01709, + "grad_norm": 0.601511173671749, + "learning_rate": 0.003, + "loss": 4.2245, + "step": 1709 + }, + { + "epoch": 0.0171, + "grad_norm": 0.6457072596477392, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 1710 + }, + { + "epoch": 0.01711, + "grad_norm": 0.6931354519414087, + "learning_rate": 0.003, + "loss": 4.206, + "step": 1711 + }, + { + "epoch": 0.01712, + "grad_norm": 0.7843884994764989, + "learning_rate": 0.003, + "loss": 4.2207, + "step": 1712 + }, + { + "epoch": 0.01713, + "grad_norm": 0.8034349282711927, + "learning_rate": 0.003, + "loss": 4.1971, + "step": 1713 + }, + { + "epoch": 0.01714, + "grad_norm": 0.8979650985806761, + "learning_rate": 0.003, + "loss": 4.2412, + "step": 1714 + }, + { + "epoch": 0.01715, + "grad_norm": 0.9199811287834594, + "learning_rate": 0.003, + "loss": 4.2496, + "step": 1715 + }, + { + "epoch": 0.01716, + "grad_norm": 0.8606879750752993, + "learning_rate": 0.003, + "loss": 4.2283, + "step": 1716 + }, + { + "epoch": 0.01717, + "grad_norm": 0.9133738914940963, + "learning_rate": 0.003, + "loss": 4.2649, + "step": 1717 + }, + { + "epoch": 0.01718, + "grad_norm": 0.9751937973978706, + "learning_rate": 0.003, + "loss": 4.2385, + "step": 1718 + }, + { + "epoch": 0.01719, + "grad_norm": 0.9249947621239035, + "learning_rate": 0.003, + "loss": 4.2632, + "step": 1719 + }, + { + "epoch": 0.0172, + "grad_norm": 0.9364522247503431, + "learning_rate": 0.003, + "loss": 4.2208, + "step": 1720 + }, + { + "epoch": 0.01721, + "grad_norm": 0.8201093609998811, + "learning_rate": 0.003, + "loss": 4.2353, + "step": 1721 + }, + { + "epoch": 0.01722, + "grad_norm": 0.8214210127594672, + "learning_rate": 0.003, + "loss": 4.2288, + "step": 1722 + }, + { + "epoch": 0.01723, + "grad_norm": 0.8471893471073342, + "learning_rate": 0.003, + "loss": 4.2399, + "step": 1723 + }, + { + "epoch": 0.01724, + "grad_norm": 0.7683690946961591, + "learning_rate": 0.003, + "loss": 4.2434, + "step": 1724 + }, + { + "epoch": 0.01725, + "grad_norm": 0.8115047065658012, + "learning_rate": 0.003, + "loss": 4.2272, + "step": 1725 + }, + { + "epoch": 0.01726, + "grad_norm": 0.8092210144832711, + "learning_rate": 0.003, + "loss": 4.2174, + "step": 1726 + }, + { + "epoch": 0.01727, + "grad_norm": 0.8550625584545292, + "learning_rate": 0.003, + "loss": 4.226, + "step": 1727 + }, + { + "epoch": 0.01728, + "grad_norm": 0.8021206512196133, + "learning_rate": 0.003, + "loss": 4.2298, + "step": 1728 + }, + { + "epoch": 0.01729, + "grad_norm": 0.7363522392419929, + "learning_rate": 0.003, + "loss": 4.2336, + "step": 1729 + }, + { + "epoch": 0.0173, + "grad_norm": 0.6975599031052584, + "learning_rate": 0.003, + "loss": 4.2051, + "step": 1730 + }, + { + "epoch": 0.01731, + "grad_norm": 0.6221588088033019, + "learning_rate": 0.003, + "loss": 4.2243, + "step": 1731 + }, + { + "epoch": 0.01732, + "grad_norm": 0.612618335472448, + "learning_rate": 0.003, + "loss": 4.2315, + "step": 1732 + }, + { + "epoch": 0.01733, + "grad_norm": 0.6541617874624247, + "learning_rate": 0.003, + "loss": 4.2125, + "step": 1733 + }, + { + "epoch": 0.01734, + "grad_norm": 0.5863531048899692, + "learning_rate": 0.003, + "loss": 4.2205, + "step": 1734 + }, + { + "epoch": 0.01735, + "grad_norm": 0.5699608850382641, + "learning_rate": 0.003, + "loss": 4.1915, + "step": 1735 + }, + { + "epoch": 0.01736, + "grad_norm": 0.5871625865225932, + "learning_rate": 0.003, + "loss": 4.1898, + "step": 1736 + }, + { + "epoch": 0.01737, + "grad_norm": 0.5870381595261084, + "learning_rate": 0.003, + "loss": 4.2212, + "step": 1737 + }, + { + "epoch": 0.01738, + "grad_norm": 0.5498612952733963, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 1738 + }, + { + "epoch": 0.01739, + "grad_norm": 0.6117901077151079, + "learning_rate": 0.003, + "loss": 4.2218, + "step": 1739 + }, + { + "epoch": 0.0174, + "grad_norm": 0.6161361385546128, + "learning_rate": 0.003, + "loss": 4.2194, + "step": 1740 + }, + { + "epoch": 0.01741, + "grad_norm": 0.6356544908412293, + "learning_rate": 0.003, + "loss": 4.2288, + "step": 1741 + }, + { + "epoch": 0.01742, + "grad_norm": 0.6972113147848942, + "learning_rate": 0.003, + "loss": 4.2174, + "step": 1742 + }, + { + "epoch": 0.01743, + "grad_norm": 0.76783001607299, + "learning_rate": 0.003, + "loss": 4.2327, + "step": 1743 + }, + { + "epoch": 0.01744, + "grad_norm": 0.7771532804911003, + "learning_rate": 0.003, + "loss": 4.2177, + "step": 1744 + }, + { + "epoch": 0.01745, + "grad_norm": 0.61106822860021, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 1745 + }, + { + "epoch": 0.01746, + "grad_norm": 0.6089323419748714, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 1746 + }, + { + "epoch": 0.01747, + "grad_norm": 0.6068526864216675, + "learning_rate": 0.003, + "loss": 4.2005, + "step": 1747 + }, + { + "epoch": 0.01748, + "grad_norm": 0.6143873963395499, + "learning_rate": 0.003, + "loss": 4.194, + "step": 1748 + }, + { + "epoch": 0.01749, + "grad_norm": 0.505834715053384, + "learning_rate": 0.003, + "loss": 4.1866, + "step": 1749 + }, + { + "epoch": 0.0175, + "grad_norm": 0.4183071099190176, + "learning_rate": 0.003, + "loss": 4.1986, + "step": 1750 + }, + { + "epoch": 0.01751, + "grad_norm": 0.4200651633429046, + "learning_rate": 0.003, + "loss": 4.1916, + "step": 1751 + }, + { + "epoch": 0.01752, + "grad_norm": 0.42294753534060125, + "learning_rate": 0.003, + "loss": 4.2318, + "step": 1752 + }, + { + "epoch": 0.01753, + "grad_norm": 0.4040582052822703, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 1753 + }, + { + "epoch": 0.01754, + "grad_norm": 0.39927824671480916, + "learning_rate": 0.003, + "loss": 4.2108, + "step": 1754 + }, + { + "epoch": 0.01755, + "grad_norm": 0.39324726452057945, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 1755 + }, + { + "epoch": 0.01756, + "grad_norm": 0.40168143250637856, + "learning_rate": 0.003, + "loss": 4.227, + "step": 1756 + }, + { + "epoch": 0.01757, + "grad_norm": 0.4561700566055392, + "learning_rate": 0.003, + "loss": 4.2081, + "step": 1757 + }, + { + "epoch": 0.01758, + "grad_norm": 0.5466402873638766, + "learning_rate": 0.003, + "loss": 4.2157, + "step": 1758 + }, + { + "epoch": 0.01759, + "grad_norm": 0.6515131335626271, + "learning_rate": 0.003, + "loss": 4.1877, + "step": 1759 + }, + { + "epoch": 0.0176, + "grad_norm": 0.7221307724120931, + "learning_rate": 0.003, + "loss": 4.2153, + "step": 1760 + }, + { + "epoch": 0.01761, + "grad_norm": 0.7514386530231217, + "learning_rate": 0.003, + "loss": 4.2014, + "step": 1761 + }, + { + "epoch": 0.01762, + "grad_norm": 0.8563467654669285, + "learning_rate": 0.003, + "loss": 4.2266, + "step": 1762 + }, + { + "epoch": 0.01763, + "grad_norm": 1.0251781966749487, + "learning_rate": 0.003, + "loss": 4.2304, + "step": 1763 + }, + { + "epoch": 0.01764, + "grad_norm": 1.1022130069351062, + "learning_rate": 0.003, + "loss": 4.204, + "step": 1764 + }, + { + "epoch": 0.01765, + "grad_norm": 0.9424736955387317, + "learning_rate": 0.003, + "loss": 4.2315, + "step": 1765 + }, + { + "epoch": 0.01766, + "grad_norm": 0.9227888573535311, + "learning_rate": 0.003, + "loss": 4.2455, + "step": 1766 + }, + { + "epoch": 0.01767, + "grad_norm": 0.8152567408138972, + "learning_rate": 0.003, + "loss": 4.2043, + "step": 1767 + }, + { + "epoch": 0.01768, + "grad_norm": 0.8952526282025846, + "learning_rate": 0.003, + "loss": 4.2178, + "step": 1768 + }, + { + "epoch": 0.01769, + "grad_norm": 0.9903660885622939, + "learning_rate": 0.003, + "loss": 4.221, + "step": 1769 + }, + { + "epoch": 0.0177, + "grad_norm": 1.1434768458973932, + "learning_rate": 0.003, + "loss": 4.2196, + "step": 1770 + }, + { + "epoch": 0.01771, + "grad_norm": 1.1463152027154602, + "learning_rate": 0.003, + "loss": 4.2541, + "step": 1771 + }, + { + "epoch": 0.01772, + "grad_norm": 0.9320499048871318, + "learning_rate": 0.003, + "loss": 4.2202, + "step": 1772 + }, + { + "epoch": 0.01773, + "grad_norm": 0.9158947764816099, + "learning_rate": 0.003, + "loss": 4.2126, + "step": 1773 + }, + { + "epoch": 0.01774, + "grad_norm": 1.0234123818722582, + "learning_rate": 0.003, + "loss": 4.25, + "step": 1774 + }, + { + "epoch": 0.01775, + "grad_norm": 1.0396053755051766, + "learning_rate": 0.003, + "loss": 4.2387, + "step": 1775 + }, + { + "epoch": 0.01776, + "grad_norm": 0.8699023391257188, + "learning_rate": 0.003, + "loss": 4.2279, + "step": 1776 + }, + { + "epoch": 0.01777, + "grad_norm": 0.7504359108002385, + "learning_rate": 0.003, + "loss": 4.2409, + "step": 1777 + }, + { + "epoch": 0.01778, + "grad_norm": 0.6686463646026519, + "learning_rate": 0.003, + "loss": 4.2296, + "step": 1778 + }, + { + "epoch": 0.01779, + "grad_norm": 0.627501081983313, + "learning_rate": 0.003, + "loss": 4.2441, + "step": 1779 + }, + { + "epoch": 0.0178, + "grad_norm": 0.6074782341836313, + "learning_rate": 0.003, + "loss": 4.213, + "step": 1780 + }, + { + "epoch": 0.01781, + "grad_norm": 0.6632306275046532, + "learning_rate": 0.003, + "loss": 4.2271, + "step": 1781 + }, + { + "epoch": 0.01782, + "grad_norm": 0.6482376732581775, + "learning_rate": 0.003, + "loss": 4.2498, + "step": 1782 + }, + { + "epoch": 0.01783, + "grad_norm": 0.6647693064019089, + "learning_rate": 0.003, + "loss": 4.229, + "step": 1783 + }, + { + "epoch": 0.01784, + "grad_norm": 0.7433144987650033, + "learning_rate": 0.003, + "loss": 4.2289, + "step": 1784 + }, + { + "epoch": 0.01785, + "grad_norm": 0.8023042672883854, + "learning_rate": 0.003, + "loss": 4.1929, + "step": 1785 + }, + { + "epoch": 0.01786, + "grad_norm": 0.7590350402092263, + "learning_rate": 0.003, + "loss": 4.2189, + "step": 1786 + }, + { + "epoch": 0.01787, + "grad_norm": 0.6044391074634635, + "learning_rate": 0.003, + "loss": 4.223, + "step": 1787 + }, + { + "epoch": 0.01788, + "grad_norm": 0.5734225415780418, + "learning_rate": 0.003, + "loss": 4.2291, + "step": 1788 + }, + { + "epoch": 0.01789, + "grad_norm": 0.5619094879913861, + "learning_rate": 0.003, + "loss": 4.1969, + "step": 1789 + }, + { + "epoch": 0.0179, + "grad_norm": 0.5903172908647163, + "learning_rate": 0.003, + "loss": 4.2152, + "step": 1790 + }, + { + "epoch": 0.01791, + "grad_norm": 0.48196952921006103, + "learning_rate": 0.003, + "loss": 4.2224, + "step": 1791 + }, + { + "epoch": 0.01792, + "grad_norm": 0.451545250518652, + "learning_rate": 0.003, + "loss": 4.2239, + "step": 1792 + }, + { + "epoch": 0.01793, + "grad_norm": 0.521884061957321, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 1793 + }, + { + "epoch": 0.01794, + "grad_norm": 0.5052209264531972, + "learning_rate": 0.003, + "loss": 4.1997, + "step": 1794 + }, + { + "epoch": 0.01795, + "grad_norm": 0.4700227936734472, + "learning_rate": 0.003, + "loss": 4.1984, + "step": 1795 + }, + { + "epoch": 0.01796, + "grad_norm": 0.48384581004820526, + "learning_rate": 0.003, + "loss": 4.2111, + "step": 1796 + }, + { + "epoch": 0.01797, + "grad_norm": 0.5674789805804968, + "learning_rate": 0.003, + "loss": 4.1791, + "step": 1797 + }, + { + "epoch": 0.01798, + "grad_norm": 0.5733510313406954, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 1798 + }, + { + "epoch": 0.01799, + "grad_norm": 0.6643001766771917, + "learning_rate": 0.003, + "loss": 4.2106, + "step": 1799 + }, + { + "epoch": 0.018, + "grad_norm": 0.7693742824268138, + "learning_rate": 0.003, + "loss": 4.2019, + "step": 1800 + }, + { + "epoch": 0.01801, + "grad_norm": 0.7789742970956256, + "learning_rate": 0.003, + "loss": 4.2028, + "step": 1801 + }, + { + "epoch": 0.01802, + "grad_norm": 0.7483411999928193, + "learning_rate": 0.003, + "loss": 4.2174, + "step": 1802 + }, + { + "epoch": 0.01803, + "grad_norm": 0.6912125156211221, + "learning_rate": 0.003, + "loss": 4.2209, + "step": 1803 + }, + { + "epoch": 0.01804, + "grad_norm": 0.6110578547720885, + "learning_rate": 0.003, + "loss": 4.1902, + "step": 1804 + }, + { + "epoch": 0.01805, + "grad_norm": 0.5506478742000366, + "learning_rate": 0.003, + "loss": 4.2069, + "step": 1805 + }, + { + "epoch": 0.01806, + "grad_norm": 0.6061937579522265, + "learning_rate": 0.003, + "loss": 4.1674, + "step": 1806 + }, + { + "epoch": 0.01807, + "grad_norm": 0.5932129898926418, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 1807 + }, + { + "epoch": 0.01808, + "grad_norm": 0.5444224258882361, + "learning_rate": 0.003, + "loss": 4.1921, + "step": 1808 + }, + { + "epoch": 0.01809, + "grad_norm": 0.5399756231214402, + "learning_rate": 0.003, + "loss": 4.2143, + "step": 1809 + }, + { + "epoch": 0.0181, + "grad_norm": 0.4832914161347141, + "learning_rate": 0.003, + "loss": 4.2061, + "step": 1810 + }, + { + "epoch": 0.01811, + "grad_norm": 0.5091858383706224, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 1811 + }, + { + "epoch": 0.01812, + "grad_norm": 0.4960424818906239, + "learning_rate": 0.003, + "loss": 4.1761, + "step": 1812 + }, + { + "epoch": 0.01813, + "grad_norm": 0.4858161245409072, + "learning_rate": 0.003, + "loss": 4.2194, + "step": 1813 + }, + { + "epoch": 0.01814, + "grad_norm": 0.5128609502675707, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 1814 + }, + { + "epoch": 0.01815, + "grad_norm": 0.4696917504376051, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 1815 + }, + { + "epoch": 0.01816, + "grad_norm": 0.4276704853681577, + "learning_rate": 0.003, + "loss": 4.1969, + "step": 1816 + }, + { + "epoch": 0.01817, + "grad_norm": 0.49742714220067663, + "learning_rate": 0.003, + "loss": 4.1904, + "step": 1817 + }, + { + "epoch": 0.01818, + "grad_norm": 0.6269759675631359, + "learning_rate": 0.003, + "loss": 4.1931, + "step": 1818 + }, + { + "epoch": 0.01819, + "grad_norm": 0.7520796750217182, + "learning_rate": 0.003, + "loss": 4.1832, + "step": 1819 + }, + { + "epoch": 0.0182, + "grad_norm": 0.9014481818810978, + "learning_rate": 0.003, + "loss": 4.2101, + "step": 1820 + }, + { + "epoch": 0.01821, + "grad_norm": 0.7861765129999728, + "learning_rate": 0.003, + "loss": 4.1954, + "step": 1821 + }, + { + "epoch": 0.01822, + "grad_norm": 0.6190127192025973, + "learning_rate": 0.003, + "loss": 4.2, + "step": 1822 + }, + { + "epoch": 0.01823, + "grad_norm": 0.7778520136299403, + "learning_rate": 0.003, + "loss": 4.2211, + "step": 1823 + }, + { + "epoch": 0.01824, + "grad_norm": 0.7843362143306014, + "learning_rate": 0.003, + "loss": 4.2055, + "step": 1824 + }, + { + "epoch": 0.01825, + "grad_norm": 0.7380149583575544, + "learning_rate": 0.003, + "loss": 4.2108, + "step": 1825 + }, + { + "epoch": 0.01826, + "grad_norm": 0.7040620507144565, + "learning_rate": 0.003, + "loss": 4.2188, + "step": 1826 + }, + { + "epoch": 0.01827, + "grad_norm": 0.8207902713006542, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 1827 + }, + { + "epoch": 0.01828, + "grad_norm": 0.8054878482534694, + "learning_rate": 0.003, + "loss": 4.2093, + "step": 1828 + }, + { + "epoch": 0.01829, + "grad_norm": 0.8245790564775142, + "learning_rate": 0.003, + "loss": 4.2301, + "step": 1829 + }, + { + "epoch": 0.0183, + "grad_norm": 0.8846767870010911, + "learning_rate": 0.003, + "loss": 4.2277, + "step": 1830 + }, + { + "epoch": 0.01831, + "grad_norm": 0.8289617540456092, + "learning_rate": 0.003, + "loss": 4.2216, + "step": 1831 + }, + { + "epoch": 0.01832, + "grad_norm": 0.731274205528004, + "learning_rate": 0.003, + "loss": 4.2108, + "step": 1832 + }, + { + "epoch": 0.01833, + "grad_norm": 0.7928566591291573, + "learning_rate": 0.003, + "loss": 4.2459, + "step": 1833 + }, + { + "epoch": 0.01834, + "grad_norm": 0.7576155251040918, + "learning_rate": 0.003, + "loss": 4.2123, + "step": 1834 + }, + { + "epoch": 0.01835, + "grad_norm": 0.7931842778411774, + "learning_rate": 0.003, + "loss": 4.214, + "step": 1835 + }, + { + "epoch": 0.01836, + "grad_norm": 0.6445999325297602, + "learning_rate": 0.003, + "loss": 4.2036, + "step": 1836 + }, + { + "epoch": 0.01837, + "grad_norm": 0.680643728796186, + "learning_rate": 0.003, + "loss": 4.208, + "step": 1837 + }, + { + "epoch": 0.01838, + "grad_norm": 0.6239843687734086, + "learning_rate": 0.003, + "loss": 4.2079, + "step": 1838 + }, + { + "epoch": 0.01839, + "grad_norm": 0.6182658027468049, + "learning_rate": 0.003, + "loss": 4.2277, + "step": 1839 + }, + { + "epoch": 0.0184, + "grad_norm": 0.5890463339530954, + "learning_rate": 0.003, + "loss": 4.2247, + "step": 1840 + }, + { + "epoch": 0.01841, + "grad_norm": 0.5863839914398512, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 1841 + }, + { + "epoch": 0.01842, + "grad_norm": 0.5858091006393259, + "learning_rate": 0.003, + "loss": 4.1952, + "step": 1842 + }, + { + "epoch": 0.01843, + "grad_norm": 0.5624576041074113, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 1843 + }, + { + "epoch": 0.01844, + "grad_norm": 0.5465927738839299, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 1844 + }, + { + "epoch": 0.01845, + "grad_norm": 0.6027290912828838, + "learning_rate": 0.003, + "loss": 4.1768, + "step": 1845 + }, + { + "epoch": 0.01846, + "grad_norm": 0.7027351856984548, + "learning_rate": 0.003, + "loss": 4.2065, + "step": 1846 + }, + { + "epoch": 0.01847, + "grad_norm": 0.8199216861922868, + "learning_rate": 0.003, + "loss": 4.2072, + "step": 1847 + }, + { + "epoch": 0.01848, + "grad_norm": 0.7370764094470499, + "learning_rate": 0.003, + "loss": 4.2221, + "step": 1848 + }, + { + "epoch": 0.01849, + "grad_norm": 0.6895980947043083, + "learning_rate": 0.003, + "loss": 4.2176, + "step": 1849 + }, + { + "epoch": 0.0185, + "grad_norm": 0.6744601685013915, + "learning_rate": 0.003, + "loss": 4.2071, + "step": 1850 + }, + { + "epoch": 0.01851, + "grad_norm": 0.7592763226230413, + "learning_rate": 0.003, + "loss": 4.1894, + "step": 1851 + }, + { + "epoch": 0.01852, + "grad_norm": 0.7512384494982312, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 1852 + }, + { + "epoch": 0.01853, + "grad_norm": 0.642959606761384, + "learning_rate": 0.003, + "loss": 4.1897, + "step": 1853 + }, + { + "epoch": 0.01854, + "grad_norm": 0.6302719538000866, + "learning_rate": 0.003, + "loss": 4.2198, + "step": 1854 + }, + { + "epoch": 0.01855, + "grad_norm": 0.5865658475812112, + "learning_rate": 0.003, + "loss": 4.2046, + "step": 1855 + }, + { + "epoch": 0.01856, + "grad_norm": 0.6166190320727692, + "learning_rate": 0.003, + "loss": 4.189, + "step": 1856 + }, + { + "epoch": 0.01857, + "grad_norm": 0.7181557293855338, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 1857 + }, + { + "epoch": 0.01858, + "grad_norm": 0.8412048999626534, + "learning_rate": 0.003, + "loss": 4.1934, + "step": 1858 + }, + { + "epoch": 0.01859, + "grad_norm": 0.9023360494815195, + "learning_rate": 0.003, + "loss": 4.1912, + "step": 1859 + }, + { + "epoch": 0.0186, + "grad_norm": 0.7898380300240098, + "learning_rate": 0.003, + "loss": 4.1718, + "step": 1860 + }, + { + "epoch": 0.01861, + "grad_norm": 0.7208167303209774, + "learning_rate": 0.003, + "loss": 4.2166, + "step": 1861 + }, + { + "epoch": 0.01862, + "grad_norm": 0.7104369497590945, + "learning_rate": 0.003, + "loss": 4.179, + "step": 1862 + }, + { + "epoch": 0.01863, + "grad_norm": 0.7138975727375058, + "learning_rate": 0.003, + "loss": 4.1882, + "step": 1863 + }, + { + "epoch": 0.01864, + "grad_norm": 0.6519153740387305, + "learning_rate": 0.003, + "loss": 4.2193, + "step": 1864 + }, + { + "epoch": 0.01865, + "grad_norm": 0.7057872052617501, + "learning_rate": 0.003, + "loss": 4.1967, + "step": 1865 + }, + { + "epoch": 0.01866, + "grad_norm": 0.7579696750373845, + "learning_rate": 0.003, + "loss": 4.203, + "step": 1866 + }, + { + "epoch": 0.01867, + "grad_norm": 0.7437399572447418, + "learning_rate": 0.003, + "loss": 4.1772, + "step": 1867 + }, + { + "epoch": 0.01868, + "grad_norm": 0.7143070391507745, + "learning_rate": 0.003, + "loss": 4.1947, + "step": 1868 + }, + { + "epoch": 0.01869, + "grad_norm": 0.7078429320794878, + "learning_rate": 0.003, + "loss": 4.2102, + "step": 1869 + }, + { + "epoch": 0.0187, + "grad_norm": 0.7908878960855552, + "learning_rate": 0.003, + "loss": 4.2031, + "step": 1870 + }, + { + "epoch": 0.01871, + "grad_norm": 0.9307560381931724, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 1871 + }, + { + "epoch": 0.01872, + "grad_norm": 0.937846873989918, + "learning_rate": 0.003, + "loss": 4.2367, + "step": 1872 + }, + { + "epoch": 0.01873, + "grad_norm": 0.7588221299073229, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 1873 + }, + { + "epoch": 0.01874, + "grad_norm": 0.6728512390013429, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 1874 + }, + { + "epoch": 0.01875, + "grad_norm": 0.6514462862930852, + "learning_rate": 0.003, + "loss": 4.21, + "step": 1875 + }, + { + "epoch": 0.01876, + "grad_norm": 0.703173016887, + "learning_rate": 0.003, + "loss": 4.2153, + "step": 1876 + }, + { + "epoch": 0.01877, + "grad_norm": 0.73028206380223, + "learning_rate": 0.003, + "loss": 4.2227, + "step": 1877 + }, + { + "epoch": 0.01878, + "grad_norm": 0.7555785531530169, + "learning_rate": 0.003, + "loss": 4.2069, + "step": 1878 + }, + { + "epoch": 0.01879, + "grad_norm": 0.7934996955230512, + "learning_rate": 0.003, + "loss": 4.196, + "step": 1879 + }, + { + "epoch": 0.0188, + "grad_norm": 0.7476047962254878, + "learning_rate": 0.003, + "loss": 4.199, + "step": 1880 + }, + { + "epoch": 0.01881, + "grad_norm": 0.732632895668458, + "learning_rate": 0.003, + "loss": 4.195, + "step": 1881 + }, + { + "epoch": 0.01882, + "grad_norm": 0.7676196664715533, + "learning_rate": 0.003, + "loss": 4.2236, + "step": 1882 + }, + { + "epoch": 0.01883, + "grad_norm": 0.774042907112019, + "learning_rate": 0.003, + "loss": 4.197, + "step": 1883 + }, + { + "epoch": 0.01884, + "grad_norm": 0.6501154660141607, + "learning_rate": 0.003, + "loss": 4.1919, + "step": 1884 + }, + { + "epoch": 0.01885, + "grad_norm": 0.5559497730489893, + "learning_rate": 0.003, + "loss": 4.1659, + "step": 1885 + }, + { + "epoch": 0.01886, + "grad_norm": 0.5108169963928473, + "learning_rate": 0.003, + "loss": 4.2034, + "step": 1886 + }, + { + "epoch": 0.01887, + "grad_norm": 0.4802312262485427, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 1887 + }, + { + "epoch": 0.01888, + "grad_norm": 0.44491425211803826, + "learning_rate": 0.003, + "loss": 4.1789, + "step": 1888 + }, + { + "epoch": 0.01889, + "grad_norm": 0.4616155311993016, + "learning_rate": 0.003, + "loss": 4.1923, + "step": 1889 + }, + { + "epoch": 0.0189, + "grad_norm": 0.4932376764256785, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 1890 + }, + { + "epoch": 0.01891, + "grad_norm": 0.5809213606136114, + "learning_rate": 0.003, + "loss": 4.2219, + "step": 1891 + }, + { + "epoch": 0.01892, + "grad_norm": 0.7827791825078997, + "learning_rate": 0.003, + "loss": 4.2063, + "step": 1892 + }, + { + "epoch": 0.01893, + "grad_norm": 0.9634537072256092, + "learning_rate": 0.003, + "loss": 4.1968, + "step": 1893 + }, + { + "epoch": 0.01894, + "grad_norm": 0.8975581900781968, + "learning_rate": 0.003, + "loss": 4.2047, + "step": 1894 + }, + { + "epoch": 0.01895, + "grad_norm": 0.670439659623389, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 1895 + }, + { + "epoch": 0.01896, + "grad_norm": 0.774532069243431, + "learning_rate": 0.003, + "loss": 4.208, + "step": 1896 + }, + { + "epoch": 0.01897, + "grad_norm": 0.7575182930280945, + "learning_rate": 0.003, + "loss": 4.2077, + "step": 1897 + }, + { + "epoch": 0.01898, + "grad_norm": 0.6342553387149622, + "learning_rate": 0.003, + "loss": 4.1723, + "step": 1898 + }, + { + "epoch": 0.01899, + "grad_norm": 0.6743268097098117, + "learning_rate": 0.003, + "loss": 4.1867, + "step": 1899 + }, + { + "epoch": 0.019, + "grad_norm": 0.6675442258402475, + "learning_rate": 0.003, + "loss": 4.1804, + "step": 1900 + }, + { + "epoch": 0.01901, + "grad_norm": 0.6448176087551704, + "learning_rate": 0.003, + "loss": 4.2267, + "step": 1901 + }, + { + "epoch": 0.01902, + "grad_norm": 0.6309260453072276, + "learning_rate": 0.003, + "loss": 4.2038, + "step": 1902 + }, + { + "epoch": 0.01903, + "grad_norm": 0.5175825025006746, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 1903 + }, + { + "epoch": 0.01904, + "grad_norm": 0.44558661360578977, + "learning_rate": 0.003, + "loss": 4.2075, + "step": 1904 + }, + { + "epoch": 0.01905, + "grad_norm": 0.4519539268740215, + "learning_rate": 0.003, + "loss": 4.1828, + "step": 1905 + }, + { + "epoch": 0.01906, + "grad_norm": 0.4534257888011715, + "learning_rate": 0.003, + "loss": 4.1993, + "step": 1906 + }, + { + "epoch": 0.01907, + "grad_norm": 0.4205401145833596, + "learning_rate": 0.003, + "loss": 4.1917, + "step": 1907 + }, + { + "epoch": 0.01908, + "grad_norm": 0.4717335154892585, + "learning_rate": 0.003, + "loss": 4.2064, + "step": 1908 + }, + { + "epoch": 0.01909, + "grad_norm": 0.5307249224766913, + "learning_rate": 0.003, + "loss": 4.1715, + "step": 1909 + }, + { + "epoch": 0.0191, + "grad_norm": 0.6322393539173273, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 1910 + }, + { + "epoch": 0.01911, + "grad_norm": 0.6545216933386392, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 1911 + }, + { + "epoch": 0.01912, + "grad_norm": 0.5663347338798967, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 1912 + }, + { + "epoch": 0.01913, + "grad_norm": 0.5706339861620788, + "learning_rate": 0.003, + "loss": 4.188, + "step": 1913 + }, + { + "epoch": 0.01914, + "grad_norm": 0.6036176037925194, + "learning_rate": 0.003, + "loss": 4.2071, + "step": 1914 + }, + { + "epoch": 0.01915, + "grad_norm": 0.5703771194775653, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 1915 + }, + { + "epoch": 0.01916, + "grad_norm": 0.6012322903506473, + "learning_rate": 0.003, + "loss": 4.173, + "step": 1916 + }, + { + "epoch": 0.01917, + "grad_norm": 0.7091647735109047, + "learning_rate": 0.003, + "loss": 4.1961, + "step": 1917 + }, + { + "epoch": 0.01918, + "grad_norm": 0.8129980711183025, + "learning_rate": 0.003, + "loss": 4.1895, + "step": 1918 + }, + { + "epoch": 0.01919, + "grad_norm": 1.0293282506765802, + "learning_rate": 0.003, + "loss": 4.2041, + "step": 1919 + }, + { + "epoch": 0.0192, + "grad_norm": 0.9313730998487869, + "learning_rate": 0.003, + "loss": 4.2061, + "step": 1920 + }, + { + "epoch": 0.01921, + "grad_norm": 0.8720092800409024, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 1921 + }, + { + "epoch": 0.01922, + "grad_norm": 0.7996945150379353, + "learning_rate": 0.003, + "loss": 4.2091, + "step": 1922 + }, + { + "epoch": 0.01923, + "grad_norm": 0.8800137161869123, + "learning_rate": 0.003, + "loss": 4.2165, + "step": 1923 + }, + { + "epoch": 0.01924, + "grad_norm": 1.0182126450186646, + "learning_rate": 0.003, + "loss": 4.2271, + "step": 1924 + }, + { + "epoch": 0.01925, + "grad_norm": 0.9807294623812044, + "learning_rate": 0.003, + "loss": 4.2164, + "step": 1925 + }, + { + "epoch": 0.01926, + "grad_norm": 0.8494587539441437, + "learning_rate": 0.003, + "loss": 4.2224, + "step": 1926 + }, + { + "epoch": 0.01927, + "grad_norm": 0.7179780359821647, + "learning_rate": 0.003, + "loss": 4.2071, + "step": 1927 + }, + { + "epoch": 0.01928, + "grad_norm": 0.7281644181803769, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 1928 + }, + { + "epoch": 0.01929, + "grad_norm": 0.7065947298847162, + "learning_rate": 0.003, + "loss": 4.1992, + "step": 1929 + }, + { + "epoch": 0.0193, + "grad_norm": 0.5649201338939857, + "learning_rate": 0.003, + "loss": 4.1851, + "step": 1930 + }, + { + "epoch": 0.01931, + "grad_norm": 0.5271540347343493, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 1931 + }, + { + "epoch": 0.01932, + "grad_norm": 0.45255335182017054, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 1932 + }, + { + "epoch": 0.01933, + "grad_norm": 0.47680900276175325, + "learning_rate": 0.003, + "loss": 4.1783, + "step": 1933 + }, + { + "epoch": 0.01934, + "grad_norm": 0.5504435487639365, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 1934 + }, + { + "epoch": 0.01935, + "grad_norm": 0.6345773523818935, + "learning_rate": 0.003, + "loss": 4.1989, + "step": 1935 + }, + { + "epoch": 0.01936, + "grad_norm": 0.7991398209168664, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 1936 + }, + { + "epoch": 0.01937, + "grad_norm": 0.9725224419371103, + "learning_rate": 0.003, + "loss": 4.2069, + "step": 1937 + }, + { + "epoch": 0.01938, + "grad_norm": 0.8008224777593416, + "learning_rate": 0.003, + "loss": 4.2024, + "step": 1938 + }, + { + "epoch": 0.01939, + "grad_norm": 0.6377996497038249, + "learning_rate": 0.003, + "loss": 4.2204, + "step": 1939 + }, + { + "epoch": 0.0194, + "grad_norm": 0.7081879235874463, + "learning_rate": 0.003, + "loss": 4.1973, + "step": 1940 + }, + { + "epoch": 0.01941, + "grad_norm": 0.6250228653949582, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 1941 + }, + { + "epoch": 0.01942, + "grad_norm": 0.6717463944474127, + "learning_rate": 0.003, + "loss": 4.2019, + "step": 1942 + }, + { + "epoch": 0.01943, + "grad_norm": 0.6654540389471064, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 1943 + }, + { + "epoch": 0.01944, + "grad_norm": 0.6610991638775895, + "learning_rate": 0.003, + "loss": 4.2146, + "step": 1944 + }, + { + "epoch": 0.01945, + "grad_norm": 0.6283793764328768, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 1945 + }, + { + "epoch": 0.01946, + "grad_norm": 0.563815590170461, + "learning_rate": 0.003, + "loss": 4.175, + "step": 1946 + }, + { + "epoch": 0.01947, + "grad_norm": 0.5825718696133029, + "learning_rate": 0.003, + "loss": 4.1841, + "step": 1947 + }, + { + "epoch": 0.01948, + "grad_norm": 0.7034835120761256, + "learning_rate": 0.003, + "loss": 4.2096, + "step": 1948 + }, + { + "epoch": 0.01949, + "grad_norm": 0.9237001793340572, + "learning_rate": 0.003, + "loss": 4.209, + "step": 1949 + }, + { + "epoch": 0.0195, + "grad_norm": 0.9365393224462327, + "learning_rate": 0.003, + "loss": 4.207, + "step": 1950 + }, + { + "epoch": 0.01951, + "grad_norm": 0.8087831479581115, + "learning_rate": 0.003, + "loss": 4.194, + "step": 1951 + }, + { + "epoch": 0.01952, + "grad_norm": 0.8819853779113177, + "learning_rate": 0.003, + "loss": 4.2078, + "step": 1952 + }, + { + "epoch": 0.01953, + "grad_norm": 0.9866084074346885, + "learning_rate": 0.003, + "loss": 4.2291, + "step": 1953 + }, + { + "epoch": 0.01954, + "grad_norm": 1.037114417680343, + "learning_rate": 0.003, + "loss": 4.2303, + "step": 1954 + }, + { + "epoch": 0.01955, + "grad_norm": 0.9479637499094957, + "learning_rate": 0.003, + "loss": 4.2225, + "step": 1955 + }, + { + "epoch": 0.01956, + "grad_norm": 0.8994010519419664, + "learning_rate": 0.003, + "loss": 4.2135, + "step": 1956 + }, + { + "epoch": 0.01957, + "grad_norm": 0.8842871012128143, + "learning_rate": 0.003, + "loss": 4.1851, + "step": 1957 + }, + { + "epoch": 0.01958, + "grad_norm": 0.9376407923168335, + "learning_rate": 0.003, + "loss": 4.1829, + "step": 1958 + }, + { + "epoch": 0.01959, + "grad_norm": 0.8263866600223667, + "learning_rate": 0.003, + "loss": 4.2075, + "step": 1959 + }, + { + "epoch": 0.0196, + "grad_norm": 0.6863533699796172, + "learning_rate": 0.003, + "loss": 4.1925, + "step": 1960 + }, + { + "epoch": 0.01961, + "grad_norm": 0.6116137760392246, + "learning_rate": 0.003, + "loss": 4.191, + "step": 1961 + }, + { + "epoch": 0.01962, + "grad_norm": 0.6395227196191122, + "learning_rate": 0.003, + "loss": 4.1912, + "step": 1962 + }, + { + "epoch": 0.01963, + "grad_norm": 0.5883418831705517, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 1963 + }, + { + "epoch": 0.01964, + "grad_norm": 0.5921487580120356, + "learning_rate": 0.003, + "loss": 4.2037, + "step": 1964 + }, + { + "epoch": 0.01965, + "grad_norm": 0.5536329776274996, + "learning_rate": 0.003, + "loss": 4.1947, + "step": 1965 + }, + { + "epoch": 0.01966, + "grad_norm": 0.47388805459585637, + "learning_rate": 0.003, + "loss": 4.1925, + "step": 1966 + }, + { + "epoch": 0.01967, + "grad_norm": 0.4786574861196878, + "learning_rate": 0.003, + "loss": 4.1683, + "step": 1967 + }, + { + "epoch": 0.01968, + "grad_norm": 0.49077496594088527, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 1968 + }, + { + "epoch": 0.01969, + "grad_norm": 0.5052802929518135, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 1969 + }, + { + "epoch": 0.0197, + "grad_norm": 0.5295838660446354, + "learning_rate": 0.003, + "loss": 4.1837, + "step": 1970 + }, + { + "epoch": 0.01971, + "grad_norm": 0.5513602113626036, + "learning_rate": 0.003, + "loss": 4.1684, + "step": 1971 + }, + { + "epoch": 0.01972, + "grad_norm": 0.5210985598522027, + "learning_rate": 0.003, + "loss": 4.1701, + "step": 1972 + }, + { + "epoch": 0.01973, + "grad_norm": 0.6615280275024347, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 1973 + }, + { + "epoch": 0.01974, + "grad_norm": 0.728061728231907, + "learning_rate": 0.003, + "loss": 4.1988, + "step": 1974 + }, + { + "epoch": 0.01975, + "grad_norm": 0.7196409318491062, + "learning_rate": 0.003, + "loss": 4.1895, + "step": 1975 + }, + { + "epoch": 0.01976, + "grad_norm": 0.7476578183215388, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 1976 + }, + { + "epoch": 0.01977, + "grad_norm": 0.721989664484263, + "learning_rate": 0.003, + "loss": 4.2138, + "step": 1977 + }, + { + "epoch": 0.01978, + "grad_norm": 0.6639417357411178, + "learning_rate": 0.003, + "loss": 4.211, + "step": 1978 + }, + { + "epoch": 0.01979, + "grad_norm": 0.7965280918139568, + "learning_rate": 0.003, + "loss": 4.2066, + "step": 1979 + }, + { + "epoch": 0.0198, + "grad_norm": 0.8613955063166061, + "learning_rate": 0.003, + "loss": 4.1895, + "step": 1980 + }, + { + "epoch": 0.01981, + "grad_norm": 0.8808259730526657, + "learning_rate": 0.003, + "loss": 4.171, + "step": 1981 + }, + { + "epoch": 0.01982, + "grad_norm": 0.7293195128794533, + "learning_rate": 0.003, + "loss": 4.1948, + "step": 1982 + }, + { + "epoch": 0.01983, + "grad_norm": 0.6924854410180289, + "learning_rate": 0.003, + "loss": 4.1872, + "step": 1983 + }, + { + "epoch": 0.01984, + "grad_norm": 0.6776958639455118, + "learning_rate": 0.003, + "loss": 4.1991, + "step": 1984 + }, + { + "epoch": 0.01985, + "grad_norm": 0.6405168066364108, + "learning_rate": 0.003, + "loss": 4.1873, + "step": 1985 + }, + { + "epoch": 0.01986, + "grad_norm": 0.6675206686955611, + "learning_rate": 0.003, + "loss": 4.2151, + "step": 1986 + }, + { + "epoch": 0.01987, + "grad_norm": 0.6315266333005412, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 1987 + }, + { + "epoch": 0.01988, + "grad_norm": 0.5919848970771883, + "learning_rate": 0.003, + "loss": 4.1792, + "step": 1988 + }, + { + "epoch": 0.01989, + "grad_norm": 0.6323287062537767, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 1989 + }, + { + "epoch": 0.0199, + "grad_norm": 0.6888866985403065, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 1990 + }, + { + "epoch": 0.01991, + "grad_norm": 0.770206798498307, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 1991 + }, + { + "epoch": 0.01992, + "grad_norm": 0.7066635985645741, + "learning_rate": 0.003, + "loss": 4.2074, + "step": 1992 + }, + { + "epoch": 0.01993, + "grad_norm": 0.5958380593977555, + "learning_rate": 0.003, + "loss": 4.1861, + "step": 1993 + }, + { + "epoch": 0.01994, + "grad_norm": 0.4987218166773851, + "learning_rate": 0.003, + "loss": 4.204, + "step": 1994 + }, + { + "epoch": 0.01995, + "grad_norm": 0.5506297947485004, + "learning_rate": 0.003, + "loss": 4.1947, + "step": 1995 + }, + { + "epoch": 0.01996, + "grad_norm": 0.6196883290824272, + "learning_rate": 0.003, + "loss": 4.1663, + "step": 1996 + }, + { + "epoch": 0.01997, + "grad_norm": 0.5528178376723564, + "learning_rate": 0.003, + "loss": 4.1937, + "step": 1997 + }, + { + "epoch": 0.01998, + "grad_norm": 0.5452766022105004, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 1998 + }, + { + "epoch": 0.01999, + "grad_norm": 0.5989401536795388, + "learning_rate": 0.003, + "loss": 4.1688, + "step": 1999 + }, + { + "epoch": 0.02, + "grad_norm": 0.5871101885148012, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 2000 + }, + { + "epoch": 0.02001, + "grad_norm": 0.6439983033952634, + "learning_rate": 0.003, + "loss": 4.192, + "step": 2001 + }, + { + "epoch": 0.02002, + "grad_norm": 0.705152096434181, + "learning_rate": 0.003, + "loss": 4.1881, + "step": 2002 + }, + { + "epoch": 0.02003, + "grad_norm": 0.6697476649776103, + "learning_rate": 0.003, + "loss": 4.1877, + "step": 2003 + }, + { + "epoch": 0.02004, + "grad_norm": 0.6315450525269957, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 2004 + }, + { + "epoch": 0.02005, + "grad_norm": 0.6886524691680549, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 2005 + }, + { + "epoch": 0.02006, + "grad_norm": 0.7776759666901092, + "learning_rate": 0.003, + "loss": 4.1814, + "step": 2006 + }, + { + "epoch": 0.02007, + "grad_norm": 0.7175450543148997, + "learning_rate": 0.003, + "loss": 4.1942, + "step": 2007 + }, + { + "epoch": 0.02008, + "grad_norm": 0.6740433282603207, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 2008 + }, + { + "epoch": 0.02009, + "grad_norm": 0.6053852368029262, + "learning_rate": 0.003, + "loss": 4.1814, + "step": 2009 + }, + { + "epoch": 0.0201, + "grad_norm": 0.7400822237531193, + "learning_rate": 0.003, + "loss": 4.1997, + "step": 2010 + }, + { + "epoch": 0.02011, + "grad_norm": 0.9191058488932693, + "learning_rate": 0.003, + "loss": 4.1955, + "step": 2011 + }, + { + "epoch": 0.02012, + "grad_norm": 0.8835602271422344, + "learning_rate": 0.003, + "loss": 4.1963, + "step": 2012 + }, + { + "epoch": 0.02013, + "grad_norm": 0.8767352085076577, + "learning_rate": 0.003, + "loss": 4.2053, + "step": 2013 + }, + { + "epoch": 0.02014, + "grad_norm": 0.9141598127621, + "learning_rate": 0.003, + "loss": 4.1917, + "step": 2014 + }, + { + "epoch": 0.02015, + "grad_norm": 0.8643085839134408, + "learning_rate": 0.003, + "loss": 4.1905, + "step": 2015 + }, + { + "epoch": 0.02016, + "grad_norm": 0.826128703847821, + "learning_rate": 0.003, + "loss": 4.198, + "step": 2016 + }, + { + "epoch": 0.02017, + "grad_norm": 0.9738713965140303, + "learning_rate": 0.003, + "loss": 4.2044, + "step": 2017 + }, + { + "epoch": 0.02018, + "grad_norm": 1.0740244285823746, + "learning_rate": 0.003, + "loss": 4.2125, + "step": 2018 + }, + { + "epoch": 0.02019, + "grad_norm": 0.9545167923514007, + "learning_rate": 0.003, + "loss": 4.2147, + "step": 2019 + }, + { + "epoch": 0.0202, + "grad_norm": 0.9294839382102774, + "learning_rate": 0.003, + "loss": 4.2225, + "step": 2020 + }, + { + "epoch": 0.02021, + "grad_norm": 0.9972429801480568, + "learning_rate": 0.003, + "loss": 4.2402, + "step": 2021 + }, + { + "epoch": 0.02022, + "grad_norm": 0.9712013734706865, + "learning_rate": 0.003, + "loss": 4.2016, + "step": 2022 + }, + { + "epoch": 0.02023, + "grad_norm": 0.9479529184724994, + "learning_rate": 0.003, + "loss": 4.1766, + "step": 2023 + }, + { + "epoch": 0.02024, + "grad_norm": 0.7262535435296761, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 2024 + }, + { + "epoch": 0.02025, + "grad_norm": 0.7062431870057428, + "learning_rate": 0.003, + "loss": 4.243, + "step": 2025 + }, + { + "epoch": 0.02026, + "grad_norm": 0.6882014648662833, + "learning_rate": 0.003, + "loss": 4.22, + "step": 2026 + }, + { + "epoch": 0.02027, + "grad_norm": 0.7627434127876699, + "learning_rate": 0.003, + "loss": 4.2172, + "step": 2027 + }, + { + "epoch": 0.02028, + "grad_norm": 0.8227722936843153, + "learning_rate": 0.003, + "loss": 4.2032, + "step": 2028 + }, + { + "epoch": 0.02029, + "grad_norm": 0.8797059921588131, + "learning_rate": 0.003, + "loss": 4.2112, + "step": 2029 + }, + { + "epoch": 0.0203, + "grad_norm": 0.667440342318075, + "learning_rate": 0.003, + "loss": 4.1975, + "step": 2030 + }, + { + "epoch": 0.02031, + "grad_norm": 0.6037342326248459, + "learning_rate": 0.003, + "loss": 4.1958, + "step": 2031 + }, + { + "epoch": 0.02032, + "grad_norm": 0.597929577022335, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 2032 + }, + { + "epoch": 0.02033, + "grad_norm": 0.5784662304182125, + "learning_rate": 0.003, + "loss": 4.2091, + "step": 2033 + }, + { + "epoch": 0.02034, + "grad_norm": 0.5473032978236689, + "learning_rate": 0.003, + "loss": 4.1993, + "step": 2034 + }, + { + "epoch": 0.02035, + "grad_norm": 0.4379983714060978, + "learning_rate": 0.003, + "loss": 4.1805, + "step": 2035 + }, + { + "epoch": 0.02036, + "grad_norm": 0.441845978539556, + "learning_rate": 0.003, + "loss": 4.1813, + "step": 2036 + }, + { + "epoch": 0.02037, + "grad_norm": 0.3955320194493982, + "learning_rate": 0.003, + "loss": 4.1899, + "step": 2037 + }, + { + "epoch": 0.02038, + "grad_norm": 0.38162570795156125, + "learning_rate": 0.003, + "loss": 4.1737, + "step": 2038 + }, + { + "epoch": 0.02039, + "grad_norm": 0.3933398272384119, + "learning_rate": 0.003, + "loss": 4.1803, + "step": 2039 + }, + { + "epoch": 0.0204, + "grad_norm": 0.3849152800759688, + "learning_rate": 0.003, + "loss": 4.1839, + "step": 2040 + }, + { + "epoch": 0.02041, + "grad_norm": 0.40314953755241956, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 2041 + }, + { + "epoch": 0.02042, + "grad_norm": 0.4229133216155547, + "learning_rate": 0.003, + "loss": 4.1441, + "step": 2042 + }, + { + "epoch": 0.02043, + "grad_norm": 0.36825585827625734, + "learning_rate": 0.003, + "loss": 4.1881, + "step": 2043 + }, + { + "epoch": 0.02044, + "grad_norm": 0.3910413723031291, + "learning_rate": 0.003, + "loss": 4.161, + "step": 2044 + }, + { + "epoch": 0.02045, + "grad_norm": 0.47033000256708646, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2045 + }, + { + "epoch": 0.02046, + "grad_norm": 0.5280528216345948, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 2046 + }, + { + "epoch": 0.02047, + "grad_norm": 0.611592109875483, + "learning_rate": 0.003, + "loss": 4.1961, + "step": 2047 + }, + { + "epoch": 0.02048, + "grad_norm": 0.7788142590682364, + "learning_rate": 0.003, + "loss": 4.1814, + "step": 2048 + }, + { + "epoch": 0.02049, + "grad_norm": 0.86700421151072, + "learning_rate": 0.003, + "loss": 4.211, + "step": 2049 + }, + { + "epoch": 0.0205, + "grad_norm": 0.6903033807839286, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 2050 + }, + { + "epoch": 0.02051, + "grad_norm": 0.6816093568467978, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 2051 + }, + { + "epoch": 0.02052, + "grad_norm": 0.6461107254927798, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 2052 + }, + { + "epoch": 0.02053, + "grad_norm": 0.6080923242609303, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 2053 + }, + { + "epoch": 0.02054, + "grad_norm": 0.6595388402451854, + "learning_rate": 0.003, + "loss": 4.1954, + "step": 2054 + }, + { + "epoch": 0.02055, + "grad_norm": 0.8004674462519312, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 2055 + }, + { + "epoch": 0.02056, + "grad_norm": 0.810242140253679, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 2056 + }, + { + "epoch": 0.02057, + "grad_norm": 0.7021896821785775, + "learning_rate": 0.003, + "loss": 4.1966, + "step": 2057 + }, + { + "epoch": 0.02058, + "grad_norm": 0.7276103690856331, + "learning_rate": 0.003, + "loss": 4.172, + "step": 2058 + }, + { + "epoch": 0.02059, + "grad_norm": 0.7347285469505955, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 2059 + }, + { + "epoch": 0.0206, + "grad_norm": 0.7236883097957534, + "learning_rate": 0.003, + "loss": 4.1984, + "step": 2060 + }, + { + "epoch": 0.02061, + "grad_norm": 0.8529927568671346, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 2061 + }, + { + "epoch": 0.02062, + "grad_norm": 0.8076038753970438, + "learning_rate": 0.003, + "loss": 4.1765, + "step": 2062 + }, + { + "epoch": 0.02063, + "grad_norm": 0.8159969089384409, + "learning_rate": 0.003, + "loss": 4.1781, + "step": 2063 + }, + { + "epoch": 0.02064, + "grad_norm": 0.7974324377759511, + "learning_rate": 0.003, + "loss": 4.1904, + "step": 2064 + }, + { + "epoch": 0.02065, + "grad_norm": 0.9038297269644521, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 2065 + }, + { + "epoch": 0.02066, + "grad_norm": 0.8675285134834587, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 2066 + }, + { + "epoch": 0.02067, + "grad_norm": 0.9126353355460807, + "learning_rate": 0.003, + "loss": 4.1976, + "step": 2067 + }, + { + "epoch": 0.02068, + "grad_norm": 0.7893382087059589, + "learning_rate": 0.003, + "loss": 4.1723, + "step": 2068 + }, + { + "epoch": 0.02069, + "grad_norm": 0.6955889017759005, + "learning_rate": 0.003, + "loss": 4.1909, + "step": 2069 + }, + { + "epoch": 0.0207, + "grad_norm": 0.7051039362555491, + "learning_rate": 0.003, + "loss": 4.1666, + "step": 2070 + }, + { + "epoch": 0.02071, + "grad_norm": 0.7057450771805537, + "learning_rate": 0.003, + "loss": 4.1946, + "step": 2071 + }, + { + "epoch": 0.02072, + "grad_norm": 0.6944630799653789, + "learning_rate": 0.003, + "loss": 4.1993, + "step": 2072 + }, + { + "epoch": 0.02073, + "grad_norm": 0.6945347127033654, + "learning_rate": 0.003, + "loss": 4.1809, + "step": 2073 + }, + { + "epoch": 0.02074, + "grad_norm": 0.7147074603428184, + "learning_rate": 0.003, + "loss": 4.1867, + "step": 2074 + }, + { + "epoch": 0.02075, + "grad_norm": 0.7883118765821364, + "learning_rate": 0.003, + "loss": 4.2193, + "step": 2075 + }, + { + "epoch": 0.02076, + "grad_norm": 0.7967778811959101, + "learning_rate": 0.003, + "loss": 4.211, + "step": 2076 + }, + { + "epoch": 0.02077, + "grad_norm": 0.8643539826729795, + "learning_rate": 0.003, + "loss": 4.1829, + "step": 2077 + }, + { + "epoch": 0.02078, + "grad_norm": 0.7306892024891165, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 2078 + }, + { + "epoch": 0.02079, + "grad_norm": 0.8002825789678965, + "learning_rate": 0.003, + "loss": 4.1635, + "step": 2079 + }, + { + "epoch": 0.0208, + "grad_norm": 0.8337706568911861, + "learning_rate": 0.003, + "loss": 4.1775, + "step": 2080 + }, + { + "epoch": 0.02081, + "grad_norm": 0.7760856311517963, + "learning_rate": 0.003, + "loss": 4.2099, + "step": 2081 + }, + { + "epoch": 0.02082, + "grad_norm": 0.7247815797190178, + "learning_rate": 0.003, + "loss": 4.2001, + "step": 2082 + }, + { + "epoch": 0.02083, + "grad_norm": 0.6465413633406625, + "learning_rate": 0.003, + "loss": 4.1935, + "step": 2083 + }, + { + "epoch": 0.02084, + "grad_norm": 0.5758163815437505, + "learning_rate": 0.003, + "loss": 4.1846, + "step": 2084 + }, + { + "epoch": 0.02085, + "grad_norm": 0.5794926828670885, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 2085 + }, + { + "epoch": 0.02086, + "grad_norm": 0.6037900992919937, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 2086 + }, + { + "epoch": 0.02087, + "grad_norm": 0.6387493020325943, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 2087 + }, + { + "epoch": 0.02088, + "grad_norm": 0.6388553503180331, + "learning_rate": 0.003, + "loss": 4.1981, + "step": 2088 + }, + { + "epoch": 0.02089, + "grad_norm": 0.6848861243568777, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 2089 + }, + { + "epoch": 0.0209, + "grad_norm": 0.708404587498897, + "learning_rate": 0.003, + "loss": 4.1867, + "step": 2090 + }, + { + "epoch": 0.02091, + "grad_norm": 0.6270385456303872, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 2091 + }, + { + "epoch": 0.02092, + "grad_norm": 0.651809660380216, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 2092 + }, + { + "epoch": 0.02093, + "grad_norm": 0.5981283591633111, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 2093 + }, + { + "epoch": 0.02094, + "grad_norm": 0.546273829694926, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 2094 + }, + { + "epoch": 0.02095, + "grad_norm": 0.5346067735053792, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 2095 + }, + { + "epoch": 0.02096, + "grad_norm": 0.5116949103009285, + "learning_rate": 0.003, + "loss": 4.174, + "step": 2096 + }, + { + "epoch": 0.02097, + "grad_norm": 0.5469361242716593, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2097 + }, + { + "epoch": 0.02098, + "grad_norm": 0.6062954162895189, + "learning_rate": 0.003, + "loss": 4.173, + "step": 2098 + }, + { + "epoch": 0.02099, + "grad_norm": 0.6421105100347297, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 2099 + }, + { + "epoch": 0.021, + "grad_norm": 0.7851067242648688, + "learning_rate": 0.003, + "loss": 4.2009, + "step": 2100 + }, + { + "epoch": 0.02101, + "grad_norm": 0.9400152889909587, + "learning_rate": 0.003, + "loss": 4.2058, + "step": 2101 + }, + { + "epoch": 0.02102, + "grad_norm": 0.8482284616580508, + "learning_rate": 0.003, + "loss": 4.2076, + "step": 2102 + }, + { + "epoch": 0.02103, + "grad_norm": 0.669340434796385, + "learning_rate": 0.003, + "loss": 4.1587, + "step": 2103 + }, + { + "epoch": 0.02104, + "grad_norm": 0.6570018762001313, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 2104 + }, + { + "epoch": 0.02105, + "grad_norm": 0.6446826452345433, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 2105 + }, + { + "epoch": 0.02106, + "grad_norm": 0.6376558757445897, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 2106 + }, + { + "epoch": 0.02107, + "grad_norm": 0.5227976317469611, + "learning_rate": 0.003, + "loss": 4.1682, + "step": 2107 + }, + { + "epoch": 0.02108, + "grad_norm": 0.4773493930172518, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 2108 + }, + { + "epoch": 0.02109, + "grad_norm": 0.4542234337544374, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 2109 + }, + { + "epoch": 0.0211, + "grad_norm": 0.40089709667602264, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 2110 + }, + { + "epoch": 0.02111, + "grad_norm": 0.41517718675764137, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 2111 + }, + { + "epoch": 0.02112, + "grad_norm": 0.4272554935908712, + "learning_rate": 0.003, + "loss": 4.1595, + "step": 2112 + }, + { + "epoch": 0.02113, + "grad_norm": 0.4355570629960646, + "learning_rate": 0.003, + "loss": 4.156, + "step": 2113 + }, + { + "epoch": 0.02114, + "grad_norm": 0.5048950749068443, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 2114 + }, + { + "epoch": 0.02115, + "grad_norm": 0.5652295254722385, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 2115 + }, + { + "epoch": 0.02116, + "grad_norm": 0.7232485908783208, + "learning_rate": 0.003, + "loss": 4.175, + "step": 2116 + }, + { + "epoch": 0.02117, + "grad_norm": 0.8906364224969214, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 2117 + }, + { + "epoch": 0.02118, + "grad_norm": 0.9352235963796359, + "learning_rate": 0.003, + "loss": 4.1793, + "step": 2118 + }, + { + "epoch": 0.02119, + "grad_norm": 0.9941756204818011, + "learning_rate": 0.003, + "loss": 4.1936, + "step": 2119 + }, + { + "epoch": 0.0212, + "grad_norm": 0.8089475589381471, + "learning_rate": 0.003, + "loss": 4.1659, + "step": 2120 + }, + { + "epoch": 0.02121, + "grad_norm": 0.7941200832800023, + "learning_rate": 0.003, + "loss": 4.1566, + "step": 2121 + }, + { + "epoch": 0.02122, + "grad_norm": 0.8852375528823769, + "learning_rate": 0.003, + "loss": 4.1755, + "step": 2122 + }, + { + "epoch": 0.02123, + "grad_norm": 0.9310836249103811, + "learning_rate": 0.003, + "loss": 4.162, + "step": 2123 + }, + { + "epoch": 0.02124, + "grad_norm": 0.782140762136661, + "learning_rate": 0.003, + "loss": 4.176, + "step": 2124 + }, + { + "epoch": 0.02125, + "grad_norm": 0.6927282600457383, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 2125 + }, + { + "epoch": 0.02126, + "grad_norm": 0.7154365490953082, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2126 + }, + { + "epoch": 0.02127, + "grad_norm": 0.6523159147762717, + "learning_rate": 0.003, + "loss": 4.1916, + "step": 2127 + }, + { + "epoch": 0.02128, + "grad_norm": 0.6320464529967942, + "learning_rate": 0.003, + "loss": 4.182, + "step": 2128 + }, + { + "epoch": 0.02129, + "grad_norm": 0.6885134025293627, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 2129 + }, + { + "epoch": 0.0213, + "grad_norm": 0.7482869835262143, + "learning_rate": 0.003, + "loss": 4.1579, + "step": 2130 + }, + { + "epoch": 0.02131, + "grad_norm": 0.7264411276151157, + "learning_rate": 0.003, + "loss": 4.1855, + "step": 2131 + }, + { + "epoch": 0.02132, + "grad_norm": 0.6880950136173917, + "learning_rate": 0.003, + "loss": 4.1935, + "step": 2132 + }, + { + "epoch": 0.02133, + "grad_norm": 0.5261874570457686, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 2133 + }, + { + "epoch": 0.02134, + "grad_norm": 0.5315464263746116, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 2134 + }, + { + "epoch": 0.02135, + "grad_norm": 0.43923091756094446, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 2135 + }, + { + "epoch": 0.02136, + "grad_norm": 0.48700083119140153, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 2136 + }, + { + "epoch": 0.02137, + "grad_norm": 0.5126955001612766, + "learning_rate": 0.003, + "loss": 4.1858, + "step": 2137 + }, + { + "epoch": 0.02138, + "grad_norm": 0.6176037320363783, + "learning_rate": 0.003, + "loss": 4.1995, + "step": 2138 + }, + { + "epoch": 0.02139, + "grad_norm": 0.7208235487909802, + "learning_rate": 0.003, + "loss": 4.1899, + "step": 2139 + }, + { + "epoch": 0.0214, + "grad_norm": 0.7707292730061747, + "learning_rate": 0.003, + "loss": 4.1913, + "step": 2140 + }, + { + "epoch": 0.02141, + "grad_norm": 0.7064558422533533, + "learning_rate": 0.003, + "loss": 4.167, + "step": 2141 + }, + { + "epoch": 0.02142, + "grad_norm": 0.6923504671414542, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 2142 + }, + { + "epoch": 0.02143, + "grad_norm": 0.6551621732834115, + "learning_rate": 0.003, + "loss": 4.1739, + "step": 2143 + }, + { + "epoch": 0.02144, + "grad_norm": 0.5959473459562541, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 2144 + }, + { + "epoch": 0.02145, + "grad_norm": 0.5646977797207685, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 2145 + }, + { + "epoch": 0.02146, + "grad_norm": 0.5906862377141963, + "learning_rate": 0.003, + "loss": 4.1686, + "step": 2146 + }, + { + "epoch": 0.02147, + "grad_norm": 0.6186164788167312, + "learning_rate": 0.003, + "loss": 4.1908, + "step": 2147 + }, + { + "epoch": 0.02148, + "grad_norm": 0.6604588490147959, + "learning_rate": 0.003, + "loss": 4.2035, + "step": 2148 + }, + { + "epoch": 0.02149, + "grad_norm": 0.7223218056273694, + "learning_rate": 0.003, + "loss": 4.1886, + "step": 2149 + }, + { + "epoch": 0.0215, + "grad_norm": 0.6818732999516106, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 2150 + }, + { + "epoch": 0.02151, + "grad_norm": 0.8143587891236486, + "learning_rate": 0.003, + "loss": 4.1862, + "step": 2151 + }, + { + "epoch": 0.02152, + "grad_norm": 0.9437946413521845, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 2152 + }, + { + "epoch": 0.02153, + "grad_norm": 0.8846561798234059, + "learning_rate": 0.003, + "loss": 4.1795, + "step": 2153 + }, + { + "epoch": 0.02154, + "grad_norm": 0.9840455403840825, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 2154 + }, + { + "epoch": 0.02155, + "grad_norm": 0.868802506205752, + "learning_rate": 0.003, + "loss": 4.2039, + "step": 2155 + }, + { + "epoch": 0.02156, + "grad_norm": 0.8265645662347763, + "learning_rate": 0.003, + "loss": 4.1871, + "step": 2156 + }, + { + "epoch": 0.02157, + "grad_norm": 0.8098367128696027, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 2157 + }, + { + "epoch": 0.02158, + "grad_norm": 0.7336411646083258, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 2158 + }, + { + "epoch": 0.02159, + "grad_norm": 0.7414314515769056, + "learning_rate": 0.003, + "loss": 4.191, + "step": 2159 + }, + { + "epoch": 0.0216, + "grad_norm": 0.789371817110674, + "learning_rate": 0.003, + "loss": 4.1969, + "step": 2160 + }, + { + "epoch": 0.02161, + "grad_norm": 0.869630636766529, + "learning_rate": 0.003, + "loss": 4.1914, + "step": 2161 + }, + { + "epoch": 0.02162, + "grad_norm": 0.8529653643448841, + "learning_rate": 0.003, + "loss": 4.1965, + "step": 2162 + }, + { + "epoch": 0.02163, + "grad_norm": 0.8494552625345028, + "learning_rate": 0.003, + "loss": 4.2103, + "step": 2163 + }, + { + "epoch": 0.02164, + "grad_norm": 0.6854506453174236, + "learning_rate": 0.003, + "loss": 4.199, + "step": 2164 + }, + { + "epoch": 0.02165, + "grad_norm": 0.6874765295980679, + "learning_rate": 0.003, + "loss": 4.1608, + "step": 2165 + }, + { + "epoch": 0.02166, + "grad_norm": 0.6729378799675353, + "learning_rate": 0.003, + "loss": 4.2004, + "step": 2166 + }, + { + "epoch": 0.02167, + "grad_norm": 0.6549841279194847, + "learning_rate": 0.003, + "loss": 4.1756, + "step": 2167 + }, + { + "epoch": 0.02168, + "grad_norm": 0.6330056563132089, + "learning_rate": 0.003, + "loss": 4.1725, + "step": 2168 + }, + { + "epoch": 0.02169, + "grad_norm": 0.6890647234914371, + "learning_rate": 0.003, + "loss": 4.1729, + "step": 2169 + }, + { + "epoch": 0.0217, + "grad_norm": 0.7156443468233303, + "learning_rate": 0.003, + "loss": 4.1822, + "step": 2170 + }, + { + "epoch": 0.02171, + "grad_norm": 0.819425074049774, + "learning_rate": 0.003, + "loss": 4.1983, + "step": 2171 + }, + { + "epoch": 0.02172, + "grad_norm": 0.954379072785258, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 2172 + }, + { + "epoch": 0.02173, + "grad_norm": 0.8792451322567497, + "learning_rate": 0.003, + "loss": 4.1778, + "step": 2173 + }, + { + "epoch": 0.02174, + "grad_norm": 0.8645953746444872, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 2174 + }, + { + "epoch": 0.02175, + "grad_norm": 0.7880189873567361, + "learning_rate": 0.003, + "loss": 4.1986, + "step": 2175 + }, + { + "epoch": 0.02176, + "grad_norm": 0.8439689409908241, + "learning_rate": 0.003, + "loss": 4.1833, + "step": 2176 + }, + { + "epoch": 0.02177, + "grad_norm": 0.7775791439063611, + "learning_rate": 0.003, + "loss": 4.1857, + "step": 2177 + }, + { + "epoch": 0.02178, + "grad_norm": 0.7349880561182358, + "learning_rate": 0.003, + "loss": 4.1695, + "step": 2178 + }, + { + "epoch": 0.02179, + "grad_norm": 0.756761244769964, + "learning_rate": 0.003, + "loss": 4.1816, + "step": 2179 + }, + { + "epoch": 0.0218, + "grad_norm": 0.7656137924156894, + "learning_rate": 0.003, + "loss": 4.199, + "step": 2180 + }, + { + "epoch": 0.02181, + "grad_norm": 0.7429091674969758, + "learning_rate": 0.003, + "loss": 4.1962, + "step": 2181 + }, + { + "epoch": 0.02182, + "grad_norm": 0.6661606742959821, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 2182 + }, + { + "epoch": 0.02183, + "grad_norm": 0.585479805956627, + "learning_rate": 0.003, + "loss": 4.1874, + "step": 2183 + }, + { + "epoch": 0.02184, + "grad_norm": 0.5163479427423384, + "learning_rate": 0.003, + "loss": 4.1599, + "step": 2184 + }, + { + "epoch": 0.02185, + "grad_norm": 0.5015989065878748, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 2185 + }, + { + "epoch": 0.02186, + "grad_norm": 0.4629145117923421, + "learning_rate": 0.003, + "loss": 4.1697, + "step": 2186 + }, + { + "epoch": 0.02187, + "grad_norm": 0.4658350248376296, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 2187 + }, + { + "epoch": 0.02188, + "grad_norm": 0.5035659083342326, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 2188 + }, + { + "epoch": 0.02189, + "grad_norm": 0.6089757052974601, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 2189 + }, + { + "epoch": 0.0219, + "grad_norm": 0.6207116226673802, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 2190 + }, + { + "epoch": 0.02191, + "grad_norm": 0.5519445812257309, + "learning_rate": 0.003, + "loss": 4.1582, + "step": 2191 + }, + { + "epoch": 0.02192, + "grad_norm": 0.5165727484783152, + "learning_rate": 0.003, + "loss": 4.1868, + "step": 2192 + }, + { + "epoch": 0.02193, + "grad_norm": 0.5613296796670594, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 2193 + }, + { + "epoch": 0.02194, + "grad_norm": 0.6448991178014012, + "learning_rate": 0.003, + "loss": 4.2021, + "step": 2194 + }, + { + "epoch": 0.02195, + "grad_norm": 0.8306209034791217, + "learning_rate": 0.003, + "loss": 4.1742, + "step": 2195 + }, + { + "epoch": 0.02196, + "grad_norm": 0.8976316389921701, + "learning_rate": 0.003, + "loss": 4.191, + "step": 2196 + }, + { + "epoch": 0.02197, + "grad_norm": 0.7626102135855414, + "learning_rate": 0.003, + "loss": 4.1794, + "step": 2197 + }, + { + "epoch": 0.02198, + "grad_norm": 0.6813540276229926, + "learning_rate": 0.003, + "loss": 4.166, + "step": 2198 + }, + { + "epoch": 0.02199, + "grad_norm": 0.6831695997532532, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 2199 + }, + { + "epoch": 0.022, + "grad_norm": 0.7020969843271442, + "learning_rate": 0.003, + "loss": 4.1643, + "step": 2200 + }, + { + "epoch": 0.02201, + "grad_norm": 0.7491568432192739, + "learning_rate": 0.003, + "loss": 4.1801, + "step": 2201 + }, + { + "epoch": 0.02202, + "grad_norm": 0.6710737182979849, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 2202 + }, + { + "epoch": 0.02203, + "grad_norm": 0.5593268037532971, + "learning_rate": 0.003, + "loss": 4.1894, + "step": 2203 + }, + { + "epoch": 0.02204, + "grad_norm": 0.6418879951190652, + "learning_rate": 0.003, + "loss": 4.1849, + "step": 2204 + }, + { + "epoch": 0.02205, + "grad_norm": 0.6552738463900842, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2205 + }, + { + "epoch": 0.02206, + "grad_norm": 0.6380810530743423, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 2206 + }, + { + "epoch": 0.02207, + "grad_norm": 0.7692413477540456, + "learning_rate": 0.003, + "loss": 4.1717, + "step": 2207 + }, + { + "epoch": 0.02208, + "grad_norm": 0.8957310934771069, + "learning_rate": 0.003, + "loss": 4.157, + "step": 2208 + }, + { + "epoch": 0.02209, + "grad_norm": 0.9998279344279337, + "learning_rate": 0.003, + "loss": 4.1843, + "step": 2209 + }, + { + "epoch": 0.0221, + "grad_norm": 0.8752638810520407, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 2210 + }, + { + "epoch": 0.02211, + "grad_norm": 0.8582513198506979, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 2211 + }, + { + "epoch": 0.02212, + "grad_norm": 0.8651183213085927, + "learning_rate": 0.003, + "loss": 4.1758, + "step": 2212 + }, + { + "epoch": 0.02213, + "grad_norm": 0.758173352380571, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 2213 + }, + { + "epoch": 0.02214, + "grad_norm": 0.7923595762527091, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 2214 + }, + { + "epoch": 0.02215, + "grad_norm": 0.8522282518089913, + "learning_rate": 0.003, + "loss": 4.1811, + "step": 2215 + }, + { + "epoch": 0.02216, + "grad_norm": 0.7529422856661093, + "learning_rate": 0.003, + "loss": 4.1577, + "step": 2216 + }, + { + "epoch": 0.02217, + "grad_norm": 0.7502334626130458, + "learning_rate": 0.003, + "loss": 4.1565, + "step": 2217 + }, + { + "epoch": 0.02218, + "grad_norm": 0.7116888337918957, + "learning_rate": 0.003, + "loss": 4.185, + "step": 2218 + }, + { + "epoch": 0.02219, + "grad_norm": 0.6534101315686133, + "learning_rate": 0.003, + "loss": 4.1743, + "step": 2219 + }, + { + "epoch": 0.0222, + "grad_norm": 0.6567704163897022, + "learning_rate": 0.003, + "loss": 4.1717, + "step": 2220 + }, + { + "epoch": 0.02221, + "grad_norm": 0.5723395032111224, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 2221 + }, + { + "epoch": 0.02222, + "grad_norm": 0.5681850083902724, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2222 + }, + { + "epoch": 0.02223, + "grad_norm": 0.6564825959145019, + "learning_rate": 0.003, + "loss": 4.1724, + "step": 2223 + }, + { + "epoch": 0.02224, + "grad_norm": 0.712683429040134, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2224 + }, + { + "epoch": 0.02225, + "grad_norm": 0.6849747924946487, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 2225 + }, + { + "epoch": 0.02226, + "grad_norm": 0.6540087066775325, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 2226 + }, + { + "epoch": 0.02227, + "grad_norm": 0.6826292383532007, + "learning_rate": 0.003, + "loss": 4.166, + "step": 2227 + }, + { + "epoch": 0.02228, + "grad_norm": 0.5799818830839, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 2228 + }, + { + "epoch": 0.02229, + "grad_norm": 0.5894157536297949, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 2229 + }, + { + "epoch": 0.0223, + "grad_norm": 0.6245037255354391, + "learning_rate": 0.003, + "loss": 4.1698, + "step": 2230 + }, + { + "epoch": 0.02231, + "grad_norm": 0.6677005676334468, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 2231 + }, + { + "epoch": 0.02232, + "grad_norm": 0.6309595182965612, + "learning_rate": 0.003, + "loss": 4.1818, + "step": 2232 + }, + { + "epoch": 0.02233, + "grad_norm": 0.5749001820910085, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 2233 + }, + { + "epoch": 0.02234, + "grad_norm": 0.5771155284341811, + "learning_rate": 0.003, + "loss": 4.1691, + "step": 2234 + }, + { + "epoch": 0.02235, + "grad_norm": 0.6588273696821386, + "learning_rate": 0.003, + "loss": 4.196, + "step": 2235 + }, + { + "epoch": 0.02236, + "grad_norm": 0.7590132381295709, + "learning_rate": 0.003, + "loss": 4.1516, + "step": 2236 + }, + { + "epoch": 0.02237, + "grad_norm": 0.7623151027362074, + "learning_rate": 0.003, + "loss": 4.1875, + "step": 2237 + }, + { + "epoch": 0.02238, + "grad_norm": 0.7212724546890452, + "learning_rate": 0.003, + "loss": 4.1649, + "step": 2238 + }, + { + "epoch": 0.02239, + "grad_norm": 0.6076067350982214, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 2239 + }, + { + "epoch": 0.0224, + "grad_norm": 0.5613798169152191, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 2240 + }, + { + "epoch": 0.02241, + "grad_norm": 0.6354864223953033, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 2241 + }, + { + "epoch": 0.02242, + "grad_norm": 0.7026833475134019, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 2242 + }, + { + "epoch": 0.02243, + "grad_norm": 0.8203638094834097, + "learning_rate": 0.003, + "loss": 4.1853, + "step": 2243 + }, + { + "epoch": 0.02244, + "grad_norm": 0.8513655326229856, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 2244 + }, + { + "epoch": 0.02245, + "grad_norm": 0.8797653404012851, + "learning_rate": 0.003, + "loss": 4.1557, + "step": 2245 + }, + { + "epoch": 0.02246, + "grad_norm": 0.8294187018525063, + "learning_rate": 0.003, + "loss": 4.1865, + "step": 2246 + }, + { + "epoch": 0.02247, + "grad_norm": 0.7912400416148854, + "learning_rate": 0.003, + "loss": 4.156, + "step": 2247 + }, + { + "epoch": 0.02248, + "grad_norm": 0.7049105418667632, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 2248 + }, + { + "epoch": 0.02249, + "grad_norm": 0.5951151979412677, + "learning_rate": 0.003, + "loss": 4.1518, + "step": 2249 + }, + { + "epoch": 0.0225, + "grad_norm": 0.6395742460803966, + "learning_rate": 0.003, + "loss": 4.1675, + "step": 2250 + }, + { + "epoch": 0.02251, + "grad_norm": 0.6467480978867014, + "learning_rate": 0.003, + "loss": 4.1917, + "step": 2251 + }, + { + "epoch": 0.02252, + "grad_norm": 0.6689172158571011, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 2252 + }, + { + "epoch": 0.02253, + "grad_norm": 0.7398444104304565, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 2253 + }, + { + "epoch": 0.02254, + "grad_norm": 0.8101086781568522, + "learning_rate": 0.003, + "loss": 4.1752, + "step": 2254 + }, + { + "epoch": 0.02255, + "grad_norm": 0.7085844665571392, + "learning_rate": 0.003, + "loss": 4.157, + "step": 2255 + }, + { + "epoch": 0.02256, + "grad_norm": 0.6026772397946912, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 2256 + }, + { + "epoch": 0.02257, + "grad_norm": 0.6189957040838983, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 2257 + }, + { + "epoch": 0.02258, + "grad_norm": 0.6802591452330198, + "learning_rate": 0.003, + "loss": 4.1693, + "step": 2258 + }, + { + "epoch": 0.02259, + "grad_norm": 0.6338285955929223, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 2259 + }, + { + "epoch": 0.0226, + "grad_norm": 0.6760560537016971, + "learning_rate": 0.003, + "loss": 4.1871, + "step": 2260 + }, + { + "epoch": 0.02261, + "grad_norm": 0.6563449400208117, + "learning_rate": 0.003, + "loss": 4.162, + "step": 2261 + }, + { + "epoch": 0.02262, + "grad_norm": 0.6030476007619959, + "learning_rate": 0.003, + "loss": 4.161, + "step": 2262 + }, + { + "epoch": 0.02263, + "grad_norm": 0.7255721325380844, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 2263 + }, + { + "epoch": 0.02264, + "grad_norm": 0.7617708920633665, + "learning_rate": 0.003, + "loss": 4.171, + "step": 2264 + }, + { + "epoch": 0.02265, + "grad_norm": 0.9153277292381196, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 2265 + }, + { + "epoch": 0.02266, + "grad_norm": 0.9998858504049777, + "learning_rate": 0.003, + "loss": 4.1601, + "step": 2266 + }, + { + "epoch": 0.02267, + "grad_norm": 0.9240396378434623, + "learning_rate": 0.003, + "loss": 4.1868, + "step": 2267 + }, + { + "epoch": 0.02268, + "grad_norm": 0.761714877605015, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 2268 + }, + { + "epoch": 0.02269, + "grad_norm": 0.8730845867006987, + "learning_rate": 0.003, + "loss": 4.1842, + "step": 2269 + }, + { + "epoch": 0.0227, + "grad_norm": 0.6840331465080978, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 2270 + }, + { + "epoch": 0.02271, + "grad_norm": 0.7065260877420262, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 2271 + }, + { + "epoch": 0.02272, + "grad_norm": 0.7697104492178649, + "learning_rate": 0.003, + "loss": 4.1845, + "step": 2272 + }, + { + "epoch": 0.02273, + "grad_norm": 0.9089453564263758, + "learning_rate": 0.003, + "loss": 4.1879, + "step": 2273 + }, + { + "epoch": 0.02274, + "grad_norm": 0.922684666524011, + "learning_rate": 0.003, + "loss": 4.1771, + "step": 2274 + }, + { + "epoch": 0.02275, + "grad_norm": 0.6952806272792083, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 2275 + }, + { + "epoch": 0.02276, + "grad_norm": 0.6268140754122905, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 2276 + }, + { + "epoch": 0.02277, + "grad_norm": 0.5396215664772549, + "learning_rate": 0.003, + "loss": 4.162, + "step": 2277 + }, + { + "epoch": 0.02278, + "grad_norm": 0.5588177104516212, + "learning_rate": 0.003, + "loss": 4.1678, + "step": 2278 + }, + { + "epoch": 0.02279, + "grad_norm": 0.4777888535712354, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 2279 + }, + { + "epoch": 0.0228, + "grad_norm": 0.4626896091488448, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 2280 + }, + { + "epoch": 0.02281, + "grad_norm": 0.5161557902540315, + "learning_rate": 0.003, + "loss": 4.1604, + "step": 2281 + }, + { + "epoch": 0.02282, + "grad_norm": 0.5986868645644398, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 2282 + }, + { + "epoch": 0.02283, + "grad_norm": 0.7001958546215141, + "learning_rate": 0.003, + "loss": 4.1896, + "step": 2283 + }, + { + "epoch": 0.02284, + "grad_norm": 0.6229146935433538, + "learning_rate": 0.003, + "loss": 4.1839, + "step": 2284 + }, + { + "epoch": 0.02285, + "grad_norm": 0.52064610524233, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 2285 + }, + { + "epoch": 0.02286, + "grad_norm": 0.5472583313274128, + "learning_rate": 0.003, + "loss": 4.1836, + "step": 2286 + }, + { + "epoch": 0.02287, + "grad_norm": 0.43588049449954086, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 2287 + }, + { + "epoch": 0.02288, + "grad_norm": 0.4488910315034822, + "learning_rate": 0.003, + "loss": 4.182, + "step": 2288 + }, + { + "epoch": 0.02289, + "grad_norm": 0.4606248937804462, + "learning_rate": 0.003, + "loss": 4.1819, + "step": 2289 + }, + { + "epoch": 0.0229, + "grad_norm": 0.4519060586440639, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 2290 + }, + { + "epoch": 0.02291, + "grad_norm": 0.5302305524974421, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 2291 + }, + { + "epoch": 0.02292, + "grad_norm": 0.5642464770993318, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 2292 + }, + { + "epoch": 0.02293, + "grad_norm": 0.6841533454383677, + "learning_rate": 0.003, + "loss": 4.139, + "step": 2293 + }, + { + "epoch": 0.02294, + "grad_norm": 0.8504015466478836, + "learning_rate": 0.003, + "loss": 4.1521, + "step": 2294 + }, + { + "epoch": 0.02295, + "grad_norm": 0.8252843562391086, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 2295 + }, + { + "epoch": 0.02296, + "grad_norm": 0.665213160185038, + "learning_rate": 0.003, + "loss": 4.1746, + "step": 2296 + }, + { + "epoch": 0.02297, + "grad_norm": 0.6761854492753422, + "learning_rate": 0.003, + "loss": 4.1483, + "step": 2297 + }, + { + "epoch": 0.02298, + "grad_norm": 0.6817380369476475, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 2298 + }, + { + "epoch": 0.02299, + "grad_norm": 0.557025913483661, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 2299 + }, + { + "epoch": 0.023, + "grad_norm": 0.5203637909333125, + "learning_rate": 0.003, + "loss": 4.1751, + "step": 2300 + }, + { + "epoch": 0.02301, + "grad_norm": 0.5702409842972981, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 2301 + }, + { + "epoch": 0.02302, + "grad_norm": 0.6139587698754356, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 2302 + }, + { + "epoch": 0.02303, + "grad_norm": 0.6299069266162713, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 2303 + }, + { + "epoch": 0.02304, + "grad_norm": 0.7175990299856692, + "learning_rate": 0.003, + "loss": 4.1735, + "step": 2304 + }, + { + "epoch": 0.02305, + "grad_norm": 0.7179844195837081, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 2305 + }, + { + "epoch": 0.02306, + "grad_norm": 0.7602689533392006, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 2306 + }, + { + "epoch": 0.02307, + "grad_norm": 0.827690043660852, + "learning_rate": 0.003, + "loss": 4.1623, + "step": 2307 + }, + { + "epoch": 0.02308, + "grad_norm": 0.904152647018861, + "learning_rate": 0.003, + "loss": 4.1622, + "step": 2308 + }, + { + "epoch": 0.02309, + "grad_norm": 0.8427659380267021, + "learning_rate": 0.003, + "loss": 4.1894, + "step": 2309 + }, + { + "epoch": 0.0231, + "grad_norm": 0.7985394811691259, + "learning_rate": 0.003, + "loss": 4.1717, + "step": 2310 + }, + { + "epoch": 0.02311, + "grad_norm": 0.691274929019855, + "learning_rate": 0.003, + "loss": 4.1708, + "step": 2311 + }, + { + "epoch": 0.02312, + "grad_norm": 0.7784301593593793, + "learning_rate": 0.003, + "loss": 4.1481, + "step": 2312 + }, + { + "epoch": 0.02313, + "grad_norm": 0.6869612271173453, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 2313 + }, + { + "epoch": 0.02314, + "grad_norm": 0.6829073789580666, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 2314 + }, + { + "epoch": 0.02315, + "grad_norm": 0.7074062741013188, + "learning_rate": 0.003, + "loss": 4.1625, + "step": 2315 + }, + { + "epoch": 0.02316, + "grad_norm": 0.680387274486601, + "learning_rate": 0.003, + "loss": 4.1746, + "step": 2316 + }, + { + "epoch": 0.02317, + "grad_norm": 0.7792707849212159, + "learning_rate": 0.003, + "loss": 4.1594, + "step": 2317 + }, + { + "epoch": 0.02318, + "grad_norm": 0.8226437095400972, + "learning_rate": 0.003, + "loss": 4.1513, + "step": 2318 + }, + { + "epoch": 0.02319, + "grad_norm": 0.6715795313728831, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2319 + }, + { + "epoch": 0.0232, + "grad_norm": 0.6878542404220956, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 2320 + }, + { + "epoch": 0.02321, + "grad_norm": 0.7111510099511913, + "learning_rate": 0.003, + "loss": 4.159, + "step": 2321 + }, + { + "epoch": 0.02322, + "grad_norm": 0.7818521424895224, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 2322 + }, + { + "epoch": 0.02323, + "grad_norm": 0.8314457265055445, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 2323 + }, + { + "epoch": 0.02324, + "grad_norm": 0.8120222680920612, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 2324 + }, + { + "epoch": 0.02325, + "grad_norm": 0.8662835036259435, + "learning_rate": 0.003, + "loss": 4.1793, + "step": 2325 + }, + { + "epoch": 0.02326, + "grad_norm": 0.8953930566671354, + "learning_rate": 0.003, + "loss": 4.1978, + "step": 2326 + }, + { + "epoch": 0.02327, + "grad_norm": 0.9057739156940444, + "learning_rate": 0.003, + "loss": 4.1866, + "step": 2327 + }, + { + "epoch": 0.02328, + "grad_norm": 0.9080610106189098, + "learning_rate": 0.003, + "loss": 4.1861, + "step": 2328 + }, + { + "epoch": 0.02329, + "grad_norm": 0.9179138451329095, + "learning_rate": 0.003, + "loss": 4.1777, + "step": 2329 + }, + { + "epoch": 0.0233, + "grad_norm": 0.9872631652325534, + "learning_rate": 0.003, + "loss": 4.1761, + "step": 2330 + }, + { + "epoch": 0.02331, + "grad_norm": 0.8876721352093778, + "learning_rate": 0.003, + "loss": 4.2133, + "step": 2331 + }, + { + "epoch": 0.02332, + "grad_norm": 0.8690706198673939, + "learning_rate": 0.003, + "loss": 4.1762, + "step": 2332 + }, + { + "epoch": 0.02333, + "grad_norm": 0.7331697884498708, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 2333 + }, + { + "epoch": 0.02334, + "grad_norm": 0.7269136503407467, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 2334 + }, + { + "epoch": 0.02335, + "grad_norm": 0.7243967687261031, + "learning_rate": 0.003, + "loss": 4.1839, + "step": 2335 + }, + { + "epoch": 0.02336, + "grad_norm": 0.6334493981917083, + "learning_rate": 0.003, + "loss": 4.1736, + "step": 2336 + }, + { + "epoch": 0.02337, + "grad_norm": 0.5360700841015614, + "learning_rate": 0.003, + "loss": 4.1731, + "step": 2337 + }, + { + "epoch": 0.02338, + "grad_norm": 0.47513859907874345, + "learning_rate": 0.003, + "loss": 4.1569, + "step": 2338 + }, + { + "epoch": 0.02339, + "grad_norm": 0.48845559064930505, + "learning_rate": 0.003, + "loss": 4.154, + "step": 2339 + }, + { + "epoch": 0.0234, + "grad_norm": 0.5049304676560479, + "learning_rate": 0.003, + "loss": 4.164, + "step": 2340 + }, + { + "epoch": 0.02341, + "grad_norm": 0.49444039687567853, + "learning_rate": 0.003, + "loss": 4.138, + "step": 2341 + }, + { + "epoch": 0.02342, + "grad_norm": 0.44971202039339475, + "learning_rate": 0.003, + "loss": 4.151, + "step": 2342 + }, + { + "epoch": 0.02343, + "grad_norm": 0.4735761548639779, + "learning_rate": 0.003, + "loss": 4.1491, + "step": 2343 + }, + { + "epoch": 0.02344, + "grad_norm": 0.4908191136535766, + "learning_rate": 0.003, + "loss": 4.1467, + "step": 2344 + }, + { + "epoch": 0.02345, + "grad_norm": 0.5527599395986549, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 2345 + }, + { + "epoch": 0.02346, + "grad_norm": 0.6231997069030808, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 2346 + }, + { + "epoch": 0.02347, + "grad_norm": 0.6858961367895726, + "learning_rate": 0.003, + "loss": 4.1815, + "step": 2347 + }, + { + "epoch": 0.02348, + "grad_norm": 0.8108937526078669, + "learning_rate": 0.003, + "loss": 4.1523, + "step": 2348 + }, + { + "epoch": 0.02349, + "grad_norm": 1.0585674817855537, + "learning_rate": 0.003, + "loss": 4.1805, + "step": 2349 + }, + { + "epoch": 0.0235, + "grad_norm": 0.8845602959741193, + "learning_rate": 0.003, + "loss": 4.1759, + "step": 2350 + }, + { + "epoch": 0.02351, + "grad_norm": 0.6996358412000482, + "learning_rate": 0.003, + "loss": 4.1782, + "step": 2351 + }, + { + "epoch": 0.02352, + "grad_norm": 0.7420583586925507, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 2352 + }, + { + "epoch": 0.02353, + "grad_norm": 0.7172109070601732, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 2353 + }, + { + "epoch": 0.02354, + "grad_norm": 0.5964437578409529, + "learning_rate": 0.003, + "loss": 4.1687, + "step": 2354 + }, + { + "epoch": 0.02355, + "grad_norm": 0.5517509795400629, + "learning_rate": 0.003, + "loss": 4.1631, + "step": 2355 + }, + { + "epoch": 0.02356, + "grad_norm": 0.5364645866463458, + "learning_rate": 0.003, + "loss": 4.144, + "step": 2356 + }, + { + "epoch": 0.02357, + "grad_norm": 0.523366916844322, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 2357 + }, + { + "epoch": 0.02358, + "grad_norm": 0.46650318458047774, + "learning_rate": 0.003, + "loss": 4.1618, + "step": 2358 + }, + { + "epoch": 0.02359, + "grad_norm": 0.4588225253606432, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 2359 + }, + { + "epoch": 0.0236, + "grad_norm": 0.39104184932360553, + "learning_rate": 0.003, + "loss": 4.1744, + "step": 2360 + }, + { + "epoch": 0.02361, + "grad_norm": 0.41200606597184536, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 2361 + }, + { + "epoch": 0.02362, + "grad_norm": 0.44600201043530324, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 2362 + }, + { + "epoch": 0.02363, + "grad_norm": 0.5475729333295288, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 2363 + }, + { + "epoch": 0.02364, + "grad_norm": 0.6388100152040984, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 2364 + }, + { + "epoch": 0.02365, + "grad_norm": 0.6577283287258094, + "learning_rate": 0.003, + "loss": 4.1667, + "step": 2365 + }, + { + "epoch": 0.02366, + "grad_norm": 0.648157002451967, + "learning_rate": 0.003, + "loss": 4.1399, + "step": 2366 + }, + { + "epoch": 0.02367, + "grad_norm": 0.7087349193313575, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 2367 + }, + { + "epoch": 0.02368, + "grad_norm": 0.7860836539271487, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 2368 + }, + { + "epoch": 0.02369, + "grad_norm": 0.7250688611440719, + "learning_rate": 0.003, + "loss": 4.139, + "step": 2369 + }, + { + "epoch": 0.0237, + "grad_norm": 0.7397920357739604, + "learning_rate": 0.003, + "loss": 4.1687, + "step": 2370 + }, + { + "epoch": 0.02371, + "grad_norm": 0.6128692264075699, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 2371 + }, + { + "epoch": 0.02372, + "grad_norm": 0.6087414784500429, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 2372 + }, + { + "epoch": 0.02373, + "grad_norm": 0.5756089927991157, + "learning_rate": 0.003, + "loss": 4.127, + "step": 2373 + }, + { + "epoch": 0.02374, + "grad_norm": 0.5497734986819502, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 2374 + }, + { + "epoch": 0.02375, + "grad_norm": 0.6407324880159457, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 2375 + }, + { + "epoch": 0.02376, + "grad_norm": 0.7947750158067093, + "learning_rate": 0.003, + "loss": 4.1704, + "step": 2376 + }, + { + "epoch": 0.02377, + "grad_norm": 0.9624081040845262, + "learning_rate": 0.003, + "loss": 4.1477, + "step": 2377 + }, + { + "epoch": 0.02378, + "grad_norm": 0.9490289584100083, + "learning_rate": 0.003, + "loss": 4.1578, + "step": 2378 + }, + { + "epoch": 0.02379, + "grad_norm": 1.0177225731187802, + "learning_rate": 0.003, + "loss": 4.1529, + "step": 2379 + }, + { + "epoch": 0.0238, + "grad_norm": 0.9212562993006215, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 2380 + }, + { + "epoch": 0.02381, + "grad_norm": 1.0180830178551743, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 2381 + }, + { + "epoch": 0.02382, + "grad_norm": 1.1887638352290792, + "learning_rate": 0.003, + "loss": 4.2208, + "step": 2382 + }, + { + "epoch": 0.02383, + "grad_norm": 0.9623142745402433, + "learning_rate": 0.003, + "loss": 4.175, + "step": 2383 + }, + { + "epoch": 0.02384, + "grad_norm": 0.9996485767059176, + "learning_rate": 0.003, + "loss": 4.1655, + "step": 2384 + }, + { + "epoch": 0.02385, + "grad_norm": 1.0005230697453882, + "learning_rate": 0.003, + "loss": 4.1971, + "step": 2385 + }, + { + "epoch": 0.02386, + "grad_norm": 0.8991298449949392, + "learning_rate": 0.003, + "loss": 4.1664, + "step": 2386 + }, + { + "epoch": 0.02387, + "grad_norm": 0.9140537337734469, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 2387 + }, + { + "epoch": 0.02388, + "grad_norm": 0.8457102462087976, + "learning_rate": 0.003, + "loss": 4.221, + "step": 2388 + }, + { + "epoch": 0.02389, + "grad_norm": 0.7797281022177047, + "learning_rate": 0.003, + "loss": 4.1797, + "step": 2389 + }, + { + "epoch": 0.0239, + "grad_norm": 0.8105447862252997, + "learning_rate": 0.003, + "loss": 4.1733, + "step": 2390 + }, + { + "epoch": 0.02391, + "grad_norm": 0.7879965131242213, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 2391 + }, + { + "epoch": 0.02392, + "grad_norm": 0.7432570244684692, + "learning_rate": 0.003, + "loss": 4.1961, + "step": 2392 + }, + { + "epoch": 0.02393, + "grad_norm": 0.8242151325256147, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 2393 + }, + { + "epoch": 0.02394, + "grad_norm": 0.6795475929865452, + "learning_rate": 0.003, + "loss": 4.1668, + "step": 2394 + }, + { + "epoch": 0.02395, + "grad_norm": 0.6500813135710026, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 2395 + }, + { + "epoch": 0.02396, + "grad_norm": 0.7072760557799878, + "learning_rate": 0.003, + "loss": 4.1941, + "step": 2396 + }, + { + "epoch": 0.02397, + "grad_norm": 0.7310301965212079, + "learning_rate": 0.003, + "loss": 4.1748, + "step": 2397 + }, + { + "epoch": 0.02398, + "grad_norm": 0.7234811825265847, + "learning_rate": 0.003, + "loss": 4.1805, + "step": 2398 + }, + { + "epoch": 0.02399, + "grad_norm": 0.6512343977979285, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 2399 + }, + { + "epoch": 0.024, + "grad_norm": 0.5558371506699439, + "learning_rate": 0.003, + "loss": 4.1685, + "step": 2400 + }, + { + "epoch": 0.02401, + "grad_norm": 0.4681876356369122, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 2401 + }, + { + "epoch": 0.02402, + "grad_norm": 0.47661723869702477, + "learning_rate": 0.003, + "loss": 4.1712, + "step": 2402 + }, + { + "epoch": 0.02403, + "grad_norm": 0.4829559242535778, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 2403 + }, + { + "epoch": 0.02404, + "grad_norm": 0.6800724198255437, + "learning_rate": 0.003, + "loss": 4.1812, + "step": 2404 + }, + { + "epoch": 0.02405, + "grad_norm": 0.9248642816365124, + "learning_rate": 0.003, + "loss": 4.1969, + "step": 2405 + }, + { + "epoch": 0.02406, + "grad_norm": 1.1727881472231674, + "learning_rate": 0.003, + "loss": 4.1905, + "step": 2406 + }, + { + "epoch": 0.02407, + "grad_norm": 0.7329741974809963, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 2407 + }, + { + "epoch": 0.02408, + "grad_norm": 0.7726138363997163, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 2408 + }, + { + "epoch": 0.02409, + "grad_norm": 0.719138696906034, + "learning_rate": 0.003, + "loss": 4.1669, + "step": 2409 + }, + { + "epoch": 0.0241, + "grad_norm": 0.555949652942558, + "learning_rate": 0.003, + "loss": 4.1732, + "step": 2410 + }, + { + "epoch": 0.02411, + "grad_norm": 0.5377697517451042, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 2411 + }, + { + "epoch": 0.02412, + "grad_norm": 0.6012288683334335, + "learning_rate": 0.003, + "loss": 4.147, + "step": 2412 + }, + { + "epoch": 0.02413, + "grad_norm": 0.5105596338171011, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 2413 + }, + { + "epoch": 0.02414, + "grad_norm": 0.485763479569989, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 2414 + }, + { + "epoch": 0.02415, + "grad_norm": 0.40662288301451616, + "learning_rate": 0.003, + "loss": 4.1487, + "step": 2415 + }, + { + "epoch": 0.02416, + "grad_norm": 0.3978272900525064, + "learning_rate": 0.003, + "loss": 4.1312, + "step": 2416 + }, + { + "epoch": 0.02417, + "grad_norm": 0.37265436578822503, + "learning_rate": 0.003, + "loss": 4.1575, + "step": 2417 + }, + { + "epoch": 0.02418, + "grad_norm": 0.36540896134976886, + "learning_rate": 0.003, + "loss": 4.148, + "step": 2418 + }, + { + "epoch": 0.02419, + "grad_norm": 0.37596634629156483, + "learning_rate": 0.003, + "loss": 4.1442, + "step": 2419 + }, + { + "epoch": 0.0242, + "grad_norm": 0.3802320259983771, + "learning_rate": 0.003, + "loss": 4.1567, + "step": 2420 + }, + { + "epoch": 0.02421, + "grad_norm": 0.41571053488619564, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 2421 + }, + { + "epoch": 0.02422, + "grad_norm": 0.4285818265856571, + "learning_rate": 0.003, + "loss": 4.173, + "step": 2422 + }, + { + "epoch": 0.02423, + "grad_norm": 0.48624061885426095, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 2423 + }, + { + "epoch": 0.02424, + "grad_norm": 0.5959185371409056, + "learning_rate": 0.003, + "loss": 4.1614, + "step": 2424 + }, + { + "epoch": 0.02425, + "grad_norm": 0.6229653459244197, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 2425 + }, + { + "epoch": 0.02426, + "grad_norm": 0.6411403339689572, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 2426 + }, + { + "epoch": 0.02427, + "grad_norm": 0.7189014576661708, + "learning_rate": 0.003, + "loss": 4.1827, + "step": 2427 + }, + { + "epoch": 0.02428, + "grad_norm": 0.7417243215195037, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 2428 + }, + { + "epoch": 0.02429, + "grad_norm": 0.7648380869850353, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 2429 + }, + { + "epoch": 0.0243, + "grad_norm": 0.830483868233713, + "learning_rate": 0.003, + "loss": 4.1745, + "step": 2430 + }, + { + "epoch": 0.02431, + "grad_norm": 0.7926489355067414, + "learning_rate": 0.003, + "loss": 4.1517, + "step": 2431 + }, + { + "epoch": 0.02432, + "grad_norm": 0.7608367712051879, + "learning_rate": 0.003, + "loss": 4.173, + "step": 2432 + }, + { + "epoch": 0.02433, + "grad_norm": 0.7894251146149308, + "learning_rate": 0.003, + "loss": 4.1861, + "step": 2433 + }, + { + "epoch": 0.02434, + "grad_norm": 0.7224291073363042, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 2434 + }, + { + "epoch": 0.02435, + "grad_norm": 0.747878118674312, + "learning_rate": 0.003, + "loss": 4.152, + "step": 2435 + }, + { + "epoch": 0.02436, + "grad_norm": 0.747137566594131, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 2436 + }, + { + "epoch": 0.02437, + "grad_norm": 0.5861461270232895, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 2437 + }, + { + "epoch": 0.02438, + "grad_norm": 0.695209068272505, + "learning_rate": 0.003, + "loss": 4.1607, + "step": 2438 + }, + { + "epoch": 0.02439, + "grad_norm": 0.7517625624937753, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 2439 + }, + { + "epoch": 0.0244, + "grad_norm": 0.762876594628319, + "learning_rate": 0.003, + "loss": 4.1846, + "step": 2440 + }, + { + "epoch": 0.02441, + "grad_norm": 0.8073496344275582, + "learning_rate": 0.003, + "loss": 4.1504, + "step": 2441 + }, + { + "epoch": 0.02442, + "grad_norm": 0.7951248150411244, + "learning_rate": 0.003, + "loss": 4.1836, + "step": 2442 + }, + { + "epoch": 0.02443, + "grad_norm": 0.7020771235734734, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 2443 + }, + { + "epoch": 0.02444, + "grad_norm": 0.7036802681519476, + "learning_rate": 0.003, + "loss": 4.1466, + "step": 2444 + }, + { + "epoch": 0.02445, + "grad_norm": 0.7782431022054782, + "learning_rate": 0.003, + "loss": 4.156, + "step": 2445 + }, + { + "epoch": 0.02446, + "grad_norm": 0.8768109280222425, + "learning_rate": 0.003, + "loss": 4.162, + "step": 2446 + }, + { + "epoch": 0.02447, + "grad_norm": 0.8860280168318521, + "learning_rate": 0.003, + "loss": 4.1721, + "step": 2447 + }, + { + "epoch": 0.02448, + "grad_norm": 0.8084130173823347, + "learning_rate": 0.003, + "loss": 4.1749, + "step": 2448 + }, + { + "epoch": 0.02449, + "grad_norm": 0.7919496319432228, + "learning_rate": 0.003, + "loss": 4.1681, + "step": 2449 + }, + { + "epoch": 0.0245, + "grad_norm": 0.7550563828443017, + "learning_rate": 0.003, + "loss": 4.171, + "step": 2450 + }, + { + "epoch": 0.02451, + "grad_norm": 0.7000106205173663, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 2451 + }, + { + "epoch": 0.02452, + "grad_norm": 0.7037695682912876, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 2452 + }, + { + "epoch": 0.02453, + "grad_norm": 0.8175542104548721, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 2453 + }, + { + "epoch": 0.02454, + "grad_norm": 0.8060583832916741, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 2454 + }, + { + "epoch": 0.02455, + "grad_norm": 0.8291689568387238, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2455 + }, + { + "epoch": 0.02456, + "grad_norm": 0.9834326131795995, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 2456 + }, + { + "epoch": 0.02457, + "grad_norm": 1.038602947579566, + "learning_rate": 0.003, + "loss": 4.1855, + "step": 2457 + }, + { + "epoch": 0.02458, + "grad_norm": 0.8411896951589359, + "learning_rate": 0.003, + "loss": 4.1522, + "step": 2458 + }, + { + "epoch": 0.02459, + "grad_norm": 0.8764491170718162, + "learning_rate": 0.003, + "loss": 4.1807, + "step": 2459 + }, + { + "epoch": 0.0246, + "grad_norm": 0.8235856930392861, + "learning_rate": 0.003, + "loss": 4.1862, + "step": 2460 + }, + { + "epoch": 0.02461, + "grad_norm": 0.7914755810846422, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 2461 + }, + { + "epoch": 0.02462, + "grad_norm": 0.6924934683339846, + "learning_rate": 0.003, + "loss": 4.1852, + "step": 2462 + }, + { + "epoch": 0.02463, + "grad_norm": 0.7267046618021085, + "learning_rate": 0.003, + "loss": 4.1828, + "step": 2463 + }, + { + "epoch": 0.02464, + "grad_norm": 0.7273513960616863, + "learning_rate": 0.003, + "loss": 4.1619, + "step": 2464 + }, + { + "epoch": 0.02465, + "grad_norm": 0.6908347256139015, + "learning_rate": 0.003, + "loss": 4.1789, + "step": 2465 + }, + { + "epoch": 0.02466, + "grad_norm": 0.5440404451212307, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 2466 + }, + { + "epoch": 0.02467, + "grad_norm": 0.48142907245166, + "learning_rate": 0.003, + "loss": 4.129, + "step": 2467 + }, + { + "epoch": 0.02468, + "grad_norm": 0.49753302892715556, + "learning_rate": 0.003, + "loss": 4.1713, + "step": 2468 + }, + { + "epoch": 0.02469, + "grad_norm": 0.44092067265774965, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 2469 + }, + { + "epoch": 0.0247, + "grad_norm": 0.3789674393880185, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 2470 + }, + { + "epoch": 0.02471, + "grad_norm": 0.3943974525292669, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 2471 + }, + { + "epoch": 0.02472, + "grad_norm": 0.3923267553353639, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 2472 + }, + { + "epoch": 0.02473, + "grad_norm": 0.42593583713033994, + "learning_rate": 0.003, + "loss": 4.115, + "step": 2473 + }, + { + "epoch": 0.02474, + "grad_norm": 0.4520474047910496, + "learning_rate": 0.003, + "loss": 4.146, + "step": 2474 + }, + { + "epoch": 0.02475, + "grad_norm": 0.5344288526577453, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 2475 + }, + { + "epoch": 0.02476, + "grad_norm": 0.6292976482720128, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 2476 + }, + { + "epoch": 0.02477, + "grad_norm": 0.66063786243246, + "learning_rate": 0.003, + "loss": 4.1564, + "step": 2477 + }, + { + "epoch": 0.02478, + "grad_norm": 0.628521888800231, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 2478 + }, + { + "epoch": 0.02479, + "grad_norm": 0.5793386388274427, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 2479 + }, + { + "epoch": 0.0248, + "grad_norm": 0.5948986185020736, + "learning_rate": 0.003, + "loss": 4.1605, + "step": 2480 + }, + { + "epoch": 0.02481, + "grad_norm": 0.6967787520484608, + "learning_rate": 0.003, + "loss": 4.1459, + "step": 2481 + }, + { + "epoch": 0.02482, + "grad_norm": 0.8277600361067297, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 2482 + }, + { + "epoch": 0.02483, + "grad_norm": 0.8234952353564556, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 2483 + }, + { + "epoch": 0.02484, + "grad_norm": 0.8257905213668915, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 2484 + }, + { + "epoch": 0.02485, + "grad_norm": 0.8993714294926498, + "learning_rate": 0.003, + "loss": 4.1721, + "step": 2485 + }, + { + "epoch": 0.02486, + "grad_norm": 0.8326033436616108, + "learning_rate": 0.003, + "loss": 4.1926, + "step": 2486 + }, + { + "epoch": 0.02487, + "grad_norm": 0.7529249524206648, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 2487 + }, + { + "epoch": 0.02488, + "grad_norm": 0.7126565932475263, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 2488 + }, + { + "epoch": 0.02489, + "grad_norm": 0.610203079656125, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 2489 + }, + { + "epoch": 0.0249, + "grad_norm": 0.7614901790844095, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 2490 + }, + { + "epoch": 0.02491, + "grad_norm": 0.8289703407204001, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2491 + }, + { + "epoch": 0.02492, + "grad_norm": 0.743875798496966, + "learning_rate": 0.003, + "loss": 4.1625, + "step": 2492 + }, + { + "epoch": 0.02493, + "grad_norm": 0.7682635486879051, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 2493 + }, + { + "epoch": 0.02494, + "grad_norm": 0.9219887768531273, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 2494 + }, + { + "epoch": 0.02495, + "grad_norm": 0.8830573101913496, + "learning_rate": 0.003, + "loss": 4.1657, + "step": 2495 + }, + { + "epoch": 0.02496, + "grad_norm": 0.8835839715836804, + "learning_rate": 0.003, + "loss": 4.1556, + "step": 2496 + }, + { + "epoch": 0.02497, + "grad_norm": 0.8715710679970281, + "learning_rate": 0.003, + "loss": 4.1634, + "step": 2497 + }, + { + "epoch": 0.02498, + "grad_norm": 0.7317894257693476, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 2498 + }, + { + "epoch": 0.02499, + "grad_norm": 0.8203454328345735, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 2499 + }, + { + "epoch": 0.025, + "grad_norm": 0.8654906782724638, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 2500 + }, + { + "epoch": 0.02501, + "grad_norm": 0.7323060183877268, + "learning_rate": 0.003, + "loss": 4.179, + "step": 2501 + }, + { + "epoch": 0.02502, + "grad_norm": 0.582392117680552, + "learning_rate": 0.003, + "loss": 4.1553, + "step": 2502 + }, + { + "epoch": 0.02503, + "grad_norm": 0.6091436814763177, + "learning_rate": 0.003, + "loss": 4.1636, + "step": 2503 + }, + { + "epoch": 0.02504, + "grad_norm": 0.5495845907736094, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 2504 + }, + { + "epoch": 0.02505, + "grad_norm": 0.5133362105777172, + "learning_rate": 0.003, + "loss": 4.138, + "step": 2505 + }, + { + "epoch": 0.02506, + "grad_norm": 0.5169309681279142, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 2506 + }, + { + "epoch": 0.02507, + "grad_norm": 0.6065719655319186, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 2507 + }, + { + "epoch": 0.02508, + "grad_norm": 0.5781191649097883, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 2508 + }, + { + "epoch": 0.02509, + "grad_norm": 0.4910293327621469, + "learning_rate": 0.003, + "loss": 4.136, + "step": 2509 + }, + { + "epoch": 0.0251, + "grad_norm": 0.5320031044066156, + "learning_rate": 0.003, + "loss": 4.1515, + "step": 2510 + }, + { + "epoch": 0.02511, + "grad_norm": 0.5931245641117349, + "learning_rate": 0.003, + "loss": 4.161, + "step": 2511 + }, + { + "epoch": 0.02512, + "grad_norm": 0.6656113562789021, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 2512 + }, + { + "epoch": 0.02513, + "grad_norm": 0.7074045972943324, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 2513 + }, + { + "epoch": 0.02514, + "grad_norm": 0.6725692970833035, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 2514 + }, + { + "epoch": 0.02515, + "grad_norm": 0.6792214563254976, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 2515 + }, + { + "epoch": 0.02516, + "grad_norm": 0.7140524085493386, + "learning_rate": 0.003, + "loss": 4.1428, + "step": 2516 + }, + { + "epoch": 0.02517, + "grad_norm": 0.6845997321477366, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 2517 + }, + { + "epoch": 0.02518, + "grad_norm": 0.7888218080320933, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 2518 + }, + { + "epoch": 0.02519, + "grad_norm": 0.9116083917350077, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 2519 + }, + { + "epoch": 0.0252, + "grad_norm": 0.7568133469610978, + "learning_rate": 0.003, + "loss": 4.1519, + "step": 2520 + }, + { + "epoch": 0.02521, + "grad_norm": 0.5849027216145644, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 2521 + }, + { + "epoch": 0.02522, + "grad_norm": 0.5672517169470362, + "learning_rate": 0.003, + "loss": 4.135, + "step": 2522 + }, + { + "epoch": 0.02523, + "grad_norm": 0.5215015625950845, + "learning_rate": 0.003, + "loss": 4.1352, + "step": 2523 + }, + { + "epoch": 0.02524, + "grad_norm": 0.5381024179104523, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 2524 + }, + { + "epoch": 0.02525, + "grad_norm": 0.6588506425380558, + "learning_rate": 0.003, + "loss": 4.165, + "step": 2525 + }, + { + "epoch": 0.02526, + "grad_norm": 0.8688895000859529, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 2526 + }, + { + "epoch": 0.02527, + "grad_norm": 1.0992881351085537, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 2527 + }, + { + "epoch": 0.02528, + "grad_norm": 0.8657251086311964, + "learning_rate": 0.003, + "loss": 4.1821, + "step": 2528 + }, + { + "epoch": 0.02529, + "grad_norm": 0.8533230819827688, + "learning_rate": 0.003, + "loss": 4.1512, + "step": 2529 + }, + { + "epoch": 0.0253, + "grad_norm": 0.9133367359531667, + "learning_rate": 0.003, + "loss": 4.1784, + "step": 2530 + }, + { + "epoch": 0.02531, + "grad_norm": 0.9751978802418442, + "learning_rate": 0.003, + "loss": 4.1851, + "step": 2531 + }, + { + "epoch": 0.02532, + "grad_norm": 0.8808398368059976, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 2532 + }, + { + "epoch": 0.02533, + "grad_norm": 0.812882204136674, + "learning_rate": 0.003, + "loss": 4.1877, + "step": 2533 + }, + { + "epoch": 0.02534, + "grad_norm": 0.6201400349331786, + "learning_rate": 0.003, + "loss": 4.16, + "step": 2534 + }, + { + "epoch": 0.02535, + "grad_norm": 0.6261338524312167, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2535 + }, + { + "epoch": 0.02536, + "grad_norm": 0.5914190494426448, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 2536 + }, + { + "epoch": 0.02537, + "grad_norm": 0.5985227538802093, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 2537 + }, + { + "epoch": 0.02538, + "grad_norm": 0.5564235786609277, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 2538 + }, + { + "epoch": 0.02539, + "grad_norm": 0.6121625314049407, + "learning_rate": 0.003, + "loss": 4.1652, + "step": 2539 + }, + { + "epoch": 0.0254, + "grad_norm": 0.6548540879696481, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 2540 + }, + { + "epoch": 0.02541, + "grad_norm": 0.7053057284074064, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 2541 + }, + { + "epoch": 0.02542, + "grad_norm": 0.7101556887835341, + "learning_rate": 0.003, + "loss": 4.1286, + "step": 2542 + }, + { + "epoch": 0.02543, + "grad_norm": 0.6754772454378443, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2543 + }, + { + "epoch": 0.02544, + "grad_norm": 0.6658865987327929, + "learning_rate": 0.003, + "loss": 4.145, + "step": 2544 + }, + { + "epoch": 0.02545, + "grad_norm": 0.6975583001668989, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 2545 + }, + { + "epoch": 0.02546, + "grad_norm": 0.7660109635603722, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 2546 + }, + { + "epoch": 0.02547, + "grad_norm": 0.7962006497881392, + "learning_rate": 0.003, + "loss": 4.1795, + "step": 2547 + }, + { + "epoch": 0.02548, + "grad_norm": 0.8051928165355774, + "learning_rate": 0.003, + "loss": 4.177, + "step": 2548 + }, + { + "epoch": 0.02549, + "grad_norm": 0.7367424583276029, + "learning_rate": 0.003, + "loss": 4.1656, + "step": 2549 + }, + { + "epoch": 0.0255, + "grad_norm": 0.6908727242126833, + "learning_rate": 0.003, + "loss": 4.1276, + "step": 2550 + }, + { + "epoch": 0.02551, + "grad_norm": 0.560657185849122, + "learning_rate": 0.003, + "loss": 4.1528, + "step": 2551 + }, + { + "epoch": 0.02552, + "grad_norm": 0.5148851527769897, + "learning_rate": 0.003, + "loss": 4.1213, + "step": 2552 + }, + { + "epoch": 0.02553, + "grad_norm": 0.430301186414899, + "learning_rate": 0.003, + "loss": 4.1828, + "step": 2553 + }, + { + "epoch": 0.02554, + "grad_norm": 0.5046174269848671, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 2554 + }, + { + "epoch": 0.02555, + "grad_norm": 0.46123127159126703, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 2555 + }, + { + "epoch": 0.02556, + "grad_norm": 0.45807479982386295, + "learning_rate": 0.003, + "loss": 4.1332, + "step": 2556 + }, + { + "epoch": 0.02557, + "grad_norm": 0.4789253794302061, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 2557 + }, + { + "epoch": 0.02558, + "grad_norm": 0.5953519804107527, + "learning_rate": 0.003, + "loss": 4.155, + "step": 2558 + }, + { + "epoch": 0.02559, + "grad_norm": 0.7762708669370249, + "learning_rate": 0.003, + "loss": 4.1569, + "step": 2559 + }, + { + "epoch": 0.0256, + "grad_norm": 0.9269892716474006, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 2560 + }, + { + "epoch": 0.02561, + "grad_norm": 0.7935762354738295, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 2561 + }, + { + "epoch": 0.02562, + "grad_norm": 0.5613742219703036, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 2562 + }, + { + "epoch": 0.02563, + "grad_norm": 0.620997139110311, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 2563 + }, + { + "epoch": 0.02564, + "grad_norm": 0.6260221226485523, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 2564 + }, + { + "epoch": 0.02565, + "grad_norm": 0.6519732862898392, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 2565 + }, + { + "epoch": 0.02566, + "grad_norm": 0.5807100901282953, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 2566 + }, + { + "epoch": 0.02567, + "grad_norm": 0.5901104990972573, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 2567 + }, + { + "epoch": 0.02568, + "grad_norm": 0.608480077065613, + "learning_rate": 0.003, + "loss": 4.1437, + "step": 2568 + }, + { + "epoch": 0.02569, + "grad_norm": 0.6208335544332332, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 2569 + }, + { + "epoch": 0.0257, + "grad_norm": 0.6669873558179679, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 2570 + }, + { + "epoch": 0.02571, + "grad_norm": 0.7466165598574759, + "learning_rate": 0.003, + "loss": 4.1569, + "step": 2571 + }, + { + "epoch": 0.02572, + "grad_norm": 0.7593942160416755, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 2572 + }, + { + "epoch": 0.02573, + "grad_norm": 0.8116124977641566, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 2573 + }, + { + "epoch": 0.02574, + "grad_norm": 0.8165190332663509, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 2574 + }, + { + "epoch": 0.02575, + "grad_norm": 0.7997376276347931, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 2575 + }, + { + "epoch": 0.02576, + "grad_norm": 0.8606509388677399, + "learning_rate": 0.003, + "loss": 4.124, + "step": 2576 + }, + { + "epoch": 0.02577, + "grad_norm": 0.8274185827265477, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 2577 + }, + { + "epoch": 0.02578, + "grad_norm": 0.8380609959115332, + "learning_rate": 0.003, + "loss": 4.1541, + "step": 2578 + }, + { + "epoch": 0.02579, + "grad_norm": 0.7900021411710185, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 2579 + }, + { + "epoch": 0.0258, + "grad_norm": 0.7777918852672433, + "learning_rate": 0.003, + "loss": 4.155, + "step": 2580 + }, + { + "epoch": 0.02581, + "grad_norm": 0.8333235015806116, + "learning_rate": 0.003, + "loss": 4.1764, + "step": 2581 + }, + { + "epoch": 0.02582, + "grad_norm": 0.9070003711783441, + "learning_rate": 0.003, + "loss": 4.1779, + "step": 2582 + }, + { + "epoch": 0.02583, + "grad_norm": 0.7888121706822758, + "learning_rate": 0.003, + "loss": 4.1503, + "step": 2583 + }, + { + "epoch": 0.02584, + "grad_norm": 0.6927272790493887, + "learning_rate": 0.003, + "loss": 4.1616, + "step": 2584 + }, + { + "epoch": 0.02585, + "grad_norm": 0.7960397742665343, + "learning_rate": 0.003, + "loss": 4.1455, + "step": 2585 + }, + { + "epoch": 0.02586, + "grad_norm": 0.7991616204483157, + "learning_rate": 0.003, + "loss": 4.1666, + "step": 2586 + }, + { + "epoch": 0.02587, + "grad_norm": 0.8314442215989661, + "learning_rate": 0.003, + "loss": 4.1584, + "step": 2587 + }, + { + "epoch": 0.02588, + "grad_norm": 0.8301691730936955, + "learning_rate": 0.003, + "loss": 4.1603, + "step": 2588 + }, + { + "epoch": 0.02589, + "grad_norm": 0.8541934379581658, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 2589 + }, + { + "epoch": 0.0259, + "grad_norm": 0.919508505438588, + "learning_rate": 0.003, + "loss": 4.1759, + "step": 2590 + }, + { + "epoch": 0.02591, + "grad_norm": 0.8517196219178954, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 2591 + }, + { + "epoch": 0.02592, + "grad_norm": 0.7963954176530331, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 2592 + }, + { + "epoch": 0.02593, + "grad_norm": 0.6602605507860401, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 2593 + }, + { + "epoch": 0.02594, + "grad_norm": 0.6584920925234209, + "learning_rate": 0.003, + "loss": 4.1711, + "step": 2594 + }, + { + "epoch": 0.02595, + "grad_norm": 0.6413195861867523, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 2595 + }, + { + "epoch": 0.02596, + "grad_norm": 0.5585769233496517, + "learning_rate": 0.003, + "loss": 4.1585, + "step": 2596 + }, + { + "epoch": 0.02597, + "grad_norm": 0.5531354692950633, + "learning_rate": 0.003, + "loss": 4.127, + "step": 2597 + }, + { + "epoch": 0.02598, + "grad_norm": 0.5841489894205579, + "learning_rate": 0.003, + "loss": 4.1665, + "step": 2598 + }, + { + "epoch": 0.02599, + "grad_norm": 0.5386601238229825, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 2599 + }, + { + "epoch": 0.026, + "grad_norm": 0.48249487027027893, + "learning_rate": 0.003, + "loss": 4.1544, + "step": 2600 + }, + { + "epoch": 0.02601, + "grad_norm": 0.5633711470347749, + "learning_rate": 0.003, + "loss": 4.1612, + "step": 2601 + }, + { + "epoch": 0.02602, + "grad_norm": 0.5283455753312563, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 2602 + }, + { + "epoch": 0.02603, + "grad_norm": 0.4245511813607633, + "learning_rate": 0.003, + "loss": 4.125, + "step": 2603 + }, + { + "epoch": 0.02604, + "grad_norm": 0.4587095055563627, + "learning_rate": 0.003, + "loss": 4.115, + "step": 2604 + }, + { + "epoch": 0.02605, + "grad_norm": 0.46796143155617576, + "learning_rate": 0.003, + "loss": 4.13, + "step": 2605 + }, + { + "epoch": 0.02606, + "grad_norm": 0.4516424293786292, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 2606 + }, + { + "epoch": 0.02607, + "grad_norm": 0.44215061855329385, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 2607 + }, + { + "epoch": 0.02608, + "grad_norm": 0.5353978515447827, + "learning_rate": 0.003, + "loss": 4.12, + "step": 2608 + }, + { + "epoch": 0.02609, + "grad_norm": 0.5797588396112132, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 2609 + }, + { + "epoch": 0.0261, + "grad_norm": 0.7287463194886674, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 2610 + }, + { + "epoch": 0.02611, + "grad_norm": 0.8961743338680872, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 2611 + }, + { + "epoch": 0.02612, + "grad_norm": 0.9346789409803079, + "learning_rate": 0.003, + "loss": 4.1576, + "step": 2612 + }, + { + "epoch": 0.02613, + "grad_norm": 0.792609416991834, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 2613 + }, + { + "epoch": 0.02614, + "grad_norm": 0.6956918290504753, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 2614 + }, + { + "epoch": 0.02615, + "grad_norm": 0.6912407283992354, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 2615 + }, + { + "epoch": 0.02616, + "grad_norm": 0.686973332839249, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 2616 + }, + { + "epoch": 0.02617, + "grad_norm": 0.6142692731488304, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 2617 + }, + { + "epoch": 0.02618, + "grad_norm": 0.5524427267847271, + "learning_rate": 0.003, + "loss": 4.142, + "step": 2618 + }, + { + "epoch": 0.02619, + "grad_norm": 0.5065926029133155, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 2619 + }, + { + "epoch": 0.0262, + "grad_norm": 0.5314850416827239, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 2620 + }, + { + "epoch": 0.02621, + "grad_norm": 0.6558721052067381, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 2621 + }, + { + "epoch": 0.02622, + "grad_norm": 0.6159166903853056, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 2622 + }, + { + "epoch": 0.02623, + "grad_norm": 0.5586448659094586, + "learning_rate": 0.003, + "loss": 4.1457, + "step": 2623 + }, + { + "epoch": 0.02624, + "grad_norm": 0.4617510923789607, + "learning_rate": 0.003, + "loss": 4.1803, + "step": 2624 + }, + { + "epoch": 0.02625, + "grad_norm": 0.4793745081555762, + "learning_rate": 0.003, + "loss": 4.1738, + "step": 2625 + }, + { + "epoch": 0.02626, + "grad_norm": 0.5349123420647252, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 2626 + }, + { + "epoch": 0.02627, + "grad_norm": 0.5555525418610604, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 2627 + }, + { + "epoch": 0.02628, + "grad_norm": 0.6710640559411927, + "learning_rate": 0.003, + "loss": 4.1625, + "step": 2628 + }, + { + "epoch": 0.02629, + "grad_norm": 0.8640604833690961, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 2629 + }, + { + "epoch": 0.0263, + "grad_norm": 0.8319642559617553, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 2630 + }, + { + "epoch": 0.02631, + "grad_norm": 0.8249496168740192, + "learning_rate": 0.003, + "loss": 4.16, + "step": 2631 + }, + { + "epoch": 0.02632, + "grad_norm": 0.8306415820236996, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 2632 + }, + { + "epoch": 0.02633, + "grad_norm": 0.8783672336464626, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 2633 + }, + { + "epoch": 0.02634, + "grad_norm": 0.9508643378899662, + "learning_rate": 0.003, + "loss": 4.1694, + "step": 2634 + }, + { + "epoch": 0.02635, + "grad_norm": 0.9199222639775697, + "learning_rate": 0.003, + "loss": 4.1651, + "step": 2635 + }, + { + "epoch": 0.02636, + "grad_norm": 0.7863775669942374, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 2636 + }, + { + "epoch": 0.02637, + "grad_norm": 0.7219685137900227, + "learning_rate": 0.003, + "loss": 4.1757, + "step": 2637 + }, + { + "epoch": 0.02638, + "grad_norm": 0.8502176142495371, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 2638 + }, + { + "epoch": 0.02639, + "grad_norm": 0.9138805852187049, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 2639 + }, + { + "epoch": 0.0264, + "grad_norm": 0.9716590933841531, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 2640 + }, + { + "epoch": 0.02641, + "grad_norm": 0.8479391226588503, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 2641 + }, + { + "epoch": 0.02642, + "grad_norm": 0.7817798070201066, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 2642 + }, + { + "epoch": 0.02643, + "grad_norm": 0.7815806266435678, + "learning_rate": 0.003, + "loss": 4.1415, + "step": 2643 + }, + { + "epoch": 0.02644, + "grad_norm": 0.6831537267196337, + "learning_rate": 0.003, + "loss": 4.1543, + "step": 2644 + }, + { + "epoch": 0.02645, + "grad_norm": 0.6667598749526873, + "learning_rate": 0.003, + "loss": 4.1672, + "step": 2645 + }, + { + "epoch": 0.02646, + "grad_norm": 0.7199150347001335, + "learning_rate": 0.003, + "loss": 4.133, + "step": 2646 + }, + { + "epoch": 0.02647, + "grad_norm": 0.8298627711746913, + "learning_rate": 0.003, + "loss": 4.1593, + "step": 2647 + }, + { + "epoch": 0.02648, + "grad_norm": 0.7939232125492355, + "learning_rate": 0.003, + "loss": 4.1786, + "step": 2648 + }, + { + "epoch": 0.02649, + "grad_norm": 0.5847519159195422, + "learning_rate": 0.003, + "loss": 4.1568, + "step": 2649 + }, + { + "epoch": 0.0265, + "grad_norm": 0.7057608315084718, + "learning_rate": 0.003, + "loss": 4.1201, + "step": 2650 + }, + { + "epoch": 0.02651, + "grad_norm": 0.7969636923468641, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 2651 + }, + { + "epoch": 0.02652, + "grad_norm": 0.8082848111682119, + "learning_rate": 0.003, + "loss": 4.1699, + "step": 2652 + }, + { + "epoch": 0.02653, + "grad_norm": 0.8108330809177104, + "learning_rate": 0.003, + "loss": 4.1285, + "step": 2653 + }, + { + "epoch": 0.02654, + "grad_norm": 0.73574455302129, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 2654 + }, + { + "epoch": 0.02655, + "grad_norm": 0.7777124335595014, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 2655 + }, + { + "epoch": 0.02656, + "grad_norm": 0.8001543729940896, + "learning_rate": 0.003, + "loss": 4.1223, + "step": 2656 + }, + { + "epoch": 0.02657, + "grad_norm": 0.7462722849224278, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 2657 + }, + { + "epoch": 0.02658, + "grad_norm": 0.6674659180842099, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 2658 + }, + { + "epoch": 0.02659, + "grad_norm": 0.6581089259980051, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 2659 + }, + { + "epoch": 0.0266, + "grad_norm": 0.5543097498952742, + "learning_rate": 0.003, + "loss": 4.131, + "step": 2660 + }, + { + "epoch": 0.02661, + "grad_norm": 0.5555420550528128, + "learning_rate": 0.003, + "loss": 4.1508, + "step": 2661 + }, + { + "epoch": 0.02662, + "grad_norm": 0.5211936699294397, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 2662 + }, + { + "epoch": 0.02663, + "grad_norm": 0.4737441563022719, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 2663 + }, + { + "epoch": 0.02664, + "grad_norm": 0.5577603547702016, + "learning_rate": 0.003, + "loss": 4.1215, + "step": 2664 + }, + { + "epoch": 0.02665, + "grad_norm": 0.6855853252021962, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 2665 + }, + { + "epoch": 0.02666, + "grad_norm": 0.7532957399226684, + "learning_rate": 0.003, + "loss": 4.1662, + "step": 2666 + }, + { + "epoch": 0.02667, + "grad_norm": 0.8148730979135085, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 2667 + }, + { + "epoch": 0.02668, + "grad_norm": 0.7976618329010464, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 2668 + }, + { + "epoch": 0.02669, + "grad_norm": 0.7400358093817193, + "learning_rate": 0.003, + "loss": 4.1393, + "step": 2669 + }, + { + "epoch": 0.0267, + "grad_norm": 0.7865417654339665, + "learning_rate": 0.003, + "loss": 4.1402, + "step": 2670 + }, + { + "epoch": 0.02671, + "grad_norm": 0.7269600286873952, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 2671 + }, + { + "epoch": 0.02672, + "grad_norm": 0.6657094405649409, + "learning_rate": 0.003, + "loss": 4.1443, + "step": 2672 + }, + { + "epoch": 0.02673, + "grad_norm": 0.6575213330413505, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 2673 + }, + { + "epoch": 0.02674, + "grad_norm": 0.6041640138713956, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 2674 + }, + { + "epoch": 0.02675, + "grad_norm": 0.5049727263384894, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 2675 + }, + { + "epoch": 0.02676, + "grad_norm": 0.5741374680507492, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 2676 + }, + { + "epoch": 0.02677, + "grad_norm": 0.5481194086503574, + "learning_rate": 0.003, + "loss": 4.1546, + "step": 2677 + }, + { + "epoch": 0.02678, + "grad_norm": 0.4995203590220423, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 2678 + }, + { + "epoch": 0.02679, + "grad_norm": 0.4694779292150945, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 2679 + }, + { + "epoch": 0.0268, + "grad_norm": 0.410806251273339, + "learning_rate": 0.003, + "loss": 4.1319, + "step": 2680 + }, + { + "epoch": 0.02681, + "grad_norm": 0.39193093743312024, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 2681 + }, + { + "epoch": 0.02682, + "grad_norm": 0.43475326352374855, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 2682 + }, + { + "epoch": 0.02683, + "grad_norm": 0.46937066962289054, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 2683 + }, + { + "epoch": 0.02684, + "grad_norm": 0.5387722044363351, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 2684 + }, + { + "epoch": 0.02685, + "grad_norm": 0.5969082447311024, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 2685 + }, + { + "epoch": 0.02686, + "grad_norm": 0.6584545240643431, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 2686 + }, + { + "epoch": 0.02687, + "grad_norm": 0.6352592103785321, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 2687 + }, + { + "epoch": 0.02688, + "grad_norm": 0.640407097795565, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 2688 + }, + { + "epoch": 0.02689, + "grad_norm": 0.6088138501938627, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 2689 + }, + { + "epoch": 0.0269, + "grad_norm": 0.7914214321890414, + "learning_rate": 0.003, + "loss": 4.144, + "step": 2690 + }, + { + "epoch": 0.02691, + "grad_norm": 1.0475505674838492, + "learning_rate": 0.003, + "loss": 4.1503, + "step": 2691 + }, + { + "epoch": 0.02692, + "grad_norm": 1.09889289903992, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 2692 + }, + { + "epoch": 0.02693, + "grad_norm": 0.7324561888334064, + "learning_rate": 0.003, + "loss": 4.1602, + "step": 2693 + }, + { + "epoch": 0.02694, + "grad_norm": 0.7027494176504083, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 2694 + }, + { + "epoch": 0.02695, + "grad_norm": 0.6342746970055929, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 2695 + }, + { + "epoch": 0.02696, + "grad_norm": 0.6665461114225348, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 2696 + }, + { + "epoch": 0.02697, + "grad_norm": 0.7278916312951845, + "learning_rate": 0.003, + "loss": 4.154, + "step": 2697 + }, + { + "epoch": 0.02698, + "grad_norm": 0.7787909239516192, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 2698 + }, + { + "epoch": 0.02699, + "grad_norm": 0.8836227046949343, + "learning_rate": 0.003, + "loss": 4.1632, + "step": 2699 + }, + { + "epoch": 0.027, + "grad_norm": 0.9118008328600917, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 2700 + }, + { + "epoch": 0.02701, + "grad_norm": 1.117237043878659, + "learning_rate": 0.003, + "loss": 4.1471, + "step": 2701 + }, + { + "epoch": 0.02702, + "grad_norm": 1.055756255993552, + "learning_rate": 0.003, + "loss": 4.1928, + "step": 2702 + }, + { + "epoch": 0.02703, + "grad_norm": 0.8177757417077984, + "learning_rate": 0.003, + "loss": 4.1828, + "step": 2703 + }, + { + "epoch": 0.02704, + "grad_norm": 0.8596703413889796, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 2704 + }, + { + "epoch": 0.02705, + "grad_norm": 0.8535161380163439, + "learning_rate": 0.003, + "loss": 4.1814, + "step": 2705 + }, + { + "epoch": 0.02706, + "grad_norm": 0.9659607306902468, + "learning_rate": 0.003, + "loss": 4.1533, + "step": 2706 + }, + { + "epoch": 0.02707, + "grad_norm": 0.9534642181669635, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 2707 + }, + { + "epoch": 0.02708, + "grad_norm": 0.8819926702804946, + "learning_rate": 0.003, + "loss": 4.1839, + "step": 2708 + }, + { + "epoch": 0.02709, + "grad_norm": 0.8341889119521959, + "learning_rate": 0.003, + "loss": 4.1498, + "step": 2709 + }, + { + "epoch": 0.0271, + "grad_norm": 0.8203465032094468, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 2710 + }, + { + "epoch": 0.02711, + "grad_norm": 0.7100401760580335, + "learning_rate": 0.003, + "loss": 4.1389, + "step": 2711 + }, + { + "epoch": 0.02712, + "grad_norm": 0.7068998330315834, + "learning_rate": 0.003, + "loss": 4.1499, + "step": 2712 + }, + { + "epoch": 0.02713, + "grad_norm": 0.7217216820883305, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 2713 + }, + { + "epoch": 0.02714, + "grad_norm": 0.7495769401012529, + "learning_rate": 0.003, + "loss": 4.1433, + "step": 2714 + }, + { + "epoch": 0.02715, + "grad_norm": 0.721135936817452, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 2715 + }, + { + "epoch": 0.02716, + "grad_norm": 0.6506027681741677, + "learning_rate": 0.003, + "loss": 4.1445, + "step": 2716 + }, + { + "epoch": 0.02717, + "grad_norm": 0.6669526413956082, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 2717 + }, + { + "epoch": 0.02718, + "grad_norm": 0.5633827874170128, + "learning_rate": 0.003, + "loss": 4.1604, + "step": 2718 + }, + { + "epoch": 0.02719, + "grad_norm": 0.4774270588916974, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 2719 + }, + { + "epoch": 0.0272, + "grad_norm": 0.46494508060084494, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 2720 + }, + { + "epoch": 0.02721, + "grad_norm": 0.41965798015402206, + "learning_rate": 0.003, + "loss": 4.132, + "step": 2721 + }, + { + "epoch": 0.02722, + "grad_norm": 0.40511078871192474, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 2722 + }, + { + "epoch": 0.02723, + "grad_norm": 0.3811271645582291, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 2723 + }, + { + "epoch": 0.02724, + "grad_norm": 0.37156244396727683, + "learning_rate": 0.003, + "loss": 4.1359, + "step": 2724 + }, + { + "epoch": 0.02725, + "grad_norm": 0.3879154674297152, + "learning_rate": 0.003, + "loss": 4.127, + "step": 2725 + }, + { + "epoch": 0.02726, + "grad_norm": 0.39593656565174934, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 2726 + }, + { + "epoch": 0.02727, + "grad_norm": 0.35677598647030584, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 2727 + }, + { + "epoch": 0.02728, + "grad_norm": 0.4083471465139489, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 2728 + }, + { + "epoch": 0.02729, + "grad_norm": 0.5166822235833873, + "learning_rate": 0.003, + "loss": 4.1489, + "step": 2729 + }, + { + "epoch": 0.0273, + "grad_norm": 0.6141441357599978, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 2730 + }, + { + "epoch": 0.02731, + "grad_norm": 0.6769146491120129, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 2731 + }, + { + "epoch": 0.02732, + "grad_norm": 0.6831890167942476, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 2732 + }, + { + "epoch": 0.02733, + "grad_norm": 0.6928624331582912, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 2733 + }, + { + "epoch": 0.02734, + "grad_norm": 0.8066796752318611, + "learning_rate": 0.003, + "loss": 4.107, + "step": 2734 + }, + { + "epoch": 0.02735, + "grad_norm": 0.8149444038754412, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 2735 + }, + { + "epoch": 0.02736, + "grad_norm": 0.9352910435235678, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 2736 + }, + { + "epoch": 0.02737, + "grad_norm": 1.0878188768837687, + "learning_rate": 0.003, + "loss": 4.1336, + "step": 2737 + }, + { + "epoch": 0.02738, + "grad_norm": 0.9785425363084683, + "learning_rate": 0.003, + "loss": 4.1637, + "step": 2738 + }, + { + "epoch": 0.02739, + "grad_norm": 0.7662690259696622, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 2739 + }, + { + "epoch": 0.0274, + "grad_norm": 0.6523497965235603, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 2740 + }, + { + "epoch": 0.02741, + "grad_norm": 0.5973844845147926, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 2741 + }, + { + "epoch": 0.02742, + "grad_norm": 0.5819006920062658, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 2742 + }, + { + "epoch": 0.02743, + "grad_norm": 0.6721498523545438, + "learning_rate": 0.003, + "loss": 4.1528, + "step": 2743 + }, + { + "epoch": 0.02744, + "grad_norm": 0.7300682624413913, + "learning_rate": 0.003, + "loss": 4.1392, + "step": 2744 + }, + { + "epoch": 0.02745, + "grad_norm": 0.715734484490805, + "learning_rate": 0.003, + "loss": 4.1484, + "step": 2745 + }, + { + "epoch": 0.02746, + "grad_norm": 0.7914480456919996, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 2746 + }, + { + "epoch": 0.02747, + "grad_norm": 0.8556964837503391, + "learning_rate": 0.003, + "loss": 4.1633, + "step": 2747 + }, + { + "epoch": 0.02748, + "grad_norm": 0.7536361109889692, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 2748 + }, + { + "epoch": 0.02749, + "grad_norm": 0.78073390311764, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 2749 + }, + { + "epoch": 0.0275, + "grad_norm": 0.7116445265532969, + "learning_rate": 0.003, + "loss": 4.1518, + "step": 2750 + }, + { + "epoch": 0.02751, + "grad_norm": 0.724589910610394, + "learning_rate": 0.003, + "loss": 4.1661, + "step": 2751 + }, + { + "epoch": 0.02752, + "grad_norm": 0.7065719140333095, + "learning_rate": 0.003, + "loss": 4.154, + "step": 2752 + }, + { + "epoch": 0.02753, + "grad_norm": 0.6640172302274574, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 2753 + }, + { + "epoch": 0.02754, + "grad_norm": 0.791074384921737, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 2754 + }, + { + "epoch": 0.02755, + "grad_norm": 0.868926766374716, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 2755 + }, + { + "epoch": 0.02756, + "grad_norm": 0.8063418318827459, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 2756 + }, + { + "epoch": 0.02757, + "grad_norm": 0.6544663568834579, + "learning_rate": 0.003, + "loss": 4.1586, + "step": 2757 + }, + { + "epoch": 0.02758, + "grad_norm": 0.7029692980295897, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 2758 + }, + { + "epoch": 0.02759, + "grad_norm": 0.674500639776644, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 2759 + }, + { + "epoch": 0.0276, + "grad_norm": 0.7060188591522503, + "learning_rate": 0.003, + "loss": 4.143, + "step": 2760 + }, + { + "epoch": 0.02761, + "grad_norm": 0.738534864819568, + "learning_rate": 0.003, + "loss": 4.1642, + "step": 2761 + }, + { + "epoch": 0.02762, + "grad_norm": 0.659967663491127, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 2762 + }, + { + "epoch": 0.02763, + "grad_norm": 0.6199285873189208, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 2763 + }, + { + "epoch": 0.02764, + "grad_norm": 0.6745825025111755, + "learning_rate": 0.003, + "loss": 4.132, + "step": 2764 + }, + { + "epoch": 0.02765, + "grad_norm": 0.7249937483915444, + "learning_rate": 0.003, + "loss": 4.138, + "step": 2765 + }, + { + "epoch": 0.02766, + "grad_norm": 0.6129690011150926, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 2766 + }, + { + "epoch": 0.02767, + "grad_norm": 0.6107438426303683, + "learning_rate": 0.003, + "loss": 4.1596, + "step": 2767 + }, + { + "epoch": 0.02768, + "grad_norm": 0.5493772562758883, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 2768 + }, + { + "epoch": 0.02769, + "grad_norm": 0.527030416961108, + "learning_rate": 0.003, + "loss": 4.1497, + "step": 2769 + }, + { + "epoch": 0.0277, + "grad_norm": 0.5613697625411952, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 2770 + }, + { + "epoch": 0.02771, + "grad_norm": 0.49606940418885315, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 2771 + }, + { + "epoch": 0.02772, + "grad_norm": 0.5340726847694284, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 2772 + }, + { + "epoch": 0.02773, + "grad_norm": 0.5630042053963976, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 2773 + }, + { + "epoch": 0.02774, + "grad_norm": 0.573524403737376, + "learning_rate": 0.003, + "loss": 4.1535, + "step": 2774 + }, + { + "epoch": 0.02775, + "grad_norm": 0.6813651069617616, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 2775 + }, + { + "epoch": 0.02776, + "grad_norm": 0.7422632562361271, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 2776 + }, + { + "epoch": 0.02777, + "grad_norm": 0.6834667304690145, + "learning_rate": 0.003, + "loss": 4.103, + "step": 2777 + }, + { + "epoch": 0.02778, + "grad_norm": 0.5760656464348943, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 2778 + }, + { + "epoch": 0.02779, + "grad_norm": 0.6444880287249539, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 2779 + }, + { + "epoch": 0.0278, + "grad_norm": 0.749888908987373, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 2780 + }, + { + "epoch": 0.02781, + "grad_norm": 0.7307643574331208, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 2781 + }, + { + "epoch": 0.02782, + "grad_norm": 0.6771317283846435, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 2782 + }, + { + "epoch": 0.02783, + "grad_norm": 0.6529335307661079, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 2783 + }, + { + "epoch": 0.02784, + "grad_norm": 0.723028364106339, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 2784 + }, + { + "epoch": 0.02785, + "grad_norm": 0.6848288469291863, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 2785 + }, + { + "epoch": 0.02786, + "grad_norm": 0.7240507487034193, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 2786 + }, + { + "epoch": 0.02787, + "grad_norm": 0.7032011113862562, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 2787 + }, + { + "epoch": 0.02788, + "grad_norm": 0.688044555087296, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 2788 + }, + { + "epoch": 0.02789, + "grad_norm": 0.6456728914254233, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 2789 + }, + { + "epoch": 0.0279, + "grad_norm": 0.6924107822260622, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 2790 + }, + { + "epoch": 0.02791, + "grad_norm": 0.6934189560440472, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 2791 + }, + { + "epoch": 0.02792, + "grad_norm": 0.7094014857878168, + "learning_rate": 0.003, + "loss": 4.1362, + "step": 2792 + }, + { + "epoch": 0.02793, + "grad_norm": 0.7389340164037368, + "learning_rate": 0.003, + "loss": 4.134, + "step": 2793 + }, + { + "epoch": 0.02794, + "grad_norm": 0.6127615122860384, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 2794 + }, + { + "epoch": 0.02795, + "grad_norm": 0.6692744466707173, + "learning_rate": 0.003, + "loss": 4.1409, + "step": 2795 + }, + { + "epoch": 0.02796, + "grad_norm": 0.5857761910891065, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 2796 + }, + { + "epoch": 0.02797, + "grad_norm": 0.4759667061876108, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 2797 + }, + { + "epoch": 0.02798, + "grad_norm": 0.5427328011775863, + "learning_rate": 0.003, + "loss": 4.1398, + "step": 2798 + }, + { + "epoch": 0.02799, + "grad_norm": 0.6018438011777394, + "learning_rate": 0.003, + "loss": 4.1262, + "step": 2799 + }, + { + "epoch": 0.028, + "grad_norm": 0.7397876650820135, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 2800 + }, + { + "epoch": 0.02801, + "grad_norm": 0.8243785793578178, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 2801 + }, + { + "epoch": 0.02802, + "grad_norm": 0.9434507107930815, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 2802 + }, + { + "epoch": 0.02803, + "grad_norm": 1.0540743689647354, + "learning_rate": 0.003, + "loss": 4.1574, + "step": 2803 + }, + { + "epoch": 0.02804, + "grad_norm": 1.2001150836528194, + "learning_rate": 0.003, + "loss": 4.1447, + "step": 2804 + }, + { + "epoch": 0.02805, + "grad_norm": 0.8942350042178676, + "learning_rate": 0.003, + "loss": 4.168, + "step": 2805 + }, + { + "epoch": 0.02806, + "grad_norm": 0.8216755183599684, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 2806 + }, + { + "epoch": 0.02807, + "grad_norm": 0.852160162758772, + "learning_rate": 0.003, + "loss": 4.1502, + "step": 2807 + }, + { + "epoch": 0.02808, + "grad_norm": 0.8737137669736023, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 2808 + }, + { + "epoch": 0.02809, + "grad_norm": 0.8415150409926513, + "learning_rate": 0.003, + "loss": 4.1702, + "step": 2809 + }, + { + "epoch": 0.0281, + "grad_norm": 0.9200296787037161, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 2810 + }, + { + "epoch": 0.02811, + "grad_norm": 0.8935535267441544, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 2811 + }, + { + "epoch": 0.02812, + "grad_norm": 0.8825585842287229, + "learning_rate": 0.003, + "loss": 4.1617, + "step": 2812 + }, + { + "epoch": 0.02813, + "grad_norm": 0.7235553242795063, + "learning_rate": 0.003, + "loss": 4.1448, + "step": 2813 + }, + { + "epoch": 0.02814, + "grad_norm": 0.6116687799618377, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 2814 + }, + { + "epoch": 0.02815, + "grad_norm": 0.6519703377655225, + "learning_rate": 0.003, + "loss": 4.178, + "step": 2815 + }, + { + "epoch": 0.02816, + "grad_norm": 0.6471767650351875, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 2816 + }, + { + "epoch": 0.02817, + "grad_norm": 0.6729379157407176, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 2817 + }, + { + "epoch": 0.02818, + "grad_norm": 0.6590132047625883, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 2818 + }, + { + "epoch": 0.02819, + "grad_norm": 0.6155103890489352, + "learning_rate": 0.003, + "loss": 4.146, + "step": 2819 + }, + { + "epoch": 0.0282, + "grad_norm": 0.6547119856777294, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 2820 + }, + { + "epoch": 0.02821, + "grad_norm": 0.7024315034082068, + "learning_rate": 0.003, + "loss": 4.1588, + "step": 2821 + }, + { + "epoch": 0.02822, + "grad_norm": 0.6752201506554464, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 2822 + }, + { + "epoch": 0.02823, + "grad_norm": 0.7207634885141747, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 2823 + }, + { + "epoch": 0.02824, + "grad_norm": 0.7099065820240217, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 2824 + }, + { + "epoch": 0.02825, + "grad_norm": 0.7380956712506125, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 2825 + }, + { + "epoch": 0.02826, + "grad_norm": 0.7896466713602348, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 2826 + }, + { + "epoch": 0.02827, + "grad_norm": 0.6938102549439281, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 2827 + }, + { + "epoch": 0.02828, + "grad_norm": 0.6683108375100919, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 2828 + }, + { + "epoch": 0.02829, + "grad_norm": 0.6979815634873964, + "learning_rate": 0.003, + "loss": 4.1727, + "step": 2829 + }, + { + "epoch": 0.0283, + "grad_norm": 0.5956877870461438, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 2830 + }, + { + "epoch": 0.02831, + "grad_norm": 0.5622298844948989, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 2831 + }, + { + "epoch": 0.02832, + "grad_norm": 0.587824453887482, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 2832 + }, + { + "epoch": 0.02833, + "grad_norm": 0.6622875210307357, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 2833 + }, + { + "epoch": 0.02834, + "grad_norm": 0.8665464077386911, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 2834 + }, + { + "epoch": 0.02835, + "grad_norm": 1.0264247411196268, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 2835 + }, + { + "epoch": 0.02836, + "grad_norm": 1.047672643824818, + "learning_rate": 0.003, + "loss": 4.1613, + "step": 2836 + }, + { + "epoch": 0.02837, + "grad_norm": 0.7736558581241362, + "learning_rate": 0.003, + "loss": 4.1247, + "step": 2837 + }, + { + "epoch": 0.02838, + "grad_norm": 0.6668958414020719, + "learning_rate": 0.003, + "loss": 4.091, + "step": 2838 + }, + { + "epoch": 0.02839, + "grad_norm": 0.6130179397103852, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 2839 + }, + { + "epoch": 0.0284, + "grad_norm": 0.5832331109848543, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 2840 + }, + { + "epoch": 0.02841, + "grad_norm": 0.561338338486519, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 2841 + }, + { + "epoch": 0.02842, + "grad_norm": 0.5576009464207491, + "learning_rate": 0.003, + "loss": 4.1175, + "step": 2842 + }, + { + "epoch": 0.02843, + "grad_norm": 0.529180646318587, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 2843 + }, + { + "epoch": 0.02844, + "grad_norm": 0.5310943153933595, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 2844 + }, + { + "epoch": 0.02845, + "grad_norm": 0.619003165124954, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 2845 + }, + { + "epoch": 0.02846, + "grad_norm": 0.5751981499915155, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 2846 + }, + { + "epoch": 0.02847, + "grad_norm": 0.5644823092326103, + "learning_rate": 0.003, + "loss": 4.1645, + "step": 2847 + }, + { + "epoch": 0.02848, + "grad_norm": 0.4766905538578811, + "learning_rate": 0.003, + "loss": 4.118, + "step": 2848 + }, + { + "epoch": 0.02849, + "grad_norm": 0.4730279660799319, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 2849 + }, + { + "epoch": 0.0285, + "grad_norm": 0.4862310178433145, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 2850 + }, + { + "epoch": 0.02851, + "grad_norm": 0.4444282499798774, + "learning_rate": 0.003, + "loss": 4.1449, + "step": 2851 + }, + { + "epoch": 0.02852, + "grad_norm": 0.5119251742433475, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 2852 + }, + { + "epoch": 0.02853, + "grad_norm": 0.543036144934772, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 2853 + }, + { + "epoch": 0.02854, + "grad_norm": 0.5997570730451834, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 2854 + }, + { + "epoch": 0.02855, + "grad_norm": 0.6454882733166921, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 2855 + }, + { + "epoch": 0.02856, + "grad_norm": 0.6988672826686806, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 2856 + }, + { + "epoch": 0.02857, + "grad_norm": 0.7074349768918466, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 2857 + }, + { + "epoch": 0.02858, + "grad_norm": 0.7412673488294538, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 2858 + }, + { + "epoch": 0.02859, + "grad_norm": 0.8102934829686708, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 2859 + }, + { + "epoch": 0.0286, + "grad_norm": 0.7746505956101851, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 2860 + }, + { + "epoch": 0.02861, + "grad_norm": 0.752523650350801, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 2861 + }, + { + "epoch": 0.02862, + "grad_norm": 0.6143460753973949, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 2862 + }, + { + "epoch": 0.02863, + "grad_norm": 0.7376408004996478, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 2863 + }, + { + "epoch": 0.02864, + "grad_norm": 0.7940734119740299, + "learning_rate": 0.003, + "loss": 4.123, + "step": 2864 + }, + { + "epoch": 0.02865, + "grad_norm": 0.8435691569693393, + "learning_rate": 0.003, + "loss": 4.1571, + "step": 2865 + }, + { + "epoch": 0.02866, + "grad_norm": 0.8922637961049935, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 2866 + }, + { + "epoch": 0.02867, + "grad_norm": 0.9133786616627491, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 2867 + }, + { + "epoch": 0.02868, + "grad_norm": 0.7907008263271671, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 2868 + }, + { + "epoch": 0.02869, + "grad_norm": 0.8279068392057516, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 2869 + }, + { + "epoch": 0.0287, + "grad_norm": 0.7251262273068938, + "learning_rate": 0.003, + "loss": 4.1492, + "step": 2870 + }, + { + "epoch": 0.02871, + "grad_norm": 0.6487151209805817, + "learning_rate": 0.003, + "loss": 4.1308, + "step": 2871 + }, + { + "epoch": 0.02872, + "grad_norm": 0.6560662243479557, + "learning_rate": 0.003, + "loss": 4.165, + "step": 2872 + }, + { + "epoch": 0.02873, + "grad_norm": 0.7813788666355885, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 2873 + }, + { + "epoch": 0.02874, + "grad_norm": 0.8171936837522956, + "learning_rate": 0.003, + "loss": 4.1552, + "step": 2874 + }, + { + "epoch": 0.02875, + "grad_norm": 0.7930932018349085, + "learning_rate": 0.003, + "loss": 4.1631, + "step": 2875 + }, + { + "epoch": 0.02876, + "grad_norm": 0.7044440796090409, + "learning_rate": 0.003, + "loss": 4.1418, + "step": 2876 + }, + { + "epoch": 0.02877, + "grad_norm": 0.7336680793330352, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 2877 + }, + { + "epoch": 0.02878, + "grad_norm": 0.800156609826111, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 2878 + }, + { + "epoch": 0.02879, + "grad_norm": 0.8079625308040846, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 2879 + }, + { + "epoch": 0.0288, + "grad_norm": 0.7812166094808956, + "learning_rate": 0.003, + "loss": 4.126, + "step": 2880 + }, + { + "epoch": 0.02881, + "grad_norm": 0.7550076751502341, + "learning_rate": 0.003, + "loss": 4.1367, + "step": 2881 + }, + { + "epoch": 0.02882, + "grad_norm": 0.8306963413015411, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 2882 + }, + { + "epoch": 0.02883, + "grad_norm": 0.9802344240635292, + "learning_rate": 0.003, + "loss": 4.1422, + "step": 2883 + }, + { + "epoch": 0.02884, + "grad_norm": 0.9476092033234238, + "learning_rate": 0.003, + "loss": 4.1878, + "step": 2884 + }, + { + "epoch": 0.02885, + "grad_norm": 0.6612748788766601, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 2885 + }, + { + "epoch": 0.02886, + "grad_norm": 0.650321142178644, + "learning_rate": 0.003, + "loss": 4.1464, + "step": 2886 + }, + { + "epoch": 0.02887, + "grad_norm": 0.6414037849792964, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 2887 + }, + { + "epoch": 0.02888, + "grad_norm": 0.6483147698285003, + "learning_rate": 0.003, + "loss": 4.127, + "step": 2888 + }, + { + "epoch": 0.02889, + "grad_norm": 0.7143988192087037, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 2889 + }, + { + "epoch": 0.0289, + "grad_norm": 0.7249826580865393, + "learning_rate": 0.003, + "loss": 4.1501, + "step": 2890 + }, + { + "epoch": 0.02891, + "grad_norm": 0.6433766259411497, + "learning_rate": 0.003, + "loss": 4.1654, + "step": 2891 + }, + { + "epoch": 0.02892, + "grad_norm": 0.6414911637610158, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 2892 + }, + { + "epoch": 0.02893, + "grad_norm": 0.5937835515542452, + "learning_rate": 0.003, + "loss": 4.12, + "step": 2893 + }, + { + "epoch": 0.02894, + "grad_norm": 0.6423092987376106, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 2894 + }, + { + "epoch": 0.02895, + "grad_norm": 0.8192654034452247, + "learning_rate": 0.003, + "loss": 4.1769, + "step": 2895 + }, + { + "epoch": 0.02896, + "grad_norm": 0.7542855653728944, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 2896 + }, + { + "epoch": 0.02897, + "grad_norm": 0.6213200104434871, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 2897 + }, + { + "epoch": 0.02898, + "grad_norm": 0.5542641271643183, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 2898 + }, + { + "epoch": 0.02899, + "grad_norm": 0.5885012839030546, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 2899 + }, + { + "epoch": 0.029, + "grad_norm": 0.6561836901717327, + "learning_rate": 0.003, + "loss": 4.124, + "step": 2900 + }, + { + "epoch": 0.02901, + "grad_norm": 0.6174294765197629, + "learning_rate": 0.003, + "loss": 4.137, + "step": 2901 + }, + { + "epoch": 0.02902, + "grad_norm": 0.6594126091718887, + "learning_rate": 0.003, + "loss": 4.1555, + "step": 2902 + }, + { + "epoch": 0.02903, + "grad_norm": 0.7415660334257447, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 2903 + }, + { + "epoch": 0.02904, + "grad_norm": 0.6510679116134973, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 2904 + }, + { + "epoch": 0.02905, + "grad_norm": 0.5948221054802846, + "learning_rate": 0.003, + "loss": 4.1468, + "step": 2905 + }, + { + "epoch": 0.02906, + "grad_norm": 0.5971790256995725, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 2906 + }, + { + "epoch": 0.02907, + "grad_norm": 0.52973271537399, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 2907 + }, + { + "epoch": 0.02908, + "grad_norm": 0.5621746025103919, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 2908 + }, + { + "epoch": 0.02909, + "grad_norm": 0.5744582140463568, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 2909 + }, + { + "epoch": 0.0291, + "grad_norm": 0.5580267389664545, + "learning_rate": 0.003, + "loss": 4.128, + "step": 2910 + }, + { + "epoch": 0.02911, + "grad_norm": 0.5131135408743797, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 2911 + }, + { + "epoch": 0.02912, + "grad_norm": 0.5657170953707965, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 2912 + }, + { + "epoch": 0.02913, + "grad_norm": 0.5682042701647317, + "learning_rate": 0.003, + "loss": 4.1265, + "step": 2913 + }, + { + "epoch": 0.02914, + "grad_norm": 0.4461883530075565, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 2914 + }, + { + "epoch": 0.02915, + "grad_norm": 0.4563920329797094, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 2915 + }, + { + "epoch": 0.02916, + "grad_norm": 0.5577293421244759, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 2916 + }, + { + "epoch": 0.02917, + "grad_norm": 0.6452637672182776, + "learning_rate": 0.003, + "loss": 4.1461, + "step": 2917 + }, + { + "epoch": 0.02918, + "grad_norm": 0.7426519428446772, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 2918 + }, + { + "epoch": 0.02919, + "grad_norm": 0.7200306264703864, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 2919 + }, + { + "epoch": 0.0292, + "grad_norm": 0.6107819046990816, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 2920 + }, + { + "epoch": 0.02921, + "grad_norm": 0.530958071638472, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 2921 + }, + { + "epoch": 0.02922, + "grad_norm": 0.5770825294799934, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 2922 + }, + { + "epoch": 0.02923, + "grad_norm": 0.6283380422254848, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 2923 + }, + { + "epoch": 0.02924, + "grad_norm": 0.6846523574137975, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 2924 + }, + { + "epoch": 0.02925, + "grad_norm": 0.7350709671290444, + "learning_rate": 0.003, + "loss": 4.1431, + "step": 2925 + }, + { + "epoch": 0.02926, + "grad_norm": 0.8049703720528609, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 2926 + }, + { + "epoch": 0.02927, + "grad_norm": 0.9752146906903091, + "learning_rate": 0.003, + "loss": 4.1366, + "step": 2927 + }, + { + "epoch": 0.02928, + "grad_norm": 1.2027710385165162, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 2928 + }, + { + "epoch": 0.02929, + "grad_norm": 0.9137363777580481, + "learning_rate": 0.003, + "loss": 4.1486, + "step": 2929 + }, + { + "epoch": 0.0293, + "grad_norm": 0.914544810494064, + "learning_rate": 0.003, + "loss": 4.1479, + "step": 2930 + }, + { + "epoch": 0.02931, + "grad_norm": 1.0287625969953516, + "learning_rate": 0.003, + "loss": 4.1554, + "step": 2931 + }, + { + "epoch": 0.02932, + "grad_norm": 0.9917853189058099, + "learning_rate": 0.003, + "loss": 4.1452, + "step": 2932 + }, + { + "epoch": 0.02933, + "grad_norm": 0.8763901168137523, + "learning_rate": 0.003, + "loss": 4.1673, + "step": 2933 + }, + { + "epoch": 0.02934, + "grad_norm": 0.735577786783199, + "learning_rate": 0.003, + "loss": 4.1511, + "step": 2934 + }, + { + "epoch": 0.02935, + "grad_norm": 0.6627366189302382, + "learning_rate": 0.003, + "loss": 4.1346, + "step": 2935 + }, + { + "epoch": 0.02936, + "grad_norm": 0.6254387917411456, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 2936 + }, + { + "epoch": 0.02937, + "grad_norm": 0.6534121931204608, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 2937 + }, + { + "epoch": 0.02938, + "grad_norm": 0.650713517648723, + "learning_rate": 0.003, + "loss": 4.1476, + "step": 2938 + }, + { + "epoch": 0.02939, + "grad_norm": 0.6615159170277072, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 2939 + }, + { + "epoch": 0.0294, + "grad_norm": 0.62647373109481, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 2940 + }, + { + "epoch": 0.02941, + "grad_norm": 0.5291945312413562, + "learning_rate": 0.003, + "loss": 4.1524, + "step": 2941 + }, + { + "epoch": 0.02942, + "grad_norm": 0.5941823945903285, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 2942 + }, + { + "epoch": 0.02943, + "grad_norm": 0.5464093704546763, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 2943 + }, + { + "epoch": 0.02944, + "grad_norm": 0.5303894957276417, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 2944 + }, + { + "epoch": 0.02945, + "grad_norm": 0.629302957243814, + "learning_rate": 0.003, + "loss": 4.1257, + "step": 2945 + }, + { + "epoch": 0.02946, + "grad_norm": 0.7216986427052362, + "learning_rate": 0.003, + "loss": 4.1245, + "step": 2946 + }, + { + "epoch": 0.02947, + "grad_norm": 0.773136464545061, + "learning_rate": 0.003, + "loss": 4.1591, + "step": 2947 + }, + { + "epoch": 0.02948, + "grad_norm": 0.8581792881178094, + "learning_rate": 0.003, + "loss": 4.1463, + "step": 2948 + }, + { + "epoch": 0.02949, + "grad_norm": 0.7679345385149048, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 2949 + }, + { + "epoch": 0.0295, + "grad_norm": 0.8267756623287326, + "learning_rate": 0.003, + "loss": 4.1426, + "step": 2950 + }, + { + "epoch": 0.02951, + "grad_norm": 0.7612422330742484, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 2951 + }, + { + "epoch": 0.02952, + "grad_norm": 0.7322062479916658, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 2952 + }, + { + "epoch": 0.02953, + "grad_norm": 0.7904501063984217, + "learning_rate": 0.003, + "loss": 4.1328, + "step": 2953 + }, + { + "epoch": 0.02954, + "grad_norm": 0.8150360710975566, + "learning_rate": 0.003, + "loss": 4.1754, + "step": 2954 + }, + { + "epoch": 0.02955, + "grad_norm": 0.6387792277886543, + "learning_rate": 0.003, + "loss": 4.1417, + "step": 2955 + }, + { + "epoch": 0.02956, + "grad_norm": 0.6265803715231203, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 2956 + }, + { + "epoch": 0.02957, + "grad_norm": 0.6530135122360672, + "learning_rate": 0.003, + "loss": 4.1413, + "step": 2957 + }, + { + "epoch": 0.02958, + "grad_norm": 0.7323333201760076, + "learning_rate": 0.003, + "loss": 4.159, + "step": 2958 + }, + { + "epoch": 0.02959, + "grad_norm": 0.7893891920336619, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 2959 + }, + { + "epoch": 0.0296, + "grad_norm": 0.9738950017385256, + "learning_rate": 0.003, + "loss": 4.1629, + "step": 2960 + }, + { + "epoch": 0.02961, + "grad_norm": 0.9480886003061869, + "learning_rate": 0.003, + "loss": 4.1345, + "step": 2961 + }, + { + "epoch": 0.02962, + "grad_norm": 0.7754036108490701, + "learning_rate": 0.003, + "loss": 4.1383, + "step": 2962 + }, + { + "epoch": 0.02963, + "grad_norm": 0.729365514324412, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 2963 + }, + { + "epoch": 0.02964, + "grad_norm": 0.6752194662691007, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 2964 + }, + { + "epoch": 0.02965, + "grad_norm": 0.7029064516315103, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 2965 + }, + { + "epoch": 0.02966, + "grad_norm": 0.6775330621288554, + "learning_rate": 0.003, + "loss": 4.1547, + "step": 2966 + }, + { + "epoch": 0.02967, + "grad_norm": 0.5637108439064215, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 2967 + }, + { + "epoch": 0.02968, + "grad_norm": 0.48687140946780966, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 2968 + }, + { + "epoch": 0.02969, + "grad_norm": 0.5444201191933566, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 2969 + }, + { + "epoch": 0.0297, + "grad_norm": 0.5939263915198243, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 2970 + }, + { + "epoch": 0.02971, + "grad_norm": 0.5833592362245291, + "learning_rate": 0.003, + "loss": 4.1341, + "step": 2971 + }, + { + "epoch": 0.02972, + "grad_norm": 0.642339811999635, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 2972 + }, + { + "epoch": 0.02973, + "grad_norm": 0.6755234777209843, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 2973 + }, + { + "epoch": 0.02974, + "grad_norm": 0.7375944906573911, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 2974 + }, + { + "epoch": 0.02975, + "grad_norm": 0.7066341506793363, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 2975 + }, + { + "epoch": 0.02976, + "grad_norm": 0.7321898906171056, + "learning_rate": 0.003, + "loss": 4.1268, + "step": 2976 + }, + { + "epoch": 0.02977, + "grad_norm": 0.6550268523156474, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 2977 + }, + { + "epoch": 0.02978, + "grad_norm": 0.7024563931655272, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 2978 + }, + { + "epoch": 0.02979, + "grad_norm": 0.8062037165735291, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 2979 + }, + { + "epoch": 0.0298, + "grad_norm": 0.8638890231840887, + "learning_rate": 0.003, + "loss": 4.1347, + "step": 2980 + }, + { + "epoch": 0.02981, + "grad_norm": 0.8561746435152567, + "learning_rate": 0.003, + "loss": 4.1406, + "step": 2981 + }, + { + "epoch": 0.02982, + "grad_norm": 0.857616149681916, + "learning_rate": 0.003, + "loss": 4.143, + "step": 2982 + }, + { + "epoch": 0.02983, + "grad_norm": 0.8544260423256697, + "learning_rate": 0.003, + "loss": 4.1562, + "step": 2983 + }, + { + "epoch": 0.02984, + "grad_norm": 0.6670887165062284, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 2984 + }, + { + "epoch": 0.02985, + "grad_norm": 0.631424077374236, + "learning_rate": 0.003, + "loss": 4.1548, + "step": 2985 + }, + { + "epoch": 0.02986, + "grad_norm": 0.6150237051186566, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 2986 + }, + { + "epoch": 0.02987, + "grad_norm": 0.6009490083597219, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 2987 + }, + { + "epoch": 0.02988, + "grad_norm": 0.6856901545766193, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 2988 + }, + { + "epoch": 0.02989, + "grad_norm": 0.6971328900679812, + "learning_rate": 0.003, + "loss": 4.135, + "step": 2989 + }, + { + "epoch": 0.0299, + "grad_norm": 0.774097850725608, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 2990 + }, + { + "epoch": 0.02991, + "grad_norm": 0.8717977906192367, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 2991 + }, + { + "epoch": 0.02992, + "grad_norm": 0.8623658119853468, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 2992 + }, + { + "epoch": 0.02993, + "grad_norm": 1.0735677523955205, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 2993 + }, + { + "epoch": 0.02994, + "grad_norm": 0.8637296642447976, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 2994 + }, + { + "epoch": 0.02995, + "grad_norm": 0.7266640028404571, + "learning_rate": 0.003, + "loss": 4.1703, + "step": 2995 + }, + { + "epoch": 0.02996, + "grad_norm": 0.7432650045295823, + "learning_rate": 0.003, + "loss": 4.158, + "step": 2996 + }, + { + "epoch": 0.02997, + "grad_norm": 0.8060032310312305, + "learning_rate": 0.003, + "loss": 4.123, + "step": 2997 + }, + { + "epoch": 0.02998, + "grad_norm": 0.8682559997943201, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 2998 + }, + { + "epoch": 0.02999, + "grad_norm": 0.8501950539808205, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 2999 + }, + { + "epoch": 0.03, + "grad_norm": 0.7053221529954677, + "learning_rate": 0.003, + "loss": 4.16, + "step": 3000 + }, + { + "epoch": 0.03001, + "grad_norm": 0.6124443102712299, + "learning_rate": 0.003, + "loss": 4.126, + "step": 3001 + }, + { + "epoch": 0.03002, + "grad_norm": 0.6139771585331106, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 3002 + }, + { + "epoch": 0.03003, + "grad_norm": 0.6532527258316251, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3003 + }, + { + "epoch": 0.03004, + "grad_norm": 0.654278622264137, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 3004 + }, + { + "epoch": 0.03005, + "grad_norm": 0.6288378928024987, + "learning_rate": 0.003, + "loss": 4.146, + "step": 3005 + }, + { + "epoch": 0.03006, + "grad_norm": 0.584994074585291, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 3006 + }, + { + "epoch": 0.03007, + "grad_norm": 0.5366445106661158, + "learning_rate": 0.003, + "loss": 4.1074, + "step": 3007 + }, + { + "epoch": 0.03008, + "grad_norm": 0.46078204653660604, + "learning_rate": 0.003, + "loss": 4.1321, + "step": 3008 + }, + { + "epoch": 0.03009, + "grad_norm": 0.44367029679639464, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 3009 + }, + { + "epoch": 0.0301, + "grad_norm": 0.407224113006354, + "learning_rate": 0.003, + "loss": 4.127, + "step": 3010 + }, + { + "epoch": 0.03011, + "grad_norm": 0.43474103508569256, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 3011 + }, + { + "epoch": 0.03012, + "grad_norm": 0.46396616934212626, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 3012 + }, + { + "epoch": 0.03013, + "grad_norm": 0.4792456706607542, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 3013 + }, + { + "epoch": 0.03014, + "grad_norm": 0.5341940346855687, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 3014 + }, + { + "epoch": 0.03015, + "grad_norm": 0.538175902055071, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 3015 + }, + { + "epoch": 0.03016, + "grad_norm": 0.5829260362106814, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 3016 + }, + { + "epoch": 0.03017, + "grad_norm": 0.6581521346227238, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3017 + }, + { + "epoch": 0.03018, + "grad_norm": 0.769640091421768, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 3018 + }, + { + "epoch": 0.03019, + "grad_norm": 0.7982191899252973, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 3019 + }, + { + "epoch": 0.0302, + "grad_norm": 0.6974430504363406, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 3020 + }, + { + "epoch": 0.03021, + "grad_norm": 0.7317581928264297, + "learning_rate": 0.003, + "loss": 4.1405, + "step": 3021 + }, + { + "epoch": 0.03022, + "grad_norm": 0.7238636171905158, + "learning_rate": 0.003, + "loss": 4.1427, + "step": 3022 + }, + { + "epoch": 0.03023, + "grad_norm": 0.7005431651842164, + "learning_rate": 0.003, + "loss": 4.1539, + "step": 3023 + }, + { + "epoch": 0.03024, + "grad_norm": 0.691212312197827, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 3024 + }, + { + "epoch": 0.03025, + "grad_norm": 0.747828906617943, + "learning_rate": 0.003, + "loss": 4.1294, + "step": 3025 + }, + { + "epoch": 0.03026, + "grad_norm": 0.8978302984423533, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 3026 + }, + { + "epoch": 0.03027, + "grad_norm": 0.8881606554496674, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 3027 + }, + { + "epoch": 0.03028, + "grad_norm": 0.9256957009322, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 3028 + }, + { + "epoch": 0.03029, + "grad_norm": 0.7874902657267214, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 3029 + }, + { + "epoch": 0.0303, + "grad_norm": 0.6577501250055792, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 3030 + }, + { + "epoch": 0.03031, + "grad_norm": 0.5781820815886244, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 3031 + }, + { + "epoch": 0.03032, + "grad_norm": 0.6515119732971181, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 3032 + }, + { + "epoch": 0.03033, + "grad_norm": 0.6184230127538959, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 3033 + }, + { + "epoch": 0.03034, + "grad_norm": 0.6037451722184786, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 3034 + }, + { + "epoch": 0.03035, + "grad_norm": 0.7884678918699862, + "learning_rate": 0.003, + "loss": 4.1527, + "step": 3035 + }, + { + "epoch": 0.03036, + "grad_norm": 0.8850311524835377, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 3036 + }, + { + "epoch": 0.03037, + "grad_norm": 0.7695226928037275, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 3037 + }, + { + "epoch": 0.03038, + "grad_norm": 0.8630456389356111, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 3038 + }, + { + "epoch": 0.03039, + "grad_norm": 1.0230116086117664, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 3039 + }, + { + "epoch": 0.0304, + "grad_norm": 1.1203610279827287, + "learning_rate": 0.003, + "loss": 4.1506, + "step": 3040 + }, + { + "epoch": 0.03041, + "grad_norm": 0.9624810005532227, + "learning_rate": 0.003, + "loss": 4.1525, + "step": 3041 + }, + { + "epoch": 0.03042, + "grad_norm": 0.8745864982346226, + "learning_rate": 0.003, + "loss": 4.1783, + "step": 3042 + }, + { + "epoch": 0.03043, + "grad_norm": 0.7521748986107755, + "learning_rate": 0.003, + "loss": 4.124, + "step": 3043 + }, + { + "epoch": 0.03044, + "grad_norm": 0.7324795629041659, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 3044 + }, + { + "epoch": 0.03045, + "grad_norm": 0.6924142991448835, + "learning_rate": 0.003, + "loss": 4.1282, + "step": 3045 + }, + { + "epoch": 0.03046, + "grad_norm": 0.6060629990084535, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 3046 + }, + { + "epoch": 0.03047, + "grad_norm": 0.587624527305691, + "learning_rate": 0.003, + "loss": 4.1377, + "step": 3047 + }, + { + "epoch": 0.03048, + "grad_norm": 0.6232568692689779, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 3048 + }, + { + "epoch": 0.03049, + "grad_norm": 0.6894886517421532, + "learning_rate": 0.003, + "loss": 4.11, + "step": 3049 + }, + { + "epoch": 0.0305, + "grad_norm": 0.7503365157947721, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 3050 + }, + { + "epoch": 0.03051, + "grad_norm": 0.8678920586002613, + "learning_rate": 0.003, + "loss": 4.1404, + "step": 3051 + }, + { + "epoch": 0.03052, + "grad_norm": 0.8752304528311026, + "learning_rate": 0.003, + "loss": 4.1485, + "step": 3052 + }, + { + "epoch": 0.03053, + "grad_norm": 0.7907881299110537, + "learning_rate": 0.003, + "loss": 4.1446, + "step": 3053 + }, + { + "epoch": 0.03054, + "grad_norm": 0.8192443935144303, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 3054 + }, + { + "epoch": 0.03055, + "grad_norm": 0.7995722322031535, + "learning_rate": 0.003, + "loss": 4.1338, + "step": 3055 + }, + { + "epoch": 0.03056, + "grad_norm": 0.7117761253884235, + "learning_rate": 0.003, + "loss": 4.1414, + "step": 3056 + }, + { + "epoch": 0.03057, + "grad_norm": 0.6287453765382515, + "learning_rate": 0.003, + "loss": 4.1364, + "step": 3057 + }, + { + "epoch": 0.03058, + "grad_norm": 0.5236381931760589, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 3058 + }, + { + "epoch": 0.03059, + "grad_norm": 0.49346412015239594, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 3059 + }, + { + "epoch": 0.0306, + "grad_norm": 0.531819203867298, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 3060 + }, + { + "epoch": 0.03061, + "grad_norm": 0.5165993493909542, + "learning_rate": 0.003, + "loss": 4.126, + "step": 3061 + }, + { + "epoch": 0.03062, + "grad_norm": 0.549360696360483, + "learning_rate": 0.003, + "loss": 4.149, + "step": 3062 + }, + { + "epoch": 0.03063, + "grad_norm": 0.5202059591921928, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 3063 + }, + { + "epoch": 0.03064, + "grad_norm": 0.5384372757145418, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 3064 + }, + { + "epoch": 0.03065, + "grad_norm": 0.6204359493281749, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 3065 + }, + { + "epoch": 0.03066, + "grad_norm": 0.5858327565498244, + "learning_rate": 0.003, + "loss": 4.1357, + "step": 3066 + }, + { + "epoch": 0.03067, + "grad_norm": 0.4615503449616634, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 3067 + }, + { + "epoch": 0.03068, + "grad_norm": 0.5687493846057298, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 3068 + }, + { + "epoch": 0.03069, + "grad_norm": 0.7571890300599649, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 3069 + }, + { + "epoch": 0.0307, + "grad_norm": 0.9684439261975131, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3070 + }, + { + "epoch": 0.03071, + "grad_norm": 0.9773476262524128, + "learning_rate": 0.003, + "loss": 4.1273, + "step": 3071 + }, + { + "epoch": 0.03072, + "grad_norm": 0.8757423171832723, + "learning_rate": 0.003, + "loss": 4.127, + "step": 3072 + }, + { + "epoch": 0.03073, + "grad_norm": 0.7498562493695595, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 3073 + }, + { + "epoch": 0.03074, + "grad_norm": 0.7622450157686376, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 3074 + }, + { + "epoch": 0.03075, + "grad_norm": 0.7110114223967975, + "learning_rate": 0.003, + "loss": 4.1482, + "step": 3075 + }, + { + "epoch": 0.03076, + "grad_norm": 0.742775436261942, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 3076 + }, + { + "epoch": 0.03077, + "grad_norm": 0.676167918634839, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 3077 + }, + { + "epoch": 0.03078, + "grad_norm": 0.6544497343861535, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3078 + }, + { + "epoch": 0.03079, + "grad_norm": 0.6291167247174531, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 3079 + }, + { + "epoch": 0.0308, + "grad_norm": 0.6415913525881871, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 3080 + }, + { + "epoch": 0.03081, + "grad_norm": 0.7076666838760035, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 3081 + }, + { + "epoch": 0.03082, + "grad_norm": 0.779988064371264, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 3082 + }, + { + "epoch": 0.03083, + "grad_norm": 0.7845402321190362, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3083 + }, + { + "epoch": 0.03084, + "grad_norm": 0.6639542336722029, + "learning_rate": 0.003, + "loss": 4.1284, + "step": 3084 + }, + { + "epoch": 0.03085, + "grad_norm": 0.6819164544747766, + "learning_rate": 0.003, + "loss": 4.121, + "step": 3085 + }, + { + "epoch": 0.03086, + "grad_norm": 0.6820830017722485, + "learning_rate": 0.003, + "loss": 4.1434, + "step": 3086 + }, + { + "epoch": 0.03087, + "grad_norm": 0.6367600887185143, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 3087 + }, + { + "epoch": 0.03088, + "grad_norm": 0.7512443202409169, + "learning_rate": 0.003, + "loss": 4.141, + "step": 3088 + }, + { + "epoch": 0.03089, + "grad_norm": 0.7908237036004492, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 3089 + }, + { + "epoch": 0.0309, + "grad_norm": 0.7370194219461084, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3090 + }, + { + "epoch": 0.03091, + "grad_norm": 0.6529848345152789, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 3091 + }, + { + "epoch": 0.03092, + "grad_norm": 0.621325350966411, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 3092 + }, + { + "epoch": 0.03093, + "grad_norm": 0.6153261038514841, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 3093 + }, + { + "epoch": 0.03094, + "grad_norm": 0.6439261561218894, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 3094 + }, + { + "epoch": 0.03095, + "grad_norm": 0.6388307147361507, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 3095 + }, + { + "epoch": 0.03096, + "grad_norm": 0.6464786530825956, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 3096 + }, + { + "epoch": 0.03097, + "grad_norm": 0.6686657944539134, + "learning_rate": 0.003, + "loss": 4.1297, + "step": 3097 + }, + { + "epoch": 0.03098, + "grad_norm": 0.6274502748485118, + "learning_rate": 0.003, + "loss": 4.1246, + "step": 3098 + }, + { + "epoch": 0.03099, + "grad_norm": 0.5989042485666842, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 3099 + }, + { + "epoch": 0.031, + "grad_norm": 0.6747907015515466, + "learning_rate": 0.003, + "loss": 4.1157, + "step": 3100 + }, + { + "epoch": 0.03101, + "grad_norm": 0.6458202114886833, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 3101 + }, + { + "epoch": 0.03102, + "grad_norm": 0.6345607061133579, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 3102 + }, + { + "epoch": 0.03103, + "grad_norm": 0.7580472336670743, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 3103 + }, + { + "epoch": 0.03104, + "grad_norm": 0.8807116376909954, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 3104 + }, + { + "epoch": 0.03105, + "grad_norm": 0.9776071587101375, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 3105 + }, + { + "epoch": 0.03106, + "grad_norm": 0.9780028835386098, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 3106 + }, + { + "epoch": 0.03107, + "grad_norm": 0.9002666019108568, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 3107 + }, + { + "epoch": 0.03108, + "grad_norm": 0.8044719769060891, + "learning_rate": 0.003, + "loss": 4.1091, + "step": 3108 + }, + { + "epoch": 0.03109, + "grad_norm": 0.7755286757066427, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 3109 + }, + { + "epoch": 0.0311, + "grad_norm": 0.6914540961575923, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 3110 + }, + { + "epoch": 0.03111, + "grad_norm": 0.6947246605853371, + "learning_rate": 0.003, + "loss": 4.113, + "step": 3111 + }, + { + "epoch": 0.03112, + "grad_norm": 0.6894897260670371, + "learning_rate": 0.003, + "loss": 4.1343, + "step": 3112 + }, + { + "epoch": 0.03113, + "grad_norm": 0.7445737386600938, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 3113 + }, + { + "epoch": 0.03114, + "grad_norm": 0.8217574054645302, + "learning_rate": 0.003, + "loss": 4.127, + "step": 3114 + }, + { + "epoch": 0.03115, + "grad_norm": 0.8150291013016293, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 3115 + }, + { + "epoch": 0.03116, + "grad_norm": 0.776039406771769, + "learning_rate": 0.003, + "loss": 4.1151, + "step": 3116 + }, + { + "epoch": 0.03117, + "grad_norm": 0.7314741791885638, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 3117 + }, + { + "epoch": 0.03118, + "grad_norm": 0.716876149343852, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 3118 + }, + { + "epoch": 0.03119, + "grad_norm": 0.7781988415265245, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 3119 + }, + { + "epoch": 0.0312, + "grad_norm": 0.8587787606797088, + "learning_rate": 0.003, + "loss": 4.1469, + "step": 3120 + }, + { + "epoch": 0.03121, + "grad_norm": 0.7776701867972049, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 3121 + }, + { + "epoch": 0.03122, + "grad_norm": 0.6517901221927174, + "learning_rate": 0.003, + "loss": 4.1371, + "step": 3122 + }, + { + "epoch": 0.03123, + "grad_norm": 0.611011295800404, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 3123 + }, + { + "epoch": 0.03124, + "grad_norm": 0.6198028418813943, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 3124 + }, + { + "epoch": 0.03125, + "grad_norm": 0.6582222160573743, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 3125 + }, + { + "epoch": 0.03126, + "grad_norm": 0.5168188926246537, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 3126 + }, + { + "epoch": 0.03127, + "grad_norm": 0.5441632757270384, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 3127 + }, + { + "epoch": 0.03128, + "grad_norm": 0.5513891902503286, + "learning_rate": 0.003, + "loss": 4.107, + "step": 3128 + }, + { + "epoch": 0.03129, + "grad_norm": 0.5937504432335633, + "learning_rate": 0.003, + "loss": 4.1241, + "step": 3129 + }, + { + "epoch": 0.0313, + "grad_norm": 0.6875576103568961, + "learning_rate": 0.003, + "loss": 4.1331, + "step": 3130 + }, + { + "epoch": 0.03131, + "grad_norm": 0.6476143879877436, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 3131 + }, + { + "epoch": 0.03132, + "grad_norm": 0.5692364353277257, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 3132 + }, + { + "epoch": 0.03133, + "grad_norm": 0.6327666219121378, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 3133 + }, + { + "epoch": 0.03134, + "grad_norm": 0.6408561169351972, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 3134 + }, + { + "epoch": 0.03135, + "grad_norm": 0.5581220549438155, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 3135 + }, + { + "epoch": 0.03136, + "grad_norm": 0.5723395177501843, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 3136 + }, + { + "epoch": 0.03137, + "grad_norm": 0.6319020855348492, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 3137 + }, + { + "epoch": 0.03138, + "grad_norm": 0.6531409289943023, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 3138 + }, + { + "epoch": 0.03139, + "grad_norm": 0.7357745473070502, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3139 + }, + { + "epoch": 0.0314, + "grad_norm": 0.7384381595286016, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 3140 + }, + { + "epoch": 0.03141, + "grad_norm": 0.6788418950416313, + "learning_rate": 0.003, + "loss": 4.1465, + "step": 3141 + }, + { + "epoch": 0.03142, + "grad_norm": 0.7676098693850516, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 3142 + }, + { + "epoch": 0.03143, + "grad_norm": 0.7025497017025326, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 3143 + }, + { + "epoch": 0.03144, + "grad_norm": 0.6888741800565794, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 3144 + }, + { + "epoch": 0.03145, + "grad_norm": 0.6322894060721962, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 3145 + }, + { + "epoch": 0.03146, + "grad_norm": 0.6596496859438639, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3146 + }, + { + "epoch": 0.03147, + "grad_norm": 0.733264003225595, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 3147 + }, + { + "epoch": 0.03148, + "grad_norm": 0.6748371556539176, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3148 + }, + { + "epoch": 0.03149, + "grad_norm": 0.851512750970393, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 3149 + }, + { + "epoch": 0.0315, + "grad_norm": 0.8049510001144058, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 3150 + }, + { + "epoch": 0.03151, + "grad_norm": 0.79116005443752, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 3151 + }, + { + "epoch": 0.03152, + "grad_norm": 1.0376561914501656, + "learning_rate": 0.003, + "loss": 4.1403, + "step": 3152 + }, + { + "epoch": 0.03153, + "grad_norm": 1.1927909613612722, + "learning_rate": 0.003, + "loss": 4.1625, + "step": 3153 + }, + { + "epoch": 0.03154, + "grad_norm": 0.7630767355207296, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 3154 + }, + { + "epoch": 0.03155, + "grad_norm": 0.6762123885129709, + "learning_rate": 0.003, + "loss": 4.1295, + "step": 3155 + }, + { + "epoch": 0.03156, + "grad_norm": 0.6636191569872171, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 3156 + }, + { + "epoch": 0.03157, + "grad_norm": 0.6499423709377473, + "learning_rate": 0.003, + "loss": 4.1356, + "step": 3157 + }, + { + "epoch": 0.03158, + "grad_norm": 0.552206041395336, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 3158 + }, + { + "epoch": 0.03159, + "grad_norm": 0.5182750861462773, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 3159 + }, + { + "epoch": 0.0316, + "grad_norm": 0.5736268849520852, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 3160 + }, + { + "epoch": 0.03161, + "grad_norm": 0.5991039624831055, + "learning_rate": 0.003, + "loss": 4.1382, + "step": 3161 + }, + { + "epoch": 0.03162, + "grad_norm": 0.5947855026308295, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 3162 + }, + { + "epoch": 0.03163, + "grad_norm": 0.6189285110953556, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 3163 + }, + { + "epoch": 0.03164, + "grad_norm": 0.6633162626114895, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 3164 + }, + { + "epoch": 0.03165, + "grad_norm": 0.7032964547449561, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 3165 + }, + { + "epoch": 0.03166, + "grad_norm": 0.6809598557008177, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 3166 + }, + { + "epoch": 0.03167, + "grad_norm": 0.5685007785118529, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 3167 + }, + { + "epoch": 0.03168, + "grad_norm": 0.6110451444106639, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 3168 + }, + { + "epoch": 0.03169, + "grad_norm": 0.6511765155667689, + "learning_rate": 0.003, + "loss": 4.1349, + "step": 3169 + }, + { + "epoch": 0.0317, + "grad_norm": 0.6213375365540387, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 3170 + }, + { + "epoch": 0.03171, + "grad_norm": 0.5992859314044605, + "learning_rate": 0.003, + "loss": 4.1274, + "step": 3171 + }, + { + "epoch": 0.03172, + "grad_norm": 0.7257303986858412, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 3172 + }, + { + "epoch": 0.03173, + "grad_norm": 0.7045314858588687, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 3173 + }, + { + "epoch": 0.03174, + "grad_norm": 0.612406881062346, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 3174 + }, + { + "epoch": 0.03175, + "grad_norm": 0.614783966308366, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 3175 + }, + { + "epoch": 0.03176, + "grad_norm": 0.6294514934763787, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 3176 + }, + { + "epoch": 0.03177, + "grad_norm": 0.6263015270891977, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 3177 + }, + { + "epoch": 0.03178, + "grad_norm": 0.6413811720921261, + "learning_rate": 0.003, + "loss": 4.1495, + "step": 3178 + }, + { + "epoch": 0.03179, + "grad_norm": 0.6332246855150367, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 3179 + }, + { + "epoch": 0.0318, + "grad_norm": 0.6696552438679908, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 3180 + }, + { + "epoch": 0.03181, + "grad_norm": 0.7065450750059905, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 3181 + }, + { + "epoch": 0.03182, + "grad_norm": 0.7679625195572235, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 3182 + }, + { + "epoch": 0.03183, + "grad_norm": 0.8738824642888428, + "learning_rate": 0.003, + "loss": 4.1309, + "step": 3183 + }, + { + "epoch": 0.03184, + "grad_norm": 0.9064159168435794, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 3184 + }, + { + "epoch": 0.03185, + "grad_norm": 0.8550427300013329, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 3185 + }, + { + "epoch": 0.03186, + "grad_norm": 0.6551579840777808, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 3186 + }, + { + "epoch": 0.03187, + "grad_norm": 0.7094751892506671, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 3187 + }, + { + "epoch": 0.03188, + "grad_norm": 0.7344511062881034, + "learning_rate": 0.003, + "loss": 4.1203, + "step": 3188 + }, + { + "epoch": 0.03189, + "grad_norm": 0.7623868515040703, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 3189 + }, + { + "epoch": 0.0319, + "grad_norm": 0.7621303628372581, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 3190 + }, + { + "epoch": 0.03191, + "grad_norm": 0.9483714426871703, + "learning_rate": 0.003, + "loss": 4.1496, + "step": 3191 + }, + { + "epoch": 0.03192, + "grad_norm": 1.02326337686603, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 3192 + }, + { + "epoch": 0.03193, + "grad_norm": 0.9529315196214988, + "learning_rate": 0.003, + "loss": 4.1323, + "step": 3193 + }, + { + "epoch": 0.03194, + "grad_norm": 0.9316148461825665, + "learning_rate": 0.003, + "loss": 4.1598, + "step": 3194 + }, + { + "epoch": 0.03195, + "grad_norm": 0.9802297659007858, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 3195 + }, + { + "epoch": 0.03196, + "grad_norm": 1.1026234014638998, + "learning_rate": 0.003, + "loss": 4.1478, + "step": 3196 + }, + { + "epoch": 0.03197, + "grad_norm": 0.9480265649122571, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 3197 + }, + { + "epoch": 0.03198, + "grad_norm": 1.1388718376675486, + "learning_rate": 0.003, + "loss": 4.1536, + "step": 3198 + }, + { + "epoch": 0.03199, + "grad_norm": 1.0081692469141126, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 3199 + }, + { + "epoch": 0.032, + "grad_norm": 1.0517580958620087, + "learning_rate": 0.003, + "loss": 4.153, + "step": 3200 + }, + { + "epoch": 0.03201, + "grad_norm": 0.8409805916282491, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 3201 + }, + { + "epoch": 0.03202, + "grad_norm": 0.7378220920813567, + "learning_rate": 0.003, + "loss": 4.1453, + "step": 3202 + }, + { + "epoch": 0.03203, + "grad_norm": 0.7706024735219662, + "learning_rate": 0.003, + "loss": 4.1264, + "step": 3203 + }, + { + "epoch": 0.03204, + "grad_norm": 0.7926317772737794, + "learning_rate": 0.003, + "loss": 4.1458, + "step": 3204 + }, + { + "epoch": 0.03205, + "grad_norm": 0.7029254241866181, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 3205 + }, + { + "epoch": 0.03206, + "grad_norm": 0.591794603564226, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 3206 + }, + { + "epoch": 0.03207, + "grad_norm": 0.5290100761472903, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 3207 + }, + { + "epoch": 0.03208, + "grad_norm": 0.60614264303989, + "learning_rate": 0.003, + "loss": 4.1305, + "step": 3208 + }, + { + "epoch": 0.03209, + "grad_norm": 0.6842751597180756, + "learning_rate": 0.003, + "loss": 4.1657, + "step": 3209 + }, + { + "epoch": 0.0321, + "grad_norm": 0.7775258875142942, + "learning_rate": 0.003, + "loss": 4.1493, + "step": 3210 + }, + { + "epoch": 0.03211, + "grad_norm": 0.8898686097125754, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 3211 + }, + { + "epoch": 0.03212, + "grad_norm": 0.8601156764415516, + "learning_rate": 0.003, + "loss": 4.1344, + "step": 3212 + }, + { + "epoch": 0.03213, + "grad_norm": 0.6327806087127427, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3213 + }, + { + "epoch": 0.03214, + "grad_norm": 0.5270111703400867, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 3214 + }, + { + "epoch": 0.03215, + "grad_norm": 0.532626362926619, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 3215 + }, + { + "epoch": 0.03216, + "grad_norm": 0.5536165578098594, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 3216 + }, + { + "epoch": 0.03217, + "grad_norm": 0.5659471039370246, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 3217 + }, + { + "epoch": 0.03218, + "grad_norm": 0.4739424052144777, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 3218 + }, + { + "epoch": 0.03219, + "grad_norm": 0.45647876929555886, + "learning_rate": 0.003, + "loss": 4.107, + "step": 3219 + }, + { + "epoch": 0.0322, + "grad_norm": 0.5300531473518815, + "learning_rate": 0.003, + "loss": 4.119, + "step": 3220 + }, + { + "epoch": 0.03221, + "grad_norm": 0.5666259975676212, + "learning_rate": 0.003, + "loss": 4.1248, + "step": 3221 + }, + { + "epoch": 0.03222, + "grad_norm": 0.5769078015222051, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3222 + }, + { + "epoch": 0.03223, + "grad_norm": 0.6807752622848917, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 3223 + }, + { + "epoch": 0.03224, + "grad_norm": 0.6566831883335548, + "learning_rate": 0.003, + "loss": 4.081, + "step": 3224 + }, + { + "epoch": 0.03225, + "grad_norm": 0.6269821235625146, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 3225 + }, + { + "epoch": 0.03226, + "grad_norm": 0.6845667285120688, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 3226 + }, + { + "epoch": 0.03227, + "grad_norm": 0.6907992262129431, + "learning_rate": 0.003, + "loss": 4.1158, + "step": 3227 + }, + { + "epoch": 0.03228, + "grad_norm": 0.7082508546642823, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 3228 + }, + { + "epoch": 0.03229, + "grad_norm": 0.5553247224061585, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 3229 + }, + { + "epoch": 0.0323, + "grad_norm": 0.4833936225921864, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 3230 + }, + { + "epoch": 0.03231, + "grad_norm": 0.508445371164304, + "learning_rate": 0.003, + "loss": 4.114, + "step": 3231 + }, + { + "epoch": 0.03232, + "grad_norm": 0.515209314547404, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 3232 + }, + { + "epoch": 0.03233, + "grad_norm": 0.5619772875869593, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 3233 + }, + { + "epoch": 0.03234, + "grad_norm": 0.5996079981675662, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 3234 + }, + { + "epoch": 0.03235, + "grad_norm": 0.6517090954163339, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 3235 + }, + { + "epoch": 0.03236, + "grad_norm": 0.6450234906982093, + "learning_rate": 0.003, + "loss": 4.104, + "step": 3236 + }, + { + "epoch": 0.03237, + "grad_norm": 0.5550128762421827, + "learning_rate": 0.003, + "loss": 4.123, + "step": 3237 + }, + { + "epoch": 0.03238, + "grad_norm": 0.44865332646977973, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 3238 + }, + { + "epoch": 0.03239, + "grad_norm": 0.4851640824384115, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 3239 + }, + { + "epoch": 0.0324, + "grad_norm": 0.5854398306854629, + "learning_rate": 0.003, + "loss": 4.1228, + "step": 3240 + }, + { + "epoch": 0.03241, + "grad_norm": 0.5878891088924086, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 3241 + }, + { + "epoch": 0.03242, + "grad_norm": 0.6049093430695732, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 3242 + }, + { + "epoch": 0.03243, + "grad_norm": 0.5903284236245521, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 3243 + }, + { + "epoch": 0.03244, + "grad_norm": 0.6115977539570783, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 3244 + }, + { + "epoch": 0.03245, + "grad_norm": 0.7190082548141721, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 3245 + }, + { + "epoch": 0.03246, + "grad_norm": 0.7482030743366556, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 3246 + }, + { + "epoch": 0.03247, + "grad_norm": 0.8289542781973179, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 3247 + }, + { + "epoch": 0.03248, + "grad_norm": 0.8470264014053567, + "learning_rate": 0.003, + "loss": 4.1424, + "step": 3248 + }, + { + "epoch": 0.03249, + "grad_norm": 0.9980300498793666, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3249 + }, + { + "epoch": 0.0325, + "grad_norm": 0.9714365913666329, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 3250 + }, + { + "epoch": 0.03251, + "grad_norm": 1.1238874142120285, + "learning_rate": 0.003, + "loss": 4.1627, + "step": 3251 + }, + { + "epoch": 0.03252, + "grad_norm": 0.8199564469617361, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 3252 + }, + { + "epoch": 0.03253, + "grad_norm": 0.8450268186634116, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 3253 + }, + { + "epoch": 0.03254, + "grad_norm": 0.8533746390280161, + "learning_rate": 0.003, + "loss": 4.1198, + "step": 3254 + }, + { + "epoch": 0.03255, + "grad_norm": 0.7528067000570253, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3255 + }, + { + "epoch": 0.03256, + "grad_norm": 0.6385551011393206, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 3256 + }, + { + "epoch": 0.03257, + "grad_norm": 0.7031639597939752, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 3257 + }, + { + "epoch": 0.03258, + "grad_norm": 0.8271897507400796, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 3258 + }, + { + "epoch": 0.03259, + "grad_norm": 1.0050892315882949, + "learning_rate": 0.003, + "loss": 4.1407, + "step": 3259 + }, + { + "epoch": 0.0326, + "grad_norm": 0.9908981489981602, + "learning_rate": 0.003, + "loss": 4.1378, + "step": 3260 + }, + { + "epoch": 0.03261, + "grad_norm": 0.9087527160981669, + "learning_rate": 0.003, + "loss": 4.11, + "step": 3261 + }, + { + "epoch": 0.03262, + "grad_norm": 0.8187132577967141, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 3262 + }, + { + "epoch": 0.03263, + "grad_norm": 0.8689798444062119, + "learning_rate": 0.003, + "loss": 4.1439, + "step": 3263 + }, + { + "epoch": 0.03264, + "grad_norm": 0.8439047022001098, + "learning_rate": 0.003, + "loss": 4.1581, + "step": 3264 + }, + { + "epoch": 0.03265, + "grad_norm": 0.8890020599458566, + "learning_rate": 0.003, + "loss": 4.1545, + "step": 3265 + }, + { + "epoch": 0.03266, + "grad_norm": 0.8443986192708829, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 3266 + }, + { + "epoch": 0.03267, + "grad_norm": 0.6943132611741926, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 3267 + }, + { + "epoch": 0.03268, + "grad_norm": 0.6818596077689152, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 3268 + }, + { + "epoch": 0.03269, + "grad_norm": 0.6442065545268074, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 3269 + }, + { + "epoch": 0.0327, + "grad_norm": 0.6475427138691676, + "learning_rate": 0.003, + "loss": 4.1531, + "step": 3270 + }, + { + "epoch": 0.03271, + "grad_norm": 0.5839071309649821, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 3271 + }, + { + "epoch": 0.03272, + "grad_norm": 0.5588346130923799, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 3272 + }, + { + "epoch": 0.03273, + "grad_norm": 0.5429760238191433, + "learning_rate": 0.003, + "loss": 4.1173, + "step": 3273 + }, + { + "epoch": 0.03274, + "grad_norm": 0.5519539740214184, + "learning_rate": 0.003, + "loss": 4.134, + "step": 3274 + }, + { + "epoch": 0.03275, + "grad_norm": 0.5044788109597094, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 3275 + }, + { + "epoch": 0.03276, + "grad_norm": 0.5410434542093497, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 3276 + }, + { + "epoch": 0.03277, + "grad_norm": 0.5174842971937997, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 3277 + }, + { + "epoch": 0.03278, + "grad_norm": 0.5397398510984489, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 3278 + }, + { + "epoch": 0.03279, + "grad_norm": 0.5310270556745288, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 3279 + }, + { + "epoch": 0.0328, + "grad_norm": 0.5319284745301865, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 3280 + }, + { + "epoch": 0.03281, + "grad_norm": 0.5524652117491456, + "learning_rate": 0.003, + "loss": 4.1311, + "step": 3281 + }, + { + "epoch": 0.03282, + "grad_norm": 0.6685924963163522, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 3282 + }, + { + "epoch": 0.03283, + "grad_norm": 0.7948935779606678, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 3283 + }, + { + "epoch": 0.03284, + "grad_norm": 0.8845797923720062, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 3284 + }, + { + "epoch": 0.03285, + "grad_norm": 1.0237302815471576, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 3285 + }, + { + "epoch": 0.03286, + "grad_norm": 0.8871634657347365, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 3286 + }, + { + "epoch": 0.03287, + "grad_norm": 0.6589949194541027, + "learning_rate": 0.003, + "loss": 4.1289, + "step": 3287 + }, + { + "epoch": 0.03288, + "grad_norm": 0.6316707636138289, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 3288 + }, + { + "epoch": 0.03289, + "grad_norm": 0.5614054682127885, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 3289 + }, + { + "epoch": 0.0329, + "grad_norm": 0.5805560523196955, + "learning_rate": 0.003, + "loss": 4.14, + "step": 3290 + }, + { + "epoch": 0.03291, + "grad_norm": 0.5446479920144529, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3291 + }, + { + "epoch": 0.03292, + "grad_norm": 0.6319435521626124, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 3292 + }, + { + "epoch": 0.03293, + "grad_norm": 0.6965122481103596, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 3293 + }, + { + "epoch": 0.03294, + "grad_norm": 0.6965725499961043, + "learning_rate": 0.003, + "loss": 4.112, + "step": 3294 + }, + { + "epoch": 0.03295, + "grad_norm": 0.713730043155873, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 3295 + }, + { + "epoch": 0.03296, + "grad_norm": 0.7754035666178799, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 3296 + }, + { + "epoch": 0.03297, + "grad_norm": 0.7907058646333082, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 3297 + }, + { + "epoch": 0.03298, + "grad_norm": 0.7675790084798387, + "learning_rate": 0.003, + "loss": 4.1329, + "step": 3298 + }, + { + "epoch": 0.03299, + "grad_norm": 0.7780707471732844, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 3299 + }, + { + "epoch": 0.033, + "grad_norm": 0.6786231813624723, + "learning_rate": 0.003, + "loss": 4.102, + "step": 3300 + }, + { + "epoch": 0.03301, + "grad_norm": 0.6934578703759183, + "learning_rate": 0.003, + "loss": 4.135, + "step": 3301 + }, + { + "epoch": 0.03302, + "grad_norm": 0.6989193090870045, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 3302 + }, + { + "epoch": 0.03303, + "grad_norm": 0.6846564265330911, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 3303 + }, + { + "epoch": 0.03304, + "grad_norm": 0.6507328383119774, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 3304 + }, + { + "epoch": 0.03305, + "grad_norm": 0.6708718906871574, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 3305 + }, + { + "epoch": 0.03306, + "grad_norm": 0.6357733156987033, + "learning_rate": 0.003, + "loss": 4.1306, + "step": 3306 + }, + { + "epoch": 0.03307, + "grad_norm": 0.6261259341466859, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 3307 + }, + { + "epoch": 0.03308, + "grad_norm": 0.6912986058583214, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 3308 + }, + { + "epoch": 0.03309, + "grad_norm": 0.7338476936987169, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 3309 + }, + { + "epoch": 0.0331, + "grad_norm": 0.6124514330382202, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 3310 + }, + { + "epoch": 0.03311, + "grad_norm": 0.5872103273958902, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 3311 + }, + { + "epoch": 0.03312, + "grad_norm": 0.6203568310068858, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 3312 + }, + { + "epoch": 0.03313, + "grad_norm": 0.5552557630995372, + "learning_rate": 0.003, + "loss": 4.118, + "step": 3313 + }, + { + "epoch": 0.03314, + "grad_norm": 0.5291443435730824, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 3314 + }, + { + "epoch": 0.03315, + "grad_norm": 0.5668780535216992, + "learning_rate": 0.003, + "loss": 4.1103, + "step": 3315 + }, + { + "epoch": 0.03316, + "grad_norm": 0.6218583286817391, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3316 + }, + { + "epoch": 0.03317, + "grad_norm": 0.6954294964882046, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 3317 + }, + { + "epoch": 0.03318, + "grad_norm": 0.7185698539541155, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 3318 + }, + { + "epoch": 0.03319, + "grad_norm": 0.6660062925784288, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 3319 + }, + { + "epoch": 0.0332, + "grad_norm": 0.8828120328467459, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 3320 + }, + { + "epoch": 0.03321, + "grad_norm": 1.1045400160364702, + "learning_rate": 0.003, + "loss": 4.1454, + "step": 3321 + }, + { + "epoch": 0.03322, + "grad_norm": 0.9313986368893203, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 3322 + }, + { + "epoch": 0.03323, + "grad_norm": 0.8436631996856868, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 3323 + }, + { + "epoch": 0.03324, + "grad_norm": 0.793414023613159, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 3324 + }, + { + "epoch": 0.03325, + "grad_norm": 0.8058692210593774, + "learning_rate": 0.003, + "loss": 4.1381, + "step": 3325 + }, + { + "epoch": 0.03326, + "grad_norm": 0.7718759752369884, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 3326 + }, + { + "epoch": 0.03327, + "grad_norm": 0.6377307500129057, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 3327 + }, + { + "epoch": 0.03328, + "grad_norm": 0.5845333255559306, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 3328 + }, + { + "epoch": 0.03329, + "grad_norm": 0.6432664245378488, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 3329 + }, + { + "epoch": 0.0333, + "grad_norm": 0.7076417009283467, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 3330 + }, + { + "epoch": 0.03331, + "grad_norm": 0.7024104140652564, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 3331 + }, + { + "epoch": 0.03332, + "grad_norm": 0.775257090778202, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 3332 + }, + { + "epoch": 0.03333, + "grad_norm": 0.7911285827248639, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 3333 + }, + { + "epoch": 0.03334, + "grad_norm": 0.743233559209534, + "learning_rate": 0.003, + "loss": 4.1199, + "step": 3334 + }, + { + "epoch": 0.03335, + "grad_norm": 0.6783362132091875, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3335 + }, + { + "epoch": 0.03336, + "grad_norm": 0.6979858127763721, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3336 + }, + { + "epoch": 0.03337, + "grad_norm": 0.6453147133509979, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 3337 + }, + { + "epoch": 0.03338, + "grad_norm": 0.589458340925199, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 3338 + }, + { + "epoch": 0.03339, + "grad_norm": 0.6008794437305915, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 3339 + }, + { + "epoch": 0.0334, + "grad_norm": 0.5286553121798908, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 3340 + }, + { + "epoch": 0.03341, + "grad_norm": 0.4844921137965909, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 3341 + }, + { + "epoch": 0.03342, + "grad_norm": 0.5282684944251612, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 3342 + }, + { + "epoch": 0.03343, + "grad_norm": 0.5228578071551047, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 3343 + }, + { + "epoch": 0.03344, + "grad_norm": 0.5362037099167439, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 3344 + }, + { + "epoch": 0.03345, + "grad_norm": 0.6243318919509568, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 3345 + }, + { + "epoch": 0.03346, + "grad_norm": 0.8010216755297692, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 3346 + }, + { + "epoch": 0.03347, + "grad_norm": 1.0244615325473108, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 3347 + }, + { + "epoch": 0.03348, + "grad_norm": 0.8823905163660845, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 3348 + }, + { + "epoch": 0.03349, + "grad_norm": 0.7164478558725996, + "learning_rate": 0.003, + "loss": 4.1444, + "step": 3349 + }, + { + "epoch": 0.0335, + "grad_norm": 0.9083477796589686, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 3350 + }, + { + "epoch": 0.03351, + "grad_norm": 1.1253404122518653, + "learning_rate": 0.003, + "loss": 4.131, + "step": 3351 + }, + { + "epoch": 0.03352, + "grad_norm": 0.8510048664403838, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 3352 + }, + { + "epoch": 0.03353, + "grad_norm": 0.8486138897847525, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 3353 + }, + { + "epoch": 0.03354, + "grad_norm": 0.8166125988681096, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 3354 + }, + { + "epoch": 0.03355, + "grad_norm": 0.857862353759048, + "learning_rate": 0.003, + "loss": 4.1278, + "step": 3355 + }, + { + "epoch": 0.03356, + "grad_norm": 0.8584048211762879, + "learning_rate": 0.003, + "loss": 4.1537, + "step": 3356 + }, + { + "epoch": 0.03357, + "grad_norm": 0.979299832982737, + "learning_rate": 0.003, + "loss": 4.1313, + "step": 3357 + }, + { + "epoch": 0.03358, + "grad_norm": 1.039797443287391, + "learning_rate": 0.003, + "loss": 4.1559, + "step": 3358 + }, + { + "epoch": 0.03359, + "grad_norm": 0.8245295692820709, + "learning_rate": 0.003, + "loss": 4.1387, + "step": 3359 + }, + { + "epoch": 0.0336, + "grad_norm": 0.8180499509004296, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 3360 + }, + { + "epoch": 0.03361, + "grad_norm": 0.8269011372309851, + "learning_rate": 0.003, + "loss": 4.1505, + "step": 3361 + }, + { + "epoch": 0.03362, + "grad_norm": 0.6572000096032281, + "learning_rate": 0.003, + "loss": 4.1416, + "step": 3362 + }, + { + "epoch": 0.03363, + "grad_norm": 0.5034339182168021, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 3363 + }, + { + "epoch": 0.03364, + "grad_norm": 0.5842791784122281, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 3364 + }, + { + "epoch": 0.03365, + "grad_norm": 0.5402809877026816, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 3365 + }, + { + "epoch": 0.03366, + "grad_norm": 0.517439188525011, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 3366 + }, + { + "epoch": 0.03367, + "grad_norm": 0.4829482643334095, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 3367 + }, + { + "epoch": 0.03368, + "grad_norm": 0.4770719752203237, + "learning_rate": 0.003, + "loss": 4.1395, + "step": 3368 + }, + { + "epoch": 0.03369, + "grad_norm": 0.4686341954973017, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 3369 + }, + { + "epoch": 0.0337, + "grad_norm": 0.4635534349009037, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 3370 + }, + { + "epoch": 0.03371, + "grad_norm": 0.4387018994882627, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 3371 + }, + { + "epoch": 0.03372, + "grad_norm": 0.4624280577146149, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 3372 + }, + { + "epoch": 0.03373, + "grad_norm": 0.6134335245918775, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 3373 + }, + { + "epoch": 0.03374, + "grad_norm": 0.7665799931133178, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 3374 + }, + { + "epoch": 0.03375, + "grad_norm": 0.8584429511076206, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 3375 + }, + { + "epoch": 0.03376, + "grad_norm": 0.7797088300192078, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 3376 + }, + { + "epoch": 0.03377, + "grad_norm": 0.6370846643821149, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 3377 + }, + { + "epoch": 0.03378, + "grad_norm": 0.6345482998854712, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 3378 + }, + { + "epoch": 0.03379, + "grad_norm": 0.6635912081503885, + "learning_rate": 0.003, + "loss": 4.097, + "step": 3379 + }, + { + "epoch": 0.0338, + "grad_norm": 0.6712818971351722, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 3380 + }, + { + "epoch": 0.03381, + "grad_norm": 0.7373126925808033, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 3381 + }, + { + "epoch": 0.03382, + "grad_norm": 0.6443090947007198, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 3382 + }, + { + "epoch": 0.03383, + "grad_norm": 0.7092968962629292, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 3383 + }, + { + "epoch": 0.03384, + "grad_norm": 0.6558386353215722, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 3384 + }, + { + "epoch": 0.03385, + "grad_norm": 0.7156641214660155, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 3385 + }, + { + "epoch": 0.03386, + "grad_norm": 0.6417801933987963, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 3386 + }, + { + "epoch": 0.03387, + "grad_norm": 0.6184499120119181, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 3387 + }, + { + "epoch": 0.03388, + "grad_norm": 0.6752104424629746, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 3388 + }, + { + "epoch": 0.03389, + "grad_norm": 0.6657347440305229, + "learning_rate": 0.003, + "loss": 4.1376, + "step": 3389 + }, + { + "epoch": 0.0339, + "grad_norm": 0.6329423314455942, + "learning_rate": 0.003, + "loss": 4.1212, + "step": 3390 + }, + { + "epoch": 0.03391, + "grad_norm": 0.6431276024837138, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 3391 + }, + { + "epoch": 0.03392, + "grad_norm": 0.530330223228447, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 3392 + }, + { + "epoch": 0.03393, + "grad_norm": 0.6370711752800169, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 3393 + }, + { + "epoch": 0.03394, + "grad_norm": 0.6909572991664223, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 3394 + }, + { + "epoch": 0.03395, + "grad_norm": 0.6784018752428888, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 3395 + }, + { + "epoch": 0.03396, + "grad_norm": 0.6968690029372852, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 3396 + }, + { + "epoch": 0.03397, + "grad_norm": 0.8813749008827967, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 3397 + }, + { + "epoch": 0.03398, + "grad_norm": 1.151016476609897, + "learning_rate": 0.003, + "loss": 4.1472, + "step": 3398 + }, + { + "epoch": 0.03399, + "grad_norm": 0.8041467065300499, + "learning_rate": 0.003, + "loss": 4.1558, + "step": 3399 + }, + { + "epoch": 0.034, + "grad_norm": 0.8160338387022286, + "learning_rate": 0.003, + "loss": 4.1361, + "step": 3400 + }, + { + "epoch": 0.03401, + "grad_norm": 0.8960003824737174, + "learning_rate": 0.003, + "loss": 4.117, + "step": 3401 + }, + { + "epoch": 0.03402, + "grad_norm": 1.0471658001911934, + "learning_rate": 0.003, + "loss": 4.1507, + "step": 3402 + }, + { + "epoch": 0.03403, + "grad_norm": 0.8837552583544185, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 3403 + }, + { + "epoch": 0.03404, + "grad_norm": 1.0075931068484998, + "learning_rate": 0.003, + "loss": 4.1532, + "step": 3404 + }, + { + "epoch": 0.03405, + "grad_norm": 0.8035637909890666, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 3405 + }, + { + "epoch": 0.03406, + "grad_norm": 0.7872354689221261, + "learning_rate": 0.003, + "loss": 4.1172, + "step": 3406 + }, + { + "epoch": 0.03407, + "grad_norm": 0.7456688146328043, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 3407 + }, + { + "epoch": 0.03408, + "grad_norm": 0.6698179782986525, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 3408 + }, + { + "epoch": 0.03409, + "grad_norm": 0.5900764862651436, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 3409 + }, + { + "epoch": 0.0341, + "grad_norm": 0.5872723448917179, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 3410 + }, + { + "epoch": 0.03411, + "grad_norm": 0.5956523246254345, + "learning_rate": 0.003, + "loss": 4.121, + "step": 3411 + }, + { + "epoch": 0.03412, + "grad_norm": 0.5489570202759098, + "learning_rate": 0.003, + "loss": 4.1363, + "step": 3412 + }, + { + "epoch": 0.03413, + "grad_norm": 0.5551390389247347, + "learning_rate": 0.003, + "loss": 4.1438, + "step": 3413 + }, + { + "epoch": 0.03414, + "grad_norm": 0.5092370545358439, + "learning_rate": 0.003, + "loss": 4.1314, + "step": 3414 + }, + { + "epoch": 0.03415, + "grad_norm": 0.5365180138766702, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 3415 + }, + { + "epoch": 0.03416, + "grad_norm": 0.4942714844508712, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 3416 + }, + { + "epoch": 0.03417, + "grad_norm": 0.48118275556312734, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 3417 + }, + { + "epoch": 0.03418, + "grad_norm": 0.5157820409565562, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 3418 + }, + { + "epoch": 0.03419, + "grad_norm": 0.4914489892262719, + "learning_rate": 0.003, + "loss": 4.099, + "step": 3419 + }, + { + "epoch": 0.0342, + "grad_norm": 0.45217652193784796, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 3420 + }, + { + "epoch": 0.03421, + "grad_norm": 0.4178131650206621, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 3421 + }, + { + "epoch": 0.03422, + "grad_norm": 0.5134645923404247, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 3422 + }, + { + "epoch": 0.03423, + "grad_norm": 0.6173109226400934, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 3423 + }, + { + "epoch": 0.03424, + "grad_norm": 0.7529112332315744, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 3424 + }, + { + "epoch": 0.03425, + "grad_norm": 0.8475866076688389, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 3425 + }, + { + "epoch": 0.03426, + "grad_norm": 0.7875335028924936, + "learning_rate": 0.003, + "loss": 4.1169, + "step": 3426 + }, + { + "epoch": 0.03427, + "grad_norm": 0.8040208485850641, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 3427 + }, + { + "epoch": 0.03428, + "grad_norm": 0.7900388287985982, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 3428 + }, + { + "epoch": 0.03429, + "grad_norm": 0.9067586565341079, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 3429 + }, + { + "epoch": 0.0343, + "grad_norm": 0.8296888200634449, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 3430 + }, + { + "epoch": 0.03431, + "grad_norm": 0.7554329641955093, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 3431 + }, + { + "epoch": 0.03432, + "grad_norm": 0.7333286312672143, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 3432 + }, + { + "epoch": 0.03433, + "grad_norm": 0.8319091827451316, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 3433 + }, + { + "epoch": 0.03434, + "grad_norm": 0.8723689369613313, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3434 + }, + { + "epoch": 0.03435, + "grad_norm": 0.897980024570154, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 3435 + }, + { + "epoch": 0.03436, + "grad_norm": 1.0300814251539303, + "learning_rate": 0.003, + "loss": 4.1354, + "step": 3436 + }, + { + "epoch": 0.03437, + "grad_norm": 0.9920935726876953, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 3437 + }, + { + "epoch": 0.03438, + "grad_norm": 0.9271063168023738, + "learning_rate": 0.003, + "loss": 4.1384, + "step": 3438 + }, + { + "epoch": 0.03439, + "grad_norm": 0.9386929134771245, + "learning_rate": 0.003, + "loss": 4.141, + "step": 3439 + }, + { + "epoch": 0.0344, + "grad_norm": 0.779011690077272, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 3440 + }, + { + "epoch": 0.03441, + "grad_norm": 0.6605260047832235, + "learning_rate": 0.003, + "loss": 4.1317, + "step": 3441 + }, + { + "epoch": 0.03442, + "grad_norm": 0.7172307447888016, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 3442 + }, + { + "epoch": 0.03443, + "grad_norm": 0.7108754850296115, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 3443 + }, + { + "epoch": 0.03444, + "grad_norm": 0.7442238665681021, + "learning_rate": 0.003, + "loss": 4.1375, + "step": 3444 + }, + { + "epoch": 0.03445, + "grad_norm": 0.7962056664182668, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 3445 + }, + { + "epoch": 0.03446, + "grad_norm": 0.8188766196935572, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 3446 + }, + { + "epoch": 0.03447, + "grad_norm": 0.7314232392259893, + "learning_rate": 0.003, + "loss": 4.1401, + "step": 3447 + }, + { + "epoch": 0.03448, + "grad_norm": 0.6342556675921855, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 3448 + }, + { + "epoch": 0.03449, + "grad_norm": 0.5784387076944771, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 3449 + }, + { + "epoch": 0.0345, + "grad_norm": 0.6500246176907677, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 3450 + }, + { + "epoch": 0.03451, + "grad_norm": 0.5800124232077788, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 3451 + }, + { + "epoch": 0.03452, + "grad_norm": 0.471981967033663, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 3452 + }, + { + "epoch": 0.03453, + "grad_norm": 0.42057961437919367, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 3453 + }, + { + "epoch": 0.03454, + "grad_norm": 0.3786844238003117, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 3454 + }, + { + "epoch": 0.03455, + "grad_norm": 0.44404565934168133, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 3455 + }, + { + "epoch": 0.03456, + "grad_norm": 0.4555744427224493, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 3456 + }, + { + "epoch": 0.03457, + "grad_norm": 0.46006910320859196, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 3457 + }, + { + "epoch": 0.03458, + "grad_norm": 0.5442790604536267, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 3458 + }, + { + "epoch": 0.03459, + "grad_norm": 0.6047941532792832, + "learning_rate": 0.003, + "loss": 4.1275, + "step": 3459 + }, + { + "epoch": 0.0346, + "grad_norm": 0.6654253529634946, + "learning_rate": 0.003, + "loss": 4.1291, + "step": 3460 + }, + { + "epoch": 0.03461, + "grad_norm": 0.7215965927102934, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 3461 + }, + { + "epoch": 0.03462, + "grad_norm": 0.7956917069679417, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3462 + }, + { + "epoch": 0.03463, + "grad_norm": 0.7228618043468913, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 3463 + }, + { + "epoch": 0.03464, + "grad_norm": 0.5940427112526057, + "learning_rate": 0.003, + "loss": 4.098, + "step": 3464 + }, + { + "epoch": 0.03465, + "grad_norm": 0.5853913960362132, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 3465 + }, + { + "epoch": 0.03466, + "grad_norm": 0.5651394033946199, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 3466 + }, + { + "epoch": 0.03467, + "grad_norm": 0.6125848329478759, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 3467 + }, + { + "epoch": 0.03468, + "grad_norm": 0.5664768756271987, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 3468 + }, + { + "epoch": 0.03469, + "grad_norm": 0.5326738903647605, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 3469 + }, + { + "epoch": 0.0347, + "grad_norm": 0.5169001087270545, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 3470 + }, + { + "epoch": 0.03471, + "grad_norm": 0.5111602578591978, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 3471 + }, + { + "epoch": 0.03472, + "grad_norm": 0.5394248903129965, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 3472 + }, + { + "epoch": 0.03473, + "grad_norm": 0.7180722977176366, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 3473 + }, + { + "epoch": 0.03474, + "grad_norm": 0.8549514300364189, + "learning_rate": 0.003, + "loss": 4.1353, + "step": 3474 + }, + { + "epoch": 0.03475, + "grad_norm": 0.8682275353977381, + "learning_rate": 0.003, + "loss": 4.1037, + "step": 3475 + }, + { + "epoch": 0.03476, + "grad_norm": 0.9583085828154718, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 3476 + }, + { + "epoch": 0.03477, + "grad_norm": 0.9099764585536037, + "learning_rate": 0.003, + "loss": 4.1137, + "step": 3477 + }, + { + "epoch": 0.03478, + "grad_norm": 0.7282038370583279, + "learning_rate": 0.003, + "loss": 4.117, + "step": 3478 + }, + { + "epoch": 0.03479, + "grad_norm": 0.6704056583432447, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 3479 + }, + { + "epoch": 0.0348, + "grad_norm": 0.6286750007252314, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 3480 + }, + { + "epoch": 0.03481, + "grad_norm": 0.7149650967138926, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 3481 + }, + { + "epoch": 0.03482, + "grad_norm": 0.8647872925994115, + "learning_rate": 0.003, + "loss": 4.1209, + "step": 3482 + }, + { + "epoch": 0.03483, + "grad_norm": 0.8406773732656712, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 3483 + }, + { + "epoch": 0.03484, + "grad_norm": 0.8692793498605607, + "learning_rate": 0.003, + "loss": 4.1475, + "step": 3484 + }, + { + "epoch": 0.03485, + "grad_norm": 0.8557753970285434, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 3485 + }, + { + "epoch": 0.03486, + "grad_norm": 0.8081888267651361, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3486 + }, + { + "epoch": 0.03487, + "grad_norm": 0.8341642700607715, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 3487 + }, + { + "epoch": 0.03488, + "grad_norm": 0.9314329772486666, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 3488 + }, + { + "epoch": 0.03489, + "grad_norm": 0.889614060937855, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 3489 + }, + { + "epoch": 0.0349, + "grad_norm": 0.8728153421930327, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 3490 + }, + { + "epoch": 0.03491, + "grad_norm": 0.8217256313653218, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 3491 + }, + { + "epoch": 0.03492, + "grad_norm": 0.8472656237093616, + "learning_rate": 0.003, + "loss": 4.1239, + "step": 3492 + }, + { + "epoch": 0.03493, + "grad_norm": 0.911154676677477, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 3493 + }, + { + "epoch": 0.03494, + "grad_norm": 0.8604992149051935, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3494 + }, + { + "epoch": 0.03495, + "grad_norm": 0.7943120640561333, + "learning_rate": 0.003, + "loss": 4.124, + "step": 3495 + }, + { + "epoch": 0.03496, + "grad_norm": 0.8528252392605379, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 3496 + }, + { + "epoch": 0.03497, + "grad_norm": 0.8202667814489011, + "learning_rate": 0.003, + "loss": 4.1337, + "step": 3497 + }, + { + "epoch": 0.03498, + "grad_norm": 0.7532727804867784, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 3498 + }, + { + "epoch": 0.03499, + "grad_norm": 0.6342348301625571, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 3499 + }, + { + "epoch": 0.035, + "grad_norm": 0.6225754670774569, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 3500 + }, + { + "epoch": 0.03501, + "grad_norm": 0.6417198865715938, + "learning_rate": 0.003, + "loss": 4.1179, + "step": 3501 + }, + { + "epoch": 0.03502, + "grad_norm": 0.7016099536316687, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 3502 + }, + { + "epoch": 0.03503, + "grad_norm": 0.7217275979695842, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 3503 + }, + { + "epoch": 0.03504, + "grad_norm": 0.7355434624325226, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 3504 + }, + { + "epoch": 0.03505, + "grad_norm": 0.6978092550167364, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 3505 + }, + { + "epoch": 0.03506, + "grad_norm": 0.7550753861534789, + "learning_rate": 0.003, + "loss": 4.1298, + "step": 3506 + }, + { + "epoch": 0.03507, + "grad_norm": 0.7769412242807696, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 3507 + }, + { + "epoch": 0.03508, + "grad_norm": 0.7321531146263615, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 3508 + }, + { + "epoch": 0.03509, + "grad_norm": 0.6149401724420952, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 3509 + }, + { + "epoch": 0.0351, + "grad_norm": 0.5203674910818351, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 3510 + }, + { + "epoch": 0.03511, + "grad_norm": 0.5215782994430587, + "learning_rate": 0.003, + "loss": 4.1, + "step": 3511 + }, + { + "epoch": 0.03512, + "grad_norm": 0.553705329026448, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 3512 + }, + { + "epoch": 0.03513, + "grad_norm": 0.594607131220142, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 3513 + }, + { + "epoch": 0.03514, + "grad_norm": 0.5990739529020204, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 3514 + }, + { + "epoch": 0.03515, + "grad_norm": 0.5749124550340262, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 3515 + }, + { + "epoch": 0.03516, + "grad_norm": 0.5841715979805169, + "learning_rate": 0.003, + "loss": 4.1386, + "step": 3516 + }, + { + "epoch": 0.03517, + "grad_norm": 0.6124972906955293, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 3517 + }, + { + "epoch": 0.03518, + "grad_norm": 0.6175915714607444, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 3518 + }, + { + "epoch": 0.03519, + "grad_norm": 0.6463255905076363, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 3519 + }, + { + "epoch": 0.0352, + "grad_norm": 0.5874580572373398, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 3520 + }, + { + "epoch": 0.03521, + "grad_norm": 0.5191729208535423, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 3521 + }, + { + "epoch": 0.03522, + "grad_norm": 0.5767366245296268, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 3522 + }, + { + "epoch": 0.03523, + "grad_norm": 0.7172396543600251, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 3523 + }, + { + "epoch": 0.03524, + "grad_norm": 0.8664060165774367, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 3524 + }, + { + "epoch": 0.03525, + "grad_norm": 0.8616398665742534, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 3525 + }, + { + "epoch": 0.03526, + "grad_norm": 0.7419883977158586, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 3526 + }, + { + "epoch": 0.03527, + "grad_norm": 0.6840049350238313, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 3527 + }, + { + "epoch": 0.03528, + "grad_norm": 0.7195500960905333, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 3528 + }, + { + "epoch": 0.03529, + "grad_norm": 0.7094134206381539, + "learning_rate": 0.003, + "loss": 4.102, + "step": 3529 + }, + { + "epoch": 0.0353, + "grad_norm": 0.6232845058956232, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 3530 + }, + { + "epoch": 0.03531, + "grad_norm": 0.675207674156698, + "learning_rate": 0.003, + "loss": 4.1094, + "step": 3531 + }, + { + "epoch": 0.03532, + "grad_norm": 0.6121029642524871, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 3532 + }, + { + "epoch": 0.03533, + "grad_norm": 0.566553895374516, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 3533 + }, + { + "epoch": 0.03534, + "grad_norm": 0.597291210514331, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 3534 + }, + { + "epoch": 0.03535, + "grad_norm": 0.589502833082128, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 3535 + }, + { + "epoch": 0.03536, + "grad_norm": 0.6293469176136443, + "learning_rate": 0.003, + "loss": 4.075, + "step": 3536 + }, + { + "epoch": 0.03537, + "grad_norm": 0.6490867298208753, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 3537 + }, + { + "epoch": 0.03538, + "grad_norm": 0.7584151945893494, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 3538 + }, + { + "epoch": 0.03539, + "grad_norm": 0.809399903324439, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 3539 + }, + { + "epoch": 0.0354, + "grad_norm": 0.8133610737803239, + "learning_rate": 0.003, + "loss": 4.1283, + "step": 3540 + }, + { + "epoch": 0.03541, + "grad_norm": 0.7403928480817895, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 3541 + }, + { + "epoch": 0.03542, + "grad_norm": 0.6221290767159878, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 3542 + }, + { + "epoch": 0.03543, + "grad_norm": 0.6796566068519472, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 3543 + }, + { + "epoch": 0.03544, + "grad_norm": 0.7905007488129874, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 3544 + }, + { + "epoch": 0.03545, + "grad_norm": 0.7191600250796164, + "learning_rate": 0.003, + "loss": 4.08, + "step": 3545 + }, + { + "epoch": 0.03546, + "grad_norm": 0.7535466199672892, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 3546 + }, + { + "epoch": 0.03547, + "grad_norm": 0.7876369752842546, + "learning_rate": 0.003, + "loss": 4.1133, + "step": 3547 + }, + { + "epoch": 0.03548, + "grad_norm": 0.8973560453926198, + "learning_rate": 0.003, + "loss": 4.1242, + "step": 3548 + }, + { + "epoch": 0.03549, + "grad_norm": 0.9348798724920742, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 3549 + }, + { + "epoch": 0.0355, + "grad_norm": 0.9182893351176102, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 3550 + }, + { + "epoch": 0.03551, + "grad_norm": 0.7258075570130067, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 3551 + }, + { + "epoch": 0.03552, + "grad_norm": 0.5650469407701177, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 3552 + }, + { + "epoch": 0.03553, + "grad_norm": 0.5620092220243577, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 3553 + }, + { + "epoch": 0.03554, + "grad_norm": 0.6088349885998489, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 3554 + }, + { + "epoch": 0.03555, + "grad_norm": 0.6807405751922103, + "learning_rate": 0.003, + "loss": 4.1171, + "step": 3555 + }, + { + "epoch": 0.03556, + "grad_norm": 0.6940537010469384, + "learning_rate": 0.003, + "loss": 4.1, + "step": 3556 + }, + { + "epoch": 0.03557, + "grad_norm": 0.7909685218495854, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 3557 + }, + { + "epoch": 0.03558, + "grad_norm": 0.9302792295850243, + "learning_rate": 0.003, + "loss": 4.1388, + "step": 3558 + }, + { + "epoch": 0.03559, + "grad_norm": 0.8633210466881153, + "learning_rate": 0.003, + "loss": 4.1225, + "step": 3559 + }, + { + "epoch": 0.0356, + "grad_norm": 0.8646838386003994, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 3560 + }, + { + "epoch": 0.03561, + "grad_norm": 0.8329616623999954, + "learning_rate": 0.003, + "loss": 4.1514, + "step": 3561 + }, + { + "epoch": 0.03562, + "grad_norm": 0.8101687293018341, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 3562 + }, + { + "epoch": 0.03563, + "grad_norm": 0.792482272613334, + "learning_rate": 0.003, + "loss": 4.1473, + "step": 3563 + }, + { + "epoch": 0.03564, + "grad_norm": 0.7393211745220851, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 3564 + }, + { + "epoch": 0.03565, + "grad_norm": 0.7708978673287942, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 3565 + }, + { + "epoch": 0.03566, + "grad_norm": 0.843974895175368, + "learning_rate": 0.003, + "loss": 4.1236, + "step": 3566 + }, + { + "epoch": 0.03567, + "grad_norm": 0.8249138698289672, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 3567 + }, + { + "epoch": 0.03568, + "grad_norm": 0.7391482897148736, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 3568 + }, + { + "epoch": 0.03569, + "grad_norm": 0.655768565425149, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 3569 + }, + { + "epoch": 0.0357, + "grad_norm": 0.7212180200719822, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 3570 + }, + { + "epoch": 0.03571, + "grad_norm": 0.697309197694063, + "learning_rate": 0.003, + "loss": 4.1136, + "step": 3571 + }, + { + "epoch": 0.03572, + "grad_norm": 0.6452207834915866, + "learning_rate": 0.003, + "loss": 4.1322, + "step": 3572 + }, + { + "epoch": 0.03573, + "grad_norm": 0.6552426807658631, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 3573 + }, + { + "epoch": 0.03574, + "grad_norm": 0.6222445133801513, + "learning_rate": 0.003, + "loss": 4.1216, + "step": 3574 + }, + { + "epoch": 0.03575, + "grad_norm": 0.5720574208555151, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 3575 + }, + { + "epoch": 0.03576, + "grad_norm": 0.49253781076647446, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 3576 + }, + { + "epoch": 0.03577, + "grad_norm": 0.4741074552760465, + "learning_rate": 0.003, + "loss": 4.1113, + "step": 3577 + }, + { + "epoch": 0.03578, + "grad_norm": 0.4354327435437879, + "learning_rate": 0.003, + "loss": 4.1, + "step": 3578 + }, + { + "epoch": 0.03579, + "grad_norm": 0.48473058715401496, + "learning_rate": 0.003, + "loss": 4.1149, + "step": 3579 + }, + { + "epoch": 0.0358, + "grad_norm": 0.519509179948186, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 3580 + }, + { + "epoch": 0.03581, + "grad_norm": 0.568971201309002, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 3581 + }, + { + "epoch": 0.03582, + "grad_norm": 0.6251350779931762, + "learning_rate": 0.003, + "loss": 4.087, + "step": 3582 + }, + { + "epoch": 0.03583, + "grad_norm": 0.6443831077379412, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 3583 + }, + { + "epoch": 0.03584, + "grad_norm": 0.5935480717270938, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 3584 + }, + { + "epoch": 0.03585, + "grad_norm": 0.5112750882339391, + "learning_rate": 0.003, + "loss": 4.072, + "step": 3585 + }, + { + "epoch": 0.03586, + "grad_norm": 0.4878002263964438, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 3586 + }, + { + "epoch": 0.03587, + "grad_norm": 0.5382355861597571, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 3587 + }, + { + "epoch": 0.03588, + "grad_norm": 0.5251657277896673, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 3588 + }, + { + "epoch": 0.03589, + "grad_norm": 0.5736866692797807, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 3589 + }, + { + "epoch": 0.0359, + "grad_norm": 0.618021476459667, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 3590 + }, + { + "epoch": 0.03591, + "grad_norm": 0.6725307239188835, + "learning_rate": 0.003, + "loss": 4.1081, + "step": 3591 + }, + { + "epoch": 0.03592, + "grad_norm": 0.6517707886130953, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 3592 + }, + { + "epoch": 0.03593, + "grad_norm": 0.6242923682333518, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 3593 + }, + { + "epoch": 0.03594, + "grad_norm": 0.599222907451357, + "learning_rate": 0.003, + "loss": 4.093, + "step": 3594 + }, + { + "epoch": 0.03595, + "grad_norm": 0.7282948178373052, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 3595 + }, + { + "epoch": 0.03596, + "grad_norm": 0.873557319258201, + "learning_rate": 0.003, + "loss": 4.1315, + "step": 3596 + }, + { + "epoch": 0.03597, + "grad_norm": 1.0223165738118598, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 3597 + }, + { + "epoch": 0.03598, + "grad_norm": 1.1445270672742742, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 3598 + }, + { + "epoch": 0.03599, + "grad_norm": 0.9504241887166247, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 3599 + }, + { + "epoch": 0.036, + "grad_norm": 0.8239905535481999, + "learning_rate": 0.003, + "loss": 4.1369, + "step": 3600 + }, + { + "epoch": 0.03601, + "grad_norm": 0.9706059190211762, + "learning_rate": 0.003, + "loss": 4.116, + "step": 3601 + }, + { + "epoch": 0.03602, + "grad_norm": 1.0143506537258324, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 3602 + }, + { + "epoch": 0.03603, + "grad_norm": 1.123766736658672, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 3603 + }, + { + "epoch": 0.03604, + "grad_norm": 0.9726152594373947, + "learning_rate": 0.003, + "loss": 4.1348, + "step": 3604 + }, + { + "epoch": 0.03605, + "grad_norm": 0.9791909776755724, + "learning_rate": 0.003, + "loss": 4.13, + "step": 3605 + }, + { + "epoch": 0.03606, + "grad_norm": 0.8590102214343427, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 3606 + }, + { + "epoch": 0.03607, + "grad_norm": 0.8729635647144726, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 3607 + }, + { + "epoch": 0.03608, + "grad_norm": 0.7683960325709851, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 3608 + }, + { + "epoch": 0.03609, + "grad_norm": 0.7311817839920504, + "learning_rate": 0.003, + "loss": 4.1396, + "step": 3609 + }, + { + "epoch": 0.0361, + "grad_norm": 0.726065438689883, + "learning_rate": 0.003, + "loss": 4.1423, + "step": 3610 + }, + { + "epoch": 0.03611, + "grad_norm": 0.7496813843670395, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 3611 + }, + { + "epoch": 0.03612, + "grad_norm": 0.8186325937108399, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 3612 + }, + { + "epoch": 0.03613, + "grad_norm": 0.724653255352612, + "learning_rate": 0.003, + "loss": 4.1324, + "step": 3613 + }, + { + "epoch": 0.03614, + "grad_norm": 0.6814030562471027, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 3614 + }, + { + "epoch": 0.03615, + "grad_norm": 0.7514852122747766, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3615 + }, + { + "epoch": 0.03616, + "grad_norm": 0.6779351410604817, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 3616 + }, + { + "epoch": 0.03617, + "grad_norm": 0.5962340273665172, + "learning_rate": 0.003, + "loss": 4.118, + "step": 3617 + }, + { + "epoch": 0.03618, + "grad_norm": 0.5747994450502394, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 3618 + }, + { + "epoch": 0.03619, + "grad_norm": 0.6124169745731184, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 3619 + }, + { + "epoch": 0.0362, + "grad_norm": 0.679240989179329, + "learning_rate": 0.003, + "loss": 4.1116, + "step": 3620 + }, + { + "epoch": 0.03621, + "grad_norm": 0.6531466491397193, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 3621 + }, + { + "epoch": 0.03622, + "grad_norm": 0.591069143707812, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 3622 + }, + { + "epoch": 0.03623, + "grad_norm": 0.5722197113946066, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 3623 + }, + { + "epoch": 0.03624, + "grad_norm": 0.47011413381496236, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 3624 + }, + { + "epoch": 0.03625, + "grad_norm": 0.4831082167865949, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 3625 + }, + { + "epoch": 0.03626, + "grad_norm": 0.4418579922418605, + "learning_rate": 0.003, + "loss": 4.086, + "step": 3626 + }, + { + "epoch": 0.03627, + "grad_norm": 0.4706415296838731, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 3627 + }, + { + "epoch": 0.03628, + "grad_norm": 0.5848605513871726, + "learning_rate": 0.003, + "loss": 4.098, + "step": 3628 + }, + { + "epoch": 0.03629, + "grad_norm": 0.6442184074605638, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3629 + }, + { + "epoch": 0.0363, + "grad_norm": 0.7660596782495712, + "learning_rate": 0.003, + "loss": 4.104, + "step": 3630 + }, + { + "epoch": 0.03631, + "grad_norm": 0.8649083804784878, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 3631 + }, + { + "epoch": 0.03632, + "grad_norm": 0.7889307075613666, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 3632 + }, + { + "epoch": 0.03633, + "grad_norm": 0.7272834868780877, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 3633 + }, + { + "epoch": 0.03634, + "grad_norm": 0.6943195233911823, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 3634 + }, + { + "epoch": 0.03635, + "grad_norm": 0.7127615205257642, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 3635 + }, + { + "epoch": 0.03636, + "grad_norm": 0.7207477184688909, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 3636 + }, + { + "epoch": 0.03637, + "grad_norm": 0.7168747331716148, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 3637 + }, + { + "epoch": 0.03638, + "grad_norm": 0.614271838210121, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 3638 + }, + { + "epoch": 0.03639, + "grad_norm": 0.7060818429431696, + "learning_rate": 0.003, + "loss": 4.1249, + "step": 3639 + }, + { + "epoch": 0.0364, + "grad_norm": 0.8566380633392303, + "learning_rate": 0.003, + "loss": 4.1301, + "step": 3640 + }, + { + "epoch": 0.03641, + "grad_norm": 0.7681799431934659, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 3641 + }, + { + "epoch": 0.03642, + "grad_norm": 0.7631357464917338, + "learning_rate": 0.003, + "loss": 4.1304, + "step": 3642 + }, + { + "epoch": 0.03643, + "grad_norm": 0.7556758688016673, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 3643 + }, + { + "epoch": 0.03644, + "grad_norm": 0.8090486748144888, + "learning_rate": 0.003, + "loss": 4.1205, + "step": 3644 + }, + { + "epoch": 0.03645, + "grad_norm": 0.7501255355718079, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 3645 + }, + { + "epoch": 0.03646, + "grad_norm": 0.6708652982419059, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 3646 + }, + { + "epoch": 0.03647, + "grad_norm": 0.6140472583820327, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 3647 + }, + { + "epoch": 0.03648, + "grad_norm": 0.6482366392745758, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 3648 + }, + { + "epoch": 0.03649, + "grad_norm": 0.678712826850872, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3649 + }, + { + "epoch": 0.0365, + "grad_norm": 0.6143703671274712, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 3650 + }, + { + "epoch": 0.03651, + "grad_norm": 0.5790322114365879, + "learning_rate": 0.003, + "loss": 4.115, + "step": 3651 + }, + { + "epoch": 0.03652, + "grad_norm": 0.5667923377729286, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 3652 + }, + { + "epoch": 0.03653, + "grad_norm": 0.6490363634747012, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 3653 + }, + { + "epoch": 0.03654, + "grad_norm": 0.7166565810026996, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 3654 + }, + { + "epoch": 0.03655, + "grad_norm": 0.7276215358477922, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 3655 + }, + { + "epoch": 0.03656, + "grad_norm": 0.6662404081315199, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 3656 + }, + { + "epoch": 0.03657, + "grad_norm": 0.6783909600207876, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 3657 + }, + { + "epoch": 0.03658, + "grad_norm": 0.61103432355221, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 3658 + }, + { + "epoch": 0.03659, + "grad_norm": 0.669048259204836, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 3659 + }, + { + "epoch": 0.0366, + "grad_norm": 0.7124861425212368, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 3660 + }, + { + "epoch": 0.03661, + "grad_norm": 0.8932517497039868, + "learning_rate": 0.003, + "loss": 4.103, + "step": 3661 + }, + { + "epoch": 0.03662, + "grad_norm": 0.9883303532780402, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 3662 + }, + { + "epoch": 0.03663, + "grad_norm": 0.9106744901085427, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 3663 + }, + { + "epoch": 0.03664, + "grad_norm": 0.8465057568969986, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 3664 + }, + { + "epoch": 0.03665, + "grad_norm": 0.793710576530082, + "learning_rate": 0.003, + "loss": 4.1256, + "step": 3665 + }, + { + "epoch": 0.03666, + "grad_norm": 0.7709670688081924, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 3666 + }, + { + "epoch": 0.03667, + "grad_norm": 0.6703154460379485, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 3667 + }, + { + "epoch": 0.03668, + "grad_norm": 0.6304240228101521, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 3668 + }, + { + "epoch": 0.03669, + "grad_norm": 0.6734885037641916, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 3669 + }, + { + "epoch": 0.0367, + "grad_norm": 0.8868303231295198, + "learning_rate": 0.003, + "loss": 4.1243, + "step": 3670 + }, + { + "epoch": 0.03671, + "grad_norm": 1.1644127160486455, + "learning_rate": 0.003, + "loss": 4.1251, + "step": 3671 + }, + { + "epoch": 0.03672, + "grad_norm": 0.7403808814310928, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 3672 + }, + { + "epoch": 0.03673, + "grad_norm": 0.6669624679540266, + "learning_rate": 0.003, + "loss": 4.1195, + "step": 3673 + }, + { + "epoch": 0.03674, + "grad_norm": 0.6308282844606542, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 3674 + }, + { + "epoch": 0.03675, + "grad_norm": 0.6770566403888972, + "learning_rate": 0.003, + "loss": 4.1227, + "step": 3675 + }, + { + "epoch": 0.03676, + "grad_norm": 0.62745208547931, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 3676 + }, + { + "epoch": 0.03677, + "grad_norm": 0.6088162713144073, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 3677 + }, + { + "epoch": 0.03678, + "grad_norm": 0.5451953445050878, + "learning_rate": 0.003, + "loss": 4.118, + "step": 3678 + }, + { + "epoch": 0.03679, + "grad_norm": 0.5228966216992177, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 3679 + }, + { + "epoch": 0.0368, + "grad_norm": 0.4824533076280378, + "learning_rate": 0.003, + "loss": 4.096, + "step": 3680 + }, + { + "epoch": 0.03681, + "grad_norm": 0.45762950270260716, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 3681 + }, + { + "epoch": 0.03682, + "grad_norm": 0.4656810828099725, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 3682 + }, + { + "epoch": 0.03683, + "grad_norm": 0.42781573684855, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 3683 + }, + { + "epoch": 0.03684, + "grad_norm": 0.475263931284134, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 3684 + }, + { + "epoch": 0.03685, + "grad_norm": 0.4564425269505165, + "learning_rate": 0.003, + "loss": 4.075, + "step": 3685 + }, + { + "epoch": 0.03686, + "grad_norm": 0.44457267848827936, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 3686 + }, + { + "epoch": 0.03687, + "grad_norm": 0.511054509690436, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 3687 + }, + { + "epoch": 0.03688, + "grad_norm": 0.6138720751750697, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 3688 + }, + { + "epoch": 0.03689, + "grad_norm": 0.6687074685301466, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 3689 + }, + { + "epoch": 0.0369, + "grad_norm": 0.6597919766564714, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 3690 + }, + { + "epoch": 0.03691, + "grad_norm": 0.674432164519827, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 3691 + }, + { + "epoch": 0.03692, + "grad_norm": 0.7838468989440577, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3692 + }, + { + "epoch": 0.03693, + "grad_norm": 0.8275031112357245, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 3693 + }, + { + "epoch": 0.03694, + "grad_norm": 0.9337944115332818, + "learning_rate": 0.003, + "loss": 4.104, + "step": 3694 + }, + { + "epoch": 0.03695, + "grad_norm": 0.938096902892517, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 3695 + }, + { + "epoch": 0.03696, + "grad_norm": 1.0588528651554774, + "learning_rate": 0.003, + "loss": 4.1111, + "step": 3696 + }, + { + "epoch": 0.03697, + "grad_norm": 0.8880143483444022, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 3697 + }, + { + "epoch": 0.03698, + "grad_norm": 0.9810266272214871, + "learning_rate": 0.003, + "loss": 4.1374, + "step": 3698 + }, + { + "epoch": 0.03699, + "grad_norm": 0.991815275667102, + "learning_rate": 0.003, + "loss": 4.1288, + "step": 3699 + }, + { + "epoch": 0.037, + "grad_norm": 1.0722514413320396, + "learning_rate": 0.003, + "loss": 4.1232, + "step": 3700 + }, + { + "epoch": 0.03701, + "grad_norm": 0.9130589769872435, + "learning_rate": 0.003, + "loss": 4.1224, + "step": 3701 + }, + { + "epoch": 0.03702, + "grad_norm": 0.8132933284450143, + "learning_rate": 0.003, + "loss": 4.1079, + "step": 3702 + }, + { + "epoch": 0.03703, + "grad_norm": 0.7058660559910072, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 3703 + }, + { + "epoch": 0.03704, + "grad_norm": 0.6970634260954891, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3704 + }, + { + "epoch": 0.03705, + "grad_norm": 0.7033264140843061, + "learning_rate": 0.003, + "loss": 4.1358, + "step": 3705 + }, + { + "epoch": 0.03706, + "grad_norm": 0.5986390477145073, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 3706 + }, + { + "epoch": 0.03707, + "grad_norm": 0.63710100437172, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 3707 + }, + { + "epoch": 0.03708, + "grad_norm": 0.6650467521255319, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 3708 + }, + { + "epoch": 0.03709, + "grad_norm": 0.6313542020750926, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 3709 + }, + { + "epoch": 0.0371, + "grad_norm": 0.6517859206678731, + "learning_rate": 0.003, + "loss": 4.128, + "step": 3710 + }, + { + "epoch": 0.03711, + "grad_norm": 0.7543034529653859, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 3711 + }, + { + "epoch": 0.03712, + "grad_norm": 0.832110406867099, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 3712 + }, + { + "epoch": 0.03713, + "grad_norm": 0.7411810223380423, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 3713 + }, + { + "epoch": 0.03714, + "grad_norm": 0.6269049169284101, + "learning_rate": 0.003, + "loss": 4.1, + "step": 3714 + }, + { + "epoch": 0.03715, + "grad_norm": 0.5987729202390368, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 3715 + }, + { + "epoch": 0.03716, + "grad_norm": 0.5368594549819967, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 3716 + }, + { + "epoch": 0.03717, + "grad_norm": 0.5846745578431664, + "learning_rate": 0.003, + "loss": 4.0981, + "step": 3717 + }, + { + "epoch": 0.03718, + "grad_norm": 0.5437308761709829, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 3718 + }, + { + "epoch": 0.03719, + "grad_norm": 0.534396883191642, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 3719 + }, + { + "epoch": 0.0372, + "grad_norm": 0.5481104462942403, + "learning_rate": 0.003, + "loss": 4.099, + "step": 3720 + }, + { + "epoch": 0.03721, + "grad_norm": 0.580310944322297, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 3721 + }, + { + "epoch": 0.03722, + "grad_norm": 0.6204590674139819, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 3722 + }, + { + "epoch": 0.03723, + "grad_norm": 0.741699851288619, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 3723 + }, + { + "epoch": 0.03724, + "grad_norm": 0.932910086535998, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 3724 + }, + { + "epoch": 0.03725, + "grad_norm": 0.972219466292221, + "learning_rate": 0.003, + "loss": 4.123, + "step": 3725 + }, + { + "epoch": 0.03726, + "grad_norm": 0.7978461878785346, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 3726 + }, + { + "epoch": 0.03727, + "grad_norm": 0.6501485204398695, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3727 + }, + { + "epoch": 0.03728, + "grad_norm": 0.7526153981428457, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 3728 + }, + { + "epoch": 0.03729, + "grad_norm": 0.9342144894910926, + "learning_rate": 0.003, + "loss": 4.1168, + "step": 3729 + }, + { + "epoch": 0.0373, + "grad_norm": 0.9909748470409672, + "learning_rate": 0.003, + "loss": 4.1058, + "step": 3730 + }, + { + "epoch": 0.03731, + "grad_norm": 0.8349578029322722, + "learning_rate": 0.003, + "loss": 4.1287, + "step": 3731 + }, + { + "epoch": 0.03732, + "grad_norm": 0.7495133779638976, + "learning_rate": 0.003, + "loss": 4.1218, + "step": 3732 + }, + { + "epoch": 0.03733, + "grad_norm": 0.6703852436631956, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 3733 + }, + { + "epoch": 0.03734, + "grad_norm": 0.6477015713191686, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 3734 + }, + { + "epoch": 0.03735, + "grad_norm": 0.7105330316403606, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 3735 + }, + { + "epoch": 0.03736, + "grad_norm": 0.8711159615820404, + "learning_rate": 0.003, + "loss": 4.1296, + "step": 3736 + }, + { + "epoch": 0.03737, + "grad_norm": 0.9195615071918128, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 3737 + }, + { + "epoch": 0.03738, + "grad_norm": 0.8724967055991824, + "learning_rate": 0.003, + "loss": 4.1379, + "step": 3738 + }, + { + "epoch": 0.03739, + "grad_norm": 0.9018400105346501, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 3739 + }, + { + "epoch": 0.0374, + "grad_norm": 0.8677185429500046, + "learning_rate": 0.003, + "loss": 4.1235, + "step": 3740 + }, + { + "epoch": 0.03741, + "grad_norm": 0.8564855572868352, + "learning_rate": 0.003, + "loss": 4.1059, + "step": 3741 + }, + { + "epoch": 0.03742, + "grad_norm": 0.7622155234552144, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 3742 + }, + { + "epoch": 0.03743, + "grad_norm": 0.58417819453155, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 3743 + }, + { + "epoch": 0.03744, + "grad_norm": 0.5643583574815015, + "learning_rate": 0.003, + "loss": 4.1176, + "step": 3744 + }, + { + "epoch": 0.03745, + "grad_norm": 0.537490749220574, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 3745 + }, + { + "epoch": 0.03746, + "grad_norm": 0.4687860689966053, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 3746 + }, + { + "epoch": 0.03747, + "grad_norm": 0.4660663310861506, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 3747 + }, + { + "epoch": 0.03748, + "grad_norm": 0.4519448249577299, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 3748 + }, + { + "epoch": 0.03749, + "grad_norm": 0.411997317244979, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 3749 + }, + { + "epoch": 0.0375, + "grad_norm": 0.4679723052798586, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 3750 + }, + { + "epoch": 0.03751, + "grad_norm": 0.5871606618329933, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 3751 + }, + { + "epoch": 0.03752, + "grad_norm": 0.7076122035118425, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 3752 + }, + { + "epoch": 0.03753, + "grad_norm": 0.6846124878044001, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 3753 + }, + { + "epoch": 0.03754, + "grad_norm": 0.6009356562354289, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 3754 + }, + { + "epoch": 0.03755, + "grad_norm": 0.6629265431416048, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 3755 + }, + { + "epoch": 0.03756, + "grad_norm": 0.682203465404619, + "learning_rate": 0.003, + "loss": 4.073, + "step": 3756 + }, + { + "epoch": 0.03757, + "grad_norm": 0.7242065642570245, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 3757 + }, + { + "epoch": 0.03758, + "grad_norm": 0.7719626981021238, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 3758 + }, + { + "epoch": 0.03759, + "grad_norm": 0.690083960013151, + "learning_rate": 0.003, + "loss": 4.11, + "step": 3759 + }, + { + "epoch": 0.0376, + "grad_norm": 0.6326244569172677, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 3760 + }, + { + "epoch": 0.03761, + "grad_norm": 0.655503599585125, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 3761 + }, + { + "epoch": 0.03762, + "grad_norm": 0.6678409000408134, + "learning_rate": 0.003, + "loss": 4.118, + "step": 3762 + }, + { + "epoch": 0.03763, + "grad_norm": 0.7156432542714204, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 3763 + }, + { + "epoch": 0.03764, + "grad_norm": 0.689667735507266, + "learning_rate": 0.003, + "loss": 4.115, + "step": 3764 + }, + { + "epoch": 0.03765, + "grad_norm": 0.6608025819343777, + "learning_rate": 0.003, + "loss": 4.1372, + "step": 3765 + }, + { + "epoch": 0.03766, + "grad_norm": 0.6528370753447702, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 3766 + }, + { + "epoch": 0.03767, + "grad_norm": 0.5865372858592491, + "learning_rate": 0.003, + "loss": 4.1134, + "step": 3767 + }, + { + "epoch": 0.03768, + "grad_norm": 0.6461178071013718, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 3768 + }, + { + "epoch": 0.03769, + "grad_norm": 0.6706047973609784, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 3769 + }, + { + "epoch": 0.0377, + "grad_norm": 0.7506028968837304, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 3770 + }, + { + "epoch": 0.03771, + "grad_norm": 0.7649025556561673, + "learning_rate": 0.003, + "loss": 4.089, + "step": 3771 + }, + { + "epoch": 0.03772, + "grad_norm": 0.8975835757328303, + "learning_rate": 0.003, + "loss": 4.1108, + "step": 3772 + }, + { + "epoch": 0.03773, + "grad_norm": 0.9470107545745398, + "learning_rate": 0.003, + "loss": 4.1254, + "step": 3773 + }, + { + "epoch": 0.03774, + "grad_norm": 0.9690188852205667, + "learning_rate": 0.003, + "loss": 4.1115, + "step": 3774 + }, + { + "epoch": 0.03775, + "grad_norm": 0.8147540773286227, + "learning_rate": 0.003, + "loss": 4.1255, + "step": 3775 + }, + { + "epoch": 0.03776, + "grad_norm": 0.7392780405812923, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 3776 + }, + { + "epoch": 0.03777, + "grad_norm": 0.6608783526459457, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 3777 + }, + { + "epoch": 0.03778, + "grad_norm": 0.5902228784309004, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 3778 + }, + { + "epoch": 0.03779, + "grad_norm": 0.46244709323026423, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 3779 + }, + { + "epoch": 0.0378, + "grad_norm": 0.5005914510144783, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 3780 + }, + { + "epoch": 0.03781, + "grad_norm": 0.5352311528585781, + "learning_rate": 0.003, + "loss": 4.1277, + "step": 3781 + }, + { + "epoch": 0.03782, + "grad_norm": 0.5724938139574387, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 3782 + }, + { + "epoch": 0.03783, + "grad_norm": 0.6425852959568226, + "learning_rate": 0.003, + "loss": 4.103, + "step": 3783 + }, + { + "epoch": 0.03784, + "grad_norm": 0.5505067974834693, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 3784 + }, + { + "epoch": 0.03785, + "grad_norm": 0.6033034152743874, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 3785 + }, + { + "epoch": 0.03786, + "grad_norm": 0.5737023590125584, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 3786 + }, + { + "epoch": 0.03787, + "grad_norm": 0.5156477713564683, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 3787 + }, + { + "epoch": 0.03788, + "grad_norm": 0.5502090720241226, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 3788 + }, + { + "epoch": 0.03789, + "grad_norm": 0.5789178162739714, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 3789 + }, + { + "epoch": 0.0379, + "grad_norm": 0.5556348068625624, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 3790 + }, + { + "epoch": 0.03791, + "grad_norm": 0.5213442858524716, + "learning_rate": 0.003, + "loss": 4.086, + "step": 3791 + }, + { + "epoch": 0.03792, + "grad_norm": 0.5141252888493217, + "learning_rate": 0.003, + "loss": 4.1165, + "step": 3792 + }, + { + "epoch": 0.03793, + "grad_norm": 0.616970727459292, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 3793 + }, + { + "epoch": 0.03794, + "grad_norm": 0.7039733775408357, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 3794 + }, + { + "epoch": 0.03795, + "grad_norm": 0.7499844155426285, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 3795 + }, + { + "epoch": 0.03796, + "grad_norm": 0.7721932287432621, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 3796 + }, + { + "epoch": 0.03797, + "grad_norm": 0.8146662209566835, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 3797 + }, + { + "epoch": 0.03798, + "grad_norm": 0.951797000921836, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 3798 + }, + { + "epoch": 0.03799, + "grad_norm": 1.0631472769369203, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 3799 + }, + { + "epoch": 0.038, + "grad_norm": 1.0116839197726581, + "learning_rate": 0.003, + "loss": 4.124, + "step": 3800 + }, + { + "epoch": 0.03801, + "grad_norm": 0.7666459124842414, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 3801 + }, + { + "epoch": 0.03802, + "grad_norm": 0.7929227361723564, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 3802 + }, + { + "epoch": 0.03803, + "grad_norm": 0.7817477402746161, + "learning_rate": 0.003, + "loss": 4.1222, + "step": 3803 + }, + { + "epoch": 0.03804, + "grad_norm": 0.8223166996364771, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 3804 + }, + { + "epoch": 0.03805, + "grad_norm": 0.8529606701339681, + "learning_rate": 0.003, + "loss": 4.1148, + "step": 3805 + }, + { + "epoch": 0.03806, + "grad_norm": 0.9511992344180633, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 3806 + }, + { + "epoch": 0.03807, + "grad_norm": 0.967724808834046, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 3807 + }, + { + "epoch": 0.03808, + "grad_norm": 0.9638699838305632, + "learning_rate": 0.003, + "loss": 4.1303, + "step": 3808 + }, + { + "epoch": 0.03809, + "grad_norm": 0.8211655019302597, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 3809 + }, + { + "epoch": 0.0381, + "grad_norm": 0.7679928604914735, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 3810 + }, + { + "epoch": 0.03811, + "grad_norm": 0.6836405423601501, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 3811 + }, + { + "epoch": 0.03812, + "grad_norm": 0.6327955848764387, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 3812 + }, + { + "epoch": 0.03813, + "grad_norm": 0.6479509377296188, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 3813 + }, + { + "epoch": 0.03814, + "grad_norm": 0.5591901614484528, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 3814 + }, + { + "epoch": 0.03815, + "grad_norm": 0.6060170624699693, + "learning_rate": 0.003, + "loss": 4.099, + "step": 3815 + }, + { + "epoch": 0.03816, + "grad_norm": 0.7102384121635698, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 3816 + }, + { + "epoch": 0.03817, + "grad_norm": 0.7389812026151167, + "learning_rate": 0.003, + "loss": 4.081, + "step": 3817 + }, + { + "epoch": 0.03818, + "grad_norm": 0.6684840422502324, + "learning_rate": 0.003, + "loss": 4.129, + "step": 3818 + }, + { + "epoch": 0.03819, + "grad_norm": 0.6502252345097654, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 3819 + }, + { + "epoch": 0.0382, + "grad_norm": 0.6414694585447046, + "learning_rate": 0.003, + "loss": 4.1281, + "step": 3820 + }, + { + "epoch": 0.03821, + "grad_norm": 0.654133018202905, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 3821 + }, + { + "epoch": 0.03822, + "grad_norm": 0.6124763165148336, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 3822 + }, + { + "epoch": 0.03823, + "grad_norm": 0.67550142548948, + "learning_rate": 0.003, + "loss": 4.093, + "step": 3823 + }, + { + "epoch": 0.03824, + "grad_norm": 0.7240649349860498, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 3824 + }, + { + "epoch": 0.03825, + "grad_norm": 0.7803786088043716, + "learning_rate": 0.003, + "loss": 4.1085, + "step": 3825 + }, + { + "epoch": 0.03826, + "grad_norm": 0.7507756824371573, + "learning_rate": 0.003, + "loss": 4.095, + "step": 3826 + }, + { + "epoch": 0.03827, + "grad_norm": 0.7424150738379162, + "learning_rate": 0.003, + "loss": 4.1197, + "step": 3827 + }, + { + "epoch": 0.03828, + "grad_norm": 0.7061176843384847, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 3828 + }, + { + "epoch": 0.03829, + "grad_norm": 0.5253157191567489, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 3829 + }, + { + "epoch": 0.0383, + "grad_norm": 0.44225267316949235, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 3830 + }, + { + "epoch": 0.03831, + "grad_norm": 0.49964028898864765, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 3831 + }, + { + "epoch": 0.03832, + "grad_norm": 0.5112238758494883, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 3832 + }, + { + "epoch": 0.03833, + "grad_norm": 0.5569123842590502, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 3833 + }, + { + "epoch": 0.03834, + "grad_norm": 0.6078121276578177, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 3834 + }, + { + "epoch": 0.03835, + "grad_norm": 0.6846443299056576, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 3835 + }, + { + "epoch": 0.03836, + "grad_norm": 0.7232002082179908, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 3836 + }, + { + "epoch": 0.03837, + "grad_norm": 0.7964719772999455, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 3837 + }, + { + "epoch": 0.03838, + "grad_norm": 0.7673693729049601, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 3838 + }, + { + "epoch": 0.03839, + "grad_norm": 0.6591666726520761, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 3839 + }, + { + "epoch": 0.0384, + "grad_norm": 0.6674840500138008, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 3840 + }, + { + "epoch": 0.03841, + "grad_norm": 0.7756417558583624, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 3841 + }, + { + "epoch": 0.03842, + "grad_norm": 0.7952944913797769, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 3842 + }, + { + "epoch": 0.03843, + "grad_norm": 0.8602427339783966, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 3843 + }, + { + "epoch": 0.03844, + "grad_norm": 0.8366076028680528, + "learning_rate": 0.003, + "loss": 4.1047, + "step": 3844 + }, + { + "epoch": 0.03845, + "grad_norm": 0.909538987129958, + "learning_rate": 0.003, + "loss": 4.104, + "step": 3845 + }, + { + "epoch": 0.03846, + "grad_norm": 0.9599219516761457, + "learning_rate": 0.003, + "loss": 4.1271, + "step": 3846 + }, + { + "epoch": 0.03847, + "grad_norm": 0.8430131288309606, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 3847 + }, + { + "epoch": 0.03848, + "grad_norm": 0.6989869361493133, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 3848 + }, + { + "epoch": 0.03849, + "grad_norm": 0.639013322654069, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 3849 + }, + { + "epoch": 0.0385, + "grad_norm": 0.6129412907376528, + "learning_rate": 0.003, + "loss": 4.119, + "step": 3850 + }, + { + "epoch": 0.03851, + "grad_norm": 0.6283918847790374, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 3851 + }, + { + "epoch": 0.03852, + "grad_norm": 0.6416831075745711, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 3852 + }, + { + "epoch": 0.03853, + "grad_norm": 0.7716388973723896, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 3853 + }, + { + "epoch": 0.03854, + "grad_norm": 0.8354709196359086, + "learning_rate": 0.003, + "loss": 4.1101, + "step": 3854 + }, + { + "epoch": 0.03855, + "grad_norm": 0.8031922585003632, + "learning_rate": 0.003, + "loss": 4.1272, + "step": 3855 + }, + { + "epoch": 0.03856, + "grad_norm": 0.6891621461552916, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 3856 + }, + { + "epoch": 0.03857, + "grad_norm": 0.5857926773760566, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 3857 + }, + { + "epoch": 0.03858, + "grad_norm": 0.6840162297542994, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 3858 + }, + { + "epoch": 0.03859, + "grad_norm": 0.7380861521145294, + "learning_rate": 0.003, + "loss": 4.1193, + "step": 3859 + }, + { + "epoch": 0.0386, + "grad_norm": 0.7612451455243763, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 3860 + }, + { + "epoch": 0.03861, + "grad_norm": 0.6150432179347421, + "learning_rate": 0.003, + "loss": 4.117, + "step": 3861 + }, + { + "epoch": 0.03862, + "grad_norm": 0.6009907333799223, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 3862 + }, + { + "epoch": 0.03863, + "grad_norm": 0.5124347959456504, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 3863 + }, + { + "epoch": 0.03864, + "grad_norm": 0.526336250037537, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 3864 + }, + { + "epoch": 0.03865, + "grad_norm": 0.5622891241679555, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 3865 + }, + { + "epoch": 0.03866, + "grad_norm": 0.6758218369234158, + "learning_rate": 0.003, + "loss": 4.08, + "step": 3866 + }, + { + "epoch": 0.03867, + "grad_norm": 0.782032547380071, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 3867 + }, + { + "epoch": 0.03868, + "grad_norm": 0.8860994145175456, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 3868 + }, + { + "epoch": 0.03869, + "grad_norm": 0.8034466298465205, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 3869 + }, + { + "epoch": 0.0387, + "grad_norm": 0.6947389753623375, + "learning_rate": 0.003, + "loss": 4.1043, + "step": 3870 + }, + { + "epoch": 0.03871, + "grad_norm": 0.8664268796654279, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 3871 + }, + { + "epoch": 0.03872, + "grad_norm": 0.8822912745278199, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 3872 + }, + { + "epoch": 0.03873, + "grad_norm": 0.7610402177959409, + "learning_rate": 0.003, + "loss": 4.1181, + "step": 3873 + }, + { + "epoch": 0.03874, + "grad_norm": 0.6991976887689266, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 3874 + }, + { + "epoch": 0.03875, + "grad_norm": 0.657983945226929, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 3875 + }, + { + "epoch": 0.03876, + "grad_norm": 0.5642537716969014, + "learning_rate": 0.003, + "loss": 4.1191, + "step": 3876 + }, + { + "epoch": 0.03877, + "grad_norm": 0.5417015876051494, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 3877 + }, + { + "epoch": 0.03878, + "grad_norm": 0.49679060160900285, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 3878 + }, + { + "epoch": 0.03879, + "grad_norm": 0.4849075424941813, + "learning_rate": 0.003, + "loss": 4.1087, + "step": 3879 + }, + { + "epoch": 0.0388, + "grad_norm": 0.4505872751183022, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 3880 + }, + { + "epoch": 0.03881, + "grad_norm": 0.43446800193148083, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 3881 + }, + { + "epoch": 0.03882, + "grad_norm": 0.491956524791033, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 3882 + }, + { + "epoch": 0.03883, + "grad_norm": 0.5153254466389307, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 3883 + }, + { + "epoch": 0.03884, + "grad_norm": 0.6682942393410314, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 3884 + }, + { + "epoch": 0.03885, + "grad_norm": 0.7657469263827624, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 3885 + }, + { + "epoch": 0.03886, + "grad_norm": 0.8548160458684868, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 3886 + }, + { + "epoch": 0.03887, + "grad_norm": 0.826926565712721, + "learning_rate": 0.003, + "loss": 4.1033, + "step": 3887 + }, + { + "epoch": 0.03888, + "grad_norm": 0.8179478785959735, + "learning_rate": 0.003, + "loss": 4.083, + "step": 3888 + }, + { + "epoch": 0.03889, + "grad_norm": 0.841988953466336, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 3889 + }, + { + "epoch": 0.0389, + "grad_norm": 0.7738723655511407, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 3890 + }, + { + "epoch": 0.03891, + "grad_norm": 0.655704018049718, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 3891 + }, + { + "epoch": 0.03892, + "grad_norm": 0.7351500345850371, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 3892 + }, + { + "epoch": 0.03893, + "grad_norm": 0.7954197178101339, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 3893 + }, + { + "epoch": 0.03894, + "grad_norm": 0.8936038852376956, + "learning_rate": 0.003, + "loss": 4.1093, + "step": 3894 + }, + { + "epoch": 0.03895, + "grad_norm": 0.9990508267631595, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 3895 + }, + { + "epoch": 0.03896, + "grad_norm": 1.011299800690197, + "learning_rate": 0.003, + "loss": 4.1231, + "step": 3896 + }, + { + "epoch": 0.03897, + "grad_norm": 0.8869867437603319, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 3897 + }, + { + "epoch": 0.03898, + "grad_norm": 0.8414522995338386, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 3898 + }, + { + "epoch": 0.03899, + "grad_norm": 0.8885805913351335, + "learning_rate": 0.003, + "loss": 4.1206, + "step": 3899 + }, + { + "epoch": 0.039, + "grad_norm": 0.8563662731272593, + "learning_rate": 0.003, + "loss": 4.1174, + "step": 3900 + }, + { + "epoch": 0.03901, + "grad_norm": 0.8469310892724924, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 3901 + }, + { + "epoch": 0.03902, + "grad_norm": 0.9454822804703409, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 3902 + }, + { + "epoch": 0.03903, + "grad_norm": 0.827539892896266, + "learning_rate": 0.003, + "loss": 4.1429, + "step": 3903 + }, + { + "epoch": 0.03904, + "grad_norm": 0.8442280687314506, + "learning_rate": 0.003, + "loss": 4.1226, + "step": 3904 + }, + { + "epoch": 0.03905, + "grad_norm": 0.8477439489039165, + "learning_rate": 0.003, + "loss": 4.1432, + "step": 3905 + }, + { + "epoch": 0.03906, + "grad_norm": 0.8333478634360164, + "learning_rate": 0.003, + "loss": 4.1121, + "step": 3906 + }, + { + "epoch": 0.03907, + "grad_norm": 0.8446643407044293, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 3907 + }, + { + "epoch": 0.03908, + "grad_norm": 0.8069670799765999, + "learning_rate": 0.003, + "loss": 4.1055, + "step": 3908 + }, + { + "epoch": 0.03909, + "grad_norm": 0.6775264974869494, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 3909 + }, + { + "epoch": 0.0391, + "grad_norm": 0.6762418684990692, + "learning_rate": 0.003, + "loss": 4.1385, + "step": 3910 + }, + { + "epoch": 0.03911, + "grad_norm": 0.6558117254001743, + "learning_rate": 0.003, + "loss": 4.1184, + "step": 3911 + }, + { + "epoch": 0.03912, + "grad_norm": 0.5955826927918358, + "learning_rate": 0.003, + "loss": 4.1163, + "step": 3912 + }, + { + "epoch": 0.03913, + "grad_norm": 0.5481472027083387, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 3913 + }, + { + "epoch": 0.03914, + "grad_norm": 0.559834078668934, + "learning_rate": 0.003, + "loss": 4.1214, + "step": 3914 + }, + { + "epoch": 0.03915, + "grad_norm": 0.49376851034045816, + "learning_rate": 0.003, + "loss": 4.087, + "step": 3915 + }, + { + "epoch": 0.03916, + "grad_norm": 0.4842391536531005, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 3916 + }, + { + "epoch": 0.03917, + "grad_norm": 0.5535862868886617, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 3917 + }, + { + "epoch": 0.03918, + "grad_norm": 0.573738876593738, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 3918 + }, + { + "epoch": 0.03919, + "grad_norm": 0.4848469125464513, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 3919 + }, + { + "epoch": 0.0392, + "grad_norm": 0.45753334758336456, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 3920 + }, + { + "epoch": 0.03921, + "grad_norm": 0.5048770442437, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 3921 + }, + { + "epoch": 0.03922, + "grad_norm": 0.5635539082362038, + "learning_rate": 0.003, + "loss": 4.1177, + "step": 3922 + }, + { + "epoch": 0.03923, + "grad_norm": 0.6284167740622103, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 3923 + }, + { + "epoch": 0.03924, + "grad_norm": 0.6865466860443906, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 3924 + }, + { + "epoch": 0.03925, + "grad_norm": 0.8791111095685349, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 3925 + }, + { + "epoch": 0.03926, + "grad_norm": 0.9414291644655364, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 3926 + }, + { + "epoch": 0.03927, + "grad_norm": 0.8003716302900178, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 3927 + }, + { + "epoch": 0.03928, + "grad_norm": 0.667691866843528, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 3928 + }, + { + "epoch": 0.03929, + "grad_norm": 0.644690127736029, + "learning_rate": 0.003, + "loss": 4.1207, + "step": 3929 + }, + { + "epoch": 0.0393, + "grad_norm": 0.6175480216712721, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 3930 + }, + { + "epoch": 0.03931, + "grad_norm": 0.5439221874750041, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 3931 + }, + { + "epoch": 0.03932, + "grad_norm": 0.5987117175800913, + "learning_rate": 0.003, + "loss": 4.1053, + "step": 3932 + }, + { + "epoch": 0.03933, + "grad_norm": 0.5999360508339331, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 3933 + }, + { + "epoch": 0.03934, + "grad_norm": 0.6156421425624748, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 3934 + }, + { + "epoch": 0.03935, + "grad_norm": 0.6019053550617135, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 3935 + }, + { + "epoch": 0.03936, + "grad_norm": 0.6419278384646516, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 3936 + }, + { + "epoch": 0.03937, + "grad_norm": 0.7262635325719229, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 3937 + }, + { + "epoch": 0.03938, + "grad_norm": 0.7594253350462205, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 3938 + }, + { + "epoch": 0.03939, + "grad_norm": 0.8136988671490715, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 3939 + }, + { + "epoch": 0.0394, + "grad_norm": 0.9393012996367328, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 3940 + }, + { + "epoch": 0.03941, + "grad_norm": 1.0245161127828315, + "learning_rate": 0.003, + "loss": 4.109, + "step": 3941 + }, + { + "epoch": 0.03942, + "grad_norm": 0.8893260186448423, + "learning_rate": 0.003, + "loss": 4.117, + "step": 3942 + }, + { + "epoch": 0.03943, + "grad_norm": 0.7387166777856147, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 3943 + }, + { + "epoch": 0.03944, + "grad_norm": 0.6950691995442531, + "learning_rate": 0.003, + "loss": 4.1186, + "step": 3944 + }, + { + "epoch": 0.03945, + "grad_norm": 0.6567658603710798, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 3945 + }, + { + "epoch": 0.03946, + "grad_norm": 0.6005493814078853, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 3946 + }, + { + "epoch": 0.03947, + "grad_norm": 0.6432699022251411, + "learning_rate": 0.003, + "loss": 4.115, + "step": 3947 + }, + { + "epoch": 0.03948, + "grad_norm": 0.5842960223272707, + "learning_rate": 0.003, + "loss": 4.1293, + "step": 3948 + }, + { + "epoch": 0.03949, + "grad_norm": 0.6655173387066301, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 3949 + }, + { + "epoch": 0.0395, + "grad_norm": 0.7356236705346476, + "learning_rate": 0.003, + "loss": 4.108, + "step": 3950 + }, + { + "epoch": 0.03951, + "grad_norm": 0.6843550854372092, + "learning_rate": 0.003, + "loss": 4.133, + "step": 3951 + }, + { + "epoch": 0.03952, + "grad_norm": 0.6583108086525948, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 3952 + }, + { + "epoch": 0.03953, + "grad_norm": 0.6689671975385194, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 3953 + }, + { + "epoch": 0.03954, + "grad_norm": 0.7831752494700958, + "learning_rate": 0.003, + "loss": 4.099, + "step": 3954 + }, + { + "epoch": 0.03955, + "grad_norm": 0.8695354928798776, + "learning_rate": 0.003, + "loss": 4.102, + "step": 3955 + }, + { + "epoch": 0.03956, + "grad_norm": 0.8125406026039331, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 3956 + }, + { + "epoch": 0.03957, + "grad_norm": 0.6510211778401243, + "learning_rate": 0.003, + "loss": 4.1071, + "step": 3957 + }, + { + "epoch": 0.03958, + "grad_norm": 0.6937350063513976, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 3958 + }, + { + "epoch": 0.03959, + "grad_norm": 0.6066778842251537, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 3959 + }, + { + "epoch": 0.0396, + "grad_norm": 0.6299144936291535, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 3960 + }, + { + "epoch": 0.03961, + "grad_norm": 0.694970172885634, + "learning_rate": 0.003, + "loss": 4.098, + "step": 3961 + }, + { + "epoch": 0.03962, + "grad_norm": 0.7495784268221544, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 3962 + }, + { + "epoch": 0.03963, + "grad_norm": 0.686248076096437, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 3963 + }, + { + "epoch": 0.03964, + "grad_norm": 0.6587463609807164, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 3964 + }, + { + "epoch": 0.03965, + "grad_norm": 0.6875013352719327, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 3965 + }, + { + "epoch": 0.03966, + "grad_norm": 0.6541989500996823, + "learning_rate": 0.003, + "loss": 4.1318, + "step": 3966 + }, + { + "epoch": 0.03967, + "grad_norm": 0.7854846838712145, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 3967 + }, + { + "epoch": 0.03968, + "grad_norm": 0.7779789442450485, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 3968 + }, + { + "epoch": 0.03969, + "grad_norm": 0.6780009423625671, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 3969 + }, + { + "epoch": 0.0397, + "grad_norm": 0.6355660214766398, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 3970 + }, + { + "epoch": 0.03971, + "grad_norm": 0.5736504254871343, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 3971 + }, + { + "epoch": 0.03972, + "grad_norm": 0.5812997437852235, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 3972 + }, + { + "epoch": 0.03973, + "grad_norm": 0.574519507113032, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 3973 + }, + { + "epoch": 0.03974, + "grad_norm": 0.7518296105208935, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 3974 + }, + { + "epoch": 0.03975, + "grad_norm": 0.9350394394064168, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 3975 + }, + { + "epoch": 0.03976, + "grad_norm": 0.8601365185418591, + "learning_rate": 0.003, + "loss": 4.135, + "step": 3976 + }, + { + "epoch": 0.03977, + "grad_norm": 0.7162890742227964, + "learning_rate": 0.003, + "loss": 4.1142, + "step": 3977 + }, + { + "epoch": 0.03978, + "grad_norm": 0.8051580528890029, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 3978 + }, + { + "epoch": 0.03979, + "grad_norm": 0.7191250466873582, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 3979 + }, + { + "epoch": 0.0398, + "grad_norm": 0.8093652654338477, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 3980 + }, + { + "epoch": 0.03981, + "grad_norm": 0.8451914089829478, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 3981 + }, + { + "epoch": 0.03982, + "grad_norm": 1.0143738883319502, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 3982 + }, + { + "epoch": 0.03983, + "grad_norm": 1.0002698688701686, + "learning_rate": 0.003, + "loss": 4.108, + "step": 3983 + }, + { + "epoch": 0.03984, + "grad_norm": 0.9703268843456198, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 3984 + }, + { + "epoch": 0.03985, + "grad_norm": 0.9111599925957224, + "learning_rate": 0.003, + "loss": 4.1161, + "step": 3985 + }, + { + "epoch": 0.03986, + "grad_norm": 0.7894068170170474, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 3986 + }, + { + "epoch": 0.03987, + "grad_norm": 0.6935935756783185, + "learning_rate": 0.003, + "loss": 4.1104, + "step": 3987 + }, + { + "epoch": 0.03988, + "grad_norm": 0.6637559249979246, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 3988 + }, + { + "epoch": 0.03989, + "grad_norm": 0.7110609381131964, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 3989 + }, + { + "epoch": 0.0399, + "grad_norm": 0.6873474996517893, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 3990 + }, + { + "epoch": 0.03991, + "grad_norm": 0.7573410887418346, + "learning_rate": 0.003, + "loss": 4.1237, + "step": 3991 + }, + { + "epoch": 0.03992, + "grad_norm": 0.725412106822123, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 3992 + }, + { + "epoch": 0.03993, + "grad_norm": 0.7110009380761185, + "learning_rate": 0.003, + "loss": 4.098, + "step": 3993 + }, + { + "epoch": 0.03994, + "grad_norm": 0.6906053201061297, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 3994 + }, + { + "epoch": 0.03995, + "grad_norm": 0.7037986927292674, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 3995 + }, + { + "epoch": 0.03996, + "grad_norm": 0.6335775291682871, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 3996 + }, + { + "epoch": 0.03997, + "grad_norm": 0.5805421546899745, + "learning_rate": 0.003, + "loss": 4.1141, + "step": 3997 + }, + { + "epoch": 0.03998, + "grad_norm": 0.5879054467621071, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 3998 + }, + { + "epoch": 0.03999, + "grad_norm": 0.5953909821639622, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 3999 + }, + { + "epoch": 0.04, + "grad_norm": 0.5258400464155906, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 4000 + }, + { + "epoch": 0.04001, + "grad_norm": 0.4599772900711176, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 4001 + }, + { + "epoch": 0.04002, + "grad_norm": 0.42114214267545186, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 4002 + }, + { + "epoch": 0.04003, + "grad_norm": 0.498003956025627, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 4003 + }, + { + "epoch": 0.04004, + "grad_norm": 0.6185308014033168, + "learning_rate": 0.003, + "loss": 4.1107, + "step": 4004 + }, + { + "epoch": 0.04005, + "grad_norm": 0.8385114081425361, + "learning_rate": 0.003, + "loss": 4.1233, + "step": 4005 + }, + { + "epoch": 0.04006, + "grad_norm": 0.9912130742110296, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 4006 + }, + { + "epoch": 0.04007, + "grad_norm": 0.9851037142938546, + "learning_rate": 0.003, + "loss": 4.1178, + "step": 4007 + }, + { + "epoch": 0.04008, + "grad_norm": 0.7990925907321775, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 4008 + }, + { + "epoch": 0.04009, + "grad_norm": 0.6623034380828184, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 4009 + }, + { + "epoch": 0.0401, + "grad_norm": 0.794246416598112, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 4010 + }, + { + "epoch": 0.04011, + "grad_norm": 0.8085514357730338, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 4011 + }, + { + "epoch": 0.04012, + "grad_norm": 0.7774053424059025, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 4012 + }, + { + "epoch": 0.04013, + "grad_norm": 0.803176206903416, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 4013 + }, + { + "epoch": 0.04014, + "grad_norm": 0.7892251325270082, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 4014 + }, + { + "epoch": 0.04015, + "grad_norm": 0.8311134467189533, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 4015 + }, + { + "epoch": 0.04016, + "grad_norm": 0.674279853232673, + "learning_rate": 0.003, + "loss": 4.09, + "step": 4016 + }, + { + "epoch": 0.04017, + "grad_norm": 0.6918195791302054, + "learning_rate": 0.003, + "loss": 4.1302, + "step": 4017 + }, + { + "epoch": 0.04018, + "grad_norm": 0.7206191652575559, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4018 + }, + { + "epoch": 0.04019, + "grad_norm": 0.6826145903967715, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 4019 + }, + { + "epoch": 0.0402, + "grad_norm": 0.615354612851539, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 4020 + }, + { + "epoch": 0.04021, + "grad_norm": 0.7149878249907804, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 4021 + }, + { + "epoch": 0.04022, + "grad_norm": 0.7862128947199396, + "learning_rate": 0.003, + "loss": 4.0899, + "step": 4022 + }, + { + "epoch": 0.04023, + "grad_norm": 0.8450466797551548, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 4023 + }, + { + "epoch": 0.04024, + "grad_norm": 0.8549712234972304, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 4024 + }, + { + "epoch": 0.04025, + "grad_norm": 0.8523668909132229, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 4025 + }, + { + "epoch": 0.04026, + "grad_norm": 0.6802775915323805, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 4026 + }, + { + "epoch": 0.04027, + "grad_norm": 0.5818602851697385, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 4027 + }, + { + "epoch": 0.04028, + "grad_norm": 0.555803105527921, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 4028 + }, + { + "epoch": 0.04029, + "grad_norm": 0.6057419874060253, + "learning_rate": 0.003, + "loss": 4.1196, + "step": 4029 + }, + { + "epoch": 0.0403, + "grad_norm": 0.625572350427166, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 4030 + }, + { + "epoch": 0.04031, + "grad_norm": 0.5126039175682049, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 4031 + }, + { + "epoch": 0.04032, + "grad_norm": 0.4856562254262022, + "learning_rate": 0.003, + "loss": 4.1083, + "step": 4032 + }, + { + "epoch": 0.04033, + "grad_norm": 0.507550191171294, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 4033 + }, + { + "epoch": 0.04034, + "grad_norm": 0.49371243985466, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 4034 + }, + { + "epoch": 0.04035, + "grad_norm": 0.5607296896517224, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 4035 + }, + { + "epoch": 0.04036, + "grad_norm": 0.5513167775947864, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 4036 + }, + { + "epoch": 0.04037, + "grad_norm": 0.5130488568649084, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 4037 + }, + { + "epoch": 0.04038, + "grad_norm": 0.606862832960537, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 4038 + }, + { + "epoch": 0.04039, + "grad_norm": 0.825038673335366, + "learning_rate": 0.003, + "loss": 4.082, + "step": 4039 + }, + { + "epoch": 0.0404, + "grad_norm": 0.9628799541989916, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 4040 + }, + { + "epoch": 0.04041, + "grad_norm": 0.8379280533276655, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 4041 + }, + { + "epoch": 0.04042, + "grad_norm": 0.8201172773505675, + "learning_rate": 0.003, + "loss": 4.1238, + "step": 4042 + }, + { + "epoch": 0.04043, + "grad_norm": 0.8681003176205769, + "learning_rate": 0.003, + "loss": 4.1146, + "step": 4043 + }, + { + "epoch": 0.04044, + "grad_norm": 0.84433729309558, + "learning_rate": 0.003, + "loss": 4.1124, + "step": 4044 + }, + { + "epoch": 0.04045, + "grad_norm": 0.7125980065321708, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 4045 + }, + { + "epoch": 0.04046, + "grad_norm": 0.733846404859602, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 4046 + }, + { + "epoch": 0.04047, + "grad_norm": 0.6644354726139646, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 4047 + }, + { + "epoch": 0.04048, + "grad_norm": 0.6883218400217435, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 4048 + }, + { + "epoch": 0.04049, + "grad_norm": 0.716997711947669, + "learning_rate": 0.003, + "loss": 4.1253, + "step": 4049 + }, + { + "epoch": 0.0405, + "grad_norm": 0.7277990539042298, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 4050 + }, + { + "epoch": 0.04051, + "grad_norm": 0.6603200344726028, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 4051 + }, + { + "epoch": 0.04052, + "grad_norm": 0.6850195818608893, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 4052 + }, + { + "epoch": 0.04053, + "grad_norm": 0.6060536119467864, + "learning_rate": 0.003, + "loss": 4.053, + "step": 4053 + }, + { + "epoch": 0.04054, + "grad_norm": 0.5558365767975282, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 4054 + }, + { + "epoch": 0.04055, + "grad_norm": 0.4903629816540104, + "learning_rate": 0.003, + "loss": 4.065, + "step": 4055 + }, + { + "epoch": 0.04056, + "grad_norm": 0.4653666341920026, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 4056 + }, + { + "epoch": 0.04057, + "grad_norm": 0.560353702927233, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 4057 + }, + { + "epoch": 0.04058, + "grad_norm": 0.7401544657064691, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 4058 + }, + { + "epoch": 0.04059, + "grad_norm": 0.9560157073295488, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 4059 + }, + { + "epoch": 0.0406, + "grad_norm": 0.9613071307280268, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 4060 + }, + { + "epoch": 0.04061, + "grad_norm": 0.7008590925528005, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 4061 + }, + { + "epoch": 0.04062, + "grad_norm": 0.687246038914794, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4062 + }, + { + "epoch": 0.04063, + "grad_norm": 0.8170691801626544, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 4063 + }, + { + "epoch": 0.04064, + "grad_norm": 0.7709576091912467, + "learning_rate": 0.003, + "loss": 4.102, + "step": 4064 + }, + { + "epoch": 0.04065, + "grad_norm": 0.5574344184351803, + "learning_rate": 0.003, + "loss": 4.1327, + "step": 4065 + }, + { + "epoch": 0.04066, + "grad_norm": 0.6823316394665919, + "learning_rate": 0.003, + "loss": 4.0964, + "step": 4066 + }, + { + "epoch": 0.04067, + "grad_norm": 0.7089219247773015, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4067 + }, + { + "epoch": 0.04068, + "grad_norm": 0.709910358040333, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 4068 + }, + { + "epoch": 0.04069, + "grad_norm": 0.7544326538405565, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4069 + }, + { + "epoch": 0.0407, + "grad_norm": 0.7594013627504284, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 4070 + }, + { + "epoch": 0.04071, + "grad_norm": 0.7113992572698825, + "learning_rate": 0.003, + "loss": 4.069, + "step": 4071 + }, + { + "epoch": 0.04072, + "grad_norm": 0.7017843149523408, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 4072 + }, + { + "epoch": 0.04073, + "grad_norm": 0.7898604649603427, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 4073 + }, + { + "epoch": 0.04074, + "grad_norm": 0.765946026804729, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 4074 + }, + { + "epoch": 0.04075, + "grad_norm": 0.7083343421867635, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 4075 + }, + { + "epoch": 0.04076, + "grad_norm": 0.7476562014362685, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 4076 + }, + { + "epoch": 0.04077, + "grad_norm": 0.7267581922449575, + "learning_rate": 0.003, + "loss": 4.1138, + "step": 4077 + }, + { + "epoch": 0.04078, + "grad_norm": 0.6810760536652016, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 4078 + }, + { + "epoch": 0.04079, + "grad_norm": 0.6234099703171976, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 4079 + }, + { + "epoch": 0.0408, + "grad_norm": 0.6224948839556584, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 4080 + }, + { + "epoch": 0.04081, + "grad_norm": 0.5261887448117314, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 4081 + }, + { + "epoch": 0.04082, + "grad_norm": 0.5509844764652435, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4082 + }, + { + "epoch": 0.04083, + "grad_norm": 0.510541759108617, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 4083 + }, + { + "epoch": 0.04084, + "grad_norm": 0.5319041250478885, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 4084 + }, + { + "epoch": 0.04085, + "grad_norm": 0.5988408687711714, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 4085 + }, + { + "epoch": 0.04086, + "grad_norm": 0.6190406114176409, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 4086 + }, + { + "epoch": 0.04087, + "grad_norm": 0.6371746656296016, + "learning_rate": 0.003, + "loss": 4.0967, + "step": 4087 + }, + { + "epoch": 0.04088, + "grad_norm": 0.7548162986516674, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 4088 + }, + { + "epoch": 0.04089, + "grad_norm": 0.7959232107722483, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 4089 + }, + { + "epoch": 0.0409, + "grad_norm": 0.7025723256392559, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 4090 + }, + { + "epoch": 0.04091, + "grad_norm": 0.7287419645210326, + "learning_rate": 0.003, + "loss": 4.109, + "step": 4091 + }, + { + "epoch": 0.04092, + "grad_norm": 0.7954107549226457, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 4092 + }, + { + "epoch": 0.04093, + "grad_norm": 0.9060252315831411, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 4093 + }, + { + "epoch": 0.04094, + "grad_norm": 0.8807500389023475, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 4094 + }, + { + "epoch": 0.04095, + "grad_norm": 0.7720422342257688, + "learning_rate": 0.003, + "loss": 4.063, + "step": 4095 + }, + { + "epoch": 0.04096, + "grad_norm": 0.5981930370282977, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 4096 + }, + { + "epoch": 0.04097, + "grad_norm": 0.5818999950304069, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 4097 + }, + { + "epoch": 0.04098, + "grad_norm": 0.6540104450194776, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 4098 + }, + { + "epoch": 0.04099, + "grad_norm": 0.5935001433419117, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 4099 + }, + { + "epoch": 0.041, + "grad_norm": 0.6163781670710335, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 4100 + }, + { + "epoch": 0.04101, + "grad_norm": 0.7359799863247739, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 4101 + }, + { + "epoch": 0.04102, + "grad_norm": 0.6964490809968663, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 4102 + }, + { + "epoch": 0.04103, + "grad_norm": 0.6499758651516563, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 4103 + }, + { + "epoch": 0.04104, + "grad_norm": 0.5895630935010513, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 4104 + }, + { + "epoch": 0.04105, + "grad_norm": 0.6106315555834759, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 4105 + }, + { + "epoch": 0.04106, + "grad_norm": 0.5493622471975645, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 4106 + }, + { + "epoch": 0.04107, + "grad_norm": 0.6254712565749952, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 4107 + }, + { + "epoch": 0.04108, + "grad_norm": 0.7490707022170446, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 4108 + }, + { + "epoch": 0.04109, + "grad_norm": 0.87542457935642, + "learning_rate": 0.003, + "loss": 4.074, + "step": 4109 + }, + { + "epoch": 0.0411, + "grad_norm": 0.9521380697049077, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 4110 + }, + { + "epoch": 0.04111, + "grad_norm": 0.9074306256039075, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 4111 + }, + { + "epoch": 0.04112, + "grad_norm": 0.8437129221178633, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 4112 + }, + { + "epoch": 0.04113, + "grad_norm": 0.9466897813789951, + "learning_rate": 0.003, + "loss": 4.1412, + "step": 4113 + }, + { + "epoch": 0.04114, + "grad_norm": 0.9075833997843096, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 4114 + }, + { + "epoch": 0.04115, + "grad_norm": 0.6982873696680034, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 4115 + }, + { + "epoch": 0.04116, + "grad_norm": 0.6917139467221974, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 4116 + }, + { + "epoch": 0.04117, + "grad_norm": 0.7340533018546882, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 4117 + }, + { + "epoch": 0.04118, + "grad_norm": 0.8047197890636781, + "learning_rate": 0.003, + "loss": 4.1335, + "step": 4118 + }, + { + "epoch": 0.04119, + "grad_norm": 0.761222175960266, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 4119 + }, + { + "epoch": 0.0412, + "grad_norm": 0.760258393014731, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 4120 + }, + { + "epoch": 0.04121, + "grad_norm": 0.6534967114725192, + "learning_rate": 0.003, + "loss": 4.1187, + "step": 4121 + }, + { + "epoch": 0.04122, + "grad_norm": 0.6195285409820348, + "learning_rate": 0.003, + "loss": 4.09, + "step": 4122 + }, + { + "epoch": 0.04123, + "grad_norm": 0.5662275718484187, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 4123 + }, + { + "epoch": 0.04124, + "grad_norm": 0.6058056845766859, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 4124 + }, + { + "epoch": 0.04125, + "grad_norm": 0.6498152144398192, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 4125 + }, + { + "epoch": 0.04126, + "grad_norm": 0.6720042061777359, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 4126 + }, + { + "epoch": 0.04127, + "grad_norm": 0.7496314217642709, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 4127 + }, + { + "epoch": 0.04128, + "grad_norm": 0.7282248510466185, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 4128 + }, + { + "epoch": 0.04129, + "grad_norm": 0.6370349781617219, + "learning_rate": 0.003, + "loss": 4.109, + "step": 4129 + }, + { + "epoch": 0.0413, + "grad_norm": 0.6635234679985839, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 4130 + }, + { + "epoch": 0.04131, + "grad_norm": 0.5635265033736737, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 4131 + }, + { + "epoch": 0.04132, + "grad_norm": 0.5756981115816625, + "learning_rate": 0.003, + "loss": 4.063, + "step": 4132 + }, + { + "epoch": 0.04133, + "grad_norm": 0.606085146966565, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 4133 + }, + { + "epoch": 0.04134, + "grad_norm": 0.6223170399338656, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 4134 + }, + { + "epoch": 0.04135, + "grad_norm": 0.6714481604766104, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 4135 + }, + { + "epoch": 0.04136, + "grad_norm": 0.7812488196667416, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 4136 + }, + { + "epoch": 0.04137, + "grad_norm": 0.8124559861835243, + "learning_rate": 0.003, + "loss": 4.1183, + "step": 4137 + }, + { + "epoch": 0.04138, + "grad_norm": 0.8401690278466457, + "learning_rate": 0.003, + "loss": 4.1064, + "step": 4138 + }, + { + "epoch": 0.04139, + "grad_norm": 0.7291795936986546, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 4139 + }, + { + "epoch": 0.0414, + "grad_norm": 0.6678353895398822, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 4140 + }, + { + "epoch": 0.04141, + "grad_norm": 0.6861532303877272, + "learning_rate": 0.003, + "loss": 4.079, + "step": 4141 + }, + { + "epoch": 0.04142, + "grad_norm": 0.7092901856092118, + "learning_rate": 0.003, + "loss": 4.102, + "step": 4142 + }, + { + "epoch": 0.04143, + "grad_norm": 0.6634166148920975, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 4143 + }, + { + "epoch": 0.04144, + "grad_norm": 0.8412217403470517, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 4144 + }, + { + "epoch": 0.04145, + "grad_norm": 0.8938397308618653, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 4145 + }, + { + "epoch": 0.04146, + "grad_norm": 0.8223717282911892, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 4146 + }, + { + "epoch": 0.04147, + "grad_norm": 0.6673159910703338, + "learning_rate": 0.003, + "loss": 4.108, + "step": 4147 + }, + { + "epoch": 0.04148, + "grad_norm": 0.7137558359022244, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 4148 + }, + { + "epoch": 0.04149, + "grad_norm": 0.7868039224555098, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 4149 + }, + { + "epoch": 0.0415, + "grad_norm": 0.876551796737039, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 4150 + }, + { + "epoch": 0.04151, + "grad_norm": 0.8878288381285978, + "learning_rate": 0.003, + "loss": 4.114, + "step": 4151 + }, + { + "epoch": 0.04152, + "grad_norm": 0.8473192615849454, + "learning_rate": 0.003, + "loss": 4.1126, + "step": 4152 + }, + { + "epoch": 0.04153, + "grad_norm": 0.8500966943686163, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4153 + }, + { + "epoch": 0.04154, + "grad_norm": 0.943224454402093, + "learning_rate": 0.003, + "loss": 4.1147, + "step": 4154 + }, + { + "epoch": 0.04155, + "grad_norm": 0.9953913422292556, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 4155 + }, + { + "epoch": 0.04156, + "grad_norm": 0.8559594325769431, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 4156 + }, + { + "epoch": 0.04157, + "grad_norm": 0.7645314742737204, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 4157 + }, + { + "epoch": 0.04158, + "grad_norm": 0.8429313846745287, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 4158 + }, + { + "epoch": 0.04159, + "grad_norm": 0.8263319677990639, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4159 + }, + { + "epoch": 0.0416, + "grad_norm": 0.8278768097176705, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 4160 + }, + { + "epoch": 0.04161, + "grad_norm": 0.7397971570454055, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 4161 + }, + { + "epoch": 0.04162, + "grad_norm": 0.5958690465260493, + "learning_rate": 0.003, + "loss": 4.07, + "step": 4162 + }, + { + "epoch": 0.04163, + "grad_norm": 0.621769056675386, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 4163 + }, + { + "epoch": 0.04164, + "grad_norm": 0.6125372464726984, + "learning_rate": 0.003, + "loss": 4.1153, + "step": 4164 + }, + { + "epoch": 0.04165, + "grad_norm": 0.5500882359807601, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 4165 + }, + { + "epoch": 0.04166, + "grad_norm": 0.46198162860469383, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 4166 + }, + { + "epoch": 0.04167, + "grad_norm": 0.4913882031509486, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 4167 + }, + { + "epoch": 0.04168, + "grad_norm": 0.46831271446567896, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 4168 + }, + { + "epoch": 0.04169, + "grad_norm": 0.4563524144144197, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 4169 + }, + { + "epoch": 0.0417, + "grad_norm": 0.4624925737280829, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 4170 + }, + { + "epoch": 0.04171, + "grad_norm": 0.4639569535064953, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 4171 + }, + { + "epoch": 0.04172, + "grad_norm": 0.43354901887555225, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 4172 + }, + { + "epoch": 0.04173, + "grad_norm": 0.4392691048922942, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 4173 + }, + { + "epoch": 0.04174, + "grad_norm": 0.48730235622279516, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 4174 + }, + { + "epoch": 0.04175, + "grad_norm": 0.5667536453430381, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 4175 + }, + { + "epoch": 0.04176, + "grad_norm": 0.7418388733580782, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 4176 + }, + { + "epoch": 0.04177, + "grad_norm": 0.7679142226245451, + "learning_rate": 0.003, + "loss": 4.063, + "step": 4177 + }, + { + "epoch": 0.04178, + "grad_norm": 0.6598408448782825, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 4178 + }, + { + "epoch": 0.04179, + "grad_norm": 0.7512857701174904, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 4179 + }, + { + "epoch": 0.0418, + "grad_norm": 0.7685143970892345, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 4180 + }, + { + "epoch": 0.04181, + "grad_norm": 0.8422248145009823, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 4181 + }, + { + "epoch": 0.04182, + "grad_norm": 0.8023980969133233, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 4182 + }, + { + "epoch": 0.04183, + "grad_norm": 0.702663659555653, + "learning_rate": 0.003, + "loss": 4.078, + "step": 4183 + }, + { + "epoch": 0.04184, + "grad_norm": 0.6618786473422738, + "learning_rate": 0.003, + "loss": 4.1334, + "step": 4184 + }, + { + "epoch": 0.04185, + "grad_norm": 0.7375976618914372, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 4185 + }, + { + "epoch": 0.04186, + "grad_norm": 0.7707592655370573, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 4186 + }, + { + "epoch": 0.04187, + "grad_norm": 0.809915079101846, + "learning_rate": 0.003, + "loss": 4.112, + "step": 4187 + }, + { + "epoch": 0.04188, + "grad_norm": 0.739111858575425, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 4188 + }, + { + "epoch": 0.04189, + "grad_norm": 0.7537682286232873, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 4189 + }, + { + "epoch": 0.0419, + "grad_norm": 0.7576452519051043, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 4190 + }, + { + "epoch": 0.04191, + "grad_norm": 0.8014688841965723, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 4191 + }, + { + "epoch": 0.04192, + "grad_norm": 0.7316925408499877, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 4192 + }, + { + "epoch": 0.04193, + "grad_norm": 0.6156878728692767, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 4193 + }, + { + "epoch": 0.04194, + "grad_norm": 0.5998790257144186, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 4194 + }, + { + "epoch": 0.04195, + "grad_norm": 0.7178979637287529, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 4195 + }, + { + "epoch": 0.04196, + "grad_norm": 0.8768854833688874, + "learning_rate": 0.003, + "loss": 4.075, + "step": 4196 + }, + { + "epoch": 0.04197, + "grad_norm": 0.9311099760980657, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 4197 + }, + { + "epoch": 0.04198, + "grad_norm": 1.0512651617347428, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 4198 + }, + { + "epoch": 0.04199, + "grad_norm": 1.0272831740067394, + "learning_rate": 0.003, + "loss": 4.1326, + "step": 4199 + }, + { + "epoch": 0.042, + "grad_norm": 1.1198820610929303, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 4200 + }, + { + "epoch": 0.04201, + "grad_norm": 0.8889652064412181, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 4201 + }, + { + "epoch": 0.04202, + "grad_norm": 0.7853862624962952, + "learning_rate": 0.003, + "loss": 4.101, + "step": 4202 + }, + { + "epoch": 0.04203, + "grad_norm": 0.6644444229529121, + "learning_rate": 0.003, + "loss": 4.1202, + "step": 4203 + }, + { + "epoch": 0.04204, + "grad_norm": 0.7163125421216407, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 4204 + }, + { + "epoch": 0.04205, + "grad_norm": 0.6368900777451744, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 4205 + }, + { + "epoch": 0.04206, + "grad_norm": 0.7169269055522234, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 4206 + }, + { + "epoch": 0.04207, + "grad_norm": 0.6628041731427305, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 4207 + }, + { + "epoch": 0.04208, + "grad_norm": 0.5816518187368057, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 4208 + }, + { + "epoch": 0.04209, + "grad_norm": 0.5829511518810705, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 4209 + }, + { + "epoch": 0.0421, + "grad_norm": 0.6342671842182177, + "learning_rate": 0.003, + "loss": 4.058, + "step": 4210 + }, + { + "epoch": 0.04211, + "grad_norm": 0.6111976143762199, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 4211 + }, + { + "epoch": 0.04212, + "grad_norm": 0.6570410582270553, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 4212 + }, + { + "epoch": 0.04213, + "grad_norm": 0.6462946177588936, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 4213 + }, + { + "epoch": 0.04214, + "grad_norm": 0.6794802047783476, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 4214 + }, + { + "epoch": 0.04215, + "grad_norm": 0.6817070655241206, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 4215 + }, + { + "epoch": 0.04216, + "grad_norm": 0.7219777388665342, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 4216 + }, + { + "epoch": 0.04217, + "grad_norm": 0.7274895322201507, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 4217 + }, + { + "epoch": 0.04218, + "grad_norm": 0.6980472536574723, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 4218 + }, + { + "epoch": 0.04219, + "grad_norm": 0.7113432512720306, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 4219 + }, + { + "epoch": 0.0422, + "grad_norm": 0.5460815993555583, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 4220 + }, + { + "epoch": 0.04221, + "grad_norm": 0.5529910109723699, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4221 + }, + { + "epoch": 0.04222, + "grad_norm": 0.5779962066381917, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 4222 + }, + { + "epoch": 0.04223, + "grad_norm": 0.5025914070619163, + "learning_rate": 0.003, + "loss": 4.1017, + "step": 4223 + }, + { + "epoch": 0.04224, + "grad_norm": 0.5484624177477682, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 4224 + }, + { + "epoch": 0.04225, + "grad_norm": 0.6571518113771674, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 4225 + }, + { + "epoch": 0.04226, + "grad_norm": 0.6748654362977881, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 4226 + }, + { + "epoch": 0.04227, + "grad_norm": 0.5418039810816547, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 4227 + }, + { + "epoch": 0.04228, + "grad_norm": 0.6072497986030803, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 4228 + }, + { + "epoch": 0.04229, + "grad_norm": 0.7041642718393254, + "learning_rate": 0.003, + "loss": 4.0914, + "step": 4229 + }, + { + "epoch": 0.0423, + "grad_norm": 0.7619312147335537, + "learning_rate": 0.003, + "loss": 4.084, + "step": 4230 + }, + { + "epoch": 0.04231, + "grad_norm": 0.9669329888994561, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 4231 + }, + { + "epoch": 0.04232, + "grad_norm": 1.1879689622532548, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 4232 + }, + { + "epoch": 0.04233, + "grad_norm": 0.7154487535559305, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 4233 + }, + { + "epoch": 0.04234, + "grad_norm": 0.7382827235086054, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 4234 + }, + { + "epoch": 0.04235, + "grad_norm": 0.8307602440415276, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 4235 + }, + { + "epoch": 0.04236, + "grad_norm": 0.7837718765421066, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 4236 + }, + { + "epoch": 0.04237, + "grad_norm": 0.8007436934007545, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 4237 + }, + { + "epoch": 0.04238, + "grad_norm": 0.7887818150247421, + "learning_rate": 0.003, + "loss": 4.1211, + "step": 4238 + }, + { + "epoch": 0.04239, + "grad_norm": 0.6580451410215259, + "learning_rate": 0.003, + "loss": 4.097, + "step": 4239 + }, + { + "epoch": 0.0424, + "grad_norm": 0.546843376874823, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 4240 + }, + { + "epoch": 0.04241, + "grad_norm": 0.5886717662869974, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 4241 + }, + { + "epoch": 0.04242, + "grad_norm": 0.6049602412284655, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 4242 + }, + { + "epoch": 0.04243, + "grad_norm": 0.6826126726675993, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4243 + }, + { + "epoch": 0.04244, + "grad_norm": 0.5911137245491322, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 4244 + }, + { + "epoch": 0.04245, + "grad_norm": 0.5204413456367117, + "learning_rate": 0.003, + "loss": 4.078, + "step": 4245 + }, + { + "epoch": 0.04246, + "grad_norm": 0.47365304755715926, + "learning_rate": 0.003, + "loss": 4.1023, + "step": 4246 + }, + { + "epoch": 0.04247, + "grad_norm": 0.48257724991117407, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 4247 + }, + { + "epoch": 0.04248, + "grad_norm": 0.5051408347291324, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 4248 + }, + { + "epoch": 0.04249, + "grad_norm": 0.5794155131513592, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 4249 + }, + { + "epoch": 0.0425, + "grad_norm": 0.7867314120435951, + "learning_rate": 0.003, + "loss": 4.0927, + "step": 4250 + }, + { + "epoch": 0.04251, + "grad_norm": 0.9201761299078502, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 4251 + }, + { + "epoch": 0.04252, + "grad_norm": 1.012211071967095, + "learning_rate": 0.003, + "loss": 4.1118, + "step": 4252 + }, + { + "epoch": 0.04253, + "grad_norm": 0.8380231757288324, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 4253 + }, + { + "epoch": 0.04254, + "grad_norm": 0.7186349457402139, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 4254 + }, + { + "epoch": 0.04255, + "grad_norm": 0.6339738404406828, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 4255 + }, + { + "epoch": 0.04256, + "grad_norm": 0.7527994661117764, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4256 + }, + { + "epoch": 0.04257, + "grad_norm": 0.8381621808860645, + "learning_rate": 0.003, + "loss": 4.1067, + "step": 4257 + }, + { + "epoch": 0.04258, + "grad_norm": 0.8249872352383231, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 4258 + }, + { + "epoch": 0.04259, + "grad_norm": 0.6413019870762352, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 4259 + }, + { + "epoch": 0.0426, + "grad_norm": 0.643466733125968, + "learning_rate": 0.003, + "loss": 4.083, + "step": 4260 + }, + { + "epoch": 0.04261, + "grad_norm": 0.6489155354599366, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 4261 + }, + { + "epoch": 0.04262, + "grad_norm": 0.5889736647415632, + "learning_rate": 0.003, + "loss": 4.11, + "step": 4262 + }, + { + "epoch": 0.04263, + "grad_norm": 0.6282126889819096, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 4263 + }, + { + "epoch": 0.04264, + "grad_norm": 0.6788118395137857, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 4264 + }, + { + "epoch": 0.04265, + "grad_norm": 0.6804130516908362, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 4265 + }, + { + "epoch": 0.04266, + "grad_norm": 0.5720855817919384, + "learning_rate": 0.003, + "loss": 4.078, + "step": 4266 + }, + { + "epoch": 0.04267, + "grad_norm": 0.6357803717819517, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 4267 + }, + { + "epoch": 0.04268, + "grad_norm": 0.6753361517705672, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 4268 + }, + { + "epoch": 0.04269, + "grad_norm": 0.6942943508364977, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 4269 + }, + { + "epoch": 0.0427, + "grad_norm": 0.6462774593758114, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 4270 + }, + { + "epoch": 0.04271, + "grad_norm": 0.6375734941232201, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 4271 + }, + { + "epoch": 0.04272, + "grad_norm": 0.6097793125658084, + "learning_rate": 0.003, + "loss": 4.077, + "step": 4272 + }, + { + "epoch": 0.04273, + "grad_norm": 0.6119605743209566, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 4273 + }, + { + "epoch": 0.04274, + "grad_norm": 0.6847693274298686, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 4274 + }, + { + "epoch": 0.04275, + "grad_norm": 0.8432713950768773, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 4275 + }, + { + "epoch": 0.04276, + "grad_norm": 0.7361315913889281, + "learning_rate": 0.003, + "loss": 4.109, + "step": 4276 + }, + { + "epoch": 0.04277, + "grad_norm": 0.7202575384978905, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 4277 + }, + { + "epoch": 0.04278, + "grad_norm": 0.7981576374986135, + "learning_rate": 0.003, + "loss": 4.102, + "step": 4278 + }, + { + "epoch": 0.04279, + "grad_norm": 0.854159519325767, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 4279 + }, + { + "epoch": 0.0428, + "grad_norm": 0.8541539185029711, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 4280 + }, + { + "epoch": 0.04281, + "grad_norm": 0.8688494609195975, + "learning_rate": 0.003, + "loss": 4.1189, + "step": 4281 + }, + { + "epoch": 0.04282, + "grad_norm": 0.8712224395453085, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 4282 + }, + { + "epoch": 0.04283, + "grad_norm": 0.8729851059335475, + "learning_rate": 0.003, + "loss": 4.1307, + "step": 4283 + }, + { + "epoch": 0.04284, + "grad_norm": 0.9289611562803178, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 4284 + }, + { + "epoch": 0.04285, + "grad_norm": 1.0038816165737348, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 4285 + }, + { + "epoch": 0.04286, + "grad_norm": 0.9223210635321651, + "learning_rate": 0.003, + "loss": 4.1269, + "step": 4286 + }, + { + "epoch": 0.04287, + "grad_norm": 0.8724479834175934, + "learning_rate": 0.003, + "loss": 4.1034, + "step": 4287 + }, + { + "epoch": 0.04288, + "grad_norm": 0.8512823756099549, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 4288 + }, + { + "epoch": 0.04289, + "grad_norm": 0.7435096439422789, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 4289 + }, + { + "epoch": 0.0429, + "grad_norm": 0.6163882509252051, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 4290 + }, + { + "epoch": 0.04291, + "grad_norm": 0.5727886786494579, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 4291 + }, + { + "epoch": 0.04292, + "grad_norm": 0.4874323636863275, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 4292 + }, + { + "epoch": 0.04293, + "grad_norm": 0.49681317597562485, + "learning_rate": 0.003, + "loss": 4.1125, + "step": 4293 + }, + { + "epoch": 0.04294, + "grad_norm": 0.5393461035890804, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 4294 + }, + { + "epoch": 0.04295, + "grad_norm": 0.5598242336133272, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 4295 + }, + { + "epoch": 0.04296, + "grad_norm": 0.5462261338840079, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 4296 + }, + { + "epoch": 0.04297, + "grad_norm": 0.560226591869765, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 4297 + }, + { + "epoch": 0.04298, + "grad_norm": 0.5736597950711811, + "learning_rate": 0.003, + "loss": 4.095, + "step": 4298 + }, + { + "epoch": 0.04299, + "grad_norm": 0.5719547076361956, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 4299 + }, + { + "epoch": 0.043, + "grad_norm": 0.6567880539566495, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 4300 + }, + { + "epoch": 0.04301, + "grad_norm": 0.7327114464002823, + "learning_rate": 0.003, + "loss": 4.1167, + "step": 4301 + }, + { + "epoch": 0.04302, + "grad_norm": 0.9491126587480954, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 4302 + }, + { + "epoch": 0.04303, + "grad_norm": 1.0751357485483097, + "learning_rate": 0.003, + "loss": 4.114, + "step": 4303 + }, + { + "epoch": 0.04304, + "grad_norm": 0.853441528627707, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 4304 + }, + { + "epoch": 0.04305, + "grad_norm": 0.7891447604124461, + "learning_rate": 0.003, + "loss": 4.098, + "step": 4305 + }, + { + "epoch": 0.04306, + "grad_norm": 0.7705290051883236, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4306 + }, + { + "epoch": 0.04307, + "grad_norm": 0.7804190933125599, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4307 + }, + { + "epoch": 0.04308, + "grad_norm": 0.818741405895089, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 4308 + }, + { + "epoch": 0.04309, + "grad_norm": 0.8469770226448077, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 4309 + }, + { + "epoch": 0.0431, + "grad_norm": 0.8427103213267478, + "learning_rate": 0.003, + "loss": 4.11, + "step": 4310 + }, + { + "epoch": 0.04311, + "grad_norm": 0.8526851187182092, + "learning_rate": 0.003, + "loss": 4.107, + "step": 4311 + }, + { + "epoch": 0.04312, + "grad_norm": 0.8894605281697073, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 4312 + }, + { + "epoch": 0.04313, + "grad_norm": 0.777697251403963, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4313 + }, + { + "epoch": 0.04314, + "grad_norm": 0.7516845219704149, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 4314 + }, + { + "epoch": 0.04315, + "grad_norm": 0.7231748206646733, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 4315 + }, + { + "epoch": 0.04316, + "grad_norm": 0.7304128219563059, + "learning_rate": 0.003, + "loss": 4.1077, + "step": 4316 + }, + { + "epoch": 0.04317, + "grad_norm": 0.7373884724965908, + "learning_rate": 0.003, + "loss": 4.1208, + "step": 4317 + }, + { + "epoch": 0.04318, + "grad_norm": 0.7620039672233535, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 4318 + }, + { + "epoch": 0.04319, + "grad_norm": 0.6194456819960122, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 4319 + }, + { + "epoch": 0.0432, + "grad_norm": 0.5240778650777331, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 4320 + }, + { + "epoch": 0.04321, + "grad_norm": 0.6164342622056979, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 4321 + }, + { + "epoch": 0.04322, + "grad_norm": 0.6370627128469308, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 4322 + }, + { + "epoch": 0.04323, + "grad_norm": 0.7001769430690609, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 4323 + }, + { + "epoch": 0.04324, + "grad_norm": 0.707373357298591, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 4324 + }, + { + "epoch": 0.04325, + "grad_norm": 0.6099215364130716, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 4325 + }, + { + "epoch": 0.04326, + "grad_norm": 0.5207906423920813, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 4326 + }, + { + "epoch": 0.04327, + "grad_norm": 0.41326401107626454, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 4327 + }, + { + "epoch": 0.04328, + "grad_norm": 0.4465029817361841, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 4328 + }, + { + "epoch": 0.04329, + "grad_norm": 0.47130737963616937, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 4329 + }, + { + "epoch": 0.0433, + "grad_norm": 0.4632055157671912, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 4330 + }, + { + "epoch": 0.04331, + "grad_norm": 0.48350948156728496, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 4331 + }, + { + "epoch": 0.04332, + "grad_norm": 0.6018315111885648, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 4332 + }, + { + "epoch": 0.04333, + "grad_norm": 0.7157983104151012, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 4333 + }, + { + "epoch": 0.04334, + "grad_norm": 0.7367000664548994, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 4334 + }, + { + "epoch": 0.04335, + "grad_norm": 0.6776066215357341, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 4335 + }, + { + "epoch": 0.04336, + "grad_norm": 0.5859073646266592, + "learning_rate": 0.003, + "loss": 4.093, + "step": 4336 + }, + { + "epoch": 0.04337, + "grad_norm": 0.6713825019098838, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 4337 + }, + { + "epoch": 0.04338, + "grad_norm": 0.6858032169096283, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4338 + }, + { + "epoch": 0.04339, + "grad_norm": 0.7716180877663685, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 4339 + }, + { + "epoch": 0.0434, + "grad_norm": 0.7730784830194711, + "learning_rate": 0.003, + "loss": 4.072, + "step": 4340 + }, + { + "epoch": 0.04341, + "grad_norm": 0.8559847777611459, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 4341 + }, + { + "epoch": 0.04342, + "grad_norm": 0.9166373502324513, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 4342 + }, + { + "epoch": 0.04343, + "grad_norm": 1.0621855134342457, + "learning_rate": 0.003, + "loss": 4.1292, + "step": 4343 + }, + { + "epoch": 0.04344, + "grad_norm": 1.0049079301936457, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 4344 + }, + { + "epoch": 0.04345, + "grad_norm": 0.8913175571031999, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 4345 + }, + { + "epoch": 0.04346, + "grad_norm": 0.7813828602877204, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 4346 + }, + { + "epoch": 0.04347, + "grad_norm": 0.7091212761865211, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 4347 + }, + { + "epoch": 0.04348, + "grad_norm": 0.7040249709581968, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 4348 + }, + { + "epoch": 0.04349, + "grad_norm": 0.7101126681508375, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4349 + }, + { + "epoch": 0.0435, + "grad_norm": 0.6324595620509315, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 4350 + }, + { + "epoch": 0.04351, + "grad_norm": 0.5195700353468761, + "learning_rate": 0.003, + "loss": 4.0917, + "step": 4351 + }, + { + "epoch": 0.04352, + "grad_norm": 0.4524919284252752, + "learning_rate": 0.003, + "loss": 4.1049, + "step": 4352 + }, + { + "epoch": 0.04353, + "grad_norm": 0.46514433688674767, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 4353 + }, + { + "epoch": 0.04354, + "grad_norm": 0.43037837005842, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 4354 + }, + { + "epoch": 0.04355, + "grad_norm": 0.5197681358673037, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 4355 + }, + { + "epoch": 0.04356, + "grad_norm": 0.5917984833368946, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 4356 + }, + { + "epoch": 0.04357, + "grad_norm": 0.6540804473344335, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 4357 + }, + { + "epoch": 0.04358, + "grad_norm": 0.6473312367649264, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 4358 + }, + { + "epoch": 0.04359, + "grad_norm": 0.5591243854801702, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 4359 + }, + { + "epoch": 0.0436, + "grad_norm": 0.4843122280060517, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 4360 + }, + { + "epoch": 0.04361, + "grad_norm": 0.5083606329303971, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 4361 + }, + { + "epoch": 0.04362, + "grad_norm": 0.582546117388078, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 4362 + }, + { + "epoch": 0.04363, + "grad_norm": 0.5557114727186344, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 4363 + }, + { + "epoch": 0.04364, + "grad_norm": 0.561544283816565, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 4364 + }, + { + "epoch": 0.04365, + "grad_norm": 0.6109748047854819, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 4365 + }, + { + "epoch": 0.04366, + "grad_norm": 0.8165391166593751, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 4366 + }, + { + "epoch": 0.04367, + "grad_norm": 0.8884878888806635, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4367 + }, + { + "epoch": 0.04368, + "grad_norm": 0.8932227744180633, + "learning_rate": 0.003, + "loss": 4.091, + "step": 4368 + }, + { + "epoch": 0.04369, + "grad_norm": 0.9358343510342729, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 4369 + }, + { + "epoch": 0.0437, + "grad_norm": 0.901908553082516, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 4370 + }, + { + "epoch": 0.04371, + "grad_norm": 0.9342123254628153, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 4371 + }, + { + "epoch": 0.04372, + "grad_norm": 0.8954362784195181, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 4372 + }, + { + "epoch": 0.04373, + "grad_norm": 1.0170323746319703, + "learning_rate": 0.003, + "loss": 4.1188, + "step": 4373 + }, + { + "epoch": 0.04374, + "grad_norm": 1.0189758067564425, + "learning_rate": 0.003, + "loss": 4.1611, + "step": 4374 + }, + { + "epoch": 0.04375, + "grad_norm": 1.0609464482319642, + "learning_rate": 0.003, + "loss": 4.077, + "step": 4375 + }, + { + "epoch": 0.04376, + "grad_norm": 0.8849042629268745, + "learning_rate": 0.003, + "loss": 4.1144, + "step": 4376 + }, + { + "epoch": 0.04377, + "grad_norm": 0.87862164383254, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 4377 + }, + { + "epoch": 0.04378, + "grad_norm": 0.8670324017107475, + "learning_rate": 0.003, + "loss": 4.1258, + "step": 4378 + }, + { + "epoch": 0.04379, + "grad_norm": 0.897462536565258, + "learning_rate": 0.003, + "loss": 4.1421, + "step": 4379 + }, + { + "epoch": 0.0438, + "grad_norm": 0.7129600323468231, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 4380 + }, + { + "epoch": 0.04381, + "grad_norm": 0.7592826101762346, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4381 + }, + { + "epoch": 0.04382, + "grad_norm": 0.8362043111237921, + "learning_rate": 0.003, + "loss": 4.1122, + "step": 4382 + }, + { + "epoch": 0.04383, + "grad_norm": 0.8700932388308823, + "learning_rate": 0.003, + "loss": 4.0908, + "step": 4383 + }, + { + "epoch": 0.04384, + "grad_norm": 0.670572084649702, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 4384 + }, + { + "epoch": 0.04385, + "grad_norm": 0.577632317680403, + "learning_rate": 0.003, + "loss": 4.1219, + "step": 4385 + }, + { + "epoch": 0.04386, + "grad_norm": 0.5219618436399083, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 4386 + }, + { + "epoch": 0.04387, + "grad_norm": 0.5003050289804473, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 4387 + }, + { + "epoch": 0.04388, + "grad_norm": 0.5679169200133987, + "learning_rate": 0.003, + "loss": 4.076, + "step": 4388 + }, + { + "epoch": 0.04389, + "grad_norm": 0.57783562344273, + "learning_rate": 0.003, + "loss": 4.1069, + "step": 4389 + }, + { + "epoch": 0.0439, + "grad_norm": 0.5543997756507195, + "learning_rate": 0.003, + "loss": 4.1, + "step": 4390 + }, + { + "epoch": 0.04391, + "grad_norm": 0.5339461438889176, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 4391 + }, + { + "epoch": 0.04392, + "grad_norm": 0.5197021746130168, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 4392 + }, + { + "epoch": 0.04393, + "grad_norm": 0.5334772379221283, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 4393 + }, + { + "epoch": 0.04394, + "grad_norm": 0.6113600234878536, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 4394 + }, + { + "epoch": 0.04395, + "grad_norm": 0.597053191791215, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 4395 + }, + { + "epoch": 0.04396, + "grad_norm": 0.660906503348518, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 4396 + }, + { + "epoch": 0.04397, + "grad_norm": 0.8004570370487631, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 4397 + }, + { + "epoch": 0.04398, + "grad_norm": 0.8273740486772345, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 4398 + }, + { + "epoch": 0.04399, + "grad_norm": 0.6852587824492433, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 4399 + }, + { + "epoch": 0.044, + "grad_norm": 0.6770664379505986, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 4400 + }, + { + "epoch": 0.04401, + "grad_norm": 0.7756401866516791, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 4401 + }, + { + "epoch": 0.04402, + "grad_norm": 0.6368039660419332, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 4402 + }, + { + "epoch": 0.04403, + "grad_norm": 0.5549099284832866, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 4403 + }, + { + "epoch": 0.04404, + "grad_norm": 0.5709714007017144, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 4404 + }, + { + "epoch": 0.04405, + "grad_norm": 0.6169499832880386, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 4405 + }, + { + "epoch": 0.04406, + "grad_norm": 0.6253344936987876, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 4406 + }, + { + "epoch": 0.04407, + "grad_norm": 0.6874431480850114, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 4407 + }, + { + "epoch": 0.04408, + "grad_norm": 0.7105494644337589, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 4408 + }, + { + "epoch": 0.04409, + "grad_norm": 0.6640741752758204, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 4409 + }, + { + "epoch": 0.0441, + "grad_norm": 0.6449209524533978, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4410 + }, + { + "epoch": 0.04411, + "grad_norm": 0.6817058779115432, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 4411 + }, + { + "epoch": 0.04412, + "grad_norm": 0.7634037323352982, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 4412 + }, + { + "epoch": 0.04413, + "grad_norm": 0.6623005904400171, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 4413 + }, + { + "epoch": 0.04414, + "grad_norm": 0.7227111039174143, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 4414 + }, + { + "epoch": 0.04415, + "grad_norm": 0.6526181087831874, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 4415 + }, + { + "epoch": 0.04416, + "grad_norm": 0.6380473500954813, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 4416 + }, + { + "epoch": 0.04417, + "grad_norm": 0.6412650580690743, + "learning_rate": 0.003, + "loss": 4.044, + "step": 4417 + }, + { + "epoch": 0.04418, + "grad_norm": 0.6184151466186707, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 4418 + }, + { + "epoch": 0.04419, + "grad_norm": 0.624511618550086, + "learning_rate": 0.003, + "loss": 4.1036, + "step": 4419 + }, + { + "epoch": 0.0442, + "grad_norm": 0.7052897962380323, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 4420 + }, + { + "epoch": 0.04421, + "grad_norm": 0.8047186105521766, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 4421 + }, + { + "epoch": 0.04422, + "grad_norm": 0.7513040405791254, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 4422 + }, + { + "epoch": 0.04423, + "grad_norm": 0.7909280226642721, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 4423 + }, + { + "epoch": 0.04424, + "grad_norm": 0.7160853544305391, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 4424 + }, + { + "epoch": 0.04425, + "grad_norm": 0.6229635276587001, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 4425 + }, + { + "epoch": 0.04426, + "grad_norm": 0.7069448886404887, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 4426 + }, + { + "epoch": 0.04427, + "grad_norm": 0.754305936128332, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 4427 + }, + { + "epoch": 0.04428, + "grad_norm": 0.6758029386614973, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 4428 + }, + { + "epoch": 0.04429, + "grad_norm": 0.5775456486555952, + "learning_rate": 0.003, + "loss": 4.1143, + "step": 4429 + }, + { + "epoch": 0.0443, + "grad_norm": 0.5923638061982324, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 4430 + }, + { + "epoch": 0.04431, + "grad_norm": 0.6501230290637654, + "learning_rate": 0.003, + "loss": 4.0996, + "step": 4431 + }, + { + "epoch": 0.04432, + "grad_norm": 0.7029663761240024, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 4432 + }, + { + "epoch": 0.04433, + "grad_norm": 0.7340027924850606, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 4433 + }, + { + "epoch": 0.04434, + "grad_norm": 0.8613767026639428, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 4434 + }, + { + "epoch": 0.04435, + "grad_norm": 0.9162989248386401, + "learning_rate": 0.003, + "loss": 4.1016, + "step": 4435 + }, + { + "epoch": 0.04436, + "grad_norm": 0.9016494580946359, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 4436 + }, + { + "epoch": 0.04437, + "grad_norm": 0.7350677463029225, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 4437 + }, + { + "epoch": 0.04438, + "grad_norm": 0.7647870168682628, + "learning_rate": 0.003, + "loss": 4.1355, + "step": 4438 + }, + { + "epoch": 0.04439, + "grad_norm": 0.6984080906026269, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4439 + }, + { + "epoch": 0.0444, + "grad_norm": 0.6575544700213993, + "learning_rate": 0.003, + "loss": 4.054, + "step": 4440 + }, + { + "epoch": 0.04441, + "grad_norm": 0.7132948500801954, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 4441 + }, + { + "epoch": 0.04442, + "grad_norm": 0.8043446965273001, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 4442 + }, + { + "epoch": 0.04443, + "grad_norm": 0.8940531753671871, + "learning_rate": 0.003, + "loss": 4.064, + "step": 4443 + }, + { + "epoch": 0.04444, + "grad_norm": 0.7987210006570833, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 4444 + }, + { + "epoch": 0.04445, + "grad_norm": 0.8025825713231252, + "learning_rate": 0.003, + "loss": 4.096, + "step": 4445 + }, + { + "epoch": 0.04446, + "grad_norm": 0.7697050535672169, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 4446 + }, + { + "epoch": 0.04447, + "grad_norm": 0.6942792331675097, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 4447 + }, + { + "epoch": 0.04448, + "grad_norm": 0.7185179455675405, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 4448 + }, + { + "epoch": 0.04449, + "grad_norm": 0.7328813810404422, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 4449 + }, + { + "epoch": 0.0445, + "grad_norm": 0.7709578480451924, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 4450 + }, + { + "epoch": 0.04451, + "grad_norm": 0.7308338252174789, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 4451 + }, + { + "epoch": 0.04452, + "grad_norm": 0.6422081684584662, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 4452 + }, + { + "epoch": 0.04453, + "grad_norm": 0.639249393035527, + "learning_rate": 0.003, + "loss": 4.1057, + "step": 4453 + }, + { + "epoch": 0.04454, + "grad_norm": 0.6363441647866156, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 4454 + }, + { + "epoch": 0.04455, + "grad_norm": 0.652245490082349, + "learning_rate": 0.003, + "loss": 4.0941, + "step": 4455 + }, + { + "epoch": 0.04456, + "grad_norm": 0.5994733088949281, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 4456 + }, + { + "epoch": 0.04457, + "grad_norm": 0.6141742208882366, + "learning_rate": 0.003, + "loss": 4.078, + "step": 4457 + }, + { + "epoch": 0.04458, + "grad_norm": 0.6431173710373176, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 4458 + }, + { + "epoch": 0.04459, + "grad_norm": 0.5854312328478273, + "learning_rate": 0.003, + "loss": 4.1041, + "step": 4459 + }, + { + "epoch": 0.0446, + "grad_norm": 0.5991041596568328, + "learning_rate": 0.003, + "loss": 4.1005, + "step": 4460 + }, + { + "epoch": 0.04461, + "grad_norm": 0.5884233401703991, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 4461 + }, + { + "epoch": 0.04462, + "grad_norm": 0.5919443646043252, + "learning_rate": 0.003, + "loss": 4.056, + "step": 4462 + }, + { + "epoch": 0.04463, + "grad_norm": 0.5415194940896524, + "learning_rate": 0.003, + "loss": 4.068, + "step": 4463 + }, + { + "epoch": 0.04464, + "grad_norm": 0.5360015777059575, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 4464 + }, + { + "epoch": 0.04465, + "grad_norm": 0.6300439053158542, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 4465 + }, + { + "epoch": 0.04466, + "grad_norm": 0.7094522015304044, + "learning_rate": 0.003, + "loss": 4.087, + "step": 4466 + }, + { + "epoch": 0.04467, + "grad_norm": 0.6939291951969437, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 4467 + }, + { + "epoch": 0.04468, + "grad_norm": 0.6199543667626091, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 4468 + }, + { + "epoch": 0.04469, + "grad_norm": 0.6035414948465273, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 4469 + }, + { + "epoch": 0.0447, + "grad_norm": 0.7568692828171963, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 4470 + }, + { + "epoch": 0.04471, + "grad_norm": 0.7302919100887635, + "learning_rate": 0.003, + "loss": 4.1132, + "step": 4471 + }, + { + "epoch": 0.04472, + "grad_norm": 0.6887029287500837, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 4472 + }, + { + "epoch": 0.04473, + "grad_norm": 0.7622195713706735, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 4473 + }, + { + "epoch": 0.04474, + "grad_norm": 0.7894422125384837, + "learning_rate": 0.003, + "loss": 4.066, + "step": 4474 + }, + { + "epoch": 0.04475, + "grad_norm": 0.9071470506937508, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 4475 + }, + { + "epoch": 0.04476, + "grad_norm": 1.064163877723333, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 4476 + }, + { + "epoch": 0.04477, + "grad_norm": 0.9806685492584728, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 4477 + }, + { + "epoch": 0.04478, + "grad_norm": 0.7197683093501224, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 4478 + }, + { + "epoch": 0.04479, + "grad_norm": 0.7378826340900297, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4479 + }, + { + "epoch": 0.0448, + "grad_norm": 0.7198996405331812, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 4480 + }, + { + "epoch": 0.04481, + "grad_norm": 0.7885057092597215, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 4481 + }, + { + "epoch": 0.04482, + "grad_norm": 0.7352742812756904, + "learning_rate": 0.003, + "loss": 4.0978, + "step": 4482 + }, + { + "epoch": 0.04483, + "grad_norm": 0.7336680634853878, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 4483 + }, + { + "epoch": 0.04484, + "grad_norm": 0.7115845754486809, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4484 + }, + { + "epoch": 0.04485, + "grad_norm": 0.6457673787180004, + "learning_rate": 0.003, + "loss": 4.1095, + "step": 4485 + }, + { + "epoch": 0.04486, + "grad_norm": 0.7225146422529924, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 4486 + }, + { + "epoch": 0.04487, + "grad_norm": 0.7967991078840371, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 4487 + }, + { + "epoch": 0.04488, + "grad_norm": 0.7064552914849703, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 4488 + }, + { + "epoch": 0.04489, + "grad_norm": 0.6804355160513849, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 4489 + }, + { + "epoch": 0.0449, + "grad_norm": 0.6369684785558448, + "learning_rate": 0.003, + "loss": 4.096, + "step": 4490 + }, + { + "epoch": 0.04491, + "grad_norm": 0.5427231120143929, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 4491 + }, + { + "epoch": 0.04492, + "grad_norm": 0.5678830951564081, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 4492 + }, + { + "epoch": 0.04493, + "grad_norm": 0.6069689991946302, + "learning_rate": 0.003, + "loss": 4.0986, + "step": 4493 + }, + { + "epoch": 0.04494, + "grad_norm": 0.6239806380246308, + "learning_rate": 0.003, + "loss": 4.076, + "step": 4494 + }, + { + "epoch": 0.04495, + "grad_norm": 0.6686098051964097, + "learning_rate": 0.003, + "loss": 4.077, + "step": 4495 + }, + { + "epoch": 0.04496, + "grad_norm": 0.8281674050682258, + "learning_rate": 0.003, + "loss": 4.1204, + "step": 4496 + }, + { + "epoch": 0.04497, + "grad_norm": 0.8870826341443404, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 4497 + }, + { + "epoch": 0.04498, + "grad_norm": 0.9529015457034746, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 4498 + }, + { + "epoch": 0.04499, + "grad_norm": 0.9867686635241535, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 4499 + }, + { + "epoch": 0.045, + "grad_norm": 0.8794998465797702, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 4500 + }, + { + "epoch": 0.04501, + "grad_norm": 0.8609876507003483, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4501 + }, + { + "epoch": 0.04502, + "grad_norm": 0.8096150198550298, + "learning_rate": 0.003, + "loss": 4.1267, + "step": 4502 + }, + { + "epoch": 0.04503, + "grad_norm": 0.7589865609283131, + "learning_rate": 0.003, + "loss": 4.123, + "step": 4503 + }, + { + "epoch": 0.04504, + "grad_norm": 0.630220263156277, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 4504 + }, + { + "epoch": 0.04505, + "grad_norm": 0.6078130525658336, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 4505 + }, + { + "epoch": 0.04506, + "grad_norm": 0.5638704134505449, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 4506 + }, + { + "epoch": 0.04507, + "grad_norm": 0.5582573252429831, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 4507 + }, + { + "epoch": 0.04508, + "grad_norm": 0.6213405484449163, + "learning_rate": 0.003, + "loss": 4.0835, + "step": 4508 + }, + { + "epoch": 0.04509, + "grad_norm": 0.6800822195676799, + "learning_rate": 0.003, + "loss": 4.1072, + "step": 4509 + }, + { + "epoch": 0.0451, + "grad_norm": 0.7636653822042593, + "learning_rate": 0.003, + "loss": 4.1128, + "step": 4510 + }, + { + "epoch": 0.04511, + "grad_norm": 0.7433348959946258, + "learning_rate": 0.003, + "loss": 4.0956, + "step": 4511 + }, + { + "epoch": 0.04512, + "grad_norm": 0.8143204850899498, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 4512 + }, + { + "epoch": 0.04513, + "grad_norm": 0.9461578363186757, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 4513 + }, + { + "epoch": 0.04514, + "grad_norm": 0.9622034837786506, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 4514 + }, + { + "epoch": 0.04515, + "grad_norm": 0.9109451916056656, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 4515 + }, + { + "epoch": 0.04516, + "grad_norm": 0.7486022467468222, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4516 + }, + { + "epoch": 0.04517, + "grad_norm": 0.6503549323906535, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 4517 + }, + { + "epoch": 0.04518, + "grad_norm": 0.6045377923397035, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 4518 + }, + { + "epoch": 0.04519, + "grad_norm": 0.5001799412246596, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 4519 + }, + { + "epoch": 0.0452, + "grad_norm": 0.5122462242207521, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 4520 + }, + { + "epoch": 0.04521, + "grad_norm": 0.5031609235449497, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 4521 + }, + { + "epoch": 0.04522, + "grad_norm": 0.5162725958210236, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 4522 + }, + { + "epoch": 0.04523, + "grad_norm": 0.5792647669860322, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 4523 + }, + { + "epoch": 0.04524, + "grad_norm": 0.6958603120839049, + "learning_rate": 0.003, + "loss": 4.051, + "step": 4524 + }, + { + "epoch": 0.04525, + "grad_norm": 0.821336121455205, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 4525 + }, + { + "epoch": 0.04526, + "grad_norm": 0.7982722812786827, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4526 + }, + { + "epoch": 0.04527, + "grad_norm": 0.6835709987269241, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 4527 + }, + { + "epoch": 0.04528, + "grad_norm": 0.5621384476620582, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 4528 + }, + { + "epoch": 0.04529, + "grad_norm": 0.5404518249664664, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 4529 + }, + { + "epoch": 0.0453, + "grad_norm": 0.5362375259990702, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 4530 + }, + { + "epoch": 0.04531, + "grad_norm": 0.48254266609849994, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 4531 + }, + { + "epoch": 0.04532, + "grad_norm": 0.5236204573745596, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 4532 + }, + { + "epoch": 0.04533, + "grad_norm": 0.5582813589006514, + "learning_rate": 0.003, + "loss": 4.038, + "step": 4533 + }, + { + "epoch": 0.04534, + "grad_norm": 0.5596445464924504, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 4534 + }, + { + "epoch": 0.04535, + "grad_norm": 0.5354061621780647, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 4535 + }, + { + "epoch": 0.04536, + "grad_norm": 0.5425691808827817, + "learning_rate": 0.003, + "loss": 4.098, + "step": 4536 + }, + { + "epoch": 0.04537, + "grad_norm": 0.4920865378030088, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 4537 + }, + { + "epoch": 0.04538, + "grad_norm": 0.5256556066617067, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 4538 + }, + { + "epoch": 0.04539, + "grad_norm": 0.6185623732368083, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 4539 + }, + { + "epoch": 0.0454, + "grad_norm": 0.6433734048865866, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 4540 + }, + { + "epoch": 0.04541, + "grad_norm": 0.5625812558274345, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 4541 + }, + { + "epoch": 0.04542, + "grad_norm": 0.49080855168600757, + "learning_rate": 0.003, + "loss": 4.046, + "step": 4542 + }, + { + "epoch": 0.04543, + "grad_norm": 0.5381094881804505, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 4543 + }, + { + "epoch": 0.04544, + "grad_norm": 0.6759795201085677, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 4544 + }, + { + "epoch": 0.04545, + "grad_norm": 0.7141477670946375, + "learning_rate": 0.003, + "loss": 4.072, + "step": 4545 + }, + { + "epoch": 0.04546, + "grad_norm": 0.7899638278039026, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 4546 + }, + { + "epoch": 0.04547, + "grad_norm": 0.7778397845017162, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 4547 + }, + { + "epoch": 0.04548, + "grad_norm": 0.7991440481906241, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 4548 + }, + { + "epoch": 0.04549, + "grad_norm": 1.0913356820076647, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 4549 + }, + { + "epoch": 0.0455, + "grad_norm": 1.1319774714523656, + "learning_rate": 0.003, + "loss": 4.1221, + "step": 4550 + }, + { + "epoch": 0.04551, + "grad_norm": 0.9496828973751253, + "learning_rate": 0.003, + "loss": 4.1266, + "step": 4551 + }, + { + "epoch": 0.04552, + "grad_norm": 1.0032051859286752, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 4552 + }, + { + "epoch": 0.04553, + "grad_norm": 1.1120806778897974, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 4553 + }, + { + "epoch": 0.04554, + "grad_norm": 0.905389833750841, + "learning_rate": 0.003, + "loss": 4.092, + "step": 4554 + }, + { + "epoch": 0.04555, + "grad_norm": 0.8511524961488746, + "learning_rate": 0.003, + "loss": 4.1039, + "step": 4555 + }, + { + "epoch": 0.04556, + "grad_norm": 0.8499588018380295, + "learning_rate": 0.003, + "loss": 4.1054, + "step": 4556 + }, + { + "epoch": 0.04557, + "grad_norm": 0.969563286629227, + "learning_rate": 0.003, + "loss": 4.0953, + "step": 4557 + }, + { + "epoch": 0.04558, + "grad_norm": 1.0902033670365172, + "learning_rate": 0.003, + "loss": 4.1325, + "step": 4558 + }, + { + "epoch": 0.04559, + "grad_norm": 0.9549853524488205, + "learning_rate": 0.003, + "loss": 4.1408, + "step": 4559 + }, + { + "epoch": 0.0456, + "grad_norm": 0.9788547919742809, + "learning_rate": 0.003, + "loss": 4.1056, + "step": 4560 + }, + { + "epoch": 0.04561, + "grad_norm": 0.9514604347411316, + "learning_rate": 0.003, + "loss": 4.1119, + "step": 4561 + }, + { + "epoch": 0.04562, + "grad_norm": 1.0558998804223803, + "learning_rate": 0.003, + "loss": 4.142, + "step": 4562 + }, + { + "epoch": 0.04563, + "grad_norm": 0.894116137751401, + "learning_rate": 0.003, + "loss": 4.1032, + "step": 4563 + }, + { + "epoch": 0.04564, + "grad_norm": 0.8816900421869549, + "learning_rate": 0.003, + "loss": 4.1166, + "step": 4564 + }, + { + "epoch": 0.04565, + "grad_norm": 0.7699557425253979, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 4565 + }, + { + "epoch": 0.04566, + "grad_norm": 0.7091996574219259, + "learning_rate": 0.003, + "loss": 4.1123, + "step": 4566 + }, + { + "epoch": 0.04567, + "grad_norm": 0.6855538197358724, + "learning_rate": 0.003, + "loss": 4.1044, + "step": 4567 + }, + { + "epoch": 0.04568, + "grad_norm": 0.5560572923195427, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 4568 + }, + { + "epoch": 0.04569, + "grad_norm": 0.5310557177545742, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 4569 + }, + { + "epoch": 0.0457, + "grad_norm": 0.5511692746217328, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 4570 + }, + { + "epoch": 0.04571, + "grad_norm": 0.6018749665211555, + "learning_rate": 0.003, + "loss": 4.1156, + "step": 4571 + }, + { + "epoch": 0.04572, + "grad_norm": 0.7062993745890849, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 4572 + }, + { + "epoch": 0.04573, + "grad_norm": 0.6841185494600575, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 4573 + }, + { + "epoch": 0.04574, + "grad_norm": 0.5900736396275527, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 4574 + }, + { + "epoch": 0.04575, + "grad_norm": 0.5966609390777746, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4575 + }, + { + "epoch": 0.04576, + "grad_norm": 0.6387567457517981, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 4576 + }, + { + "epoch": 0.04577, + "grad_norm": 0.7087784475989509, + "learning_rate": 0.003, + "loss": 4.1084, + "step": 4577 + }, + { + "epoch": 0.04578, + "grad_norm": 0.6620202662611342, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 4578 + }, + { + "epoch": 0.04579, + "grad_norm": 0.629781704692626, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 4579 + }, + { + "epoch": 0.0458, + "grad_norm": 0.6274692271464493, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 4580 + }, + { + "epoch": 0.04581, + "grad_norm": 0.6869236176540191, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 4581 + }, + { + "epoch": 0.04582, + "grad_norm": 0.5598552748391781, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 4582 + }, + { + "epoch": 0.04583, + "grad_norm": 0.5735492635938381, + "learning_rate": 0.003, + "loss": 4.066, + "step": 4583 + }, + { + "epoch": 0.04584, + "grad_norm": 0.5322893366852923, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 4584 + }, + { + "epoch": 0.04585, + "grad_norm": 0.5933233165846572, + "learning_rate": 0.003, + "loss": 4.073, + "step": 4585 + }, + { + "epoch": 0.04586, + "grad_norm": 0.5723035495377876, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 4586 + }, + { + "epoch": 0.04587, + "grad_norm": 0.6512287440343816, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 4587 + }, + { + "epoch": 0.04588, + "grad_norm": 0.7280553751460431, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 4588 + }, + { + "epoch": 0.04589, + "grad_norm": 0.7258649192140623, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 4589 + }, + { + "epoch": 0.0459, + "grad_norm": 0.7115444584091245, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 4590 + }, + { + "epoch": 0.04591, + "grad_norm": 0.6556402360362764, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 4591 + }, + { + "epoch": 0.04592, + "grad_norm": 0.5612345736544048, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4592 + }, + { + "epoch": 0.04593, + "grad_norm": 0.5354638657041939, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 4593 + }, + { + "epoch": 0.04594, + "grad_norm": 0.5624339566215373, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 4594 + }, + { + "epoch": 0.04595, + "grad_norm": 0.52752312577219, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 4595 + }, + { + "epoch": 0.04596, + "grad_norm": 0.549523518400531, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 4596 + }, + { + "epoch": 0.04597, + "grad_norm": 0.5222431017051227, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4597 + }, + { + "epoch": 0.04598, + "grad_norm": 0.5146021513155267, + "learning_rate": 0.003, + "loss": 4.1159, + "step": 4598 + }, + { + "epoch": 0.04599, + "grad_norm": 0.4976260057589565, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 4599 + }, + { + "epoch": 0.046, + "grad_norm": 0.5666334876209531, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 4600 + }, + { + "epoch": 0.04601, + "grad_norm": 0.7177430633120828, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 4601 + }, + { + "epoch": 0.04602, + "grad_norm": 0.7811922241381717, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 4602 + }, + { + "epoch": 0.04603, + "grad_norm": 0.8911743649722192, + "learning_rate": 0.003, + "loss": 4.099, + "step": 4603 + }, + { + "epoch": 0.04604, + "grad_norm": 0.951199705119624, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 4604 + }, + { + "epoch": 0.04605, + "grad_norm": 0.7949251996132306, + "learning_rate": 0.003, + "loss": 4.1076, + "step": 4605 + }, + { + "epoch": 0.04606, + "grad_norm": 0.7821472922166627, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 4606 + }, + { + "epoch": 0.04607, + "grad_norm": 0.8986596047017483, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 4607 + }, + { + "epoch": 0.04608, + "grad_norm": 0.9972683296437083, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 4608 + }, + { + "epoch": 0.04609, + "grad_norm": 1.052686581991779, + "learning_rate": 0.003, + "loss": 4.1022, + "step": 4609 + }, + { + "epoch": 0.0461, + "grad_norm": 0.8043063252352641, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 4610 + }, + { + "epoch": 0.04611, + "grad_norm": 0.7913355244086184, + "learning_rate": 0.003, + "loss": 4.1234, + "step": 4611 + }, + { + "epoch": 0.04612, + "grad_norm": 0.9807171390572609, + "learning_rate": 0.003, + "loss": 4.1109, + "step": 4612 + }, + { + "epoch": 0.04613, + "grad_norm": 0.8575138819535689, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 4613 + }, + { + "epoch": 0.04614, + "grad_norm": 0.7791457185921051, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 4614 + }, + { + "epoch": 0.04615, + "grad_norm": 0.7181975717279373, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 4615 + }, + { + "epoch": 0.04616, + "grad_norm": 0.7096674267744997, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 4616 + }, + { + "epoch": 0.04617, + "grad_norm": 0.702262285982413, + "learning_rate": 0.003, + "loss": 4.0999, + "step": 4617 + }, + { + "epoch": 0.04618, + "grad_norm": 0.8636976722343075, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4618 + }, + { + "epoch": 0.04619, + "grad_norm": 1.0352177824968434, + "learning_rate": 0.003, + "loss": 4.099, + "step": 4619 + }, + { + "epoch": 0.0462, + "grad_norm": 0.8255114307922582, + "learning_rate": 0.003, + "loss": 4.121, + "step": 4620 + }, + { + "epoch": 0.04621, + "grad_norm": 0.7174190350053175, + "learning_rate": 0.003, + "loss": 4.1279, + "step": 4621 + }, + { + "epoch": 0.04622, + "grad_norm": 0.614242154706962, + "learning_rate": 0.003, + "loss": 4.1162, + "step": 4622 + }, + { + "epoch": 0.04623, + "grad_norm": 0.6052758988642168, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 4623 + }, + { + "epoch": 0.04624, + "grad_norm": 0.6268636931811352, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 4624 + }, + { + "epoch": 0.04625, + "grad_norm": 0.5535121216375328, + "learning_rate": 0.003, + "loss": 4.0951, + "step": 4625 + }, + { + "epoch": 0.04626, + "grad_norm": 0.5582239298694301, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 4626 + }, + { + "epoch": 0.04627, + "grad_norm": 0.5373148751695499, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 4627 + }, + { + "epoch": 0.04628, + "grad_norm": 0.46029006396137395, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 4628 + }, + { + "epoch": 0.04629, + "grad_norm": 0.46396385763405157, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 4629 + }, + { + "epoch": 0.0463, + "grad_norm": 0.4729507927624471, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 4630 + }, + { + "epoch": 0.04631, + "grad_norm": 0.5195515490649404, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 4631 + }, + { + "epoch": 0.04632, + "grad_norm": 0.5613098143428055, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 4632 + }, + { + "epoch": 0.04633, + "grad_norm": 0.6423345449287932, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 4633 + }, + { + "epoch": 0.04634, + "grad_norm": 0.8049813302073491, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 4634 + }, + { + "epoch": 0.04635, + "grad_norm": 0.973922696526027, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 4635 + }, + { + "epoch": 0.04636, + "grad_norm": 0.9453542204957492, + "learning_rate": 0.003, + "loss": 4.1252, + "step": 4636 + }, + { + "epoch": 0.04637, + "grad_norm": 0.8241199342985062, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 4637 + }, + { + "epoch": 0.04638, + "grad_norm": 0.7662258094904854, + "learning_rate": 0.003, + "loss": 4.1026, + "step": 4638 + }, + { + "epoch": 0.04639, + "grad_norm": 0.7397865786197445, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 4639 + }, + { + "epoch": 0.0464, + "grad_norm": 0.7035179128900486, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 4640 + }, + { + "epoch": 0.04641, + "grad_norm": 0.6123035611649836, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 4641 + }, + { + "epoch": 0.04642, + "grad_norm": 0.5160562983677347, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 4642 + }, + { + "epoch": 0.04643, + "grad_norm": 0.6136271731714404, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 4643 + }, + { + "epoch": 0.04644, + "grad_norm": 0.6118074576936773, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 4644 + }, + { + "epoch": 0.04645, + "grad_norm": 0.5921257686332748, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 4645 + }, + { + "epoch": 0.04646, + "grad_norm": 0.5196332091060915, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 4646 + }, + { + "epoch": 0.04647, + "grad_norm": 0.5442771205727563, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 4647 + }, + { + "epoch": 0.04648, + "grad_norm": 0.6327039568983159, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 4648 + }, + { + "epoch": 0.04649, + "grad_norm": 0.6869677429205404, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 4649 + }, + { + "epoch": 0.0465, + "grad_norm": 0.6191365834848964, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 4650 + }, + { + "epoch": 0.04651, + "grad_norm": 0.5660445918159056, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 4651 + }, + { + "epoch": 0.04652, + "grad_norm": 0.5996569428075281, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4652 + }, + { + "epoch": 0.04653, + "grad_norm": 0.5846108523338363, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 4653 + }, + { + "epoch": 0.04654, + "grad_norm": 0.5881011591995303, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 4654 + }, + { + "epoch": 0.04655, + "grad_norm": 0.6380072308464515, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 4655 + }, + { + "epoch": 0.04656, + "grad_norm": 0.6322736074156682, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 4656 + }, + { + "epoch": 0.04657, + "grad_norm": 0.5983615772554817, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 4657 + }, + { + "epoch": 0.04658, + "grad_norm": 0.7006643914315651, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 4658 + }, + { + "epoch": 0.04659, + "grad_norm": 0.7861358324202233, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 4659 + }, + { + "epoch": 0.0466, + "grad_norm": 0.7825397080460539, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 4660 + }, + { + "epoch": 0.04661, + "grad_norm": 0.8464193175728497, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 4661 + }, + { + "epoch": 0.04662, + "grad_norm": 0.8325690347227048, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 4662 + }, + { + "epoch": 0.04663, + "grad_norm": 0.9407032046757063, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4663 + }, + { + "epoch": 0.04664, + "grad_norm": 0.937459011893773, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 4664 + }, + { + "epoch": 0.04665, + "grad_norm": 0.8244315180896339, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 4665 + }, + { + "epoch": 0.04666, + "grad_norm": 0.7858469834758598, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 4666 + }, + { + "epoch": 0.04667, + "grad_norm": 0.80572427932393, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 4667 + }, + { + "epoch": 0.04668, + "grad_norm": 0.7502328470578125, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 4668 + }, + { + "epoch": 0.04669, + "grad_norm": 0.7321524361227727, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 4669 + }, + { + "epoch": 0.0467, + "grad_norm": 0.7332466267978432, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 4670 + }, + { + "epoch": 0.04671, + "grad_norm": 0.7107246142340815, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4671 + }, + { + "epoch": 0.04672, + "grad_norm": 0.7725237840828216, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 4672 + }, + { + "epoch": 0.04673, + "grad_norm": 0.7839185539405206, + "learning_rate": 0.003, + "loss": 4.0992, + "step": 4673 + }, + { + "epoch": 0.04674, + "grad_norm": 0.7746228494359627, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 4674 + }, + { + "epoch": 0.04675, + "grad_norm": 0.6248678993478205, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 4675 + }, + { + "epoch": 0.04676, + "grad_norm": 0.5769327359060665, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 4676 + }, + { + "epoch": 0.04677, + "grad_norm": 0.6227164444149332, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 4677 + }, + { + "epoch": 0.04678, + "grad_norm": 0.6117595646673357, + "learning_rate": 0.003, + "loss": 4.052, + "step": 4678 + }, + { + "epoch": 0.04679, + "grad_norm": 0.6305442209664343, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 4679 + }, + { + "epoch": 0.0468, + "grad_norm": 0.6783543628620735, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 4680 + }, + { + "epoch": 0.04681, + "grad_norm": 0.7268755013369022, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 4681 + }, + { + "epoch": 0.04682, + "grad_norm": 0.7281096983360308, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 4682 + }, + { + "epoch": 0.04683, + "grad_norm": 0.64781736754668, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 4683 + }, + { + "epoch": 0.04684, + "grad_norm": 0.667007415073709, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 4684 + }, + { + "epoch": 0.04685, + "grad_norm": 0.6540380773935125, + "learning_rate": 0.003, + "loss": 4.069, + "step": 4685 + }, + { + "epoch": 0.04686, + "grad_norm": 0.6048593188027076, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 4686 + }, + { + "epoch": 0.04687, + "grad_norm": 0.5418745028430608, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 4687 + }, + { + "epoch": 0.04688, + "grad_norm": 0.5251559039971293, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 4688 + }, + { + "epoch": 0.04689, + "grad_norm": 0.5481659851408645, + "learning_rate": 0.003, + "loss": 4.1002, + "step": 4689 + }, + { + "epoch": 0.0469, + "grad_norm": 0.6225054517005262, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 4690 + }, + { + "epoch": 0.04691, + "grad_norm": 0.6852701986600458, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 4691 + }, + { + "epoch": 0.04692, + "grad_norm": 0.758942515968189, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 4692 + }, + { + "epoch": 0.04693, + "grad_norm": 0.7889465888239833, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 4693 + }, + { + "epoch": 0.04694, + "grad_norm": 0.7848025164200313, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 4694 + }, + { + "epoch": 0.04695, + "grad_norm": 0.8103792418108647, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 4695 + }, + { + "epoch": 0.04696, + "grad_norm": 0.7863647496370023, + "learning_rate": 0.003, + "loss": 4.103, + "step": 4696 + }, + { + "epoch": 0.04697, + "grad_norm": 0.8410435008204653, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 4697 + }, + { + "epoch": 0.04698, + "grad_norm": 0.9527844186132859, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 4698 + }, + { + "epoch": 0.04699, + "grad_norm": 0.8935725319982992, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 4699 + }, + { + "epoch": 0.047, + "grad_norm": 0.9075175240690628, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 4700 + }, + { + "epoch": 0.04701, + "grad_norm": 0.9008982818686766, + "learning_rate": 0.003, + "loss": 4.1089, + "step": 4701 + }, + { + "epoch": 0.04702, + "grad_norm": 0.8350305129379764, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4702 + }, + { + "epoch": 0.04703, + "grad_norm": 0.9510111222641331, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 4703 + }, + { + "epoch": 0.04704, + "grad_norm": 0.8002841934004945, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 4704 + }, + { + "epoch": 0.04705, + "grad_norm": 0.7531244749957098, + "learning_rate": 0.003, + "loss": 4.114, + "step": 4705 + }, + { + "epoch": 0.04706, + "grad_norm": 0.764225953698126, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 4706 + }, + { + "epoch": 0.04707, + "grad_norm": 0.8125082520544259, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 4707 + }, + { + "epoch": 0.04708, + "grad_norm": 1.0517553975942069, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 4708 + }, + { + "epoch": 0.04709, + "grad_norm": 1.078343854063864, + "learning_rate": 0.003, + "loss": 4.106, + "step": 4709 + }, + { + "epoch": 0.0471, + "grad_norm": 0.8510921171403981, + "learning_rate": 0.003, + "loss": 4.1088, + "step": 4710 + }, + { + "epoch": 0.04711, + "grad_norm": 0.8310710354665064, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 4711 + }, + { + "epoch": 0.04712, + "grad_norm": 0.787793071924109, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 4712 + }, + { + "epoch": 0.04713, + "grad_norm": 0.6988763628466411, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 4713 + }, + { + "epoch": 0.04714, + "grad_norm": 0.651508372158522, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 4714 + }, + { + "epoch": 0.04715, + "grad_norm": 0.6358742698916999, + "learning_rate": 0.003, + "loss": 4.116, + "step": 4715 + }, + { + "epoch": 0.04716, + "grad_norm": 0.5849595703268815, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 4716 + }, + { + "epoch": 0.04717, + "grad_norm": 0.609882483799475, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 4717 + }, + { + "epoch": 0.04718, + "grad_norm": 0.602939696063538, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 4718 + }, + { + "epoch": 0.04719, + "grad_norm": 0.6576137481528811, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 4719 + }, + { + "epoch": 0.0472, + "grad_norm": 0.7283425609478092, + "learning_rate": 0.003, + "loss": 4.101, + "step": 4720 + }, + { + "epoch": 0.04721, + "grad_norm": 0.7773560622821873, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 4721 + }, + { + "epoch": 0.04722, + "grad_norm": 0.7789243592151769, + "learning_rate": 0.003, + "loss": 4.0831, + "step": 4722 + }, + { + "epoch": 0.04723, + "grad_norm": 0.6331731136829702, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 4723 + }, + { + "epoch": 0.04724, + "grad_norm": 0.5729751941329306, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 4724 + }, + { + "epoch": 0.04725, + "grad_norm": 0.6245176008740605, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 4725 + }, + { + "epoch": 0.04726, + "grad_norm": 0.5977538687538189, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 4726 + }, + { + "epoch": 0.04727, + "grad_norm": 0.6266324932805432, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 4727 + }, + { + "epoch": 0.04728, + "grad_norm": 0.6298684903126542, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 4728 + }, + { + "epoch": 0.04729, + "grad_norm": 0.6729448329354482, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 4729 + }, + { + "epoch": 0.0473, + "grad_norm": 0.628557701664857, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 4730 + }, + { + "epoch": 0.04731, + "grad_norm": 0.503933221286775, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 4731 + }, + { + "epoch": 0.04732, + "grad_norm": 0.552678991959826, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 4732 + }, + { + "epoch": 0.04733, + "grad_norm": 0.5164892173442864, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 4733 + }, + { + "epoch": 0.04734, + "grad_norm": 0.49950807533825825, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 4734 + }, + { + "epoch": 0.04735, + "grad_norm": 0.6170223179972694, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 4735 + }, + { + "epoch": 0.04736, + "grad_norm": 0.7082284220255964, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 4736 + }, + { + "epoch": 0.04737, + "grad_norm": 0.6697401947945215, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 4737 + }, + { + "epoch": 0.04738, + "grad_norm": 0.5572007742229066, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 4738 + }, + { + "epoch": 0.04739, + "grad_norm": 0.5406366290909363, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 4739 + }, + { + "epoch": 0.0474, + "grad_norm": 0.5582441573554152, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 4740 + }, + { + "epoch": 0.04741, + "grad_norm": 0.6097857648233336, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 4741 + }, + { + "epoch": 0.04742, + "grad_norm": 0.6932907716378426, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 4742 + }, + { + "epoch": 0.04743, + "grad_norm": 0.6900820991578511, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 4743 + }, + { + "epoch": 0.04744, + "grad_norm": 0.6487750942642598, + "learning_rate": 0.003, + "loss": 4.051, + "step": 4744 + }, + { + "epoch": 0.04745, + "grad_norm": 0.6374762412326888, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 4745 + }, + { + "epoch": 0.04746, + "grad_norm": 0.7133304491599274, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 4746 + }, + { + "epoch": 0.04747, + "grad_norm": 0.6702111945789702, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 4747 + }, + { + "epoch": 0.04748, + "grad_norm": 0.6527174239132283, + "learning_rate": 0.003, + "loss": 4.094, + "step": 4748 + }, + { + "epoch": 0.04749, + "grad_norm": 0.6637409448547621, + "learning_rate": 0.003, + "loss": 4.1004, + "step": 4749 + }, + { + "epoch": 0.0475, + "grad_norm": 0.6630758626969115, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 4750 + }, + { + "epoch": 0.04751, + "grad_norm": 0.7676694855222234, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 4751 + }, + { + "epoch": 0.04752, + "grad_norm": 0.9399780798007432, + "learning_rate": 0.003, + "loss": 4.074, + "step": 4752 + }, + { + "epoch": 0.04753, + "grad_norm": 0.9278552762695305, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 4753 + }, + { + "epoch": 0.04754, + "grad_norm": 0.7817053687343111, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 4754 + }, + { + "epoch": 0.04755, + "grad_norm": 0.7528396545947692, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 4755 + }, + { + "epoch": 0.04756, + "grad_norm": 0.8077765656776754, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 4756 + }, + { + "epoch": 0.04757, + "grad_norm": 0.8414708623531142, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 4757 + }, + { + "epoch": 0.04758, + "grad_norm": 0.8609047083251963, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 4758 + }, + { + "epoch": 0.04759, + "grad_norm": 0.8737600429256249, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 4759 + }, + { + "epoch": 0.0476, + "grad_norm": 0.9420334591519173, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 4760 + }, + { + "epoch": 0.04761, + "grad_norm": 0.847310934267217, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 4761 + }, + { + "epoch": 0.04762, + "grad_norm": 0.7040568516928781, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 4762 + }, + { + "epoch": 0.04763, + "grad_norm": 0.7677682282260441, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 4763 + }, + { + "epoch": 0.04764, + "grad_norm": 0.8514571396663456, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 4764 + }, + { + "epoch": 0.04765, + "grad_norm": 0.9430633690965472, + "learning_rate": 0.003, + "loss": 4.1244, + "step": 4765 + }, + { + "epoch": 0.04766, + "grad_norm": 0.9142673573403883, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 4766 + }, + { + "epoch": 0.04767, + "grad_norm": 0.9929392374160075, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 4767 + }, + { + "epoch": 0.04768, + "grad_norm": 0.8926210753433251, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 4768 + }, + { + "epoch": 0.04769, + "grad_norm": 0.7918860687074516, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 4769 + }, + { + "epoch": 0.0477, + "grad_norm": 0.7522057415031308, + "learning_rate": 0.003, + "loss": 4.07, + "step": 4770 + }, + { + "epoch": 0.04771, + "grad_norm": 0.7575487511094675, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 4771 + }, + { + "epoch": 0.04772, + "grad_norm": 0.6953778244451587, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 4772 + }, + { + "epoch": 0.04773, + "grad_norm": 0.6020436962946181, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 4773 + }, + { + "epoch": 0.04774, + "grad_norm": 0.6277027664826653, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 4774 + }, + { + "epoch": 0.04775, + "grad_norm": 0.5654389301072226, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 4775 + }, + { + "epoch": 0.04776, + "grad_norm": 0.561250147903765, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 4776 + }, + { + "epoch": 0.04777, + "grad_norm": 0.5183518853348038, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 4777 + }, + { + "epoch": 0.04778, + "grad_norm": 0.48846945945640075, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 4778 + }, + { + "epoch": 0.04779, + "grad_norm": 0.4820376636521312, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 4779 + }, + { + "epoch": 0.0478, + "grad_norm": 0.5081186588721828, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 4780 + }, + { + "epoch": 0.04781, + "grad_norm": 0.4901168542228479, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 4781 + }, + { + "epoch": 0.04782, + "grad_norm": 0.4028559954391356, + "learning_rate": 0.003, + "loss": 4.059, + "step": 4782 + }, + { + "epoch": 0.04783, + "grad_norm": 0.427914733383392, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 4783 + }, + { + "epoch": 0.04784, + "grad_norm": 0.454893996223861, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 4784 + }, + { + "epoch": 0.04785, + "grad_norm": 0.42517114197119327, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 4785 + }, + { + "epoch": 0.04786, + "grad_norm": 0.46753610520661504, + "learning_rate": 0.003, + "loss": 4.08, + "step": 4786 + }, + { + "epoch": 0.04787, + "grad_norm": 0.536917030148163, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 4787 + }, + { + "epoch": 0.04788, + "grad_norm": 0.6268104837137449, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 4788 + }, + { + "epoch": 0.04789, + "grad_norm": 0.6606781165952248, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 4789 + }, + { + "epoch": 0.0479, + "grad_norm": 0.6313438394739813, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 4790 + }, + { + "epoch": 0.04791, + "grad_norm": 0.7058289905675165, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 4791 + }, + { + "epoch": 0.04792, + "grad_norm": 0.7800415961656294, + "learning_rate": 0.003, + "loss": 4.0973, + "step": 4792 + }, + { + "epoch": 0.04793, + "grad_norm": 0.8658075410157954, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 4793 + }, + { + "epoch": 0.04794, + "grad_norm": 1.0196021945249285, + "learning_rate": 0.003, + "loss": 4.09, + "step": 4794 + }, + { + "epoch": 0.04795, + "grad_norm": 0.9247650562439667, + "learning_rate": 0.003, + "loss": 4.0892, + "step": 4795 + }, + { + "epoch": 0.04796, + "grad_norm": 0.871523325562872, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 4796 + }, + { + "epoch": 0.04797, + "grad_norm": 0.8692054801458177, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 4797 + }, + { + "epoch": 0.04798, + "grad_norm": 0.8030146926406986, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 4798 + }, + { + "epoch": 0.04799, + "grad_norm": 0.7375164748158682, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 4799 + }, + { + "epoch": 0.048, + "grad_norm": 0.6904641331606276, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 4800 + }, + { + "epoch": 0.04801, + "grad_norm": 0.720228478512294, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 4801 + }, + { + "epoch": 0.04802, + "grad_norm": 0.7770414242765028, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 4802 + }, + { + "epoch": 0.04803, + "grad_norm": 0.7480766180413105, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 4803 + }, + { + "epoch": 0.04804, + "grad_norm": 0.7939576174808407, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 4804 + }, + { + "epoch": 0.04805, + "grad_norm": 0.8151935602009253, + "learning_rate": 0.003, + "loss": 4.1007, + "step": 4805 + }, + { + "epoch": 0.04806, + "grad_norm": 0.7843083662371639, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 4806 + }, + { + "epoch": 0.04807, + "grad_norm": 0.7004959676879019, + "learning_rate": 0.003, + "loss": 4.0961, + "step": 4807 + }, + { + "epoch": 0.04808, + "grad_norm": 0.7789585489946028, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 4808 + }, + { + "epoch": 0.04809, + "grad_norm": 0.8254273800966622, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 4809 + }, + { + "epoch": 0.0481, + "grad_norm": 0.7875987879838314, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 4810 + }, + { + "epoch": 0.04811, + "grad_norm": 0.8179750856804625, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 4811 + }, + { + "epoch": 0.04812, + "grad_norm": 0.7814142074292407, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 4812 + }, + { + "epoch": 0.04813, + "grad_norm": 0.6979738713381015, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 4813 + }, + { + "epoch": 0.04814, + "grad_norm": 0.752761940665461, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 4814 + }, + { + "epoch": 0.04815, + "grad_norm": 0.7048630887240973, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 4815 + }, + { + "epoch": 0.04816, + "grad_norm": 0.7692961964666495, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 4816 + }, + { + "epoch": 0.04817, + "grad_norm": 0.7299206003044024, + "learning_rate": 0.003, + "loss": 4.061, + "step": 4817 + }, + { + "epoch": 0.04818, + "grad_norm": 0.6284864897372815, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 4818 + }, + { + "epoch": 0.04819, + "grad_norm": 0.5240360643822964, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 4819 + }, + { + "epoch": 0.0482, + "grad_norm": 0.5422219452698507, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 4820 + }, + { + "epoch": 0.04821, + "grad_norm": 0.5656830695903975, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 4821 + }, + { + "epoch": 0.04822, + "grad_norm": 0.6905601681230291, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 4822 + }, + { + "epoch": 0.04823, + "grad_norm": 0.7519015397996957, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 4823 + }, + { + "epoch": 0.04824, + "grad_norm": 0.7745737056459601, + "learning_rate": 0.003, + "loss": 4.058, + "step": 4824 + }, + { + "epoch": 0.04825, + "grad_norm": 0.8179310141911902, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 4825 + }, + { + "epoch": 0.04826, + "grad_norm": 0.7774672220264023, + "learning_rate": 0.003, + "loss": 4.075, + "step": 4826 + }, + { + "epoch": 0.04827, + "grad_norm": 0.8196097228698328, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 4827 + }, + { + "epoch": 0.04828, + "grad_norm": 0.8618639178438036, + "learning_rate": 0.003, + "loss": 4.095, + "step": 4828 + }, + { + "epoch": 0.04829, + "grad_norm": 0.7492214001729087, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 4829 + }, + { + "epoch": 0.0483, + "grad_norm": 0.7125320895277254, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 4830 + }, + { + "epoch": 0.04831, + "grad_norm": 0.6131966701173335, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 4831 + }, + { + "epoch": 0.04832, + "grad_norm": 0.5344875994225183, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 4832 + }, + { + "epoch": 0.04833, + "grad_norm": 0.6705953552259101, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 4833 + }, + { + "epoch": 0.04834, + "grad_norm": 0.6406276538582305, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 4834 + }, + { + "epoch": 0.04835, + "grad_norm": 0.6438689776896058, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 4835 + }, + { + "epoch": 0.04836, + "grad_norm": 0.7553718003571447, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 4836 + }, + { + "epoch": 0.04837, + "grad_norm": 0.8338050175436581, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 4837 + }, + { + "epoch": 0.04838, + "grad_norm": 0.8362296240535967, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 4838 + }, + { + "epoch": 0.04839, + "grad_norm": 0.774177990513827, + "learning_rate": 0.003, + "loss": 4.079, + "step": 4839 + }, + { + "epoch": 0.0484, + "grad_norm": 0.6411460101426353, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 4840 + }, + { + "epoch": 0.04841, + "grad_norm": 0.7421428678439225, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 4841 + }, + { + "epoch": 0.04842, + "grad_norm": 0.758219379419753, + "learning_rate": 0.003, + "loss": 4.0963, + "step": 4842 + }, + { + "epoch": 0.04843, + "grad_norm": 0.7091412140912875, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 4843 + }, + { + "epoch": 0.04844, + "grad_norm": 0.6570042970392534, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 4844 + }, + { + "epoch": 0.04845, + "grad_norm": 0.677637297426444, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 4845 + }, + { + "epoch": 0.04846, + "grad_norm": 0.7380695891193019, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 4846 + }, + { + "epoch": 0.04847, + "grad_norm": 0.7869818825786099, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 4847 + }, + { + "epoch": 0.04848, + "grad_norm": 0.9431123819987169, + "learning_rate": 0.003, + "loss": 4.108, + "step": 4848 + }, + { + "epoch": 0.04849, + "grad_norm": 0.9007548797525372, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 4849 + }, + { + "epoch": 0.0485, + "grad_norm": 0.8145135470840875, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 4850 + }, + { + "epoch": 0.04851, + "grad_norm": 0.8487610103607066, + "learning_rate": 0.003, + "loss": 4.1019, + "step": 4851 + }, + { + "epoch": 0.04852, + "grad_norm": 0.7427431641360362, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 4852 + }, + { + "epoch": 0.04853, + "grad_norm": 0.7183108899904691, + "learning_rate": 0.003, + "loss": 4.091, + "step": 4853 + }, + { + "epoch": 0.04854, + "grad_norm": 0.6616674896124763, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 4854 + }, + { + "epoch": 0.04855, + "grad_norm": 0.7134767687116306, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 4855 + }, + { + "epoch": 0.04856, + "grad_norm": 0.7463454260793, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 4856 + }, + { + "epoch": 0.04857, + "grad_norm": 0.787626086040093, + "learning_rate": 0.003, + "loss": 4.11, + "step": 4857 + }, + { + "epoch": 0.04858, + "grad_norm": 0.8361758641564052, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 4858 + }, + { + "epoch": 0.04859, + "grad_norm": 0.7829099377406131, + "learning_rate": 0.003, + "loss": 4.1129, + "step": 4859 + }, + { + "epoch": 0.0486, + "grad_norm": 0.7028981787907939, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 4860 + }, + { + "epoch": 0.04861, + "grad_norm": 0.6568846803120677, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 4861 + }, + { + "epoch": 0.04862, + "grad_norm": 0.6730755857369936, + "learning_rate": 0.003, + "loss": 4.084, + "step": 4862 + }, + { + "epoch": 0.04863, + "grad_norm": 0.6765221351913335, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 4863 + }, + { + "epoch": 0.04864, + "grad_norm": 0.6326322762494955, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 4864 + }, + { + "epoch": 0.04865, + "grad_norm": 0.6202935067369691, + "learning_rate": 0.003, + "loss": 4.0947, + "step": 4865 + }, + { + "epoch": 0.04866, + "grad_norm": 0.573447528692605, + "learning_rate": 0.003, + "loss": 4.0979, + "step": 4866 + }, + { + "epoch": 0.04867, + "grad_norm": 0.47159322555928723, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 4867 + }, + { + "epoch": 0.04868, + "grad_norm": 0.45789997627017254, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 4868 + }, + { + "epoch": 0.04869, + "grad_norm": 0.4284903785966079, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 4869 + }, + { + "epoch": 0.0487, + "grad_norm": 0.46635070677658913, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 4870 + }, + { + "epoch": 0.04871, + "grad_norm": 0.46878529502152205, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 4871 + }, + { + "epoch": 0.04872, + "grad_norm": 0.500754565392243, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 4872 + }, + { + "epoch": 0.04873, + "grad_norm": 0.6246387346001312, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 4873 + }, + { + "epoch": 0.04874, + "grad_norm": 0.7487902287154543, + "learning_rate": 0.003, + "loss": 4.073, + "step": 4874 + }, + { + "epoch": 0.04875, + "grad_norm": 0.795439812955094, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 4875 + }, + { + "epoch": 0.04876, + "grad_norm": 0.800617194154536, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 4876 + }, + { + "epoch": 0.04877, + "grad_norm": 0.7286926487336958, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 4877 + }, + { + "epoch": 0.04878, + "grad_norm": 0.7151020185340132, + "learning_rate": 0.003, + "loss": 4.056, + "step": 4878 + }, + { + "epoch": 0.04879, + "grad_norm": 0.7381209275996884, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 4879 + }, + { + "epoch": 0.0488, + "grad_norm": 0.8303375598896894, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 4880 + }, + { + "epoch": 0.04881, + "grad_norm": 0.9210744745554789, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 4881 + }, + { + "epoch": 0.04882, + "grad_norm": 0.8744252489749892, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 4882 + }, + { + "epoch": 0.04883, + "grad_norm": 0.8475189206698116, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 4883 + }, + { + "epoch": 0.04884, + "grad_norm": 0.914170806905025, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 4884 + }, + { + "epoch": 0.04885, + "grad_norm": 1.0532259854899952, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 4885 + }, + { + "epoch": 0.04886, + "grad_norm": 0.9647215312444578, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 4886 + }, + { + "epoch": 0.04887, + "grad_norm": 0.9766180530251676, + "learning_rate": 0.003, + "loss": 4.1048, + "step": 4887 + }, + { + "epoch": 0.04888, + "grad_norm": 0.977800916003621, + "learning_rate": 0.003, + "loss": 4.1106, + "step": 4888 + }, + { + "epoch": 0.04889, + "grad_norm": 1.0184146047042075, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 4889 + }, + { + "epoch": 0.0489, + "grad_norm": 0.9553804612595735, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 4890 + }, + { + "epoch": 0.04891, + "grad_norm": 0.7999103384592228, + "learning_rate": 0.003, + "loss": 4.102, + "step": 4891 + }, + { + "epoch": 0.04892, + "grad_norm": 0.7086626336274199, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 4892 + }, + { + "epoch": 0.04893, + "grad_norm": 0.6792888485428937, + "learning_rate": 0.003, + "loss": 4.0974, + "step": 4893 + }, + { + "epoch": 0.04894, + "grad_norm": 0.6175574098526109, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 4894 + }, + { + "epoch": 0.04895, + "grad_norm": 0.6436333244999423, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 4895 + }, + { + "epoch": 0.04896, + "grad_norm": 0.6374075092040796, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 4896 + }, + { + "epoch": 0.04897, + "grad_norm": 0.6074912195718345, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 4897 + }, + { + "epoch": 0.04898, + "grad_norm": 0.5659377033170091, + "learning_rate": 0.003, + "loss": 4.084, + "step": 4898 + }, + { + "epoch": 0.04899, + "grad_norm": 0.5904232518144595, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 4899 + }, + { + "epoch": 0.049, + "grad_norm": 0.5774204549092443, + "learning_rate": 0.003, + "loss": 4.099, + "step": 4900 + }, + { + "epoch": 0.04901, + "grad_norm": 0.5798525982998979, + "learning_rate": 0.003, + "loss": 4.058, + "step": 4901 + }, + { + "epoch": 0.04902, + "grad_norm": 0.5402988556036922, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 4902 + }, + { + "epoch": 0.04903, + "grad_norm": 0.5054483970189053, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 4903 + }, + { + "epoch": 0.04904, + "grad_norm": 0.4528668181989311, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 4904 + }, + { + "epoch": 0.04905, + "grad_norm": 0.48306762847507456, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 4905 + }, + { + "epoch": 0.04906, + "grad_norm": 0.4978227629505785, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 4906 + }, + { + "epoch": 0.04907, + "grad_norm": 0.5121849477830709, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 4907 + }, + { + "epoch": 0.04908, + "grad_norm": 0.5254098562833066, + "learning_rate": 0.003, + "loss": 4.063, + "step": 4908 + }, + { + "epoch": 0.04909, + "grad_norm": 0.5887025515549111, + "learning_rate": 0.003, + "loss": 4.068, + "step": 4909 + }, + { + "epoch": 0.0491, + "grad_norm": 0.6006000018807112, + "learning_rate": 0.003, + "loss": 4.061, + "step": 4910 + }, + { + "epoch": 0.04911, + "grad_norm": 0.5889656932504279, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 4911 + }, + { + "epoch": 0.04912, + "grad_norm": 0.6026645153725222, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 4912 + }, + { + "epoch": 0.04913, + "grad_norm": 0.6545438658528691, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 4913 + }, + { + "epoch": 0.04914, + "grad_norm": 0.8030874653933107, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 4914 + }, + { + "epoch": 0.04915, + "grad_norm": 0.8796534057435395, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 4915 + }, + { + "epoch": 0.04916, + "grad_norm": 0.7096794087992727, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 4916 + }, + { + "epoch": 0.04917, + "grad_norm": 0.6303871145752368, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 4917 + }, + { + "epoch": 0.04918, + "grad_norm": 0.6040936086406694, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 4918 + }, + { + "epoch": 0.04919, + "grad_norm": 0.5530193970084879, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 4919 + }, + { + "epoch": 0.0492, + "grad_norm": 0.5835626034981001, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 4920 + }, + { + "epoch": 0.04921, + "grad_norm": 0.6469314675335991, + "learning_rate": 0.003, + "loss": 4.088, + "step": 4921 + }, + { + "epoch": 0.04922, + "grad_norm": 0.6650667617818437, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 4922 + }, + { + "epoch": 0.04923, + "grad_norm": 0.7100934942428434, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 4923 + }, + { + "epoch": 0.04924, + "grad_norm": 0.724555712814058, + "learning_rate": 0.003, + "loss": 4.097, + "step": 4924 + }, + { + "epoch": 0.04925, + "grad_norm": 0.7817157219374906, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 4925 + }, + { + "epoch": 0.04926, + "grad_norm": 0.8619856198385645, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 4926 + }, + { + "epoch": 0.04927, + "grad_norm": 0.9248223896031712, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 4927 + }, + { + "epoch": 0.04928, + "grad_norm": 0.8402850305657008, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 4928 + }, + { + "epoch": 0.04929, + "grad_norm": 0.8421291747291888, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 4929 + }, + { + "epoch": 0.0493, + "grad_norm": 0.8175645223456335, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 4930 + }, + { + "epoch": 0.04931, + "grad_norm": 0.82707135913752, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 4931 + }, + { + "epoch": 0.04932, + "grad_norm": 0.8604370669986373, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 4932 + }, + { + "epoch": 0.04933, + "grad_norm": 0.8109880401088999, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 4933 + }, + { + "epoch": 0.04934, + "grad_norm": 0.7793957169990202, + "learning_rate": 0.003, + "loss": 4.111, + "step": 4934 + }, + { + "epoch": 0.04935, + "grad_norm": 0.7966564263192238, + "learning_rate": 0.003, + "loss": 4.1127, + "step": 4935 + }, + { + "epoch": 0.04936, + "grad_norm": 0.7629477988146514, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 4936 + }, + { + "epoch": 0.04937, + "grad_norm": 0.7753291682632341, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 4937 + }, + { + "epoch": 0.04938, + "grad_norm": 0.9178662582376598, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 4938 + }, + { + "epoch": 0.04939, + "grad_norm": 0.9352420318794955, + "learning_rate": 0.003, + "loss": 4.1086, + "step": 4939 + }, + { + "epoch": 0.0494, + "grad_norm": 0.7741772677131917, + "learning_rate": 0.003, + "loss": 4.1078, + "step": 4940 + }, + { + "epoch": 0.04941, + "grad_norm": 0.8475200734851549, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 4941 + }, + { + "epoch": 0.04942, + "grad_norm": 0.9584286460124734, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 4942 + }, + { + "epoch": 0.04943, + "grad_norm": 0.9315118608445326, + "learning_rate": 0.003, + "loss": 4.1185, + "step": 4943 + }, + { + "epoch": 0.04944, + "grad_norm": 0.8551189672891217, + "learning_rate": 0.003, + "loss": 4.1194, + "step": 4944 + }, + { + "epoch": 0.04945, + "grad_norm": 0.7174703522453804, + "learning_rate": 0.003, + "loss": 4.1029, + "step": 4945 + }, + { + "epoch": 0.04946, + "grad_norm": 0.7620497639744843, + "learning_rate": 0.003, + "loss": 4.1316, + "step": 4946 + }, + { + "epoch": 0.04947, + "grad_norm": 0.7378804860775325, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 4947 + }, + { + "epoch": 0.04948, + "grad_norm": 0.7794629039718619, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 4948 + }, + { + "epoch": 0.04949, + "grad_norm": 0.7985533666777282, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 4949 + }, + { + "epoch": 0.0495, + "grad_norm": 0.6455659585530127, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 4950 + }, + { + "epoch": 0.04951, + "grad_norm": 0.602022307053787, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 4951 + }, + { + "epoch": 0.04952, + "grad_norm": 0.6190732092088534, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 4952 + }, + { + "epoch": 0.04953, + "grad_norm": 0.578396804313404, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 4953 + }, + { + "epoch": 0.04954, + "grad_norm": 0.4942819343500649, + "learning_rate": 0.003, + "loss": 4.1, + "step": 4954 + }, + { + "epoch": 0.04955, + "grad_norm": 0.5628681778610025, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 4955 + }, + { + "epoch": 0.04956, + "grad_norm": 0.6396232968580461, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 4956 + }, + { + "epoch": 0.04957, + "grad_norm": 0.6006935108163937, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 4957 + }, + { + "epoch": 0.04958, + "grad_norm": 0.5459051828491507, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 4958 + }, + { + "epoch": 0.04959, + "grad_norm": 0.5104247811717014, + "learning_rate": 0.003, + "loss": 4.074, + "step": 4959 + }, + { + "epoch": 0.0496, + "grad_norm": 0.5834789174151195, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 4960 + }, + { + "epoch": 0.04961, + "grad_norm": 0.680486590256994, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 4961 + }, + { + "epoch": 0.04962, + "grad_norm": 0.7811370856942803, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 4962 + }, + { + "epoch": 0.04963, + "grad_norm": 0.9496148034945776, + "learning_rate": 0.003, + "loss": 4.116, + "step": 4963 + }, + { + "epoch": 0.04964, + "grad_norm": 0.8262893961841065, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 4964 + }, + { + "epoch": 0.04965, + "grad_norm": 0.5871665512907862, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 4965 + }, + { + "epoch": 0.04966, + "grad_norm": 0.6127735149114402, + "learning_rate": 0.003, + "loss": 4.05, + "step": 4966 + }, + { + "epoch": 0.04967, + "grad_norm": 0.698032553274959, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 4967 + }, + { + "epoch": 0.04968, + "grad_norm": 0.6413274733448541, + "learning_rate": 0.003, + "loss": 4.1009, + "step": 4968 + }, + { + "epoch": 0.04969, + "grad_norm": 0.5617666399608898, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 4969 + }, + { + "epoch": 0.0497, + "grad_norm": 0.6353274285676013, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 4970 + }, + { + "epoch": 0.04971, + "grad_norm": 0.6760555298648803, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 4971 + }, + { + "epoch": 0.04972, + "grad_norm": 0.7466945000341595, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 4972 + }, + { + "epoch": 0.04973, + "grad_norm": 0.717266234153821, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 4973 + }, + { + "epoch": 0.04974, + "grad_norm": 0.6936011071900033, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 4974 + }, + { + "epoch": 0.04975, + "grad_norm": 0.7243386921199736, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 4975 + }, + { + "epoch": 0.04976, + "grad_norm": 0.8374518428685792, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 4976 + }, + { + "epoch": 0.04977, + "grad_norm": 0.7745142393290544, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 4977 + }, + { + "epoch": 0.04978, + "grad_norm": 0.8200466751448288, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 4978 + }, + { + "epoch": 0.04979, + "grad_norm": 0.7222581693519674, + "learning_rate": 0.003, + "loss": 4.0931, + "step": 4979 + }, + { + "epoch": 0.0498, + "grad_norm": 0.6614661861504726, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 4980 + }, + { + "epoch": 0.04981, + "grad_norm": 0.6065296394987594, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 4981 + }, + { + "epoch": 0.04982, + "grad_norm": 0.5567277792979317, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 4982 + }, + { + "epoch": 0.04983, + "grad_norm": 0.5535485016820637, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 4983 + }, + { + "epoch": 0.04984, + "grad_norm": 0.6098925481217158, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 4984 + }, + { + "epoch": 0.04985, + "grad_norm": 0.5491431886548412, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 4985 + }, + { + "epoch": 0.04986, + "grad_norm": 0.44700521159948303, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 4986 + }, + { + "epoch": 0.04987, + "grad_norm": 0.49073903501190747, + "learning_rate": 0.003, + "loss": 4.047, + "step": 4987 + }, + { + "epoch": 0.04988, + "grad_norm": 0.5771393889667359, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 4988 + }, + { + "epoch": 0.04989, + "grad_norm": 0.6975969194414504, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 4989 + }, + { + "epoch": 0.0499, + "grad_norm": 0.7801139973684481, + "learning_rate": 0.003, + "loss": 4.065, + "step": 4990 + }, + { + "epoch": 0.04991, + "grad_norm": 0.7165901276488985, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 4991 + }, + { + "epoch": 0.04992, + "grad_norm": 0.5955647238253913, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 4992 + }, + { + "epoch": 0.04993, + "grad_norm": 0.582276527468128, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 4993 + }, + { + "epoch": 0.04994, + "grad_norm": 0.6274979531061908, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 4994 + }, + { + "epoch": 0.04995, + "grad_norm": 0.6541273329069077, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 4995 + }, + { + "epoch": 0.04996, + "grad_norm": 0.6941834464050866, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 4996 + }, + { + "epoch": 0.04997, + "grad_norm": 0.857996347011742, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 4997 + }, + { + "epoch": 0.04998, + "grad_norm": 0.9041465489466798, + "learning_rate": 0.003, + "loss": 4.1052, + "step": 4998 + }, + { + "epoch": 0.04999, + "grad_norm": 0.7738620841525107, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 4999 + }, + { + "epoch": 0.05, + "grad_norm": 0.7221082107316732, + "learning_rate": 0.003, + "loss": 4.1045, + "step": 5000 + }, + { + "epoch": 0.05001, + "grad_norm": 0.7207715169373524, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 5001 + }, + { + "epoch": 0.05002, + "grad_norm": 0.7678290236063013, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 5002 + }, + { + "epoch": 0.05003, + "grad_norm": 0.8075716990260404, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 5003 + }, + { + "epoch": 0.05004, + "grad_norm": 1.021866748996428, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 5004 + }, + { + "epoch": 0.05005, + "grad_norm": 1.1556568222208223, + "learning_rate": 0.003, + "loss": 4.1139, + "step": 5005 + }, + { + "epoch": 0.05006, + "grad_norm": 0.9425690463288231, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 5006 + }, + { + "epoch": 0.05007, + "grad_norm": 1.0277825772571685, + "learning_rate": 0.003, + "loss": 4.1096, + "step": 5007 + }, + { + "epoch": 0.05008, + "grad_norm": 1.0147371255224404, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 5008 + }, + { + "epoch": 0.05009, + "grad_norm": 0.7847088070401724, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 5009 + }, + { + "epoch": 0.0501, + "grad_norm": 0.7664678827332061, + "learning_rate": 0.003, + "loss": 4.0972, + "step": 5010 + }, + { + "epoch": 0.05011, + "grad_norm": 0.6514850145401839, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 5011 + }, + { + "epoch": 0.05012, + "grad_norm": 0.5668287812808577, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 5012 + }, + { + "epoch": 0.05013, + "grad_norm": 0.4917546030833427, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 5013 + }, + { + "epoch": 0.05014, + "grad_norm": 0.503669640471517, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 5014 + }, + { + "epoch": 0.05015, + "grad_norm": 0.5378061559804919, + "learning_rate": 0.003, + "loss": 4.091, + "step": 5015 + }, + { + "epoch": 0.05016, + "grad_norm": 0.5941934461915904, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 5016 + }, + { + "epoch": 0.05017, + "grad_norm": 0.6939960565664077, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 5017 + }, + { + "epoch": 0.05018, + "grad_norm": 0.7763714674829162, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 5018 + }, + { + "epoch": 0.05019, + "grad_norm": 0.7075378548776008, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 5019 + }, + { + "epoch": 0.0502, + "grad_norm": 0.6474647750538696, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 5020 + }, + { + "epoch": 0.05021, + "grad_norm": 0.5829202184444883, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 5021 + }, + { + "epoch": 0.05022, + "grad_norm": 0.6312364650132364, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 5022 + }, + { + "epoch": 0.05023, + "grad_norm": 0.6812129904359995, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 5023 + }, + { + "epoch": 0.05024, + "grad_norm": 0.668465595600859, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 5024 + }, + { + "epoch": 0.05025, + "grad_norm": 0.6479909661325042, + "learning_rate": 0.003, + "loss": 4.09, + "step": 5025 + }, + { + "epoch": 0.05026, + "grad_norm": 0.693394733944601, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 5026 + }, + { + "epoch": 0.05027, + "grad_norm": 0.7233013674939288, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 5027 + }, + { + "epoch": 0.05028, + "grad_norm": 0.7201078663451931, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 5028 + }, + { + "epoch": 0.05029, + "grad_norm": 0.7328952975201708, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 5029 + }, + { + "epoch": 0.0503, + "grad_norm": 0.7357735119303881, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 5030 + }, + { + "epoch": 0.05031, + "grad_norm": 0.7334599238045354, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 5031 + }, + { + "epoch": 0.05032, + "grad_norm": 0.6077007141501842, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 5032 + }, + { + "epoch": 0.05033, + "grad_norm": 0.6120484527654543, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 5033 + }, + { + "epoch": 0.05034, + "grad_norm": 0.6079303224053032, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 5034 + }, + { + "epoch": 0.05035, + "grad_norm": 0.6434514165729239, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 5035 + }, + { + "epoch": 0.05036, + "grad_norm": 0.5498593187788509, + "learning_rate": 0.003, + "loss": 4.056, + "step": 5036 + }, + { + "epoch": 0.05037, + "grad_norm": 0.5643584524301097, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 5037 + }, + { + "epoch": 0.05038, + "grad_norm": 0.5936025251821794, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 5038 + }, + { + "epoch": 0.05039, + "grad_norm": 0.5456350584257579, + "learning_rate": 0.003, + "loss": 4.073, + "step": 5039 + }, + { + "epoch": 0.0504, + "grad_norm": 0.6165655687042918, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 5040 + }, + { + "epoch": 0.05041, + "grad_norm": 0.6966996186465338, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 5041 + }, + { + "epoch": 0.05042, + "grad_norm": 0.7105431534106783, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5042 + }, + { + "epoch": 0.05043, + "grad_norm": 0.6913372053851102, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 5043 + }, + { + "epoch": 0.05044, + "grad_norm": 0.8499199140786121, + "learning_rate": 0.003, + "loss": 4.091, + "step": 5044 + }, + { + "epoch": 0.05045, + "grad_norm": 0.9837580087932947, + "learning_rate": 0.003, + "loss": 4.1155, + "step": 5045 + }, + { + "epoch": 0.05046, + "grad_norm": 1.0860427965785973, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 5046 + }, + { + "epoch": 0.05047, + "grad_norm": 0.8647487115371897, + "learning_rate": 0.003, + "loss": 4.087, + "step": 5047 + }, + { + "epoch": 0.05048, + "grad_norm": 0.9198477632114738, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 5048 + }, + { + "epoch": 0.05049, + "grad_norm": 0.8940857915471107, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 5049 + }, + { + "epoch": 0.0505, + "grad_norm": 0.9847075368877446, + "learning_rate": 0.003, + "loss": 4.1456, + "step": 5050 + }, + { + "epoch": 0.05051, + "grad_norm": 1.022169199897865, + "learning_rate": 0.003, + "loss": 4.1112, + "step": 5051 + }, + { + "epoch": 0.05052, + "grad_norm": 0.8169233060661102, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 5052 + }, + { + "epoch": 0.05053, + "grad_norm": 0.7414097932020427, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 5053 + }, + { + "epoch": 0.05054, + "grad_norm": 0.7444151111133538, + "learning_rate": 0.003, + "loss": 4.105, + "step": 5054 + }, + { + "epoch": 0.05055, + "grad_norm": 0.6541108372539527, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 5055 + }, + { + "epoch": 0.05056, + "grad_norm": 0.6726595085244714, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 5056 + }, + { + "epoch": 0.05057, + "grad_norm": 0.7222550445740372, + "learning_rate": 0.003, + "loss": 4.0984, + "step": 5057 + }, + { + "epoch": 0.05058, + "grad_norm": 0.709669541822485, + "learning_rate": 0.003, + "loss": 4.0977, + "step": 5058 + }, + { + "epoch": 0.05059, + "grad_norm": 0.6381164751793498, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 5059 + }, + { + "epoch": 0.0506, + "grad_norm": 0.6418773588116435, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 5060 + }, + { + "epoch": 0.05061, + "grad_norm": 0.7585930725150771, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 5061 + }, + { + "epoch": 0.05062, + "grad_norm": 0.719968882695951, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 5062 + }, + { + "epoch": 0.05063, + "grad_norm": 0.6991393845421153, + "learning_rate": 0.003, + "loss": 4.044, + "step": 5063 + }, + { + "epoch": 0.05064, + "grad_norm": 0.7129916992731382, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 5064 + }, + { + "epoch": 0.05065, + "grad_norm": 0.743153993906533, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 5065 + }, + { + "epoch": 0.05066, + "grad_norm": 0.7010660887268799, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5066 + }, + { + "epoch": 0.05067, + "grad_norm": 0.7115379154142746, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 5067 + }, + { + "epoch": 0.05068, + "grad_norm": 0.6158148867993987, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 5068 + }, + { + "epoch": 0.05069, + "grad_norm": 0.5763552198867099, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 5069 + }, + { + "epoch": 0.0507, + "grad_norm": 0.5122385175265607, + "learning_rate": 0.003, + "loss": 4.088, + "step": 5070 + }, + { + "epoch": 0.05071, + "grad_norm": 0.4630465826539238, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 5071 + }, + { + "epoch": 0.05072, + "grad_norm": 0.49913185855404474, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 5072 + }, + { + "epoch": 0.05073, + "grad_norm": 0.4763577977857625, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 5073 + }, + { + "epoch": 0.05074, + "grad_norm": 0.42614720445005816, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 5074 + }, + { + "epoch": 0.05075, + "grad_norm": 0.46181274770079256, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 5075 + }, + { + "epoch": 0.05076, + "grad_norm": 0.48966142691435327, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 5076 + }, + { + "epoch": 0.05077, + "grad_norm": 0.5575694198691047, + "learning_rate": 0.003, + "loss": 4.0841, + "step": 5077 + }, + { + "epoch": 0.05078, + "grad_norm": 0.6507313786062453, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 5078 + }, + { + "epoch": 0.05079, + "grad_norm": 0.6466461887871029, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 5079 + }, + { + "epoch": 0.0508, + "grad_norm": 0.628595845416593, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 5080 + }, + { + "epoch": 0.05081, + "grad_norm": 0.6561976390372858, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 5081 + }, + { + "epoch": 0.05082, + "grad_norm": 0.7472739164943476, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 5082 + }, + { + "epoch": 0.05083, + "grad_norm": 0.776324684866391, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 5083 + }, + { + "epoch": 0.05084, + "grad_norm": 0.8514758852167847, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 5084 + }, + { + "epoch": 0.05085, + "grad_norm": 0.900844104584562, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 5085 + }, + { + "epoch": 0.05086, + "grad_norm": 1.0868345284875638, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 5086 + }, + { + "epoch": 0.05087, + "grad_norm": 0.9078382576688434, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 5087 + }, + { + "epoch": 0.05088, + "grad_norm": 0.6838460502920969, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 5088 + }, + { + "epoch": 0.05089, + "grad_norm": 0.7101286475784213, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 5089 + }, + { + "epoch": 0.0509, + "grad_norm": 0.7763999319317019, + "learning_rate": 0.003, + "loss": 4.1024, + "step": 5090 + }, + { + "epoch": 0.05091, + "grad_norm": 0.7988586255605415, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 5091 + }, + { + "epoch": 0.05092, + "grad_norm": 0.7627523657308218, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 5092 + }, + { + "epoch": 0.05093, + "grad_norm": 0.7217210799024187, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 5093 + }, + { + "epoch": 0.05094, + "grad_norm": 0.6893448273383638, + "learning_rate": 0.003, + "loss": 4.041, + "step": 5094 + }, + { + "epoch": 0.05095, + "grad_norm": 0.6316145239711671, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 5095 + }, + { + "epoch": 0.05096, + "grad_norm": 0.7091189904492731, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 5096 + }, + { + "epoch": 0.05097, + "grad_norm": 0.8333921720060847, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 5097 + }, + { + "epoch": 0.05098, + "grad_norm": 0.7507751795820745, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 5098 + }, + { + "epoch": 0.05099, + "grad_norm": 0.6110153242918009, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 5099 + }, + { + "epoch": 0.051, + "grad_norm": 0.5382179594179979, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 5100 + }, + { + "epoch": 0.05101, + "grad_norm": 0.5460282315353179, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 5101 + }, + { + "epoch": 0.05102, + "grad_norm": 0.5672101062591516, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 5102 + }, + { + "epoch": 0.05103, + "grad_norm": 0.6753897534246182, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 5103 + }, + { + "epoch": 0.05104, + "grad_norm": 0.7697411050220518, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 5104 + }, + { + "epoch": 0.05105, + "grad_norm": 0.7981111757369809, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 5105 + }, + { + "epoch": 0.05106, + "grad_norm": 0.7151773080659729, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 5106 + }, + { + "epoch": 0.05107, + "grad_norm": 0.713335861878738, + "learning_rate": 0.003, + "loss": 4.054, + "step": 5107 + }, + { + "epoch": 0.05108, + "grad_norm": 0.6972283110695205, + "learning_rate": 0.003, + "loss": 4.1028, + "step": 5108 + }, + { + "epoch": 0.05109, + "grad_norm": 0.7736294848506601, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 5109 + }, + { + "epoch": 0.0511, + "grad_norm": 0.8628574576614368, + "learning_rate": 0.003, + "loss": 4.083, + "step": 5110 + }, + { + "epoch": 0.05111, + "grad_norm": 0.7681044560680576, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 5111 + }, + { + "epoch": 0.05112, + "grad_norm": 0.9403320834387587, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 5112 + }, + { + "epoch": 0.05113, + "grad_norm": 0.9154558930752019, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 5113 + }, + { + "epoch": 0.05114, + "grad_norm": 0.7887772719586122, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 5114 + }, + { + "epoch": 0.05115, + "grad_norm": 0.824193607035564, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 5115 + }, + { + "epoch": 0.05116, + "grad_norm": 0.6176260228337916, + "learning_rate": 0.003, + "loss": 4.086, + "step": 5116 + }, + { + "epoch": 0.05117, + "grad_norm": 0.5043072078793339, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 5117 + }, + { + "epoch": 0.05118, + "grad_norm": 0.451989836137091, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 5118 + }, + { + "epoch": 0.05119, + "grad_norm": 0.49455903722296113, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 5119 + }, + { + "epoch": 0.0512, + "grad_norm": 0.5154017193677266, + "learning_rate": 0.003, + "loss": 4.035, + "step": 5120 + }, + { + "epoch": 0.05121, + "grad_norm": 0.6493427063702185, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 5121 + }, + { + "epoch": 0.05122, + "grad_norm": 0.7554068873295963, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 5122 + }, + { + "epoch": 0.05123, + "grad_norm": 0.8915263282194953, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 5123 + }, + { + "epoch": 0.05124, + "grad_norm": 0.8473970015636592, + "learning_rate": 0.003, + "loss": 4.1105, + "step": 5124 + }, + { + "epoch": 0.05125, + "grad_norm": 0.8317554259883203, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 5125 + }, + { + "epoch": 0.05126, + "grad_norm": 0.7941120909157412, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 5126 + }, + { + "epoch": 0.05127, + "grad_norm": 0.7475302239093903, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 5127 + }, + { + "epoch": 0.05128, + "grad_norm": 0.8009141999180037, + "learning_rate": 0.003, + "loss": 4.0909, + "step": 5128 + }, + { + "epoch": 0.05129, + "grad_norm": 0.777292113974254, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 5129 + }, + { + "epoch": 0.0513, + "grad_norm": 0.8070571523960868, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 5130 + }, + { + "epoch": 0.05131, + "grad_norm": 0.9332241882699787, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 5131 + }, + { + "epoch": 0.05132, + "grad_norm": 1.0378085030471271, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 5132 + }, + { + "epoch": 0.05133, + "grad_norm": 1.0020091873772643, + "learning_rate": 0.003, + "loss": 4.086, + "step": 5133 + }, + { + "epoch": 0.05134, + "grad_norm": 0.8306644245851558, + "learning_rate": 0.003, + "loss": 4.0975, + "step": 5134 + }, + { + "epoch": 0.05135, + "grad_norm": 0.7041122273322672, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 5135 + }, + { + "epoch": 0.05136, + "grad_norm": 0.6953761267220954, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 5136 + }, + { + "epoch": 0.05137, + "grad_norm": 0.6704590279922362, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 5137 + }, + { + "epoch": 0.05138, + "grad_norm": 0.590958777551943, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5138 + }, + { + "epoch": 0.05139, + "grad_norm": 0.5603112710910837, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 5139 + }, + { + "epoch": 0.0514, + "grad_norm": 0.5870567253186774, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 5140 + }, + { + "epoch": 0.05141, + "grad_norm": 0.7256810258540175, + "learning_rate": 0.003, + "loss": 4.1229, + "step": 5141 + }, + { + "epoch": 0.05142, + "grad_norm": 0.8662062062184371, + "learning_rate": 0.003, + "loss": 4.071, + "step": 5142 + }, + { + "epoch": 0.05143, + "grad_norm": 0.8816655966444198, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 5143 + }, + { + "epoch": 0.05144, + "grad_norm": 0.8384660999268876, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 5144 + }, + { + "epoch": 0.05145, + "grad_norm": 0.7895905639964732, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 5145 + }, + { + "epoch": 0.05146, + "grad_norm": 0.6238234004210469, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 5146 + }, + { + "epoch": 0.05147, + "grad_norm": 0.5827565818729636, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 5147 + }, + { + "epoch": 0.05148, + "grad_norm": 0.5607047195955596, + "learning_rate": 0.003, + "loss": 4.078, + "step": 5148 + }, + { + "epoch": 0.05149, + "grad_norm": 0.5436712663789659, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 5149 + }, + { + "epoch": 0.0515, + "grad_norm": 0.5503222191573438, + "learning_rate": 0.003, + "loss": 4.0859, + "step": 5150 + }, + { + "epoch": 0.05151, + "grad_norm": 0.5810742467588578, + "learning_rate": 0.003, + "loss": 4.047, + "step": 5151 + }, + { + "epoch": 0.05152, + "grad_norm": 0.5591465702875923, + "learning_rate": 0.003, + "loss": 4.0845, + "step": 5152 + }, + { + "epoch": 0.05153, + "grad_norm": 0.5323362322821417, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5153 + }, + { + "epoch": 0.05154, + "grad_norm": 0.5373681444619369, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 5154 + }, + { + "epoch": 0.05155, + "grad_norm": 0.6219865369829133, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5155 + }, + { + "epoch": 0.05156, + "grad_norm": 0.6912922535267688, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 5156 + }, + { + "epoch": 0.05157, + "grad_norm": 0.7264789497169568, + "learning_rate": 0.003, + "loss": 4.071, + "step": 5157 + }, + { + "epoch": 0.05158, + "grad_norm": 0.6280408454397876, + "learning_rate": 0.003, + "loss": 4.065, + "step": 5158 + }, + { + "epoch": 0.05159, + "grad_norm": 0.641377350936863, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 5159 + }, + { + "epoch": 0.0516, + "grad_norm": 0.7023039070803929, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 5160 + }, + { + "epoch": 0.05161, + "grad_norm": 0.6290899720214803, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 5161 + }, + { + "epoch": 0.05162, + "grad_norm": 0.6553950060607368, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 5162 + }, + { + "epoch": 0.05163, + "grad_norm": 0.7379989909446212, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 5163 + }, + { + "epoch": 0.05164, + "grad_norm": 0.8562662935725894, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 5164 + }, + { + "epoch": 0.05165, + "grad_norm": 0.8176759123643486, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 5165 + }, + { + "epoch": 0.05166, + "grad_norm": 0.8203011699110355, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 5166 + }, + { + "epoch": 0.05167, + "grad_norm": 0.792356617789425, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 5167 + }, + { + "epoch": 0.05168, + "grad_norm": 0.792234001022072, + "learning_rate": 0.003, + "loss": 4.1152, + "step": 5168 + }, + { + "epoch": 0.05169, + "grad_norm": 0.8083430145781528, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 5169 + }, + { + "epoch": 0.0517, + "grad_norm": 0.8383371701457759, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 5170 + }, + { + "epoch": 0.05171, + "grad_norm": 0.7014479348552264, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 5171 + }, + { + "epoch": 0.05172, + "grad_norm": 0.7225176906399309, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 5172 + }, + { + "epoch": 0.05173, + "grad_norm": 0.6959872263407636, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 5173 + }, + { + "epoch": 0.05174, + "grad_norm": 0.6771774531298075, + "learning_rate": 0.003, + "loss": 4.084, + "step": 5174 + }, + { + "epoch": 0.05175, + "grad_norm": 0.6133668672518019, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 5175 + }, + { + "epoch": 0.05176, + "grad_norm": 0.5559522030531587, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 5176 + }, + { + "epoch": 0.05177, + "grad_norm": 0.5549235617650451, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 5177 + }, + { + "epoch": 0.05178, + "grad_norm": 0.538775523981983, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 5178 + }, + { + "epoch": 0.05179, + "grad_norm": 0.5585516701777474, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 5179 + }, + { + "epoch": 0.0518, + "grad_norm": 0.6380321421721828, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5180 + }, + { + "epoch": 0.05181, + "grad_norm": 0.7882136523280638, + "learning_rate": 0.003, + "loss": 4.071, + "step": 5181 + }, + { + "epoch": 0.05182, + "grad_norm": 1.147254318409289, + "learning_rate": 0.003, + "loss": 4.1011, + "step": 5182 + }, + { + "epoch": 0.05183, + "grad_norm": 0.9441048293485047, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5183 + }, + { + "epoch": 0.05184, + "grad_norm": 0.8130737788176512, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 5184 + }, + { + "epoch": 0.05185, + "grad_norm": 0.7866488153080101, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 5185 + }, + { + "epoch": 0.05186, + "grad_norm": 0.5640199174703157, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 5186 + }, + { + "epoch": 0.05187, + "grad_norm": 0.5137677003370857, + "learning_rate": 0.003, + "loss": 4.1046, + "step": 5187 + }, + { + "epoch": 0.05188, + "grad_norm": 0.4986472122679641, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 5188 + }, + { + "epoch": 0.05189, + "grad_norm": 0.5654611920971662, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5189 + }, + { + "epoch": 0.0519, + "grad_norm": 0.556204628133283, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5190 + }, + { + "epoch": 0.05191, + "grad_norm": 0.5437529440283245, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 5191 + }, + { + "epoch": 0.05192, + "grad_norm": 0.5967258421452346, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 5192 + }, + { + "epoch": 0.05193, + "grad_norm": 0.6917799808669597, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 5193 + }, + { + "epoch": 0.05194, + "grad_norm": 0.74165138098978, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 5194 + }, + { + "epoch": 0.05195, + "grad_norm": 0.7716864670524971, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 5195 + }, + { + "epoch": 0.05196, + "grad_norm": 0.8562473873224635, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 5196 + }, + { + "epoch": 0.05197, + "grad_norm": 0.9509391071165898, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 5197 + }, + { + "epoch": 0.05198, + "grad_norm": 0.8787744857211688, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 5198 + }, + { + "epoch": 0.05199, + "grad_norm": 0.7914494568615887, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 5199 + }, + { + "epoch": 0.052, + "grad_norm": 0.7290968979382663, + "learning_rate": 0.003, + "loss": 4.077, + "step": 5200 + }, + { + "epoch": 0.05201, + "grad_norm": 0.7934903405328128, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 5201 + }, + { + "epoch": 0.05202, + "grad_norm": 0.9194405333245952, + "learning_rate": 0.003, + "loss": 4.067, + "step": 5202 + }, + { + "epoch": 0.05203, + "grad_norm": 0.8861687616148332, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 5203 + }, + { + "epoch": 0.05204, + "grad_norm": 0.7297711312167907, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 5204 + }, + { + "epoch": 0.05205, + "grad_norm": 0.678316716270082, + "learning_rate": 0.003, + "loss": 4.0906, + "step": 5205 + }, + { + "epoch": 0.05206, + "grad_norm": 0.6649786677807524, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 5206 + }, + { + "epoch": 0.05207, + "grad_norm": 0.7664244369297829, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 5207 + }, + { + "epoch": 0.05208, + "grad_norm": 0.6749316469348212, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5208 + }, + { + "epoch": 0.05209, + "grad_norm": 0.5397058423337406, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 5209 + }, + { + "epoch": 0.0521, + "grad_norm": 0.5335948316709165, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 5210 + }, + { + "epoch": 0.05211, + "grad_norm": 0.5217090148004745, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 5211 + }, + { + "epoch": 0.05212, + "grad_norm": 0.628460696523494, + "learning_rate": 0.003, + "loss": 4.108, + "step": 5212 + }, + { + "epoch": 0.05213, + "grad_norm": 0.7430176043760545, + "learning_rate": 0.003, + "loss": 4.066, + "step": 5213 + }, + { + "epoch": 0.05214, + "grad_norm": 0.7992033518100482, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 5214 + }, + { + "epoch": 0.05215, + "grad_norm": 0.8156656543624262, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 5215 + }, + { + "epoch": 0.05216, + "grad_norm": 0.7576224135783045, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 5216 + }, + { + "epoch": 0.05217, + "grad_norm": 0.8159177677256783, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 5217 + }, + { + "epoch": 0.05218, + "grad_norm": 0.8750936369054375, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 5218 + }, + { + "epoch": 0.05219, + "grad_norm": 0.716234477651202, + "learning_rate": 0.003, + "loss": 4.0894, + "step": 5219 + }, + { + "epoch": 0.0522, + "grad_norm": 0.6774419733516789, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 5220 + }, + { + "epoch": 0.05221, + "grad_norm": 0.7390618513287948, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 5221 + }, + { + "epoch": 0.05222, + "grad_norm": 0.7069214107228244, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 5222 + }, + { + "epoch": 0.05223, + "grad_norm": 0.7593808148420265, + "learning_rate": 0.003, + "loss": 4.0944, + "step": 5223 + }, + { + "epoch": 0.05224, + "grad_norm": 0.8086126445495831, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 5224 + }, + { + "epoch": 0.05225, + "grad_norm": 0.8533357438534811, + "learning_rate": 0.003, + "loss": 4.0969, + "step": 5225 + }, + { + "epoch": 0.05226, + "grad_norm": 0.8017383119106186, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 5226 + }, + { + "epoch": 0.05227, + "grad_norm": 0.8408764838890277, + "learning_rate": 0.003, + "loss": 4.1102, + "step": 5227 + }, + { + "epoch": 0.05228, + "grad_norm": 0.8478661181281025, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 5228 + }, + { + "epoch": 0.05229, + "grad_norm": 0.7314872081129211, + "learning_rate": 0.003, + "loss": 4.054, + "step": 5229 + }, + { + "epoch": 0.0523, + "grad_norm": 0.6471765173013521, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 5230 + }, + { + "epoch": 0.05231, + "grad_norm": 0.6247248669558405, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5231 + }, + { + "epoch": 0.05232, + "grad_norm": 0.6593628097038716, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 5232 + }, + { + "epoch": 0.05233, + "grad_norm": 0.7074201299897884, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 5233 + }, + { + "epoch": 0.05234, + "grad_norm": 0.7787829375154793, + "learning_rate": 0.003, + "loss": 4.101, + "step": 5234 + }, + { + "epoch": 0.05235, + "grad_norm": 0.73045699098065, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 5235 + }, + { + "epoch": 0.05236, + "grad_norm": 0.5317841986991921, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 5236 + }, + { + "epoch": 0.05237, + "grad_norm": 0.5110351474013536, + "learning_rate": 0.003, + "loss": 4.056, + "step": 5237 + }, + { + "epoch": 0.05238, + "grad_norm": 0.6203808372177038, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 5238 + }, + { + "epoch": 0.05239, + "grad_norm": 0.6138803888375408, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 5239 + }, + { + "epoch": 0.0524, + "grad_norm": 0.6307861105870085, + "learning_rate": 0.003, + "loss": 4.0788, + "step": 5240 + }, + { + "epoch": 0.05241, + "grad_norm": 0.5735467081798893, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 5241 + }, + { + "epoch": 0.05242, + "grad_norm": 0.5940240867577873, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 5242 + }, + { + "epoch": 0.05243, + "grad_norm": 0.6157057005746668, + "learning_rate": 0.003, + "loss": 4.072, + "step": 5243 + }, + { + "epoch": 0.05244, + "grad_norm": 0.7070049561715044, + "learning_rate": 0.003, + "loss": 4.05, + "step": 5244 + }, + { + "epoch": 0.05245, + "grad_norm": 0.8362066173401371, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 5245 + }, + { + "epoch": 0.05246, + "grad_norm": 0.796420112342633, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 5246 + }, + { + "epoch": 0.05247, + "grad_norm": 0.7328189372046566, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 5247 + }, + { + "epoch": 0.05248, + "grad_norm": 0.7308608242850233, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 5248 + }, + { + "epoch": 0.05249, + "grad_norm": 0.7447650650944236, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 5249 + }, + { + "epoch": 0.0525, + "grad_norm": 0.7485265862205066, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5250 + }, + { + "epoch": 0.05251, + "grad_norm": 0.664302149874978, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 5251 + }, + { + "epoch": 0.05252, + "grad_norm": 0.6812983476986793, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 5252 + }, + { + "epoch": 0.05253, + "grad_norm": 0.548794522319428, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 5253 + }, + { + "epoch": 0.05254, + "grad_norm": 0.5169779120335845, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 5254 + }, + { + "epoch": 0.05255, + "grad_norm": 0.525188127864816, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 5255 + }, + { + "epoch": 0.05256, + "grad_norm": 0.5161537077115214, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 5256 + }, + { + "epoch": 0.05257, + "grad_norm": 0.6042543648695965, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 5257 + }, + { + "epoch": 0.05258, + "grad_norm": 0.671010212922069, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 5258 + }, + { + "epoch": 0.05259, + "grad_norm": 0.6360151070182973, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 5259 + }, + { + "epoch": 0.0526, + "grad_norm": 0.591299001704799, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 5260 + }, + { + "epoch": 0.05261, + "grad_norm": 0.6304188804633112, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 5261 + }, + { + "epoch": 0.05262, + "grad_norm": 0.6581104585438666, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 5262 + }, + { + "epoch": 0.05263, + "grad_norm": 0.7160923349835606, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 5263 + }, + { + "epoch": 0.05264, + "grad_norm": 0.7171792304984013, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 5264 + }, + { + "epoch": 0.05265, + "grad_norm": 0.6212819143446443, + "learning_rate": 0.003, + "loss": 4.047, + "step": 5265 + }, + { + "epoch": 0.05266, + "grad_norm": 0.5734525315671152, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 5266 + }, + { + "epoch": 0.05267, + "grad_norm": 0.7361106955426965, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 5267 + }, + { + "epoch": 0.05268, + "grad_norm": 0.7291217444153477, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 5268 + }, + { + "epoch": 0.05269, + "grad_norm": 0.5904376641582534, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 5269 + }, + { + "epoch": 0.0527, + "grad_norm": 0.6273525177946541, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 5270 + }, + { + "epoch": 0.05271, + "grad_norm": 0.6911707242462707, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 5271 + }, + { + "epoch": 0.05272, + "grad_norm": 0.8195201648471655, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 5272 + }, + { + "epoch": 0.05273, + "grad_norm": 0.9356857883916216, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 5273 + }, + { + "epoch": 0.05274, + "grad_norm": 0.9755629010734023, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 5274 + }, + { + "epoch": 0.05275, + "grad_norm": 0.9887337769527103, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 5275 + }, + { + "epoch": 0.05276, + "grad_norm": 1.0294586474084828, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 5276 + }, + { + "epoch": 0.05277, + "grad_norm": 0.7132541080208686, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 5277 + }, + { + "epoch": 0.05278, + "grad_norm": 0.5510532627183646, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 5278 + }, + { + "epoch": 0.05279, + "grad_norm": 0.6328762038986853, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 5279 + }, + { + "epoch": 0.0528, + "grad_norm": 0.6905199071622662, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 5280 + }, + { + "epoch": 0.05281, + "grad_norm": 0.7442840140341745, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 5281 + }, + { + "epoch": 0.05282, + "grad_norm": 0.7206300669027469, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 5282 + }, + { + "epoch": 0.05283, + "grad_norm": 0.6653637512172188, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 5283 + }, + { + "epoch": 0.05284, + "grad_norm": 0.6976313037046374, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 5284 + }, + { + "epoch": 0.05285, + "grad_norm": 0.8381425008063161, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 5285 + }, + { + "epoch": 0.05286, + "grad_norm": 0.8884560665877108, + "learning_rate": 0.003, + "loss": 4.106, + "step": 5286 + }, + { + "epoch": 0.05287, + "grad_norm": 0.9589555132060451, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 5287 + }, + { + "epoch": 0.05288, + "grad_norm": 1.046047830077395, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 5288 + }, + { + "epoch": 0.05289, + "grad_norm": 0.8601288917576163, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 5289 + }, + { + "epoch": 0.0529, + "grad_norm": 0.956190570874191, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 5290 + }, + { + "epoch": 0.05291, + "grad_norm": 0.8678089351640984, + "learning_rate": 0.003, + "loss": 4.1038, + "step": 5291 + }, + { + "epoch": 0.05292, + "grad_norm": 0.8627559768849714, + "learning_rate": 0.003, + "loss": 4.0943, + "step": 5292 + }, + { + "epoch": 0.05293, + "grad_norm": 0.9767158811872634, + "learning_rate": 0.003, + "loss": 4.0988, + "step": 5293 + }, + { + "epoch": 0.05294, + "grad_norm": 1.0694339695822357, + "learning_rate": 0.003, + "loss": 4.09, + "step": 5294 + }, + { + "epoch": 0.05295, + "grad_norm": 1.0905677012469588, + "learning_rate": 0.003, + "loss": 4.1097, + "step": 5295 + }, + { + "epoch": 0.05296, + "grad_norm": 0.9599364705549474, + "learning_rate": 0.003, + "loss": 4.084, + "step": 5296 + }, + { + "epoch": 0.05297, + "grad_norm": 0.9392228429088171, + "learning_rate": 0.003, + "loss": 4.0871, + "step": 5297 + }, + { + "epoch": 0.05298, + "grad_norm": 0.6911480927160727, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 5298 + }, + { + "epoch": 0.05299, + "grad_norm": 0.6983683894033775, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 5299 + }, + { + "epoch": 0.053, + "grad_norm": 0.6506445202153651, + "learning_rate": 0.003, + "loss": 4.0887, + "step": 5300 + }, + { + "epoch": 0.05301, + "grad_norm": 0.5923387894147693, + "learning_rate": 0.003, + "loss": 4.098, + "step": 5301 + }, + { + "epoch": 0.05302, + "grad_norm": 0.6326677007140554, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 5302 + }, + { + "epoch": 0.05303, + "grad_norm": 0.6342545337728676, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5303 + }, + { + "epoch": 0.05304, + "grad_norm": 0.6238057442154212, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 5304 + }, + { + "epoch": 0.05305, + "grad_norm": 0.5897516742910698, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 5305 + }, + { + "epoch": 0.05306, + "grad_norm": 0.5785142777887178, + "learning_rate": 0.003, + "loss": 4.08, + "step": 5306 + }, + { + "epoch": 0.05307, + "grad_norm": 0.599206530532653, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 5307 + }, + { + "epoch": 0.05308, + "grad_norm": 0.5818577381991036, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 5308 + }, + { + "epoch": 0.05309, + "grad_norm": 0.6272281610348261, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 5309 + }, + { + "epoch": 0.0531, + "grad_norm": 0.6151930048700504, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 5310 + }, + { + "epoch": 0.05311, + "grad_norm": 0.6136535580762066, + "learning_rate": 0.003, + "loss": 4.1, + "step": 5311 + }, + { + "epoch": 0.05312, + "grad_norm": 0.6958417556878302, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 5312 + }, + { + "epoch": 0.05313, + "grad_norm": 0.8386876260050926, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 5313 + }, + { + "epoch": 0.05314, + "grad_norm": 0.8084792327393534, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 5314 + }, + { + "epoch": 0.05315, + "grad_norm": 0.8592909787777022, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 5315 + }, + { + "epoch": 0.05316, + "grad_norm": 0.8992535701524492, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 5316 + }, + { + "epoch": 0.05317, + "grad_norm": 0.8452928272303352, + "learning_rate": 0.003, + "loss": 4.087, + "step": 5317 + }, + { + "epoch": 0.05318, + "grad_norm": 0.6993175560225076, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 5318 + }, + { + "epoch": 0.05319, + "grad_norm": 0.7814003442613279, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 5319 + }, + { + "epoch": 0.0532, + "grad_norm": 0.7283295309505183, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 5320 + }, + { + "epoch": 0.05321, + "grad_norm": 0.7015404791335473, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 5321 + }, + { + "epoch": 0.05322, + "grad_norm": 0.6089542885026478, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5322 + }, + { + "epoch": 0.05323, + "grad_norm": 0.6274762449612891, + "learning_rate": 0.003, + "loss": 4.1082, + "step": 5323 + }, + { + "epoch": 0.05324, + "grad_norm": 0.5877300221496111, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 5324 + }, + { + "epoch": 0.05325, + "grad_norm": 0.5075887034897073, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 5325 + }, + { + "epoch": 0.05326, + "grad_norm": 0.49135486794996275, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 5326 + }, + { + "epoch": 0.05327, + "grad_norm": 0.4851445551244002, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 5327 + }, + { + "epoch": 0.05328, + "grad_norm": 0.5535311936364707, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5328 + }, + { + "epoch": 0.05329, + "grad_norm": 0.5683664180442813, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5329 + }, + { + "epoch": 0.0533, + "grad_norm": 0.6384632326671968, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5330 + }, + { + "epoch": 0.05331, + "grad_norm": 0.543468136567455, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 5331 + }, + { + "epoch": 0.05332, + "grad_norm": 0.4590883549866317, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 5332 + }, + { + "epoch": 0.05333, + "grad_norm": 0.5219314294529123, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 5333 + }, + { + "epoch": 0.05334, + "grad_norm": 0.5852527821773745, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 5334 + }, + { + "epoch": 0.05335, + "grad_norm": 0.6519328194560547, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 5335 + }, + { + "epoch": 0.05336, + "grad_norm": 0.6222970682832165, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 5336 + }, + { + "epoch": 0.05337, + "grad_norm": 0.6006038232891676, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 5337 + }, + { + "epoch": 0.05338, + "grad_norm": 0.6484427973971445, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 5338 + }, + { + "epoch": 0.05339, + "grad_norm": 0.7623512827525476, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 5339 + }, + { + "epoch": 0.0534, + "grad_norm": 0.841624936424957, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 5340 + }, + { + "epoch": 0.05341, + "grad_norm": 0.8848550399912398, + "learning_rate": 0.003, + "loss": 4.076, + "step": 5341 + }, + { + "epoch": 0.05342, + "grad_norm": 0.9414927875230429, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 5342 + }, + { + "epoch": 0.05343, + "grad_norm": 0.7824113032514459, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5343 + }, + { + "epoch": 0.05344, + "grad_norm": 0.6502379071535319, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 5344 + }, + { + "epoch": 0.05345, + "grad_norm": 0.6633580881574906, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 5345 + }, + { + "epoch": 0.05346, + "grad_norm": 0.7023149197502314, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 5346 + }, + { + "epoch": 0.05347, + "grad_norm": 0.7167845221696809, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 5347 + }, + { + "epoch": 0.05348, + "grad_norm": 0.7708487042463833, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 5348 + }, + { + "epoch": 0.05349, + "grad_norm": 0.6818098878203668, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 5349 + }, + { + "epoch": 0.0535, + "grad_norm": 0.5491676098323799, + "learning_rate": 0.003, + "loss": 4.052, + "step": 5350 + }, + { + "epoch": 0.05351, + "grad_norm": 0.5613109084835443, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 5351 + }, + { + "epoch": 0.05352, + "grad_norm": 0.5147690601332807, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 5352 + }, + { + "epoch": 0.05353, + "grad_norm": 0.5197875553767887, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 5353 + }, + { + "epoch": 0.05354, + "grad_norm": 0.5486399448815402, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 5354 + }, + { + "epoch": 0.05355, + "grad_norm": 0.5682627709425156, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 5355 + }, + { + "epoch": 0.05356, + "grad_norm": 0.6891306004095482, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 5356 + }, + { + "epoch": 0.05357, + "grad_norm": 0.861304258850334, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 5357 + }, + { + "epoch": 0.05358, + "grad_norm": 0.9008113078686671, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 5358 + }, + { + "epoch": 0.05359, + "grad_norm": 0.9628436828996151, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 5359 + }, + { + "epoch": 0.0536, + "grad_norm": 1.051679251640114, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5360 + }, + { + "epoch": 0.05361, + "grad_norm": 0.9865175430176507, + "learning_rate": 0.003, + "loss": 4.088, + "step": 5361 + }, + { + "epoch": 0.05362, + "grad_norm": 1.0297680615779028, + "learning_rate": 0.003, + "loss": 4.1098, + "step": 5362 + }, + { + "epoch": 0.05363, + "grad_norm": 1.1645946733243755, + "learning_rate": 0.003, + "loss": 4.1217, + "step": 5363 + }, + { + "epoch": 0.05364, + "grad_norm": 0.81838887062069, + "learning_rate": 0.003, + "loss": 4.1012, + "step": 5364 + }, + { + "epoch": 0.05365, + "grad_norm": 0.8250720367271674, + "learning_rate": 0.003, + "loss": 4.0949, + "step": 5365 + }, + { + "epoch": 0.05366, + "grad_norm": 0.7888098525596474, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 5366 + }, + { + "epoch": 0.05367, + "grad_norm": 0.6064113181618798, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 5367 + }, + { + "epoch": 0.05368, + "grad_norm": 0.5938165614409546, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 5368 + }, + { + "epoch": 0.05369, + "grad_norm": 0.6190203588379648, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 5369 + }, + { + "epoch": 0.0537, + "grad_norm": 0.7154834676736821, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 5370 + }, + { + "epoch": 0.05371, + "grad_norm": 0.8181288306360902, + "learning_rate": 0.003, + "loss": 4.1003, + "step": 5371 + }, + { + "epoch": 0.05372, + "grad_norm": 0.8352369107250618, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 5372 + }, + { + "epoch": 0.05373, + "grad_norm": 0.8319189021739666, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 5373 + }, + { + "epoch": 0.05374, + "grad_norm": 0.7816087339025192, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 5374 + }, + { + "epoch": 0.05375, + "grad_norm": 0.6514305471977693, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 5375 + }, + { + "epoch": 0.05376, + "grad_norm": 0.6538341299728873, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 5376 + }, + { + "epoch": 0.05377, + "grad_norm": 0.6366241070912692, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 5377 + }, + { + "epoch": 0.05378, + "grad_norm": 0.6231860361185158, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 5378 + }, + { + "epoch": 0.05379, + "grad_norm": 0.533638527516239, + "learning_rate": 0.003, + "loss": 4.0825, + "step": 5379 + }, + { + "epoch": 0.0538, + "grad_norm": 0.48542200149838755, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 5380 + }, + { + "epoch": 0.05381, + "grad_norm": 0.5228170502401537, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 5381 + }, + { + "epoch": 0.05382, + "grad_norm": 0.5280377074688869, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 5382 + }, + { + "epoch": 0.05383, + "grad_norm": 0.5426640826775109, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 5383 + }, + { + "epoch": 0.05384, + "grad_norm": 0.5835591349541897, + "learning_rate": 0.003, + "loss": 4.055, + "step": 5384 + }, + { + "epoch": 0.05385, + "grad_norm": 0.5866500026854709, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5385 + }, + { + "epoch": 0.05386, + "grad_norm": 0.5292692111223407, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 5386 + }, + { + "epoch": 0.05387, + "grad_norm": 0.5091508168805366, + "learning_rate": 0.003, + "loss": 4.045, + "step": 5387 + }, + { + "epoch": 0.05388, + "grad_norm": 0.5185903920419926, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 5388 + }, + { + "epoch": 0.05389, + "grad_norm": 0.566044613877968, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 5389 + }, + { + "epoch": 0.0539, + "grad_norm": 0.6250568024796709, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 5390 + }, + { + "epoch": 0.05391, + "grad_norm": 0.7488803181264032, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 5391 + }, + { + "epoch": 0.05392, + "grad_norm": 0.8082433088962183, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 5392 + }, + { + "epoch": 0.05393, + "grad_norm": 0.8835905542137441, + "learning_rate": 0.003, + "loss": 4.087, + "step": 5393 + }, + { + "epoch": 0.05394, + "grad_norm": 0.8324896458248346, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 5394 + }, + { + "epoch": 0.05395, + "grad_norm": 0.8072118571774737, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 5395 + }, + { + "epoch": 0.05396, + "grad_norm": 0.8580443040580278, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 5396 + }, + { + "epoch": 0.05397, + "grad_norm": 0.8198480860002422, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 5397 + }, + { + "epoch": 0.05398, + "grad_norm": 0.7495581844872068, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 5398 + }, + { + "epoch": 0.05399, + "grad_norm": 0.7984076502516, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 5399 + }, + { + "epoch": 0.054, + "grad_norm": 0.7944711416213244, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 5400 + }, + { + "epoch": 0.05401, + "grad_norm": 0.8698186766840902, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 5401 + }, + { + "epoch": 0.05402, + "grad_norm": 0.8048452888050558, + "learning_rate": 0.003, + "loss": 4.0934, + "step": 5402 + }, + { + "epoch": 0.05403, + "grad_norm": 0.7736943487833503, + "learning_rate": 0.003, + "loss": 4.078, + "step": 5403 + }, + { + "epoch": 0.05404, + "grad_norm": 0.775806529410665, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 5404 + }, + { + "epoch": 0.05405, + "grad_norm": 0.7293596516972377, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 5405 + }, + { + "epoch": 0.05406, + "grad_norm": 0.7223031418481001, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 5406 + }, + { + "epoch": 0.05407, + "grad_norm": 0.7052254607628525, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5407 + }, + { + "epoch": 0.05408, + "grad_norm": 0.6648019848597027, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 5408 + }, + { + "epoch": 0.05409, + "grad_norm": 0.6052107146065081, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 5409 + }, + { + "epoch": 0.0541, + "grad_norm": 0.7090737500854986, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 5410 + }, + { + "epoch": 0.05411, + "grad_norm": 0.722770184553403, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 5411 + }, + { + "epoch": 0.05412, + "grad_norm": 0.6273640793589315, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 5412 + }, + { + "epoch": 0.05413, + "grad_norm": 0.5403618794176429, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 5413 + }, + { + "epoch": 0.05414, + "grad_norm": 0.5575823374669531, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 5414 + }, + { + "epoch": 0.05415, + "grad_norm": 0.5139509532641996, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 5415 + }, + { + "epoch": 0.05416, + "grad_norm": 0.5448807366875486, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 5416 + }, + { + "epoch": 0.05417, + "grad_norm": 0.604132696221083, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 5417 + }, + { + "epoch": 0.05418, + "grad_norm": 0.6324974519765385, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 5418 + }, + { + "epoch": 0.05419, + "grad_norm": 0.7703538383196891, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 5419 + }, + { + "epoch": 0.0542, + "grad_norm": 0.9906788006369472, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 5420 + }, + { + "epoch": 0.05421, + "grad_norm": 0.9496197458355968, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 5421 + }, + { + "epoch": 0.05422, + "grad_norm": 0.6546327416057268, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 5422 + }, + { + "epoch": 0.05423, + "grad_norm": 0.5631952440649028, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 5423 + }, + { + "epoch": 0.05424, + "grad_norm": 0.5504502197973413, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 5424 + }, + { + "epoch": 0.05425, + "grad_norm": 0.5566960170091257, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 5425 + }, + { + "epoch": 0.05426, + "grad_norm": 0.607720559432063, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 5426 + }, + { + "epoch": 0.05427, + "grad_norm": 0.6126109341012084, + "learning_rate": 0.003, + "loss": 4.047, + "step": 5427 + }, + { + "epoch": 0.05428, + "grad_norm": 0.656397792474732, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 5428 + }, + { + "epoch": 0.05429, + "grad_norm": 0.8022788262200853, + "learning_rate": 0.003, + "loss": 4.07, + "step": 5429 + }, + { + "epoch": 0.0543, + "grad_norm": 0.9207485656171693, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 5430 + }, + { + "epoch": 0.05431, + "grad_norm": 0.9680702788245453, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5431 + }, + { + "epoch": 0.05432, + "grad_norm": 0.7636811690867753, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 5432 + }, + { + "epoch": 0.05433, + "grad_norm": 0.75020059231116, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 5433 + }, + { + "epoch": 0.05434, + "grad_norm": 0.8374912672895445, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 5434 + }, + { + "epoch": 0.05435, + "grad_norm": 0.9435378327169933, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 5435 + }, + { + "epoch": 0.05436, + "grad_norm": 0.8282725342780659, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 5436 + }, + { + "epoch": 0.05437, + "grad_norm": 0.896848933807723, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 5437 + }, + { + "epoch": 0.05438, + "grad_norm": 0.7636460983483394, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5438 + }, + { + "epoch": 0.05439, + "grad_norm": 0.7199152628371309, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 5439 + }, + { + "epoch": 0.0544, + "grad_norm": 0.7576835280462306, + "learning_rate": 0.003, + "loss": 4.066, + "step": 5440 + }, + { + "epoch": 0.05441, + "grad_norm": 0.8268869352922588, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 5441 + }, + { + "epoch": 0.05442, + "grad_norm": 0.7747071753943117, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 5442 + }, + { + "epoch": 0.05443, + "grad_norm": 0.7487173504261683, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 5443 + }, + { + "epoch": 0.05444, + "grad_norm": 0.80340941944865, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 5444 + }, + { + "epoch": 0.05445, + "grad_norm": 0.8300830202017991, + "learning_rate": 0.003, + "loss": 4.067, + "step": 5445 + }, + { + "epoch": 0.05446, + "grad_norm": 0.9087289684175303, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 5446 + }, + { + "epoch": 0.05447, + "grad_norm": 0.7721205904558164, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 5447 + }, + { + "epoch": 0.05448, + "grad_norm": 0.7357954824313645, + "learning_rate": 0.003, + "loss": 4.0989, + "step": 5448 + }, + { + "epoch": 0.05449, + "grad_norm": 0.7688143904774357, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 5449 + }, + { + "epoch": 0.0545, + "grad_norm": 0.6945423364239631, + "learning_rate": 0.003, + "loss": 4.0826, + "step": 5450 + }, + { + "epoch": 0.05451, + "grad_norm": 0.7266432372664289, + "learning_rate": 0.003, + "loss": 4.1075, + "step": 5451 + }, + { + "epoch": 0.05452, + "grad_norm": 0.6959400388093868, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 5452 + }, + { + "epoch": 0.05453, + "grad_norm": 0.7338509549698786, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 5453 + }, + { + "epoch": 0.05454, + "grad_norm": 0.830272221140898, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 5454 + }, + { + "epoch": 0.05455, + "grad_norm": 0.8959114600057669, + "learning_rate": 0.003, + "loss": 4.1031, + "step": 5455 + }, + { + "epoch": 0.05456, + "grad_norm": 0.770273265069565, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 5456 + }, + { + "epoch": 0.05457, + "grad_norm": 0.5895267166539582, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 5457 + }, + { + "epoch": 0.05458, + "grad_norm": 0.5626729226097319, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 5458 + }, + { + "epoch": 0.05459, + "grad_norm": 0.5485457236655853, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 5459 + }, + { + "epoch": 0.0546, + "grad_norm": 0.6374638015061, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 5460 + }, + { + "epoch": 0.05461, + "grad_norm": 0.6068958847372934, + "learning_rate": 0.003, + "loss": 4.065, + "step": 5461 + }, + { + "epoch": 0.05462, + "grad_norm": 0.5705631257886107, + "learning_rate": 0.003, + "loss": 4.0907, + "step": 5462 + }, + { + "epoch": 0.05463, + "grad_norm": 0.5314389434674568, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 5463 + }, + { + "epoch": 0.05464, + "grad_norm": 0.4546542212981364, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 5464 + }, + { + "epoch": 0.05465, + "grad_norm": 0.44987284156069623, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 5465 + }, + { + "epoch": 0.05466, + "grad_norm": 0.4190529778827669, + "learning_rate": 0.003, + "loss": 4.0858, + "step": 5466 + }, + { + "epoch": 0.05467, + "grad_norm": 0.4075487508694463, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 5467 + }, + { + "epoch": 0.05468, + "grad_norm": 0.4552371739780703, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 5468 + }, + { + "epoch": 0.05469, + "grad_norm": 0.5169176268727639, + "learning_rate": 0.003, + "loss": 4.035, + "step": 5469 + }, + { + "epoch": 0.0547, + "grad_norm": 0.6052874958232901, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 5470 + }, + { + "epoch": 0.05471, + "grad_norm": 0.6631425678515417, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 5471 + }, + { + "epoch": 0.05472, + "grad_norm": 0.663426038753315, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 5472 + }, + { + "epoch": 0.05473, + "grad_norm": 0.6173943572099592, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 5473 + }, + { + "epoch": 0.05474, + "grad_norm": 0.7129030731687003, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 5474 + }, + { + "epoch": 0.05475, + "grad_norm": 0.7217056128323484, + "learning_rate": 0.003, + "loss": 4.104, + "step": 5475 + }, + { + "epoch": 0.05476, + "grad_norm": 0.6561107422234796, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 5476 + }, + { + "epoch": 0.05477, + "grad_norm": 0.5643842615119122, + "learning_rate": 0.003, + "loss": 4.05, + "step": 5477 + }, + { + "epoch": 0.05478, + "grad_norm": 0.6369600225968711, + "learning_rate": 0.003, + "loss": 4.052, + "step": 5478 + }, + { + "epoch": 0.05479, + "grad_norm": 0.6347388928762046, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 5479 + }, + { + "epoch": 0.0548, + "grad_norm": 0.6779156116702382, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 5480 + }, + { + "epoch": 0.05481, + "grad_norm": 0.7045357883837723, + "learning_rate": 0.003, + "loss": 4.055, + "step": 5481 + }, + { + "epoch": 0.05482, + "grad_norm": 0.7105703799872861, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 5482 + }, + { + "epoch": 0.05483, + "grad_norm": 0.7515185349964999, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 5483 + }, + { + "epoch": 0.05484, + "grad_norm": 0.8202805320368314, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 5484 + }, + { + "epoch": 0.05485, + "grad_norm": 0.8189060373320468, + "learning_rate": 0.003, + "loss": 4.0867, + "step": 5485 + }, + { + "epoch": 0.05486, + "grad_norm": 0.758494282329547, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 5486 + }, + { + "epoch": 0.05487, + "grad_norm": 0.6806571972005048, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 5487 + }, + { + "epoch": 0.05488, + "grad_norm": 0.5966215832767368, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 5488 + }, + { + "epoch": 0.05489, + "grad_norm": 0.607776747825417, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 5489 + }, + { + "epoch": 0.0549, + "grad_norm": 0.701683291319652, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 5490 + }, + { + "epoch": 0.05491, + "grad_norm": 0.75629761977149, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 5491 + }, + { + "epoch": 0.05492, + "grad_norm": 0.844985418807082, + "learning_rate": 0.003, + "loss": 4.0915, + "step": 5492 + }, + { + "epoch": 0.05493, + "grad_norm": 0.8526084163584916, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 5493 + }, + { + "epoch": 0.05494, + "grad_norm": 0.8107819734450187, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 5494 + }, + { + "epoch": 0.05495, + "grad_norm": 0.7607057721270047, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 5495 + }, + { + "epoch": 0.05496, + "grad_norm": 0.8388906009897551, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 5496 + }, + { + "epoch": 0.05497, + "grad_norm": 0.91460494667039, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 5497 + }, + { + "epoch": 0.05498, + "grad_norm": 0.945296870577468, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5498 + }, + { + "epoch": 0.05499, + "grad_norm": 1.0476883496226277, + "learning_rate": 0.003, + "loss": 4.064, + "step": 5499 + }, + { + "epoch": 0.055, + "grad_norm": 1.0023388971747953, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 5500 + }, + { + "epoch": 0.05501, + "grad_norm": 0.9433085125850219, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 5501 + }, + { + "epoch": 0.05502, + "grad_norm": 0.8876297258065301, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 5502 + }, + { + "epoch": 0.05503, + "grad_norm": 0.9909884438192074, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 5503 + }, + { + "epoch": 0.05504, + "grad_norm": 0.9707340599167434, + "learning_rate": 0.003, + "loss": 4.1299, + "step": 5504 + }, + { + "epoch": 0.05505, + "grad_norm": 0.8756524378282399, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 5505 + }, + { + "epoch": 0.05506, + "grad_norm": 0.7272545780415907, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5506 + }, + { + "epoch": 0.05507, + "grad_norm": 0.7874527455685988, + "learning_rate": 0.003, + "loss": 4.1099, + "step": 5507 + }, + { + "epoch": 0.05508, + "grad_norm": 0.8209374313714933, + "learning_rate": 0.003, + "loss": 4.0898, + "step": 5508 + }, + { + "epoch": 0.05509, + "grad_norm": 0.7326759926345884, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 5509 + }, + { + "epoch": 0.0551, + "grad_norm": 0.691254149676227, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 5510 + }, + { + "epoch": 0.05511, + "grad_norm": 0.6967434519895602, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 5511 + }, + { + "epoch": 0.05512, + "grad_norm": 0.6467429346973879, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 5512 + }, + { + "epoch": 0.05513, + "grad_norm": 0.669869560437162, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 5513 + }, + { + "epoch": 0.05514, + "grad_norm": 0.7530386220814562, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 5514 + }, + { + "epoch": 0.05515, + "grad_norm": 0.8371503755516563, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 5515 + }, + { + "epoch": 0.05516, + "grad_norm": 0.785733165964444, + "learning_rate": 0.003, + "loss": 4.101, + "step": 5516 + }, + { + "epoch": 0.05517, + "grad_norm": 0.7120097825295922, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 5517 + }, + { + "epoch": 0.05518, + "grad_norm": 0.6508817184729726, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 5518 + }, + { + "epoch": 0.05519, + "grad_norm": 0.5531063316573906, + "learning_rate": 0.003, + "loss": 4.0993, + "step": 5519 + }, + { + "epoch": 0.0552, + "grad_norm": 0.6684648251765309, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 5520 + }, + { + "epoch": 0.05521, + "grad_norm": 0.688751484997001, + "learning_rate": 0.003, + "loss": 4.086, + "step": 5521 + }, + { + "epoch": 0.05522, + "grad_norm": 0.7138340354058982, + "learning_rate": 0.003, + "loss": 4.1008, + "step": 5522 + }, + { + "epoch": 0.05523, + "grad_norm": 0.675194951468239, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 5523 + }, + { + "epoch": 0.05524, + "grad_norm": 0.6258706291369309, + "learning_rate": 0.003, + "loss": 4.086, + "step": 5524 + }, + { + "epoch": 0.05525, + "grad_norm": 0.5814483805535222, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 5525 + }, + { + "epoch": 0.05526, + "grad_norm": 0.6163771834824167, + "learning_rate": 0.003, + "loss": 4.056, + "step": 5526 + }, + { + "epoch": 0.05527, + "grad_norm": 0.6540769761398355, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 5527 + }, + { + "epoch": 0.05528, + "grad_norm": 0.6525009792233064, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 5528 + }, + { + "epoch": 0.05529, + "grad_norm": 0.5655882696183056, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 5529 + }, + { + "epoch": 0.0553, + "grad_norm": 0.6374973602849326, + "learning_rate": 0.003, + "loss": 4.088, + "step": 5530 + }, + { + "epoch": 0.05531, + "grad_norm": 0.6283116616405962, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 5531 + }, + { + "epoch": 0.05532, + "grad_norm": 0.6470334326128022, + "learning_rate": 0.003, + "loss": 4.058, + "step": 5532 + }, + { + "epoch": 0.05533, + "grad_norm": 0.7236889640577742, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 5533 + }, + { + "epoch": 0.05534, + "grad_norm": 0.5801930578534529, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 5534 + }, + { + "epoch": 0.05535, + "grad_norm": 0.6426749823546561, + "learning_rate": 0.003, + "loss": 4.077, + "step": 5535 + }, + { + "epoch": 0.05536, + "grad_norm": 0.6705890089831985, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 5536 + }, + { + "epoch": 0.05537, + "grad_norm": 0.6783865620861989, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 5537 + }, + { + "epoch": 0.05538, + "grad_norm": 0.5993810813354216, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 5538 + }, + { + "epoch": 0.05539, + "grad_norm": 0.5667840853403507, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 5539 + }, + { + "epoch": 0.0554, + "grad_norm": 0.5358452358784082, + "learning_rate": 0.003, + "loss": 4.1025, + "step": 5540 + }, + { + "epoch": 0.05541, + "grad_norm": 0.6289529704013197, + "learning_rate": 0.003, + "loss": 4.0801, + "step": 5541 + }, + { + "epoch": 0.05542, + "grad_norm": 0.7296330382057663, + "learning_rate": 0.003, + "loss": 4.054, + "step": 5542 + }, + { + "epoch": 0.05543, + "grad_norm": 0.6939656177429516, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 5543 + }, + { + "epoch": 0.05544, + "grad_norm": 0.6733504019956823, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 5544 + }, + { + "epoch": 0.05545, + "grad_norm": 0.6927445279737104, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5545 + }, + { + "epoch": 0.05546, + "grad_norm": 0.6955058635455645, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 5546 + }, + { + "epoch": 0.05547, + "grad_norm": 0.6563461644209156, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 5547 + }, + { + "epoch": 0.05548, + "grad_norm": 0.7362401231352, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 5548 + }, + { + "epoch": 0.05549, + "grad_norm": 0.7589337997661514, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 5549 + }, + { + "epoch": 0.0555, + "grad_norm": 0.7548421212644028, + "learning_rate": 0.003, + "loss": 4.065, + "step": 5550 + }, + { + "epoch": 0.05551, + "grad_norm": 0.7562161847167199, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 5551 + }, + { + "epoch": 0.05552, + "grad_norm": 0.8658842499828437, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 5552 + }, + { + "epoch": 0.05553, + "grad_norm": 0.8405862444034264, + "learning_rate": 0.003, + "loss": 4.0783, + "step": 5553 + }, + { + "epoch": 0.05554, + "grad_norm": 0.8240971044141429, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 5554 + }, + { + "epoch": 0.05555, + "grad_norm": 0.6367263324365597, + "learning_rate": 0.003, + "loss": 4.1014, + "step": 5555 + }, + { + "epoch": 0.05556, + "grad_norm": 0.7352704738956582, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 5556 + }, + { + "epoch": 0.05557, + "grad_norm": 0.7885125614922174, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 5557 + }, + { + "epoch": 0.05558, + "grad_norm": 0.8903240413174089, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 5558 + }, + { + "epoch": 0.05559, + "grad_norm": 0.9765998556816055, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 5559 + }, + { + "epoch": 0.0556, + "grad_norm": 0.9433676528582834, + "learning_rate": 0.003, + "loss": 4.1051, + "step": 5560 + }, + { + "epoch": 0.05561, + "grad_norm": 0.8900360027439277, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 5561 + }, + { + "epoch": 0.05562, + "grad_norm": 0.7590825133852874, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 5562 + }, + { + "epoch": 0.05563, + "grad_norm": 0.8451390434921501, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 5563 + }, + { + "epoch": 0.05564, + "grad_norm": 0.8793922166614564, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 5564 + }, + { + "epoch": 0.05565, + "grad_norm": 0.770394694573438, + "learning_rate": 0.003, + "loss": 4.1068, + "step": 5565 + }, + { + "epoch": 0.05566, + "grad_norm": 0.7843201026097724, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 5566 + }, + { + "epoch": 0.05567, + "grad_norm": 0.7026913170117073, + "learning_rate": 0.003, + "loss": 4.064, + "step": 5567 + }, + { + "epoch": 0.05568, + "grad_norm": 0.6494404677457177, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 5568 + }, + { + "epoch": 0.05569, + "grad_norm": 0.6537814093966875, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 5569 + }, + { + "epoch": 0.0557, + "grad_norm": 0.7193249290361016, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 5570 + }, + { + "epoch": 0.05571, + "grad_norm": 0.6890282691969432, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 5571 + }, + { + "epoch": 0.05572, + "grad_norm": 0.72963007161485, + "learning_rate": 0.003, + "loss": 4.053, + "step": 5572 + }, + { + "epoch": 0.05573, + "grad_norm": 0.6958158938732198, + "learning_rate": 0.003, + "loss": 4.0942, + "step": 5573 + }, + { + "epoch": 0.05574, + "grad_norm": 0.6950304962993357, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 5574 + }, + { + "epoch": 0.05575, + "grad_norm": 0.6274047587157795, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 5575 + }, + { + "epoch": 0.05576, + "grad_norm": 0.5903400005138694, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 5576 + }, + { + "epoch": 0.05577, + "grad_norm": 0.5603106980225232, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 5577 + }, + { + "epoch": 0.05578, + "grad_norm": 0.46084094938472203, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 5578 + }, + { + "epoch": 0.05579, + "grad_norm": 0.5083266631612878, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 5579 + }, + { + "epoch": 0.0558, + "grad_norm": 0.42489183930001295, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 5580 + }, + { + "epoch": 0.05581, + "grad_norm": 0.3970608524747849, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 5581 + }, + { + "epoch": 0.05582, + "grad_norm": 0.4194653899019576, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5582 + }, + { + "epoch": 0.05583, + "grad_norm": 0.483824723070654, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 5583 + }, + { + "epoch": 0.05584, + "grad_norm": 0.4948352087010697, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 5584 + }, + { + "epoch": 0.05585, + "grad_norm": 0.5113974503620272, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 5585 + }, + { + "epoch": 0.05586, + "grad_norm": 0.5695422218205397, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 5586 + }, + { + "epoch": 0.05587, + "grad_norm": 0.6667106487963592, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 5587 + }, + { + "epoch": 0.05588, + "grad_norm": 0.8258249918662851, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 5588 + }, + { + "epoch": 0.05589, + "grad_norm": 0.90071042553815, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 5589 + }, + { + "epoch": 0.0559, + "grad_norm": 0.8560348713578145, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 5590 + }, + { + "epoch": 0.05591, + "grad_norm": 0.7683195079809262, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 5591 + }, + { + "epoch": 0.05592, + "grad_norm": 0.7903574167337759, + "learning_rate": 0.003, + "loss": 4.078, + "step": 5592 + }, + { + "epoch": 0.05593, + "grad_norm": 0.8604483636054021, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 5593 + }, + { + "epoch": 0.05594, + "grad_norm": 0.8351496776023127, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 5594 + }, + { + "epoch": 0.05595, + "grad_norm": 0.7920951886213106, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 5595 + }, + { + "epoch": 0.05596, + "grad_norm": 0.7094805855794248, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 5596 + }, + { + "epoch": 0.05597, + "grad_norm": 0.7434590958806525, + "learning_rate": 0.003, + "loss": 4.0865, + "step": 5597 + }, + { + "epoch": 0.05598, + "grad_norm": 0.7520033822688419, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 5598 + }, + { + "epoch": 0.05599, + "grad_norm": 0.6825977668894181, + "learning_rate": 0.003, + "loss": 4.0922, + "step": 5599 + }, + { + "epoch": 0.056, + "grad_norm": 0.6362518925948988, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 5600 + }, + { + "epoch": 0.05601, + "grad_norm": 0.6100400047422697, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 5601 + }, + { + "epoch": 0.05602, + "grad_norm": 0.6052702711495338, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 5602 + }, + { + "epoch": 0.05603, + "grad_norm": 0.6545867437337354, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 5603 + }, + { + "epoch": 0.05604, + "grad_norm": 0.7680933187578438, + "learning_rate": 0.003, + "loss": 4.0901, + "step": 5604 + }, + { + "epoch": 0.05605, + "grad_norm": 0.6885660570553119, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 5605 + }, + { + "epoch": 0.05606, + "grad_norm": 0.5891536372420663, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 5606 + }, + { + "epoch": 0.05607, + "grad_norm": 0.5613909280920695, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 5607 + }, + { + "epoch": 0.05608, + "grad_norm": 0.6289014307164557, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 5608 + }, + { + "epoch": 0.05609, + "grad_norm": 0.7124201215130834, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 5609 + }, + { + "epoch": 0.0561, + "grad_norm": 0.951179547223622, + "learning_rate": 0.003, + "loss": 4.104, + "step": 5610 + }, + { + "epoch": 0.05611, + "grad_norm": 1.2141657499969052, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 5611 + }, + { + "epoch": 0.05612, + "grad_norm": 0.6940751388795054, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5612 + }, + { + "epoch": 0.05613, + "grad_norm": 0.7380532082881979, + "learning_rate": 0.003, + "loss": 4.0888, + "step": 5613 + }, + { + "epoch": 0.05614, + "grad_norm": 0.97244244686036, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5614 + }, + { + "epoch": 0.05615, + "grad_norm": 0.8931592877767203, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 5615 + }, + { + "epoch": 0.05616, + "grad_norm": 0.7569693341780283, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 5616 + }, + { + "epoch": 0.05617, + "grad_norm": 0.7046851234925131, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 5617 + }, + { + "epoch": 0.05618, + "grad_norm": 0.689509182412502, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 5618 + }, + { + "epoch": 0.05619, + "grad_norm": 0.6561207277931308, + "learning_rate": 0.003, + "loss": 4.1164, + "step": 5619 + }, + { + "epoch": 0.0562, + "grad_norm": 0.7371491583367267, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 5620 + }, + { + "epoch": 0.05621, + "grad_norm": 0.69236518008264, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 5621 + }, + { + "epoch": 0.05622, + "grad_norm": 0.5584745691424656, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 5622 + }, + { + "epoch": 0.05623, + "grad_norm": 0.5671879118025603, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 5623 + }, + { + "epoch": 0.05624, + "grad_norm": 0.552371888326441, + "learning_rate": 0.003, + "loss": 4.0935, + "step": 5624 + }, + { + "epoch": 0.05625, + "grad_norm": 0.5365036274740907, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 5625 + }, + { + "epoch": 0.05626, + "grad_norm": 0.6571159205004364, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 5626 + }, + { + "epoch": 0.05627, + "grad_norm": 0.6941213725803296, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 5627 + }, + { + "epoch": 0.05628, + "grad_norm": 0.6795244680654942, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 5628 + }, + { + "epoch": 0.05629, + "grad_norm": 0.6006894847038436, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 5629 + }, + { + "epoch": 0.0563, + "grad_norm": 0.6575579372346501, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 5630 + }, + { + "epoch": 0.05631, + "grad_norm": 0.6084355339983706, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 5631 + }, + { + "epoch": 0.05632, + "grad_norm": 0.5424277795317403, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 5632 + }, + { + "epoch": 0.05633, + "grad_norm": 0.5385341807097478, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 5633 + }, + { + "epoch": 0.05634, + "grad_norm": 0.49971746540181633, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 5634 + }, + { + "epoch": 0.05635, + "grad_norm": 0.5299147453304455, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 5635 + }, + { + "epoch": 0.05636, + "grad_norm": 0.5039363983889078, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 5636 + }, + { + "epoch": 0.05637, + "grad_norm": 0.4882376547730458, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 5637 + }, + { + "epoch": 0.05638, + "grad_norm": 0.4580062891633645, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 5638 + }, + { + "epoch": 0.05639, + "grad_norm": 0.4704096199740475, + "learning_rate": 0.003, + "loss": 4.052, + "step": 5639 + }, + { + "epoch": 0.0564, + "grad_norm": 0.486070922650092, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 5640 + }, + { + "epoch": 0.05641, + "grad_norm": 0.5566858208536091, + "learning_rate": 0.003, + "loss": 4.012, + "step": 5641 + }, + { + "epoch": 0.05642, + "grad_norm": 0.6760282088966324, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 5642 + }, + { + "epoch": 0.05643, + "grad_norm": 0.8460582422985491, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 5643 + }, + { + "epoch": 0.05644, + "grad_norm": 0.8914752527601362, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 5644 + }, + { + "epoch": 0.05645, + "grad_norm": 0.8258813975499965, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 5645 + }, + { + "epoch": 0.05646, + "grad_norm": 0.6648514978264412, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 5646 + }, + { + "epoch": 0.05647, + "grad_norm": 0.7092276697249866, + "learning_rate": 0.003, + "loss": 4.062, + "step": 5647 + }, + { + "epoch": 0.05648, + "grad_norm": 0.8792523670045349, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 5648 + }, + { + "epoch": 0.05649, + "grad_norm": 0.9152742032499014, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 5649 + }, + { + "epoch": 0.0565, + "grad_norm": 0.8966973101090298, + "learning_rate": 0.003, + "loss": 4.0945, + "step": 5650 + }, + { + "epoch": 0.05651, + "grad_norm": 0.8893476270493953, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5651 + }, + { + "epoch": 0.05652, + "grad_norm": 0.9629938774506874, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 5652 + }, + { + "epoch": 0.05653, + "grad_norm": 0.9061824465147015, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 5653 + }, + { + "epoch": 0.05654, + "grad_norm": 0.8901092787607353, + "learning_rate": 0.003, + "loss": 4.1013, + "step": 5654 + }, + { + "epoch": 0.05655, + "grad_norm": 0.7740607360295262, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 5655 + }, + { + "epoch": 0.05656, + "grad_norm": 0.8186718478714081, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 5656 + }, + { + "epoch": 0.05657, + "grad_norm": 0.7063959683586212, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 5657 + }, + { + "epoch": 0.05658, + "grad_norm": 0.6005701617883997, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 5658 + }, + { + "epoch": 0.05659, + "grad_norm": 0.6142160635884345, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 5659 + }, + { + "epoch": 0.0566, + "grad_norm": 0.6198082232727031, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 5660 + }, + { + "epoch": 0.05661, + "grad_norm": 0.62402435922723, + "learning_rate": 0.003, + "loss": 4.1114, + "step": 5661 + }, + { + "epoch": 0.05662, + "grad_norm": 0.6895730319307702, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 5662 + }, + { + "epoch": 0.05663, + "grad_norm": 0.8919985384259705, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 5663 + }, + { + "epoch": 0.05664, + "grad_norm": 1.1444796751268222, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 5664 + }, + { + "epoch": 0.05665, + "grad_norm": 0.8301844467961512, + "learning_rate": 0.003, + "loss": 4.072, + "step": 5665 + }, + { + "epoch": 0.05666, + "grad_norm": 0.8094391535774507, + "learning_rate": 0.003, + "loss": 4.0853, + "step": 5666 + }, + { + "epoch": 0.05667, + "grad_norm": 0.7225459522645806, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 5667 + }, + { + "epoch": 0.05668, + "grad_norm": 0.8987834331424942, + "learning_rate": 0.003, + "loss": 4.1015, + "step": 5668 + }, + { + "epoch": 0.05669, + "grad_norm": 0.9694523129694718, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 5669 + }, + { + "epoch": 0.0567, + "grad_norm": 0.9344299357422136, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 5670 + }, + { + "epoch": 0.05671, + "grad_norm": 0.8587572578554192, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 5671 + }, + { + "epoch": 0.05672, + "grad_norm": 0.7310928015057686, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 5672 + }, + { + "epoch": 0.05673, + "grad_norm": 0.7522905825164684, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 5673 + }, + { + "epoch": 0.05674, + "grad_norm": 0.845571489718161, + "learning_rate": 0.003, + "loss": 4.0874, + "step": 5674 + }, + { + "epoch": 0.05675, + "grad_norm": 0.7229345745163799, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 5675 + }, + { + "epoch": 0.05676, + "grad_norm": 0.6526625416642365, + "learning_rate": 0.003, + "loss": 4.0819, + "step": 5676 + }, + { + "epoch": 0.05677, + "grad_norm": 0.6922075811570839, + "learning_rate": 0.003, + "loss": 4.0775, + "step": 5677 + }, + { + "epoch": 0.05678, + "grad_norm": 0.7204670253283796, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 5678 + }, + { + "epoch": 0.05679, + "grad_norm": 0.6809757046425243, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 5679 + }, + { + "epoch": 0.0568, + "grad_norm": 0.742458069868555, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 5680 + }, + { + "epoch": 0.05681, + "grad_norm": 0.8177209305040204, + "learning_rate": 0.003, + "loss": 4.068, + "step": 5681 + }, + { + "epoch": 0.05682, + "grad_norm": 0.8147837779366496, + "learning_rate": 0.003, + "loss": 4.0849, + "step": 5682 + }, + { + "epoch": 0.05683, + "grad_norm": 0.7660366460627734, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 5683 + }, + { + "epoch": 0.05684, + "grad_norm": 0.7546017550926526, + "learning_rate": 0.003, + "loss": 4.051, + "step": 5684 + }, + { + "epoch": 0.05685, + "grad_norm": 0.6828842646118928, + "learning_rate": 0.003, + "loss": 4.0897, + "step": 5685 + }, + { + "epoch": 0.05686, + "grad_norm": 0.5809707443334167, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 5686 + }, + { + "epoch": 0.05687, + "grad_norm": 0.5542346944098858, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 5687 + }, + { + "epoch": 0.05688, + "grad_norm": 0.4546810561878202, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 5688 + }, + { + "epoch": 0.05689, + "grad_norm": 0.5164104402020321, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 5689 + }, + { + "epoch": 0.0569, + "grad_norm": 0.5089305320959258, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 5690 + }, + { + "epoch": 0.05691, + "grad_norm": 0.60261214739639, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 5691 + }, + { + "epoch": 0.05692, + "grad_norm": 0.6336380239827692, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 5692 + }, + { + "epoch": 0.05693, + "grad_norm": 0.6583614978614867, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 5693 + }, + { + "epoch": 0.05694, + "grad_norm": 0.7169533494863379, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 5694 + }, + { + "epoch": 0.05695, + "grad_norm": 0.738594832749471, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 5695 + }, + { + "epoch": 0.05696, + "grad_norm": 0.7202129713060067, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 5696 + }, + { + "epoch": 0.05697, + "grad_norm": 0.7499939413204101, + "learning_rate": 0.003, + "loss": 4.0913, + "step": 5697 + }, + { + "epoch": 0.05698, + "grad_norm": 0.6755444397546944, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 5698 + }, + { + "epoch": 0.05699, + "grad_norm": 0.6199096715557924, + "learning_rate": 0.003, + "loss": 4.068, + "step": 5699 + }, + { + "epoch": 0.057, + "grad_norm": 0.6639713701355346, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 5700 + }, + { + "epoch": 0.05701, + "grad_norm": 0.6425007907614158, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 5701 + }, + { + "epoch": 0.05702, + "grad_norm": 0.5838816978825616, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 5702 + }, + { + "epoch": 0.05703, + "grad_norm": 0.5377038137838553, + "learning_rate": 0.003, + "loss": 4.0923, + "step": 5703 + }, + { + "epoch": 0.05704, + "grad_norm": 0.607662327343465, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 5704 + }, + { + "epoch": 0.05705, + "grad_norm": 0.6803269298552374, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 5705 + }, + { + "epoch": 0.05706, + "grad_norm": 0.6895840283045747, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 5706 + }, + { + "epoch": 0.05707, + "grad_norm": 0.7195664536330343, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 5707 + }, + { + "epoch": 0.05708, + "grad_norm": 0.820862703149936, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 5708 + }, + { + "epoch": 0.05709, + "grad_norm": 0.8698030253514494, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 5709 + }, + { + "epoch": 0.0571, + "grad_norm": 0.7605059293178158, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 5710 + }, + { + "epoch": 0.05711, + "grad_norm": 0.7781167896251296, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 5711 + }, + { + "epoch": 0.05712, + "grad_norm": 0.8601041593025995, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 5712 + }, + { + "epoch": 0.05713, + "grad_norm": 0.8344672359527461, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 5713 + }, + { + "epoch": 0.05714, + "grad_norm": 0.7601499741076896, + "learning_rate": 0.003, + "loss": 4.1135, + "step": 5714 + }, + { + "epoch": 0.05715, + "grad_norm": 0.8441678618430742, + "learning_rate": 0.003, + "loss": 4.1154, + "step": 5715 + }, + { + "epoch": 0.05716, + "grad_norm": 0.8822917554571973, + "learning_rate": 0.003, + "loss": 4.0869, + "step": 5716 + }, + { + "epoch": 0.05717, + "grad_norm": 1.0190427293285422, + "learning_rate": 0.003, + "loss": 4.1073, + "step": 5717 + }, + { + "epoch": 0.05718, + "grad_norm": 0.9170096567475269, + "learning_rate": 0.003, + "loss": 4.087, + "step": 5718 + }, + { + "epoch": 0.05719, + "grad_norm": 0.8354320541155821, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 5719 + }, + { + "epoch": 0.0572, + "grad_norm": 0.8140001760636584, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 5720 + }, + { + "epoch": 0.05721, + "grad_norm": 0.6813271173590675, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 5721 + }, + { + "epoch": 0.05722, + "grad_norm": 0.711573635108766, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 5722 + }, + { + "epoch": 0.05723, + "grad_norm": 0.7714802549025447, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 5723 + }, + { + "epoch": 0.05724, + "grad_norm": 0.7453698530466337, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 5724 + }, + { + "epoch": 0.05725, + "grad_norm": 0.673683276534276, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 5725 + }, + { + "epoch": 0.05726, + "grad_norm": 0.5698071869044569, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 5726 + }, + { + "epoch": 0.05727, + "grad_norm": 0.573801031558788, + "learning_rate": 0.003, + "loss": 4.05, + "step": 5727 + }, + { + "epoch": 0.05728, + "grad_norm": 0.5050862350710272, + "learning_rate": 0.003, + "loss": 4.081, + "step": 5728 + }, + { + "epoch": 0.05729, + "grad_norm": 0.5285683446564919, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 5729 + }, + { + "epoch": 0.0573, + "grad_norm": 0.5703395179866583, + "learning_rate": 0.003, + "loss": 4.03, + "step": 5730 + }, + { + "epoch": 0.05731, + "grad_norm": 0.5702995299643973, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 5731 + }, + { + "epoch": 0.05732, + "grad_norm": 0.5179190738895529, + "learning_rate": 0.003, + "loss": 4.075, + "step": 5732 + }, + { + "epoch": 0.05733, + "grad_norm": 0.4916230329932676, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 5733 + }, + { + "epoch": 0.05734, + "grad_norm": 0.5328851607616292, + "learning_rate": 0.003, + "loss": 4.0791, + "step": 5734 + }, + { + "epoch": 0.05735, + "grad_norm": 0.6325713342801266, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 5735 + }, + { + "epoch": 0.05736, + "grad_norm": 0.6523496013789525, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 5736 + }, + { + "epoch": 0.05737, + "grad_norm": 0.6912926413091381, + "learning_rate": 0.003, + "loss": 4.063, + "step": 5737 + }, + { + "epoch": 0.05738, + "grad_norm": 0.6737913453654937, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 5738 + }, + { + "epoch": 0.05739, + "grad_norm": 0.6785460550757706, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 5739 + }, + { + "epoch": 0.0574, + "grad_norm": 0.6669417426280696, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 5740 + }, + { + "epoch": 0.05741, + "grad_norm": 0.7212894209615361, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 5741 + }, + { + "epoch": 0.05742, + "grad_norm": 0.7237035365123983, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 5742 + }, + { + "epoch": 0.05743, + "grad_norm": 0.7201919416611785, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 5743 + }, + { + "epoch": 0.05744, + "grad_norm": 0.705042833557332, + "learning_rate": 0.003, + "loss": 4.064, + "step": 5744 + }, + { + "epoch": 0.05745, + "grad_norm": 0.6778025756186739, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 5745 + }, + { + "epoch": 0.05746, + "grad_norm": 0.632507164539623, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 5746 + }, + { + "epoch": 0.05747, + "grad_norm": 0.6315079747018355, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 5747 + }, + { + "epoch": 0.05748, + "grad_norm": 0.6392421677120411, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 5748 + }, + { + "epoch": 0.05749, + "grad_norm": 0.6332077090837185, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 5749 + }, + { + "epoch": 0.0575, + "grad_norm": 0.6917445338918489, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 5750 + }, + { + "epoch": 0.05751, + "grad_norm": 0.7111256398957205, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 5751 + }, + { + "epoch": 0.05752, + "grad_norm": 0.733099136744556, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 5752 + }, + { + "epoch": 0.05753, + "grad_norm": 0.7535131925083675, + "learning_rate": 0.003, + "loss": 4.062, + "step": 5753 + }, + { + "epoch": 0.05754, + "grad_norm": 0.7681075709584388, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 5754 + }, + { + "epoch": 0.05755, + "grad_norm": 0.8104557508845512, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 5755 + }, + { + "epoch": 0.05756, + "grad_norm": 0.8112749372660343, + "learning_rate": 0.003, + "loss": 4.0875, + "step": 5756 + }, + { + "epoch": 0.05757, + "grad_norm": 0.8106833334438651, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 5757 + }, + { + "epoch": 0.05758, + "grad_norm": 0.7610150788821725, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 5758 + }, + { + "epoch": 0.05759, + "grad_norm": 0.8135765127070101, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 5759 + }, + { + "epoch": 0.0576, + "grad_norm": 0.8194213965225279, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 5760 + }, + { + "epoch": 0.05761, + "grad_norm": 0.8139812410732657, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 5761 + }, + { + "epoch": 0.05762, + "grad_norm": 0.8843151532693655, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 5762 + }, + { + "epoch": 0.05763, + "grad_norm": 0.8306876011572301, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 5763 + }, + { + "epoch": 0.05764, + "grad_norm": 0.7579430809940337, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 5764 + }, + { + "epoch": 0.05765, + "grad_norm": 0.7310194060956013, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 5765 + }, + { + "epoch": 0.05766, + "grad_norm": 0.7047631430819997, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 5766 + }, + { + "epoch": 0.05767, + "grad_norm": 0.6161096434781133, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 5767 + }, + { + "epoch": 0.05768, + "grad_norm": 0.6143178405110209, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 5768 + }, + { + "epoch": 0.05769, + "grad_norm": 0.6703890316086005, + "learning_rate": 0.003, + "loss": 4.065, + "step": 5769 + }, + { + "epoch": 0.0577, + "grad_norm": 0.5719583002011325, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 5770 + }, + { + "epoch": 0.05771, + "grad_norm": 0.5040515300999273, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 5771 + }, + { + "epoch": 0.05772, + "grad_norm": 0.5635605843131573, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 5772 + }, + { + "epoch": 0.05773, + "grad_norm": 0.6741437851394501, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 5773 + }, + { + "epoch": 0.05774, + "grad_norm": 0.7765249979428243, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 5774 + }, + { + "epoch": 0.05775, + "grad_norm": 0.8127508904768855, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 5775 + }, + { + "epoch": 0.05776, + "grad_norm": 0.7812716481216692, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 5776 + }, + { + "epoch": 0.05777, + "grad_norm": 0.6858760794105866, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 5777 + }, + { + "epoch": 0.05778, + "grad_norm": 0.659643423724406, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 5778 + }, + { + "epoch": 0.05779, + "grad_norm": 0.706185252721032, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 5779 + }, + { + "epoch": 0.0578, + "grad_norm": 0.7249107051990246, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 5780 + }, + { + "epoch": 0.05781, + "grad_norm": 0.7234884885920239, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 5781 + }, + { + "epoch": 0.05782, + "grad_norm": 0.8553675925897402, + "learning_rate": 0.003, + "loss": 4.0891, + "step": 5782 + }, + { + "epoch": 0.05783, + "grad_norm": 0.9652582024674506, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 5783 + }, + { + "epoch": 0.05784, + "grad_norm": 0.9670558767951674, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 5784 + }, + { + "epoch": 0.05785, + "grad_norm": 0.9250522293158245, + "learning_rate": 0.003, + "loss": 4.0938, + "step": 5785 + }, + { + "epoch": 0.05786, + "grad_norm": 0.9363557161068089, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 5786 + }, + { + "epoch": 0.05787, + "grad_norm": 0.8316816776816057, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 5787 + }, + { + "epoch": 0.05788, + "grad_norm": 0.7322883934985114, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 5788 + }, + { + "epoch": 0.05789, + "grad_norm": 0.6505428962349837, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 5789 + }, + { + "epoch": 0.0579, + "grad_norm": 0.5846888750424312, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 5790 + }, + { + "epoch": 0.05791, + "grad_norm": 0.6183049082327203, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 5791 + }, + { + "epoch": 0.05792, + "grad_norm": 0.657859450929066, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 5792 + }, + { + "epoch": 0.05793, + "grad_norm": 0.7376159706764901, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 5793 + }, + { + "epoch": 0.05794, + "grad_norm": 0.6985076860125068, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 5794 + }, + { + "epoch": 0.05795, + "grad_norm": 0.6963193535211363, + "learning_rate": 0.003, + "loss": 4.1021, + "step": 5795 + }, + { + "epoch": 0.05796, + "grad_norm": 0.6674441676496599, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 5796 + }, + { + "epoch": 0.05797, + "grad_norm": 0.6952392913991753, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 5797 + }, + { + "epoch": 0.05798, + "grad_norm": 0.779313959172018, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 5798 + }, + { + "epoch": 0.05799, + "grad_norm": 0.8209733395524956, + "learning_rate": 0.003, + "loss": 4.085, + "step": 5799 + }, + { + "epoch": 0.058, + "grad_norm": 0.76695749860437, + "learning_rate": 0.003, + "loss": 4.0828, + "step": 5800 + }, + { + "epoch": 0.05801, + "grad_norm": 0.7070077833676575, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 5801 + }, + { + "epoch": 0.05802, + "grad_norm": 0.7058401451049405, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 5802 + }, + { + "epoch": 0.05803, + "grad_norm": 0.8271693946816497, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 5803 + }, + { + "epoch": 0.05804, + "grad_norm": 0.9163297011270649, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 5804 + }, + { + "epoch": 0.05805, + "grad_norm": 0.8589467219917905, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 5805 + }, + { + "epoch": 0.05806, + "grad_norm": 0.750861527421322, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 5806 + }, + { + "epoch": 0.05807, + "grad_norm": 0.6587925706080741, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 5807 + }, + { + "epoch": 0.05808, + "grad_norm": 0.6941645162755149, + "learning_rate": 0.003, + "loss": 4.0933, + "step": 5808 + }, + { + "epoch": 0.05809, + "grad_norm": 0.7583175476196371, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 5809 + }, + { + "epoch": 0.0581, + "grad_norm": 0.7798275594063941, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 5810 + }, + { + "epoch": 0.05811, + "grad_norm": 0.6885950071240249, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 5811 + }, + { + "epoch": 0.05812, + "grad_norm": 0.5691132785937718, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 5812 + }, + { + "epoch": 0.05813, + "grad_norm": 0.5666186722201968, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 5813 + }, + { + "epoch": 0.05814, + "grad_norm": 0.6393959902689906, + "learning_rate": 0.003, + "loss": 4.049, + "step": 5814 + }, + { + "epoch": 0.05815, + "grad_norm": 0.6120921892424449, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 5815 + }, + { + "epoch": 0.05816, + "grad_norm": 0.5154885435633421, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 5816 + }, + { + "epoch": 0.05817, + "grad_norm": 0.49648287725301343, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5817 + }, + { + "epoch": 0.05818, + "grad_norm": 0.5914995159012274, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 5818 + }, + { + "epoch": 0.05819, + "grad_norm": 0.6135319781218952, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 5819 + }, + { + "epoch": 0.0582, + "grad_norm": 0.6038118783805694, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 5820 + }, + { + "epoch": 0.05821, + "grad_norm": 0.6014573037304078, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 5821 + }, + { + "epoch": 0.05822, + "grad_norm": 0.6799471703459983, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 5822 + }, + { + "epoch": 0.05823, + "grad_norm": 0.7750497513963867, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 5823 + }, + { + "epoch": 0.05824, + "grad_norm": 0.8645783403607199, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 5824 + }, + { + "epoch": 0.05825, + "grad_norm": 0.8020432838282351, + "learning_rate": 0.003, + "loss": 4.0998, + "step": 5825 + }, + { + "epoch": 0.05826, + "grad_norm": 0.7479187108153651, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 5826 + }, + { + "epoch": 0.05827, + "grad_norm": 0.922570834138522, + "learning_rate": 0.003, + "loss": 4.0805, + "step": 5827 + }, + { + "epoch": 0.05828, + "grad_norm": 0.8721366088542801, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 5828 + }, + { + "epoch": 0.05829, + "grad_norm": 0.8298549971351495, + "learning_rate": 0.003, + "loss": 4.079, + "step": 5829 + }, + { + "epoch": 0.0583, + "grad_norm": 0.7226182073363325, + "learning_rate": 0.003, + "loss": 4.021, + "step": 5830 + }, + { + "epoch": 0.05831, + "grad_norm": 0.6296628264268008, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 5831 + }, + { + "epoch": 0.05832, + "grad_norm": 0.6345002594548288, + "learning_rate": 0.003, + "loss": 4.083, + "step": 5832 + }, + { + "epoch": 0.05833, + "grad_norm": 0.6320741347722838, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 5833 + }, + { + "epoch": 0.05834, + "grad_norm": 0.6450712979361416, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 5834 + }, + { + "epoch": 0.05835, + "grad_norm": 0.6344142467666736, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 5835 + }, + { + "epoch": 0.05836, + "grad_norm": 0.6080887792240663, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 5836 + }, + { + "epoch": 0.05837, + "grad_norm": 0.6076793697940542, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 5837 + }, + { + "epoch": 0.05838, + "grad_norm": 0.637467792715844, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 5838 + }, + { + "epoch": 0.05839, + "grad_norm": 0.6458613459999138, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 5839 + }, + { + "epoch": 0.0584, + "grad_norm": 0.7148820243831299, + "learning_rate": 0.003, + "loss": 4.093, + "step": 5840 + }, + { + "epoch": 0.05841, + "grad_norm": 0.7676231216241524, + "learning_rate": 0.003, + "loss": 4.047, + "step": 5841 + }, + { + "epoch": 0.05842, + "grad_norm": 0.8205428067162354, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 5842 + }, + { + "epoch": 0.05843, + "grad_norm": 1.0771960660131323, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 5843 + }, + { + "epoch": 0.05844, + "grad_norm": 0.9387802597659208, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 5844 + }, + { + "epoch": 0.05845, + "grad_norm": 0.8429024245889917, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 5845 + }, + { + "epoch": 0.05846, + "grad_norm": 0.7667982359034143, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 5846 + }, + { + "epoch": 0.05847, + "grad_norm": 0.7956142374590347, + "learning_rate": 0.003, + "loss": 4.0912, + "step": 5847 + }, + { + "epoch": 0.05848, + "grad_norm": 0.7153526972896489, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 5848 + }, + { + "epoch": 0.05849, + "grad_norm": 0.7467334437285761, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 5849 + }, + { + "epoch": 0.0585, + "grad_norm": 0.7678084289915316, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 5850 + }, + { + "epoch": 0.05851, + "grad_norm": 0.734449947111904, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5851 + }, + { + "epoch": 0.05852, + "grad_norm": 0.6684903457106877, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 5852 + }, + { + "epoch": 0.05853, + "grad_norm": 0.6957263132179012, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 5853 + }, + { + "epoch": 0.05854, + "grad_norm": 0.6530151851046444, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 5854 + }, + { + "epoch": 0.05855, + "grad_norm": 0.677729588329254, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 5855 + }, + { + "epoch": 0.05856, + "grad_norm": 0.6845650880530647, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 5856 + }, + { + "epoch": 0.05857, + "grad_norm": 0.6210217858102421, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 5857 + }, + { + "epoch": 0.05858, + "grad_norm": 0.5887625484236905, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 5858 + }, + { + "epoch": 0.05859, + "grad_norm": 0.5995633392458275, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 5859 + }, + { + "epoch": 0.0586, + "grad_norm": 0.6361868233789073, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 5860 + }, + { + "epoch": 0.05861, + "grad_norm": 0.6247628268207585, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 5861 + }, + { + "epoch": 0.05862, + "grad_norm": 0.6084163914913894, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 5862 + }, + { + "epoch": 0.05863, + "grad_norm": 0.6208357238797572, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 5863 + }, + { + "epoch": 0.05864, + "grad_norm": 0.5977781209261245, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 5864 + }, + { + "epoch": 0.05865, + "grad_norm": 0.5065551415288274, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 5865 + }, + { + "epoch": 0.05866, + "grad_norm": 0.4436863207086825, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 5866 + }, + { + "epoch": 0.05867, + "grad_norm": 0.5409720897487795, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 5867 + }, + { + "epoch": 0.05868, + "grad_norm": 0.5705548298153398, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 5868 + }, + { + "epoch": 0.05869, + "grad_norm": 0.5954260409269247, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 5869 + }, + { + "epoch": 0.0587, + "grad_norm": 0.6904373255294518, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 5870 + }, + { + "epoch": 0.05871, + "grad_norm": 0.8411981292318947, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 5871 + }, + { + "epoch": 0.05872, + "grad_norm": 0.9532220603407672, + "learning_rate": 0.003, + "loss": 4.058, + "step": 5872 + }, + { + "epoch": 0.05873, + "grad_norm": 1.0465859756942064, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 5873 + }, + { + "epoch": 0.05874, + "grad_norm": 0.9920564472472155, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 5874 + }, + { + "epoch": 0.05875, + "grad_norm": 0.9386053242956398, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 5875 + }, + { + "epoch": 0.05876, + "grad_norm": 0.7648842939285814, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 5876 + }, + { + "epoch": 0.05877, + "grad_norm": 0.7445157626804496, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 5877 + }, + { + "epoch": 0.05878, + "grad_norm": 0.7050978526136601, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 5878 + }, + { + "epoch": 0.05879, + "grad_norm": 0.7835532743980347, + "learning_rate": 0.003, + "loss": 4.0997, + "step": 5879 + }, + { + "epoch": 0.0588, + "grad_norm": 0.8102522022154618, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 5880 + }, + { + "epoch": 0.05881, + "grad_norm": 0.7910503707023705, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 5881 + }, + { + "epoch": 0.05882, + "grad_norm": 0.6899769773525232, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 5882 + }, + { + "epoch": 0.05883, + "grad_norm": 0.7752130046286964, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 5883 + }, + { + "epoch": 0.05884, + "grad_norm": 0.8935725320144741, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 5884 + }, + { + "epoch": 0.05885, + "grad_norm": 0.8146747953936463, + "learning_rate": 0.003, + "loss": 4.0846, + "step": 5885 + }, + { + "epoch": 0.05886, + "grad_norm": 0.775828308904284, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 5886 + }, + { + "epoch": 0.05887, + "grad_norm": 0.7316953610323731, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 5887 + }, + { + "epoch": 0.05888, + "grad_norm": 0.677889133906649, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 5888 + }, + { + "epoch": 0.05889, + "grad_norm": 0.6675629643020854, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 5889 + }, + { + "epoch": 0.0589, + "grad_norm": 0.6327025137591554, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 5890 + }, + { + "epoch": 0.05891, + "grad_norm": 0.6498885114201365, + "learning_rate": 0.003, + "loss": 4.058, + "step": 5891 + }, + { + "epoch": 0.05892, + "grad_norm": 0.5873315221313729, + "learning_rate": 0.003, + "loss": 4.0884, + "step": 5892 + }, + { + "epoch": 0.05893, + "grad_norm": 0.5352780508095509, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 5893 + }, + { + "epoch": 0.05894, + "grad_norm": 0.568754269018644, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 5894 + }, + { + "epoch": 0.05895, + "grad_norm": 0.6442606742946965, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 5895 + }, + { + "epoch": 0.05896, + "grad_norm": 0.6399954066581772, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 5896 + }, + { + "epoch": 0.05897, + "grad_norm": 0.6833153752522355, + "learning_rate": 0.003, + "loss": 4.083, + "step": 5897 + }, + { + "epoch": 0.05898, + "grad_norm": 0.6851863763931254, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 5898 + }, + { + "epoch": 0.05899, + "grad_norm": 0.7133533626300476, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 5899 + }, + { + "epoch": 0.059, + "grad_norm": 0.740241979832513, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 5900 + }, + { + "epoch": 0.05901, + "grad_norm": 0.7052191645378342, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 5901 + }, + { + "epoch": 0.05902, + "grad_norm": 0.7885763407484976, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 5902 + }, + { + "epoch": 0.05903, + "grad_norm": 0.8452425085652833, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 5903 + }, + { + "epoch": 0.05904, + "grad_norm": 0.8532064303149844, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 5904 + }, + { + "epoch": 0.05905, + "grad_norm": 0.9770065933311903, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 5905 + }, + { + "epoch": 0.05906, + "grad_norm": 0.9361517713279305, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 5906 + }, + { + "epoch": 0.05907, + "grad_norm": 0.7109817203594382, + "learning_rate": 0.003, + "loss": 4.054, + "step": 5907 + }, + { + "epoch": 0.05908, + "grad_norm": 0.6179187865352406, + "learning_rate": 0.003, + "loss": 4.075, + "step": 5908 + }, + { + "epoch": 0.05909, + "grad_norm": 0.5943188321089226, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 5909 + }, + { + "epoch": 0.0591, + "grad_norm": 0.5836752437674285, + "learning_rate": 0.003, + "loss": 4.051, + "step": 5910 + }, + { + "epoch": 0.05911, + "grad_norm": 0.5075495627180322, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 5911 + }, + { + "epoch": 0.05912, + "grad_norm": 0.538417041118169, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 5912 + }, + { + "epoch": 0.05913, + "grad_norm": 0.5124851133164192, + "learning_rate": 0.003, + "loss": 4.1001, + "step": 5913 + }, + { + "epoch": 0.05914, + "grad_norm": 0.43038916754526674, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 5914 + }, + { + "epoch": 0.05915, + "grad_norm": 0.3638012477833408, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 5915 + }, + { + "epoch": 0.05916, + "grad_norm": 0.4160411304640959, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 5916 + }, + { + "epoch": 0.05917, + "grad_norm": 0.5302914359553863, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 5917 + }, + { + "epoch": 0.05918, + "grad_norm": 0.758457514927116, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 5918 + }, + { + "epoch": 0.05919, + "grad_norm": 0.9887164801356736, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 5919 + }, + { + "epoch": 0.0592, + "grad_norm": 0.996859290438551, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 5920 + }, + { + "epoch": 0.05921, + "grad_norm": 0.7597218956280568, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 5921 + }, + { + "epoch": 0.05922, + "grad_norm": 0.8870945497936235, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 5922 + }, + { + "epoch": 0.05923, + "grad_norm": 0.968662547246274, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 5923 + }, + { + "epoch": 0.05924, + "grad_norm": 0.8171324522257728, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 5924 + }, + { + "epoch": 0.05925, + "grad_norm": 0.8256806594722593, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 5925 + }, + { + "epoch": 0.05926, + "grad_norm": 0.8986582132449209, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 5926 + }, + { + "epoch": 0.05927, + "grad_norm": 0.8894681636986603, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 5927 + }, + { + "epoch": 0.05928, + "grad_norm": 0.9463669098421829, + "learning_rate": 0.003, + "loss": 4.0827, + "step": 5928 + }, + { + "epoch": 0.05929, + "grad_norm": 0.9599377959070252, + "learning_rate": 0.003, + "loss": 4.0936, + "step": 5929 + }, + { + "epoch": 0.0593, + "grad_norm": 1.0033569736760652, + "learning_rate": 0.003, + "loss": 4.1006, + "step": 5930 + }, + { + "epoch": 0.05931, + "grad_norm": 0.9840567377011477, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 5931 + }, + { + "epoch": 0.05932, + "grad_norm": 1.0012047346259751, + "learning_rate": 0.003, + "loss": 4.1092, + "step": 5932 + }, + { + "epoch": 0.05933, + "grad_norm": 0.8876267807165497, + "learning_rate": 0.003, + "loss": 4.0759, + "step": 5933 + }, + { + "epoch": 0.05934, + "grad_norm": 0.8508104544897132, + "learning_rate": 0.003, + "loss": 4.1145, + "step": 5934 + }, + { + "epoch": 0.05935, + "grad_norm": 0.8445594617278674, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 5935 + }, + { + "epoch": 0.05936, + "grad_norm": 0.7464232420084959, + "learning_rate": 0.003, + "loss": 4.0926, + "step": 5936 + }, + { + "epoch": 0.05937, + "grad_norm": 0.6812618033166751, + "learning_rate": 0.003, + "loss": 4.0878, + "step": 5937 + }, + { + "epoch": 0.05938, + "grad_norm": 0.7633266717251682, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 5938 + }, + { + "epoch": 0.05939, + "grad_norm": 0.8101898839692672, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 5939 + }, + { + "epoch": 0.0594, + "grad_norm": 0.816722220896553, + "learning_rate": 0.003, + "loss": 4.098, + "step": 5940 + }, + { + "epoch": 0.05941, + "grad_norm": 0.6284146095892529, + "learning_rate": 0.003, + "loss": 4.0918, + "step": 5941 + }, + { + "epoch": 0.05942, + "grad_norm": 0.5790929546123015, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 5942 + }, + { + "epoch": 0.05943, + "grad_norm": 0.5798916060112217, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 5943 + }, + { + "epoch": 0.05944, + "grad_norm": 0.5380379593411962, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 5944 + }, + { + "epoch": 0.05945, + "grad_norm": 0.5163516475084299, + "learning_rate": 0.003, + "loss": 4.056, + "step": 5945 + }, + { + "epoch": 0.05946, + "grad_norm": 0.5394001110581252, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 5946 + }, + { + "epoch": 0.05947, + "grad_norm": 0.5853463020190633, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 5947 + }, + { + "epoch": 0.05948, + "grad_norm": 0.5906626426412858, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 5948 + }, + { + "epoch": 0.05949, + "grad_norm": 0.624679759931772, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 5949 + }, + { + "epoch": 0.0595, + "grad_norm": 0.6959908833850434, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 5950 + }, + { + "epoch": 0.05951, + "grad_norm": 0.7277738014065029, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 5951 + }, + { + "epoch": 0.05952, + "grad_norm": 0.6429364291929107, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 5952 + }, + { + "epoch": 0.05953, + "grad_norm": 0.555054363348053, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 5953 + }, + { + "epoch": 0.05954, + "grad_norm": 0.5069426149003401, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 5954 + }, + { + "epoch": 0.05955, + "grad_norm": 0.5318659312548288, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 5955 + }, + { + "epoch": 0.05956, + "grad_norm": 0.4023789213366441, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 5956 + }, + { + "epoch": 0.05957, + "grad_norm": 0.3904184394694437, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 5957 + }, + { + "epoch": 0.05958, + "grad_norm": 0.41480221619044405, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 5958 + }, + { + "epoch": 0.05959, + "grad_norm": 0.4608922512945011, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 5959 + }, + { + "epoch": 0.0596, + "grad_norm": 0.5703465749933081, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 5960 + }, + { + "epoch": 0.05961, + "grad_norm": 0.6644884974320395, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 5961 + }, + { + "epoch": 0.05962, + "grad_norm": 0.6649937575956929, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 5962 + }, + { + "epoch": 0.05963, + "grad_norm": 0.653064654024882, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 5963 + }, + { + "epoch": 0.05964, + "grad_norm": 0.7894302548954674, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 5964 + }, + { + "epoch": 0.05965, + "grad_norm": 0.8906873783453355, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 5965 + }, + { + "epoch": 0.05966, + "grad_norm": 0.7865761989581169, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 5966 + }, + { + "epoch": 0.05967, + "grad_norm": 0.6950155908732526, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 5967 + }, + { + "epoch": 0.05968, + "grad_norm": 0.7545416711441219, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 5968 + }, + { + "epoch": 0.05969, + "grad_norm": 0.678340667532674, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 5969 + }, + { + "epoch": 0.0597, + "grad_norm": 0.5831006774804711, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 5970 + }, + { + "epoch": 0.05971, + "grad_norm": 0.6539254327941352, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 5971 + }, + { + "epoch": 0.05972, + "grad_norm": 0.7697467299627136, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 5972 + }, + { + "epoch": 0.05973, + "grad_norm": 0.7299287183707025, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 5973 + }, + { + "epoch": 0.05974, + "grad_norm": 0.7635482908041388, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 5974 + }, + { + "epoch": 0.05975, + "grad_norm": 0.8303448763480615, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 5975 + }, + { + "epoch": 0.05976, + "grad_norm": 0.9592326299034352, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 5976 + }, + { + "epoch": 0.05977, + "grad_norm": 0.9664691330103383, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 5977 + }, + { + "epoch": 0.05978, + "grad_norm": 0.8929921930552236, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 5978 + }, + { + "epoch": 0.05979, + "grad_norm": 0.8592759683937536, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 5979 + }, + { + "epoch": 0.0598, + "grad_norm": 0.9385924131061661, + "learning_rate": 0.003, + "loss": 4.062, + "step": 5980 + }, + { + "epoch": 0.05981, + "grad_norm": 0.9739065527065641, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 5981 + }, + { + "epoch": 0.05982, + "grad_norm": 0.9363915032565914, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 5982 + }, + { + "epoch": 0.05983, + "grad_norm": 0.9651959077052061, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 5983 + }, + { + "epoch": 0.05984, + "grad_norm": 0.832810000246513, + "learning_rate": 0.003, + "loss": 4.067, + "step": 5984 + }, + { + "epoch": 0.05985, + "grad_norm": 0.8136898772620698, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 5985 + }, + { + "epoch": 0.05986, + "grad_norm": 0.7046061597199408, + "learning_rate": 0.003, + "loss": 4.0921, + "step": 5986 + }, + { + "epoch": 0.05987, + "grad_norm": 0.6906371950406377, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 5987 + }, + { + "epoch": 0.05988, + "grad_norm": 0.7603214128080036, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 5988 + }, + { + "epoch": 0.05989, + "grad_norm": 0.7302876614703385, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 5989 + }, + { + "epoch": 0.0599, + "grad_norm": 0.6321192160417236, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 5990 + }, + { + "epoch": 0.05991, + "grad_norm": 0.6053093654068407, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 5991 + }, + { + "epoch": 0.05992, + "grad_norm": 0.6188232762660111, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 5992 + }, + { + "epoch": 0.05993, + "grad_norm": 0.5454539568657537, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 5993 + }, + { + "epoch": 0.05994, + "grad_norm": 0.5558410897460727, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 5994 + }, + { + "epoch": 0.05995, + "grad_norm": 0.5843312097620589, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 5995 + }, + { + "epoch": 0.05996, + "grad_norm": 0.5773146578119047, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 5996 + }, + { + "epoch": 0.05997, + "grad_norm": 0.6434639736179467, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 5997 + }, + { + "epoch": 0.05998, + "grad_norm": 0.5957494167373613, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 5998 + }, + { + "epoch": 0.05999, + "grad_norm": 0.6081539666629457, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 5999 + }, + { + "epoch": 0.06, + "grad_norm": 0.557929984035511, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 6000 + }, + { + "epoch": 0.06001, + "grad_norm": 0.5227847410552245, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 6001 + }, + { + "epoch": 0.06002, + "grad_norm": 0.5209580056772323, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 6002 + }, + { + "epoch": 0.06003, + "grad_norm": 0.49604047081579417, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 6003 + }, + { + "epoch": 0.06004, + "grad_norm": 0.567025022683691, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 6004 + }, + { + "epoch": 0.06005, + "grad_norm": 0.519159265238865, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6005 + }, + { + "epoch": 0.06006, + "grad_norm": 0.5275498633242567, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 6006 + }, + { + "epoch": 0.06007, + "grad_norm": 0.6007455843541345, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 6007 + }, + { + "epoch": 0.06008, + "grad_norm": 0.6154818315371587, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 6008 + }, + { + "epoch": 0.06009, + "grad_norm": 0.6218886411970437, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 6009 + }, + { + "epoch": 0.0601, + "grad_norm": 0.66010590101245, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 6010 + }, + { + "epoch": 0.06011, + "grad_norm": 0.7417542934976062, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 6011 + }, + { + "epoch": 0.06012, + "grad_norm": 0.7002228840573359, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 6012 + }, + { + "epoch": 0.06013, + "grad_norm": 0.6004150050864936, + "learning_rate": 0.003, + "loss": 4.057, + "step": 6013 + }, + { + "epoch": 0.06014, + "grad_norm": 0.6816247181158233, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 6014 + }, + { + "epoch": 0.06015, + "grad_norm": 0.7481873297711161, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 6015 + }, + { + "epoch": 0.06016, + "grad_norm": 0.7624185756518325, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 6016 + }, + { + "epoch": 0.06017, + "grad_norm": 0.8091748147126783, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 6017 + }, + { + "epoch": 0.06018, + "grad_norm": 0.7560445366447108, + "learning_rate": 0.003, + "loss": 4.093, + "step": 6018 + }, + { + "epoch": 0.06019, + "grad_norm": 0.8159440823541698, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 6019 + }, + { + "epoch": 0.0602, + "grad_norm": 0.7374802221448221, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 6020 + }, + { + "epoch": 0.06021, + "grad_norm": 0.7260723594155657, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 6021 + }, + { + "epoch": 0.06022, + "grad_norm": 0.6987762438680518, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 6022 + }, + { + "epoch": 0.06023, + "grad_norm": 0.738224299095221, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 6023 + }, + { + "epoch": 0.06024, + "grad_norm": 0.7800935163681785, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 6024 + }, + { + "epoch": 0.06025, + "grad_norm": 0.6437570123900622, + "learning_rate": 0.003, + "loss": 4.0966, + "step": 6025 + }, + { + "epoch": 0.06026, + "grad_norm": 0.6288321067261896, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 6026 + }, + { + "epoch": 0.06027, + "grad_norm": 0.5861127212279555, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 6027 + }, + { + "epoch": 0.06028, + "grad_norm": 0.6051054303482403, + "learning_rate": 0.003, + "loss": 4.0797, + "step": 6028 + }, + { + "epoch": 0.06029, + "grad_norm": 0.5441239172070078, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 6029 + }, + { + "epoch": 0.0603, + "grad_norm": 0.49795727083937485, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 6030 + }, + { + "epoch": 0.06031, + "grad_norm": 0.5314315122641824, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 6031 + }, + { + "epoch": 0.06032, + "grad_norm": 0.6138060718211107, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 6032 + }, + { + "epoch": 0.06033, + "grad_norm": 0.6710650098429001, + "learning_rate": 0.003, + "loss": 4.067, + "step": 6033 + }, + { + "epoch": 0.06034, + "grad_norm": 0.6696336382914826, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 6034 + }, + { + "epoch": 0.06035, + "grad_norm": 0.8882846388287199, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 6035 + }, + { + "epoch": 0.06036, + "grad_norm": 1.121979484482661, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 6036 + }, + { + "epoch": 0.06037, + "grad_norm": 0.8379898111931297, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 6037 + }, + { + "epoch": 0.06038, + "grad_norm": 0.7802144931498387, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6038 + }, + { + "epoch": 0.06039, + "grad_norm": 0.7032582063035804, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 6039 + }, + { + "epoch": 0.0604, + "grad_norm": 0.7335369649224487, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 6040 + }, + { + "epoch": 0.06041, + "grad_norm": 0.7151164974332928, + "learning_rate": 0.003, + "loss": 4.0822, + "step": 6041 + }, + { + "epoch": 0.06042, + "grad_norm": 0.8228404366945576, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 6042 + }, + { + "epoch": 0.06043, + "grad_norm": 0.8502825779803321, + "learning_rate": 0.003, + "loss": 4.07, + "step": 6043 + }, + { + "epoch": 0.06044, + "grad_norm": 0.7408341984511058, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 6044 + }, + { + "epoch": 0.06045, + "grad_norm": 0.7336566685173229, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 6045 + }, + { + "epoch": 0.06046, + "grad_norm": 0.8173735731231058, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 6046 + }, + { + "epoch": 0.06047, + "grad_norm": 0.943028861920409, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 6047 + }, + { + "epoch": 0.06048, + "grad_norm": 0.8125964384603215, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6048 + }, + { + "epoch": 0.06049, + "grad_norm": 0.7858319678078841, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6049 + }, + { + "epoch": 0.0605, + "grad_norm": 0.7405244514450401, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 6050 + }, + { + "epoch": 0.06051, + "grad_norm": 0.7408205018095884, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 6051 + }, + { + "epoch": 0.06052, + "grad_norm": 0.7657603554370147, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 6052 + }, + { + "epoch": 0.06053, + "grad_norm": 0.7557387817416936, + "learning_rate": 0.003, + "loss": 4.044, + "step": 6053 + }, + { + "epoch": 0.06054, + "grad_norm": 0.6938295099182639, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 6054 + }, + { + "epoch": 0.06055, + "grad_norm": 0.6649304335265552, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 6055 + }, + { + "epoch": 0.06056, + "grad_norm": 0.7685147620610964, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 6056 + }, + { + "epoch": 0.06057, + "grad_norm": 0.7844351298814363, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 6057 + }, + { + "epoch": 0.06058, + "grad_norm": 0.6449936495042284, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 6058 + }, + { + "epoch": 0.06059, + "grad_norm": 0.6094506300928993, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 6059 + }, + { + "epoch": 0.0606, + "grad_norm": 0.6666381936609098, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 6060 + }, + { + "epoch": 0.06061, + "grad_norm": 0.5754160744810928, + "learning_rate": 0.003, + "loss": 4.069, + "step": 6061 + }, + { + "epoch": 0.06062, + "grad_norm": 0.5393788523771272, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 6062 + }, + { + "epoch": 0.06063, + "grad_norm": 0.5363460220494113, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 6063 + }, + { + "epoch": 0.06064, + "grad_norm": 0.5459077824337315, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 6064 + }, + { + "epoch": 0.06065, + "grad_norm": 0.4982351665568225, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 6065 + }, + { + "epoch": 0.06066, + "grad_norm": 0.5831355871088383, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 6066 + }, + { + "epoch": 0.06067, + "grad_norm": 0.6883666500467263, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 6067 + }, + { + "epoch": 0.06068, + "grad_norm": 0.7768904066359895, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 6068 + }, + { + "epoch": 0.06069, + "grad_norm": 0.7154765052828583, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 6069 + }, + { + "epoch": 0.0607, + "grad_norm": 0.7170877657291906, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6070 + }, + { + "epoch": 0.06071, + "grad_norm": 0.877544704736836, + "learning_rate": 0.003, + "loss": 4.039, + "step": 6071 + }, + { + "epoch": 0.06072, + "grad_norm": 0.8538517339163024, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 6072 + }, + { + "epoch": 0.06073, + "grad_norm": 0.7815558956587654, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 6073 + }, + { + "epoch": 0.06074, + "grad_norm": 0.6356040706317584, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 6074 + }, + { + "epoch": 0.06075, + "grad_norm": 0.6539954284277101, + "learning_rate": 0.003, + "loss": 4.0787, + "step": 6075 + }, + { + "epoch": 0.06076, + "grad_norm": 0.6075996423194496, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 6076 + }, + { + "epoch": 0.06077, + "grad_norm": 0.6450123270117637, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6077 + }, + { + "epoch": 0.06078, + "grad_norm": 0.6068522272710506, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 6078 + }, + { + "epoch": 0.06079, + "grad_norm": 0.6749799112393907, + "learning_rate": 0.003, + "loss": 4.06, + "step": 6079 + }, + { + "epoch": 0.0608, + "grad_norm": 0.7466812401671813, + "learning_rate": 0.003, + "loss": 4.061, + "step": 6080 + }, + { + "epoch": 0.06081, + "grad_norm": 0.7927002855467632, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 6081 + }, + { + "epoch": 0.06082, + "grad_norm": 0.8925493295539645, + "learning_rate": 0.003, + "loss": 4.09, + "step": 6082 + }, + { + "epoch": 0.06083, + "grad_norm": 0.8717428970139589, + "learning_rate": 0.003, + "loss": 4.074, + "step": 6083 + }, + { + "epoch": 0.06084, + "grad_norm": 0.8954516999988178, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 6084 + }, + { + "epoch": 0.06085, + "grad_norm": 0.9628426089350753, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 6085 + }, + { + "epoch": 0.06086, + "grad_norm": 0.8912975298581354, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 6086 + }, + { + "epoch": 0.06087, + "grad_norm": 0.7697176664205366, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 6087 + }, + { + "epoch": 0.06088, + "grad_norm": 0.6537885348957085, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 6088 + }, + { + "epoch": 0.06089, + "grad_norm": 0.6420878355964799, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 6089 + }, + { + "epoch": 0.0609, + "grad_norm": 0.6491909446073066, + "learning_rate": 0.003, + "loss": 4.0902, + "step": 6090 + }, + { + "epoch": 0.06091, + "grad_norm": 0.6915560637563698, + "learning_rate": 0.003, + "loss": 4.064, + "step": 6091 + }, + { + "epoch": 0.06092, + "grad_norm": 0.6760732818725483, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 6092 + }, + { + "epoch": 0.06093, + "grad_norm": 0.5971175413797949, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6093 + }, + { + "epoch": 0.06094, + "grad_norm": 0.5755115286066814, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 6094 + }, + { + "epoch": 0.06095, + "grad_norm": 0.5874846681819571, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 6095 + }, + { + "epoch": 0.06096, + "grad_norm": 0.7019376773885887, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 6096 + }, + { + "epoch": 0.06097, + "grad_norm": 0.767282029241582, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 6097 + }, + { + "epoch": 0.06098, + "grad_norm": 0.747175412992683, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 6098 + }, + { + "epoch": 0.06099, + "grad_norm": 0.7165415062478152, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 6099 + }, + { + "epoch": 0.061, + "grad_norm": 0.7170240862605034, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 6100 + }, + { + "epoch": 0.06101, + "grad_norm": 0.6198639654246494, + "learning_rate": 0.003, + "loss": 4.0842, + "step": 6101 + }, + { + "epoch": 0.06102, + "grad_norm": 0.6117275951014963, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 6102 + }, + { + "epoch": 0.06103, + "grad_norm": 0.6640723875125251, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 6103 + }, + { + "epoch": 0.06104, + "grad_norm": 0.6979047074375939, + "learning_rate": 0.003, + "loss": 4.049, + "step": 6104 + }, + { + "epoch": 0.06105, + "grad_norm": 0.7629586635939833, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 6105 + }, + { + "epoch": 0.06106, + "grad_norm": 0.7457634325919515, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 6106 + }, + { + "epoch": 0.06107, + "grad_norm": 0.6519303543523637, + "learning_rate": 0.003, + "loss": 4.0861, + "step": 6107 + }, + { + "epoch": 0.06108, + "grad_norm": 0.5933517948299097, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 6108 + }, + { + "epoch": 0.06109, + "grad_norm": 0.6412753985954568, + "learning_rate": 0.003, + "loss": 4.026, + "step": 6109 + }, + { + "epoch": 0.0611, + "grad_norm": 0.6832415741809712, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 6110 + }, + { + "epoch": 0.06111, + "grad_norm": 0.6923037030769662, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 6111 + }, + { + "epoch": 0.06112, + "grad_norm": 0.7580918882753322, + "learning_rate": 0.003, + "loss": 4.059, + "step": 6112 + }, + { + "epoch": 0.06113, + "grad_norm": 0.8388182870535142, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 6113 + }, + { + "epoch": 0.06114, + "grad_norm": 0.8276989484611054, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 6114 + }, + { + "epoch": 0.06115, + "grad_norm": 0.803686996389499, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 6115 + }, + { + "epoch": 0.06116, + "grad_norm": 0.8074349125174471, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 6116 + }, + { + "epoch": 0.06117, + "grad_norm": 0.9263268512640556, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6117 + }, + { + "epoch": 0.06118, + "grad_norm": 1.023418808326581, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 6118 + }, + { + "epoch": 0.06119, + "grad_norm": 1.0076087533418916, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 6119 + }, + { + "epoch": 0.0612, + "grad_norm": 0.8461498153427612, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 6120 + }, + { + "epoch": 0.06121, + "grad_norm": 0.853965333354219, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 6121 + }, + { + "epoch": 0.06122, + "grad_norm": 0.8566235471058572, + "learning_rate": 0.003, + "loss": 4.078, + "step": 6122 + }, + { + "epoch": 0.06123, + "grad_norm": 0.9906382066866388, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 6123 + }, + { + "epoch": 0.06124, + "grad_norm": 0.9356187788307065, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 6124 + }, + { + "epoch": 0.06125, + "grad_norm": 0.8468553369185784, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 6125 + }, + { + "epoch": 0.06126, + "grad_norm": 0.7145214082628322, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 6126 + }, + { + "epoch": 0.06127, + "grad_norm": 0.6508359177332966, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 6127 + }, + { + "epoch": 0.06128, + "grad_norm": 0.6217861197339226, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 6128 + }, + { + "epoch": 0.06129, + "grad_norm": 0.6257922123422462, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 6129 + }, + { + "epoch": 0.0613, + "grad_norm": 0.7169443798575936, + "learning_rate": 0.003, + "loss": 4.076, + "step": 6130 + }, + { + "epoch": 0.06131, + "grad_norm": 0.6816099174255882, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 6131 + }, + { + "epoch": 0.06132, + "grad_norm": 0.5834441649126184, + "learning_rate": 0.003, + "loss": 4.067, + "step": 6132 + }, + { + "epoch": 0.06133, + "grad_norm": 0.5579483189061786, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 6133 + }, + { + "epoch": 0.06134, + "grad_norm": 0.5187600542697383, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 6134 + }, + { + "epoch": 0.06135, + "grad_norm": 0.5042789494318808, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 6135 + }, + { + "epoch": 0.06136, + "grad_norm": 0.45744900406102773, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 6136 + }, + { + "epoch": 0.06137, + "grad_norm": 0.44367353772320356, + "learning_rate": 0.003, + "loss": 4.048, + "step": 6137 + }, + { + "epoch": 0.06138, + "grad_norm": 0.45327443797339007, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 6138 + }, + { + "epoch": 0.06139, + "grad_norm": 0.4778568770889814, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 6139 + }, + { + "epoch": 0.0614, + "grad_norm": 0.5901999444858329, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 6140 + }, + { + "epoch": 0.06141, + "grad_norm": 0.7019912051056786, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 6141 + }, + { + "epoch": 0.06142, + "grad_norm": 0.7844461444902411, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 6142 + }, + { + "epoch": 0.06143, + "grad_norm": 0.860485543399918, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 6143 + }, + { + "epoch": 0.06144, + "grad_norm": 0.8166466389717842, + "learning_rate": 0.003, + "loss": 4.046, + "step": 6144 + }, + { + "epoch": 0.06145, + "grad_norm": 0.7390794495646494, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 6145 + }, + { + "epoch": 0.06146, + "grad_norm": 0.753791612641437, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 6146 + }, + { + "epoch": 0.06147, + "grad_norm": 0.834260116899666, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 6147 + }, + { + "epoch": 0.06148, + "grad_norm": 0.8022096782506793, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 6148 + }, + { + "epoch": 0.06149, + "grad_norm": 0.8144682406232034, + "learning_rate": 0.003, + "loss": 4.054, + "step": 6149 + }, + { + "epoch": 0.0615, + "grad_norm": 0.8914398375168914, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 6150 + }, + { + "epoch": 0.06151, + "grad_norm": 0.9402974453678897, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 6151 + }, + { + "epoch": 0.06152, + "grad_norm": 0.9090090337039639, + "learning_rate": 0.003, + "loss": 4.062, + "step": 6152 + }, + { + "epoch": 0.06153, + "grad_norm": 0.7861846957681318, + "learning_rate": 0.003, + "loss": 4.0876, + "step": 6153 + }, + { + "epoch": 0.06154, + "grad_norm": 0.7817243963574592, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 6154 + }, + { + "epoch": 0.06155, + "grad_norm": 0.7888800338737242, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 6155 + }, + { + "epoch": 0.06156, + "grad_norm": 0.7735096455930125, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 6156 + }, + { + "epoch": 0.06157, + "grad_norm": 0.7267537924257372, + "learning_rate": 0.003, + "loss": 4.0785, + "step": 6157 + }, + { + "epoch": 0.06158, + "grad_norm": 0.7823125017046669, + "learning_rate": 0.003, + "loss": 4.1042, + "step": 6158 + }, + { + "epoch": 0.06159, + "grad_norm": 0.8005439759133111, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 6159 + }, + { + "epoch": 0.0616, + "grad_norm": 0.72785329759981, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 6160 + }, + { + "epoch": 0.06161, + "grad_norm": 0.7413441176723736, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 6161 + }, + { + "epoch": 0.06162, + "grad_norm": 0.7059112527174068, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 6162 + }, + { + "epoch": 0.06163, + "grad_norm": 0.6459243874828301, + "learning_rate": 0.003, + "loss": 4.096, + "step": 6163 + }, + { + "epoch": 0.06164, + "grad_norm": 0.625190575443019, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 6164 + }, + { + "epoch": 0.06165, + "grad_norm": 0.6903103373671604, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 6165 + }, + { + "epoch": 0.06166, + "grad_norm": 0.7056908909839628, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 6166 + }, + { + "epoch": 0.06167, + "grad_norm": 0.8367468508012779, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 6167 + }, + { + "epoch": 0.06168, + "grad_norm": 0.9123721068665754, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 6168 + }, + { + "epoch": 0.06169, + "grad_norm": 0.8781004962311973, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 6169 + }, + { + "epoch": 0.0617, + "grad_norm": 0.8124337684119513, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 6170 + }, + { + "epoch": 0.06171, + "grad_norm": 0.6688257946536776, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 6171 + }, + { + "epoch": 0.06172, + "grad_norm": 0.6487101647880741, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 6172 + }, + { + "epoch": 0.06173, + "grad_norm": 0.624814221895412, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 6173 + }, + { + "epoch": 0.06174, + "grad_norm": 0.6196730039761758, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 6174 + }, + { + "epoch": 0.06175, + "grad_norm": 0.5446768804985062, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 6175 + }, + { + "epoch": 0.06176, + "grad_norm": 0.5253233303668841, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 6176 + }, + { + "epoch": 0.06177, + "grad_norm": 0.5450423165675929, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 6177 + }, + { + "epoch": 0.06178, + "grad_norm": 0.5507111527408102, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 6178 + }, + { + "epoch": 0.06179, + "grad_norm": 0.5359907157173006, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 6179 + }, + { + "epoch": 0.0618, + "grad_norm": 0.5299879140384445, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 6180 + }, + { + "epoch": 0.06181, + "grad_norm": 0.5154042220581599, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 6181 + }, + { + "epoch": 0.06182, + "grad_norm": 0.5994683381767214, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 6182 + }, + { + "epoch": 0.06183, + "grad_norm": 0.6816733376852764, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 6183 + }, + { + "epoch": 0.06184, + "grad_norm": 0.6492853599060219, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 6184 + }, + { + "epoch": 0.06185, + "grad_norm": 0.5506070689473229, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 6185 + }, + { + "epoch": 0.06186, + "grad_norm": 0.6587598553130263, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 6186 + }, + { + "epoch": 0.06187, + "grad_norm": 0.717889576578481, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 6187 + }, + { + "epoch": 0.06188, + "grad_norm": 0.793341905498176, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 6188 + }, + { + "epoch": 0.06189, + "grad_norm": 0.7571167153473217, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 6189 + }, + { + "epoch": 0.0619, + "grad_norm": 0.6267058334619902, + "learning_rate": 0.003, + "loss": 4.0833, + "step": 6190 + }, + { + "epoch": 0.06191, + "grad_norm": 0.6609446191170651, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 6191 + }, + { + "epoch": 0.06192, + "grad_norm": 0.7514636540123598, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 6192 + }, + { + "epoch": 0.06193, + "grad_norm": 0.7851200580246341, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 6193 + }, + { + "epoch": 0.06194, + "grad_norm": 0.7601294623291285, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 6194 + }, + { + "epoch": 0.06195, + "grad_norm": 0.7420716899824206, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6195 + }, + { + "epoch": 0.06196, + "grad_norm": 0.6872803076682844, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 6196 + }, + { + "epoch": 0.06197, + "grad_norm": 0.6903668269181976, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 6197 + }, + { + "epoch": 0.06198, + "grad_norm": 0.6669020882719598, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 6198 + }, + { + "epoch": 0.06199, + "grad_norm": 0.7075014709934362, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 6199 + }, + { + "epoch": 0.062, + "grad_norm": 0.7123768142179934, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 6200 + }, + { + "epoch": 0.06201, + "grad_norm": 0.6597545997882802, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 6201 + }, + { + "epoch": 0.06202, + "grad_norm": 0.7382079260614234, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 6202 + }, + { + "epoch": 0.06203, + "grad_norm": 0.7496814157101167, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 6203 + }, + { + "epoch": 0.06204, + "grad_norm": 0.743239107456832, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 6204 + }, + { + "epoch": 0.06205, + "grad_norm": 0.8245445395173092, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 6205 + }, + { + "epoch": 0.06206, + "grad_norm": 0.8235056293562754, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 6206 + }, + { + "epoch": 0.06207, + "grad_norm": 0.744227639677771, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 6207 + }, + { + "epoch": 0.06208, + "grad_norm": 0.6683580031508113, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 6208 + }, + { + "epoch": 0.06209, + "grad_norm": 0.5926908376892637, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 6209 + }, + { + "epoch": 0.0621, + "grad_norm": 0.565190347858888, + "learning_rate": 0.003, + "loss": 4.0937, + "step": 6210 + }, + { + "epoch": 0.06211, + "grad_norm": 0.6757244716075078, + "learning_rate": 0.003, + "loss": 4.06, + "step": 6211 + }, + { + "epoch": 0.06212, + "grad_norm": 0.7636544528361237, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 6212 + }, + { + "epoch": 0.06213, + "grad_norm": 0.9489563689989222, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 6213 + }, + { + "epoch": 0.06214, + "grad_norm": 1.0773038335427632, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 6214 + }, + { + "epoch": 0.06215, + "grad_norm": 0.8410647238994581, + "learning_rate": 0.003, + "loss": 4.0745, + "step": 6215 + }, + { + "epoch": 0.06216, + "grad_norm": 0.830816175776546, + "learning_rate": 0.003, + "loss": 4.1066, + "step": 6216 + }, + { + "epoch": 0.06217, + "grad_norm": 0.83688847332168, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 6217 + }, + { + "epoch": 0.06218, + "grad_norm": 0.8437624588138829, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 6218 + }, + { + "epoch": 0.06219, + "grad_norm": 0.7756032770280755, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 6219 + }, + { + "epoch": 0.0622, + "grad_norm": 0.8060046290599426, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 6220 + }, + { + "epoch": 0.06221, + "grad_norm": 0.8120345275270059, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 6221 + }, + { + "epoch": 0.06222, + "grad_norm": 0.7700791856175148, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 6222 + }, + { + "epoch": 0.06223, + "grad_norm": 0.7262231348928465, + "learning_rate": 0.003, + "loss": 4.074, + "step": 6223 + }, + { + "epoch": 0.06224, + "grad_norm": 0.7618825620829891, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 6224 + }, + { + "epoch": 0.06225, + "grad_norm": 0.7659071983864602, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 6225 + }, + { + "epoch": 0.06226, + "grad_norm": 0.6750949867531024, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 6226 + }, + { + "epoch": 0.06227, + "grad_norm": 0.6141064064389489, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 6227 + }, + { + "epoch": 0.06228, + "grad_norm": 0.499011439148516, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6228 + }, + { + "epoch": 0.06229, + "grad_norm": 0.5308263576985826, + "learning_rate": 0.003, + "loss": 4.044, + "step": 6229 + }, + { + "epoch": 0.0623, + "grad_norm": 0.6153613208660836, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 6230 + }, + { + "epoch": 0.06231, + "grad_norm": 0.7146778059261917, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 6231 + }, + { + "epoch": 0.06232, + "grad_norm": 0.7589057088961252, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 6232 + }, + { + "epoch": 0.06233, + "grad_norm": 0.752290657334441, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 6233 + }, + { + "epoch": 0.06234, + "grad_norm": 0.6682072930406131, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 6234 + }, + { + "epoch": 0.06235, + "grad_norm": 0.5838170554522981, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 6235 + }, + { + "epoch": 0.06236, + "grad_norm": 0.5734587136546955, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 6236 + }, + { + "epoch": 0.06237, + "grad_norm": 0.4972201719879128, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 6237 + }, + { + "epoch": 0.06238, + "grad_norm": 0.5084275865225812, + "learning_rate": 0.003, + "loss": 4.05, + "step": 6238 + }, + { + "epoch": 0.06239, + "grad_norm": 0.5732991595727963, + "learning_rate": 0.003, + "loss": 4.031, + "step": 6239 + }, + { + "epoch": 0.0624, + "grad_norm": 0.705609873990105, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 6240 + }, + { + "epoch": 0.06241, + "grad_norm": 0.7701665580049867, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 6241 + }, + { + "epoch": 0.06242, + "grad_norm": 0.882873840384011, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 6242 + }, + { + "epoch": 0.06243, + "grad_norm": 0.9414839998832126, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 6243 + }, + { + "epoch": 0.06244, + "grad_norm": 1.0026793816565995, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 6244 + }, + { + "epoch": 0.06245, + "grad_norm": 0.8405314873608366, + "learning_rate": 0.003, + "loss": 4.0711, + "step": 6245 + }, + { + "epoch": 0.06246, + "grad_norm": 0.7722268383245824, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 6246 + }, + { + "epoch": 0.06247, + "grad_norm": 0.7227903263189711, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 6247 + }, + { + "epoch": 0.06248, + "grad_norm": 0.6642436882917294, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 6248 + }, + { + "epoch": 0.06249, + "grad_norm": 0.6168362465709406, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 6249 + }, + { + "epoch": 0.0625, + "grad_norm": 0.5997083323322405, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 6250 + }, + { + "epoch": 0.06251, + "grad_norm": 0.6108912830584505, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 6251 + }, + { + "epoch": 0.06252, + "grad_norm": 0.5991910657322239, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 6252 + }, + { + "epoch": 0.06253, + "grad_norm": 0.674934546983959, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 6253 + }, + { + "epoch": 0.06254, + "grad_norm": 0.7364517055268108, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 6254 + }, + { + "epoch": 0.06255, + "grad_norm": 0.769209147618945, + "learning_rate": 0.003, + "loss": 4.068, + "step": 6255 + }, + { + "epoch": 0.06256, + "grad_norm": 0.7110812924811214, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6256 + }, + { + "epoch": 0.06257, + "grad_norm": 0.627587247162009, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 6257 + }, + { + "epoch": 0.06258, + "grad_norm": 0.6344571908482386, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 6258 + }, + { + "epoch": 0.06259, + "grad_norm": 0.6362831075020776, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 6259 + }, + { + "epoch": 0.0626, + "grad_norm": 0.6972109315752782, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 6260 + }, + { + "epoch": 0.06261, + "grad_norm": 0.7999622988151236, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 6261 + }, + { + "epoch": 0.06262, + "grad_norm": 0.7309086582544324, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 6262 + }, + { + "epoch": 0.06263, + "grad_norm": 0.6860520624023821, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 6263 + }, + { + "epoch": 0.06264, + "grad_norm": 0.7596134209950258, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 6264 + }, + { + "epoch": 0.06265, + "grad_norm": 0.659108280279751, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 6265 + }, + { + "epoch": 0.06266, + "grad_norm": 0.5186672608849651, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6266 + }, + { + "epoch": 0.06267, + "grad_norm": 0.599452230737259, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 6267 + }, + { + "epoch": 0.06268, + "grad_norm": 0.7235207620185489, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 6268 + }, + { + "epoch": 0.06269, + "grad_norm": 0.7854951238216421, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 6269 + }, + { + "epoch": 0.0627, + "grad_norm": 0.8394671207268496, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 6270 + }, + { + "epoch": 0.06271, + "grad_norm": 0.9055077696388577, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 6271 + }, + { + "epoch": 0.06272, + "grad_norm": 0.9280843679497038, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 6272 + }, + { + "epoch": 0.06273, + "grad_norm": 0.876371467145136, + "learning_rate": 0.003, + "loss": 4.0868, + "step": 6273 + }, + { + "epoch": 0.06274, + "grad_norm": 0.7317414783734545, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 6274 + }, + { + "epoch": 0.06275, + "grad_norm": 0.6749463124240829, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 6275 + }, + { + "epoch": 0.06276, + "grad_norm": 0.6862692622659947, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 6276 + }, + { + "epoch": 0.06277, + "grad_norm": 0.6812030700587688, + "learning_rate": 0.003, + "loss": 4.0717, + "step": 6277 + }, + { + "epoch": 0.06278, + "grad_norm": 0.6995983827846336, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 6278 + }, + { + "epoch": 0.06279, + "grad_norm": 0.5608509985141449, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 6279 + }, + { + "epoch": 0.0628, + "grad_norm": 0.5658269498841932, + "learning_rate": 0.003, + "loss": 4.05, + "step": 6280 + }, + { + "epoch": 0.06281, + "grad_norm": 0.611643096587109, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 6281 + }, + { + "epoch": 0.06282, + "grad_norm": 0.602227042496892, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 6282 + }, + { + "epoch": 0.06283, + "grad_norm": 0.5904763440272163, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 6283 + }, + { + "epoch": 0.06284, + "grad_norm": 0.6757198457489071, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 6284 + }, + { + "epoch": 0.06285, + "grad_norm": 0.6821458859643692, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 6285 + }, + { + "epoch": 0.06286, + "grad_norm": 0.6372382162500136, + "learning_rate": 0.003, + "loss": 4.045, + "step": 6286 + }, + { + "epoch": 0.06287, + "grad_norm": 0.5932593850527702, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 6287 + }, + { + "epoch": 0.06288, + "grad_norm": 0.6096061298108356, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 6288 + }, + { + "epoch": 0.06289, + "grad_norm": 0.6238784562462693, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 6289 + }, + { + "epoch": 0.0629, + "grad_norm": 0.5597303309340962, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 6290 + }, + { + "epoch": 0.06291, + "grad_norm": 0.5441192539992968, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 6291 + }, + { + "epoch": 0.06292, + "grad_norm": 0.45504346541947616, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6292 + }, + { + "epoch": 0.06293, + "grad_norm": 0.5509539821776427, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 6293 + }, + { + "epoch": 0.06294, + "grad_norm": 0.5961619808774158, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 6294 + }, + { + "epoch": 0.06295, + "grad_norm": 0.6694707370744185, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 6295 + }, + { + "epoch": 0.06296, + "grad_norm": 0.7814821998888777, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 6296 + }, + { + "epoch": 0.06297, + "grad_norm": 0.8668370468312114, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 6297 + }, + { + "epoch": 0.06298, + "grad_norm": 0.7460051195647898, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6298 + }, + { + "epoch": 0.06299, + "grad_norm": 0.6009126581190146, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 6299 + }, + { + "epoch": 0.063, + "grad_norm": 0.5799863896621738, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 6300 + }, + { + "epoch": 0.06301, + "grad_norm": 0.6184868562065018, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 6301 + }, + { + "epoch": 0.06302, + "grad_norm": 0.702826737419538, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 6302 + }, + { + "epoch": 0.06303, + "grad_norm": 0.9337387486055504, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 6303 + }, + { + "epoch": 0.06304, + "grad_norm": 0.939138065901479, + "learning_rate": 0.003, + "loss": 4.0877, + "step": 6304 + }, + { + "epoch": 0.06305, + "grad_norm": 0.9209750715645206, + "learning_rate": 0.003, + "loss": 4.0965, + "step": 6305 + }, + { + "epoch": 0.06306, + "grad_norm": 1.0736935127225395, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 6306 + }, + { + "epoch": 0.06307, + "grad_norm": 0.9511152570614599, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 6307 + }, + { + "epoch": 0.06308, + "grad_norm": 0.9123867884912619, + "learning_rate": 0.003, + "loss": 4.0799, + "step": 6308 + }, + { + "epoch": 0.06309, + "grad_norm": 0.8390875868656612, + "learning_rate": 0.003, + "loss": 4.07, + "step": 6309 + }, + { + "epoch": 0.0631, + "grad_norm": 0.946314694081621, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 6310 + }, + { + "epoch": 0.06311, + "grad_norm": 1.0134868217208317, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 6311 + }, + { + "epoch": 0.06312, + "grad_norm": 1.0530180450744513, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 6312 + }, + { + "epoch": 0.06313, + "grad_norm": 0.8638088747293382, + "learning_rate": 0.003, + "loss": 4.0916, + "step": 6313 + }, + { + "epoch": 0.06314, + "grad_norm": 0.840809188590512, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 6314 + }, + { + "epoch": 0.06315, + "grad_norm": 0.8909165672841762, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 6315 + }, + { + "epoch": 0.06316, + "grad_norm": 0.8528401064762795, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 6316 + }, + { + "epoch": 0.06317, + "grad_norm": 0.819726215866848, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 6317 + }, + { + "epoch": 0.06318, + "grad_norm": 0.7145922783056121, + "learning_rate": 0.003, + "loss": 4.1117, + "step": 6318 + }, + { + "epoch": 0.06319, + "grad_norm": 0.6847237735682379, + "learning_rate": 0.003, + "loss": 4.11, + "step": 6319 + }, + { + "epoch": 0.0632, + "grad_norm": 0.6512671311277951, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 6320 + }, + { + "epoch": 0.06321, + "grad_norm": 0.6303863045155709, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 6321 + }, + { + "epoch": 0.06322, + "grad_norm": 0.5612971777375453, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 6322 + }, + { + "epoch": 0.06323, + "grad_norm": 0.5437844336320795, + "learning_rate": 0.003, + "loss": 4.055, + "step": 6323 + }, + { + "epoch": 0.06324, + "grad_norm": 0.516996981025749, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 6324 + }, + { + "epoch": 0.06325, + "grad_norm": 0.48316781011922927, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 6325 + }, + { + "epoch": 0.06326, + "grad_norm": 0.5095086179714272, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 6326 + }, + { + "epoch": 0.06327, + "grad_norm": 0.4704446438989928, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 6327 + }, + { + "epoch": 0.06328, + "grad_norm": 0.464144962740958, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 6328 + }, + { + "epoch": 0.06329, + "grad_norm": 0.5090824002577229, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 6329 + }, + { + "epoch": 0.0633, + "grad_norm": 0.5587653659050095, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 6330 + }, + { + "epoch": 0.06331, + "grad_norm": 0.5834666188400783, + "learning_rate": 0.003, + "loss": 4.036, + "step": 6331 + }, + { + "epoch": 0.06332, + "grad_norm": 0.5978758577925115, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 6332 + }, + { + "epoch": 0.06333, + "grad_norm": 0.5309153740010379, + "learning_rate": 0.003, + "loss": 4.039, + "step": 6333 + }, + { + "epoch": 0.06334, + "grad_norm": 0.5046160309402133, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 6334 + }, + { + "epoch": 0.06335, + "grad_norm": 0.5854531806090001, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 6335 + }, + { + "epoch": 0.06336, + "grad_norm": 0.6217412492804516, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 6336 + }, + { + "epoch": 0.06337, + "grad_norm": 0.6366742187835619, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 6337 + }, + { + "epoch": 0.06338, + "grad_norm": 0.7611140894755741, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 6338 + }, + { + "epoch": 0.06339, + "grad_norm": 0.9882579876349252, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 6339 + }, + { + "epoch": 0.0634, + "grad_norm": 0.9875038968772513, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 6340 + }, + { + "epoch": 0.06341, + "grad_norm": 0.9106371080834992, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 6341 + }, + { + "epoch": 0.06342, + "grad_norm": 0.8651583690797738, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6342 + }, + { + "epoch": 0.06343, + "grad_norm": 0.8187406441949702, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 6343 + }, + { + "epoch": 0.06344, + "grad_norm": 0.7988158937162896, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 6344 + }, + { + "epoch": 0.06345, + "grad_norm": 0.687447130567446, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 6345 + }, + { + "epoch": 0.06346, + "grad_norm": 0.6836509296813537, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 6346 + }, + { + "epoch": 0.06347, + "grad_norm": 0.602167193506003, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 6347 + }, + { + "epoch": 0.06348, + "grad_norm": 0.6302011048648151, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 6348 + }, + { + "epoch": 0.06349, + "grad_norm": 0.6618589989171766, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 6349 + }, + { + "epoch": 0.0635, + "grad_norm": 0.6490573512543693, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 6350 + }, + { + "epoch": 0.06351, + "grad_norm": 0.7070645549014211, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 6351 + }, + { + "epoch": 0.06352, + "grad_norm": 0.8243864460260366, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 6352 + }, + { + "epoch": 0.06353, + "grad_norm": 0.8112381920938149, + "learning_rate": 0.003, + "loss": 4.067, + "step": 6353 + }, + { + "epoch": 0.06354, + "grad_norm": 0.7745487764122767, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 6354 + }, + { + "epoch": 0.06355, + "grad_norm": 0.735458520019424, + "learning_rate": 0.003, + "loss": 4.07, + "step": 6355 + }, + { + "epoch": 0.06356, + "grad_norm": 0.7581308667508398, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6356 + }, + { + "epoch": 0.06357, + "grad_norm": 0.9052897414123441, + "learning_rate": 0.003, + "loss": 4.086, + "step": 6357 + }, + { + "epoch": 0.06358, + "grad_norm": 0.9665347964876907, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 6358 + }, + { + "epoch": 0.06359, + "grad_norm": 0.9135139985785528, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 6359 + }, + { + "epoch": 0.0636, + "grad_norm": 0.8545737512188917, + "learning_rate": 0.003, + "loss": 4.0879, + "step": 6360 + }, + { + "epoch": 0.06361, + "grad_norm": 0.779747826568656, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 6361 + }, + { + "epoch": 0.06362, + "grad_norm": 0.6129789562968603, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 6362 + }, + { + "epoch": 0.06363, + "grad_norm": 0.6360672931333222, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 6363 + }, + { + "epoch": 0.06364, + "grad_norm": 0.6523765173277813, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 6364 + }, + { + "epoch": 0.06365, + "grad_norm": 0.6448193093278599, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 6365 + }, + { + "epoch": 0.06366, + "grad_norm": 0.7444675440582652, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 6366 + }, + { + "epoch": 0.06367, + "grad_norm": 0.707209914735337, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 6367 + }, + { + "epoch": 0.06368, + "grad_norm": 0.6265593326734563, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 6368 + }, + { + "epoch": 0.06369, + "grad_norm": 0.6074014320772793, + "learning_rate": 0.003, + "loss": 4.0903, + "step": 6369 + }, + { + "epoch": 0.0637, + "grad_norm": 0.5898488946053692, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 6370 + }, + { + "epoch": 0.06371, + "grad_norm": 0.5200473638438476, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 6371 + }, + { + "epoch": 0.06372, + "grad_norm": 0.4878716528665913, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 6372 + }, + { + "epoch": 0.06373, + "grad_norm": 0.5566385870283186, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 6373 + }, + { + "epoch": 0.06374, + "grad_norm": 0.6362550157703785, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 6374 + }, + { + "epoch": 0.06375, + "grad_norm": 0.7606026342290857, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 6375 + }, + { + "epoch": 0.06376, + "grad_norm": 0.8076545901138845, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 6376 + }, + { + "epoch": 0.06377, + "grad_norm": 0.7824655778991425, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 6377 + }, + { + "epoch": 0.06378, + "grad_norm": 0.8277480958740822, + "learning_rate": 0.003, + "loss": 4.073, + "step": 6378 + }, + { + "epoch": 0.06379, + "grad_norm": 0.8049207631945982, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 6379 + }, + { + "epoch": 0.0638, + "grad_norm": 0.8373586875805686, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 6380 + }, + { + "epoch": 0.06381, + "grad_norm": 0.797764375138758, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 6381 + }, + { + "epoch": 0.06382, + "grad_norm": 0.7419189026157499, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 6382 + }, + { + "epoch": 0.06383, + "grad_norm": 0.7208705722510353, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 6383 + }, + { + "epoch": 0.06384, + "grad_norm": 0.7092357890410904, + "learning_rate": 0.003, + "loss": 4.072, + "step": 6384 + }, + { + "epoch": 0.06385, + "grad_norm": 0.6975075316291944, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 6385 + }, + { + "epoch": 0.06386, + "grad_norm": 0.6422261901598125, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 6386 + }, + { + "epoch": 0.06387, + "grad_norm": 0.5920764710876266, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 6387 + }, + { + "epoch": 0.06388, + "grad_norm": 0.5561249679612766, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6388 + }, + { + "epoch": 0.06389, + "grad_norm": 0.6475537534995942, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 6389 + }, + { + "epoch": 0.0639, + "grad_norm": 0.6966222152392437, + "learning_rate": 0.003, + "loss": 4.051, + "step": 6390 + }, + { + "epoch": 0.06391, + "grad_norm": 0.6291281160529176, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 6391 + }, + { + "epoch": 0.06392, + "grad_norm": 0.5827978414579209, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 6392 + }, + { + "epoch": 0.06393, + "grad_norm": 0.5593317420943201, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 6393 + }, + { + "epoch": 0.06394, + "grad_norm": 0.5532042146592672, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 6394 + }, + { + "epoch": 0.06395, + "grad_norm": 0.5914285964028465, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 6395 + }, + { + "epoch": 0.06396, + "grad_norm": 0.5357877019286361, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 6396 + }, + { + "epoch": 0.06397, + "grad_norm": 0.4924792307686586, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 6397 + }, + { + "epoch": 0.06398, + "grad_norm": 0.5490276137941461, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 6398 + }, + { + "epoch": 0.06399, + "grad_norm": 0.6373217178594652, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 6399 + }, + { + "epoch": 0.064, + "grad_norm": 0.6952233715647519, + "learning_rate": 0.003, + "loss": 4.053, + "step": 6400 + }, + { + "epoch": 0.06401, + "grad_norm": 0.7328202839914232, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 6401 + }, + { + "epoch": 0.06402, + "grad_norm": 0.8033053712762801, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 6402 + }, + { + "epoch": 0.06403, + "grad_norm": 0.7958901208570356, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 6403 + }, + { + "epoch": 0.06404, + "grad_norm": 0.6826818086226633, + "learning_rate": 0.003, + "loss": 4.0954, + "step": 6404 + }, + { + "epoch": 0.06405, + "grad_norm": 0.6404992629160978, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 6405 + }, + { + "epoch": 0.06406, + "grad_norm": 0.6317129663857884, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 6406 + }, + { + "epoch": 0.06407, + "grad_norm": 0.8057484050404142, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 6407 + }, + { + "epoch": 0.06408, + "grad_norm": 0.8224424488424654, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 6408 + }, + { + "epoch": 0.06409, + "grad_norm": 0.7055888788752528, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 6409 + }, + { + "epoch": 0.0641, + "grad_norm": 0.6859969201609134, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 6410 + }, + { + "epoch": 0.06411, + "grad_norm": 0.6904707820733437, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6411 + }, + { + "epoch": 0.06412, + "grad_norm": 0.5167211488198764, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 6412 + }, + { + "epoch": 0.06413, + "grad_norm": 0.6019017175402134, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 6413 + }, + { + "epoch": 0.06414, + "grad_norm": 0.6471837378745416, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 6414 + }, + { + "epoch": 0.06415, + "grad_norm": 0.6593030566384788, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 6415 + }, + { + "epoch": 0.06416, + "grad_norm": 0.6869820180653547, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 6416 + }, + { + "epoch": 0.06417, + "grad_norm": 0.7043573100714681, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 6417 + }, + { + "epoch": 0.06418, + "grad_norm": 0.8016681084164472, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 6418 + }, + { + "epoch": 0.06419, + "grad_norm": 0.8930338861412029, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 6419 + }, + { + "epoch": 0.0642, + "grad_norm": 0.8840189336482596, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 6420 + }, + { + "epoch": 0.06421, + "grad_norm": 0.8470817757753566, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 6421 + }, + { + "epoch": 0.06422, + "grad_norm": 0.8916630679468114, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 6422 + }, + { + "epoch": 0.06423, + "grad_norm": 0.9224052779442101, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 6423 + }, + { + "epoch": 0.06424, + "grad_norm": 0.8595438119985029, + "learning_rate": 0.003, + "loss": 4.043, + "step": 6424 + }, + { + "epoch": 0.06425, + "grad_norm": 0.9461745578866211, + "learning_rate": 0.003, + "loss": 4.071, + "step": 6425 + }, + { + "epoch": 0.06426, + "grad_norm": 0.9597164167471725, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 6426 + }, + { + "epoch": 0.06427, + "grad_norm": 1.065485570034077, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 6427 + }, + { + "epoch": 0.06428, + "grad_norm": 0.9698455262447494, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 6428 + }, + { + "epoch": 0.06429, + "grad_norm": 0.8963219074158595, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 6429 + }, + { + "epoch": 0.0643, + "grad_norm": 0.823811760941885, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6430 + }, + { + "epoch": 0.06431, + "grad_norm": 0.7335090550049281, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 6431 + }, + { + "epoch": 0.06432, + "grad_norm": 0.5716327362326611, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 6432 + }, + { + "epoch": 0.06433, + "grad_norm": 0.540458609342651, + "learning_rate": 0.003, + "loss": 4.083, + "step": 6433 + }, + { + "epoch": 0.06434, + "grad_norm": 0.556099701778752, + "learning_rate": 0.003, + "loss": 4.028, + "step": 6434 + }, + { + "epoch": 0.06435, + "grad_norm": 0.5941947679007499, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 6435 + }, + { + "epoch": 0.06436, + "grad_norm": 0.6049894624991385, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 6436 + }, + { + "epoch": 0.06437, + "grad_norm": 0.5770686538638992, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 6437 + }, + { + "epoch": 0.06438, + "grad_norm": 0.5743891802370877, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 6438 + }, + { + "epoch": 0.06439, + "grad_norm": 0.5540692692309691, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 6439 + }, + { + "epoch": 0.0644, + "grad_norm": 0.5699343275480232, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 6440 + }, + { + "epoch": 0.06441, + "grad_norm": 0.6290097556687986, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 6441 + }, + { + "epoch": 0.06442, + "grad_norm": 0.5609468261094211, + "learning_rate": 0.003, + "loss": 4.043, + "step": 6442 + }, + { + "epoch": 0.06443, + "grad_norm": 0.48890249673530356, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 6443 + }, + { + "epoch": 0.06444, + "grad_norm": 0.5687392536635725, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 6444 + }, + { + "epoch": 0.06445, + "grad_norm": 0.6466225753769135, + "learning_rate": 0.003, + "loss": 4.057, + "step": 6445 + }, + { + "epoch": 0.06446, + "grad_norm": 0.7500478332929661, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 6446 + }, + { + "epoch": 0.06447, + "grad_norm": 0.7478572509803384, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 6447 + }, + { + "epoch": 0.06448, + "grad_norm": 0.7156367910600131, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 6448 + }, + { + "epoch": 0.06449, + "grad_norm": 0.8527754534253787, + "learning_rate": 0.003, + "loss": 4.0959, + "step": 6449 + }, + { + "epoch": 0.0645, + "grad_norm": 0.8948278946502339, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 6450 + }, + { + "epoch": 0.06451, + "grad_norm": 0.8417515874143715, + "learning_rate": 0.003, + "loss": 4.0929, + "step": 6451 + }, + { + "epoch": 0.06452, + "grad_norm": 0.8114166422336707, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6452 + }, + { + "epoch": 0.06453, + "grad_norm": 0.7280862825250426, + "learning_rate": 0.003, + "loss": 4.0925, + "step": 6453 + }, + { + "epoch": 0.06454, + "grad_norm": 0.8112910665352214, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 6454 + }, + { + "epoch": 0.06455, + "grad_norm": 0.7618080520601063, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 6455 + }, + { + "epoch": 0.06456, + "grad_norm": 0.7967250952351331, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 6456 + }, + { + "epoch": 0.06457, + "grad_norm": 0.8342513537138647, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 6457 + }, + { + "epoch": 0.06458, + "grad_norm": 0.7542284214351345, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 6458 + }, + { + "epoch": 0.06459, + "grad_norm": 0.7010353823459967, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 6459 + }, + { + "epoch": 0.0646, + "grad_norm": 0.7285951531116313, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 6460 + }, + { + "epoch": 0.06461, + "grad_norm": 0.6845937452994477, + "learning_rate": 0.003, + "loss": 4.0946, + "step": 6461 + }, + { + "epoch": 0.06462, + "grad_norm": 0.6765161935346288, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 6462 + }, + { + "epoch": 0.06463, + "grad_norm": 0.7307143564022078, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 6463 + }, + { + "epoch": 0.06464, + "grad_norm": 0.7771854288361449, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 6464 + }, + { + "epoch": 0.06465, + "grad_norm": 0.7541089717781073, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 6465 + }, + { + "epoch": 0.06466, + "grad_norm": 0.7878868807269837, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 6466 + }, + { + "epoch": 0.06467, + "grad_norm": 0.9852819713214327, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 6467 + }, + { + "epoch": 0.06468, + "grad_norm": 0.989449846589375, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 6468 + }, + { + "epoch": 0.06469, + "grad_norm": 0.9522741967760807, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 6469 + }, + { + "epoch": 0.0647, + "grad_norm": 0.8697606544563544, + "learning_rate": 0.003, + "loss": 4.0852, + "step": 6470 + }, + { + "epoch": 0.06471, + "grad_norm": 0.8528983119813925, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 6471 + }, + { + "epoch": 0.06472, + "grad_norm": 0.8174807965443759, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 6472 + }, + { + "epoch": 0.06473, + "grad_norm": 0.7729480920444346, + "learning_rate": 0.003, + "loss": 4.078, + "step": 6473 + }, + { + "epoch": 0.06474, + "grad_norm": 0.7398300296063443, + "learning_rate": 0.003, + "loss": 4.0962, + "step": 6474 + }, + { + "epoch": 0.06475, + "grad_norm": 0.6853391663937839, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 6475 + }, + { + "epoch": 0.06476, + "grad_norm": 0.6825095595288727, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 6476 + }, + { + "epoch": 0.06477, + "grad_norm": 0.6304207768638315, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 6477 + }, + { + "epoch": 0.06478, + "grad_norm": 0.5271129428403287, + "learning_rate": 0.003, + "loss": 4.03, + "step": 6478 + }, + { + "epoch": 0.06479, + "grad_norm": 0.5078373221162751, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 6479 + }, + { + "epoch": 0.0648, + "grad_norm": 0.5250361938719068, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 6480 + }, + { + "epoch": 0.06481, + "grad_norm": 0.4730545217737081, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 6481 + }, + { + "epoch": 0.06482, + "grad_norm": 0.44750624116234056, + "learning_rate": 0.003, + "loss": 4.045, + "step": 6482 + }, + { + "epoch": 0.06483, + "grad_norm": 0.39733202193790107, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 6483 + }, + { + "epoch": 0.06484, + "grad_norm": 0.4139285208653727, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 6484 + }, + { + "epoch": 0.06485, + "grad_norm": 0.47930786005142895, + "learning_rate": 0.003, + "loss": 4.0729, + "step": 6485 + }, + { + "epoch": 0.06486, + "grad_norm": 0.44602638988948634, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 6486 + }, + { + "epoch": 0.06487, + "grad_norm": 0.45328655354637104, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6487 + }, + { + "epoch": 0.06488, + "grad_norm": 0.48214474644053756, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 6488 + }, + { + "epoch": 0.06489, + "grad_norm": 0.4919891569893605, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 6489 + }, + { + "epoch": 0.0649, + "grad_norm": 0.461124877451797, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 6490 + }, + { + "epoch": 0.06491, + "grad_norm": 0.506716254470609, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 6491 + }, + { + "epoch": 0.06492, + "grad_norm": 0.6360150112907528, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 6492 + }, + { + "epoch": 0.06493, + "grad_norm": 0.7778541734528054, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 6493 + }, + { + "epoch": 0.06494, + "grad_norm": 0.7838790198626554, + "learning_rate": 0.003, + "loss": 4.061, + "step": 6494 + }, + { + "epoch": 0.06495, + "grad_norm": 0.710782860140501, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 6495 + }, + { + "epoch": 0.06496, + "grad_norm": 0.7525070808447979, + "learning_rate": 0.003, + "loss": 4.019, + "step": 6496 + }, + { + "epoch": 0.06497, + "grad_norm": 0.7970271146332446, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 6497 + }, + { + "epoch": 0.06498, + "grad_norm": 0.8403781192748999, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 6498 + }, + { + "epoch": 0.06499, + "grad_norm": 0.8657484057569527, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 6499 + }, + { + "epoch": 0.065, + "grad_norm": 0.9045214569764982, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 6500 + }, + { + "epoch": 0.06501, + "grad_norm": 0.7534286470631377, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 6501 + }, + { + "epoch": 0.06502, + "grad_norm": 0.6820815131608521, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 6502 + }, + { + "epoch": 0.06503, + "grad_norm": 0.6395812005207934, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 6503 + }, + { + "epoch": 0.06504, + "grad_norm": 0.5805801678225314, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 6504 + }, + { + "epoch": 0.06505, + "grad_norm": 0.5874992175952523, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 6505 + }, + { + "epoch": 0.06506, + "grad_norm": 0.5903991893590559, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 6506 + }, + { + "epoch": 0.06507, + "grad_norm": 0.6727146329207334, + "learning_rate": 0.003, + "loss": 4.011, + "step": 6507 + }, + { + "epoch": 0.06508, + "grad_norm": 0.7470291932026575, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 6508 + }, + { + "epoch": 0.06509, + "grad_norm": 0.8636944960208432, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 6509 + }, + { + "epoch": 0.0651, + "grad_norm": 0.9509012632561171, + "learning_rate": 0.003, + "loss": 4.0862, + "step": 6510 + }, + { + "epoch": 0.06511, + "grad_norm": 1.0209362110535056, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 6511 + }, + { + "epoch": 0.06512, + "grad_norm": 0.9396389767649208, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 6512 + }, + { + "epoch": 0.06513, + "grad_norm": 0.8863510446414791, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 6513 + }, + { + "epoch": 0.06514, + "grad_norm": 0.88843779128409, + "learning_rate": 0.003, + "loss": 4.063, + "step": 6514 + }, + { + "epoch": 0.06515, + "grad_norm": 0.7905878355878253, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 6515 + }, + { + "epoch": 0.06516, + "grad_norm": 0.7247521791928306, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 6516 + }, + { + "epoch": 0.06517, + "grad_norm": 0.7508843456475495, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 6517 + }, + { + "epoch": 0.06518, + "grad_norm": 0.8499771695760552, + "learning_rate": 0.003, + "loss": 4.0863, + "step": 6518 + }, + { + "epoch": 0.06519, + "grad_norm": 0.9604273899869331, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 6519 + }, + { + "epoch": 0.0652, + "grad_norm": 0.9032252898281361, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 6520 + }, + { + "epoch": 0.06521, + "grad_norm": 0.9300105004115965, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 6521 + }, + { + "epoch": 0.06522, + "grad_norm": 0.9029258540712023, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 6522 + }, + { + "epoch": 0.06523, + "grad_norm": 0.8298203165125059, + "learning_rate": 0.003, + "loss": 4.068, + "step": 6523 + }, + { + "epoch": 0.06524, + "grad_norm": 0.8412505796236491, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 6524 + }, + { + "epoch": 0.06525, + "grad_norm": 0.793004203273733, + "learning_rate": 0.003, + "loss": 4.0813, + "step": 6525 + }, + { + "epoch": 0.06526, + "grad_norm": 0.7350334029742837, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 6526 + }, + { + "epoch": 0.06527, + "grad_norm": 0.681313098868743, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 6527 + }, + { + "epoch": 0.06528, + "grad_norm": 0.6361139323912299, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 6528 + }, + { + "epoch": 0.06529, + "grad_norm": 0.5751576917170754, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 6529 + }, + { + "epoch": 0.0653, + "grad_norm": 0.49677309350818466, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 6530 + }, + { + "epoch": 0.06531, + "grad_norm": 0.4897954481873861, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 6531 + }, + { + "epoch": 0.06532, + "grad_norm": 0.46610457125063426, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 6532 + }, + { + "epoch": 0.06533, + "grad_norm": 0.46853727109781146, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 6533 + }, + { + "epoch": 0.06534, + "grad_norm": 0.47506669983371863, + "learning_rate": 0.003, + "loss": 4.057, + "step": 6534 + }, + { + "epoch": 0.06535, + "grad_norm": 0.49297039594180225, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 6535 + }, + { + "epoch": 0.06536, + "grad_norm": 0.5959346456734156, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 6536 + }, + { + "epoch": 0.06537, + "grad_norm": 0.7356684833975462, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 6537 + }, + { + "epoch": 0.06538, + "grad_norm": 0.8799947058605172, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 6538 + }, + { + "epoch": 0.06539, + "grad_norm": 0.8832291307396867, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 6539 + }, + { + "epoch": 0.0654, + "grad_norm": 0.8357171766608269, + "learning_rate": 0.003, + "loss": 4.0976, + "step": 6540 + }, + { + "epoch": 0.06541, + "grad_norm": 0.7579626486054626, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 6541 + }, + { + "epoch": 0.06542, + "grad_norm": 0.6128732127331447, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 6542 + }, + { + "epoch": 0.06543, + "grad_norm": 0.6063980555795855, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 6543 + }, + { + "epoch": 0.06544, + "grad_norm": 0.6623830040032186, + "learning_rate": 0.003, + "loss": 4.056, + "step": 6544 + }, + { + "epoch": 0.06545, + "grad_norm": 0.6526469866095655, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 6545 + }, + { + "epoch": 0.06546, + "grad_norm": 0.5649705421081817, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 6546 + }, + { + "epoch": 0.06547, + "grad_norm": 0.5000723493633451, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 6547 + }, + { + "epoch": 0.06548, + "grad_norm": 0.560824918181252, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 6548 + }, + { + "epoch": 0.06549, + "grad_norm": 0.5730745771976296, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 6549 + }, + { + "epoch": 0.0655, + "grad_norm": 0.6386929012728552, + "learning_rate": 0.003, + "loss": 4.027, + "step": 6550 + }, + { + "epoch": 0.06551, + "grad_norm": 0.7624183346705704, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 6551 + }, + { + "epoch": 0.06552, + "grad_norm": 0.7862836321280505, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 6552 + }, + { + "epoch": 0.06553, + "grad_norm": 0.7732868511586495, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 6553 + }, + { + "epoch": 0.06554, + "grad_norm": 0.7991123610460387, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 6554 + }, + { + "epoch": 0.06555, + "grad_norm": 0.7119403367817155, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 6555 + }, + { + "epoch": 0.06556, + "grad_norm": 0.6024496341196729, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 6556 + }, + { + "epoch": 0.06557, + "grad_norm": 0.5998054894387628, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 6557 + }, + { + "epoch": 0.06558, + "grad_norm": 0.6574374427986027, + "learning_rate": 0.003, + "loss": 4.075, + "step": 6558 + }, + { + "epoch": 0.06559, + "grad_norm": 0.6839651026197617, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6559 + }, + { + "epoch": 0.0656, + "grad_norm": 0.7025037835789826, + "learning_rate": 0.003, + "loss": 4.053, + "step": 6560 + }, + { + "epoch": 0.06561, + "grad_norm": 0.9171977906786489, + "learning_rate": 0.003, + "loss": 4.0786, + "step": 6561 + }, + { + "epoch": 0.06562, + "grad_norm": 0.9912775009825152, + "learning_rate": 0.003, + "loss": 4.042, + "step": 6562 + }, + { + "epoch": 0.06563, + "grad_norm": 0.7977312974185572, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 6563 + }, + { + "epoch": 0.06564, + "grad_norm": 0.6938342655105021, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 6564 + }, + { + "epoch": 0.06565, + "grad_norm": 0.6633276189423883, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 6565 + }, + { + "epoch": 0.06566, + "grad_norm": 0.5718346858620951, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 6566 + }, + { + "epoch": 0.06567, + "grad_norm": 0.621209340892742, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 6567 + }, + { + "epoch": 0.06568, + "grad_norm": 0.6166563581296096, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 6568 + }, + { + "epoch": 0.06569, + "grad_norm": 0.5803433645966917, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 6569 + }, + { + "epoch": 0.0657, + "grad_norm": 0.5718002435973948, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 6570 + }, + { + "epoch": 0.06571, + "grad_norm": 0.5054885491451991, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 6571 + }, + { + "epoch": 0.06572, + "grad_norm": 0.5576816982309836, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 6572 + }, + { + "epoch": 0.06573, + "grad_norm": 0.6377836794934589, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 6573 + }, + { + "epoch": 0.06574, + "grad_norm": 0.6613539815009216, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 6574 + }, + { + "epoch": 0.06575, + "grad_norm": 0.765092207809946, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 6575 + }, + { + "epoch": 0.06576, + "grad_norm": 0.8899161068359215, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 6576 + }, + { + "epoch": 0.06577, + "grad_norm": 0.8594500632650697, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 6577 + }, + { + "epoch": 0.06578, + "grad_norm": 0.7106297854597906, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 6578 + }, + { + "epoch": 0.06579, + "grad_norm": 0.6525256532338872, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 6579 + }, + { + "epoch": 0.0658, + "grad_norm": 0.6093055546577442, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 6580 + }, + { + "epoch": 0.06581, + "grad_norm": 0.6755030956444104, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 6581 + }, + { + "epoch": 0.06582, + "grad_norm": 0.7464847148267115, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 6582 + }, + { + "epoch": 0.06583, + "grad_norm": 0.7864378034971011, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 6583 + }, + { + "epoch": 0.06584, + "grad_norm": 0.8829209255089074, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 6584 + }, + { + "epoch": 0.06585, + "grad_norm": 0.9526472377106489, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 6585 + }, + { + "epoch": 0.06586, + "grad_norm": 1.0296141159762646, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 6586 + }, + { + "epoch": 0.06587, + "grad_norm": 1.0408357276081976, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 6587 + }, + { + "epoch": 0.06588, + "grad_norm": 0.9614313651657951, + "learning_rate": 0.003, + "loss": 4.1065, + "step": 6588 + }, + { + "epoch": 0.06589, + "grad_norm": 0.9205666760199722, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 6589 + }, + { + "epoch": 0.0659, + "grad_norm": 0.7104773586163216, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 6590 + }, + { + "epoch": 0.06591, + "grad_norm": 0.6893091524593887, + "learning_rate": 0.003, + "loss": 4.117, + "step": 6591 + }, + { + "epoch": 0.06592, + "grad_norm": 0.5902474713819571, + "learning_rate": 0.003, + "loss": 4.0766, + "step": 6592 + }, + { + "epoch": 0.06593, + "grad_norm": 0.5426810063670913, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 6593 + }, + { + "epoch": 0.06594, + "grad_norm": 0.5983413957054662, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 6594 + }, + { + "epoch": 0.06595, + "grad_norm": 0.6163618971044713, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 6595 + }, + { + "epoch": 0.06596, + "grad_norm": 0.565684789447081, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 6596 + }, + { + "epoch": 0.06597, + "grad_norm": 0.5014352963989382, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 6597 + }, + { + "epoch": 0.06598, + "grad_norm": 0.492856315489037, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 6598 + }, + { + "epoch": 0.06599, + "grad_norm": 0.48385791309278686, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 6599 + }, + { + "epoch": 0.066, + "grad_norm": 0.5060080270728584, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 6600 + }, + { + "epoch": 0.06601, + "grad_norm": 0.48903368108016115, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 6601 + }, + { + "epoch": 0.06602, + "grad_norm": 0.5004627241171659, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 6602 + }, + { + "epoch": 0.06603, + "grad_norm": 0.4989311463588693, + "learning_rate": 0.003, + "loss": 4.04, + "step": 6603 + }, + { + "epoch": 0.06604, + "grad_norm": 0.492886434715741, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 6604 + }, + { + "epoch": 0.06605, + "grad_norm": 0.5994858837988067, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 6605 + }, + { + "epoch": 0.06606, + "grad_norm": 0.7764332220286436, + "learning_rate": 0.003, + "loss": 4.048, + "step": 6606 + }, + { + "epoch": 0.06607, + "grad_norm": 0.9771869105331121, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 6607 + }, + { + "epoch": 0.06608, + "grad_norm": 0.9502696250282155, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 6608 + }, + { + "epoch": 0.06609, + "grad_norm": 0.720043085527814, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 6609 + }, + { + "epoch": 0.0661, + "grad_norm": 0.7433393007661245, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 6610 + }, + { + "epoch": 0.06611, + "grad_norm": 0.7992498817022772, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 6611 + }, + { + "epoch": 0.06612, + "grad_norm": 0.7571856431455781, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 6612 + }, + { + "epoch": 0.06613, + "grad_norm": 0.7970549304744902, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 6613 + }, + { + "epoch": 0.06614, + "grad_norm": 0.824882214983017, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 6614 + }, + { + "epoch": 0.06615, + "grad_norm": 0.8549526311436051, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 6615 + }, + { + "epoch": 0.06616, + "grad_norm": 0.8317291614793756, + "learning_rate": 0.003, + "loss": 4.1027, + "step": 6616 + }, + { + "epoch": 0.06617, + "grad_norm": 0.8792796904757433, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 6617 + }, + { + "epoch": 0.06618, + "grad_norm": 0.7851122664760067, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 6618 + }, + { + "epoch": 0.06619, + "grad_norm": 0.8161562004422934, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6619 + }, + { + "epoch": 0.0662, + "grad_norm": 0.763839030368717, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 6620 + }, + { + "epoch": 0.06621, + "grad_norm": 0.6514588200431028, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 6621 + }, + { + "epoch": 0.06622, + "grad_norm": 0.7103758721136519, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 6622 + }, + { + "epoch": 0.06623, + "grad_norm": 0.793432436740194, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 6623 + }, + { + "epoch": 0.06624, + "grad_norm": 0.7262194928633572, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 6624 + }, + { + "epoch": 0.06625, + "grad_norm": 0.5873265444774995, + "learning_rate": 0.003, + "loss": 4.058, + "step": 6625 + }, + { + "epoch": 0.06626, + "grad_norm": 0.6084928388160927, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 6626 + }, + { + "epoch": 0.06627, + "grad_norm": 0.5711088804076748, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 6627 + }, + { + "epoch": 0.06628, + "grad_norm": 0.6484160142344234, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 6628 + }, + { + "epoch": 0.06629, + "grad_norm": 0.6407615738001963, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 6629 + }, + { + "epoch": 0.0663, + "grad_norm": 0.594946788669324, + "learning_rate": 0.003, + "loss": 4.0767, + "step": 6630 + }, + { + "epoch": 0.06631, + "grad_norm": 0.6273070491799196, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 6631 + }, + { + "epoch": 0.06632, + "grad_norm": 0.6059632791250404, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 6632 + }, + { + "epoch": 0.06633, + "grad_norm": 0.5373297595756357, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 6633 + }, + { + "epoch": 0.06634, + "grad_norm": 0.5083870365599727, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 6634 + }, + { + "epoch": 0.06635, + "grad_norm": 0.4986740486988127, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 6635 + }, + { + "epoch": 0.06636, + "grad_norm": 0.429977217365301, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 6636 + }, + { + "epoch": 0.06637, + "grad_norm": 0.47427760361099763, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 6637 + }, + { + "epoch": 0.06638, + "grad_norm": 0.5967936174207519, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 6638 + }, + { + "epoch": 0.06639, + "grad_norm": 0.6627661092188125, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 6639 + }, + { + "epoch": 0.0664, + "grad_norm": 0.7127572180746081, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 6640 + }, + { + "epoch": 0.06641, + "grad_norm": 0.7113816080717646, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 6641 + }, + { + "epoch": 0.06642, + "grad_norm": 0.6844295703766224, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 6642 + }, + { + "epoch": 0.06643, + "grad_norm": 0.7179577860715133, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 6643 + }, + { + "epoch": 0.06644, + "grad_norm": 0.7445108322166758, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 6644 + }, + { + "epoch": 0.06645, + "grad_norm": 0.7316510901412201, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 6645 + }, + { + "epoch": 0.06646, + "grad_norm": 0.7294842503897733, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 6646 + }, + { + "epoch": 0.06647, + "grad_norm": 0.7783483633762008, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 6647 + }, + { + "epoch": 0.06648, + "grad_norm": 0.7258605724866444, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 6648 + }, + { + "epoch": 0.06649, + "grad_norm": 0.6478833448519665, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 6649 + }, + { + "epoch": 0.0665, + "grad_norm": 0.6986331060561505, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 6650 + }, + { + "epoch": 0.06651, + "grad_norm": 0.5954422404898819, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 6651 + }, + { + "epoch": 0.06652, + "grad_norm": 0.6349889803739441, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 6652 + }, + { + "epoch": 0.06653, + "grad_norm": 0.8523800204576891, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 6653 + }, + { + "epoch": 0.06654, + "grad_norm": 0.8964420098024196, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 6654 + }, + { + "epoch": 0.06655, + "grad_norm": 0.9081660178772296, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 6655 + }, + { + "epoch": 0.06656, + "grad_norm": 1.0194000581635678, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 6656 + }, + { + "epoch": 0.06657, + "grad_norm": 1.1109958944084373, + "learning_rate": 0.003, + "loss": 4.0776, + "step": 6657 + }, + { + "epoch": 0.06658, + "grad_norm": 0.9077557277393085, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 6658 + }, + { + "epoch": 0.06659, + "grad_norm": 0.8383532047850902, + "learning_rate": 0.003, + "loss": 4.0782, + "step": 6659 + }, + { + "epoch": 0.0666, + "grad_norm": 0.6881419056783588, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 6660 + }, + { + "epoch": 0.06661, + "grad_norm": 0.7123563425948003, + "learning_rate": 0.003, + "loss": 4.052, + "step": 6661 + }, + { + "epoch": 0.06662, + "grad_norm": 0.74978741290443, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 6662 + }, + { + "epoch": 0.06663, + "grad_norm": 0.7567452901534678, + "learning_rate": 0.003, + "loss": 4.1131, + "step": 6663 + }, + { + "epoch": 0.06664, + "grad_norm": 0.6773782682153159, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 6664 + }, + { + "epoch": 0.06665, + "grad_norm": 0.6596420317817953, + "learning_rate": 0.003, + "loss": 4.0952, + "step": 6665 + }, + { + "epoch": 0.06666, + "grad_norm": 0.7469327106701796, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 6666 + }, + { + "epoch": 0.06667, + "grad_norm": 0.7300447174322503, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 6667 + }, + { + "epoch": 0.06668, + "grad_norm": 0.7428979324911045, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 6668 + }, + { + "epoch": 0.06669, + "grad_norm": 0.6824668084121378, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 6669 + }, + { + "epoch": 0.0667, + "grad_norm": 0.7647483456657251, + "learning_rate": 0.003, + "loss": 4.069, + "step": 6670 + }, + { + "epoch": 0.06671, + "grad_norm": 0.7746427596530956, + "learning_rate": 0.003, + "loss": 4.066, + "step": 6671 + }, + { + "epoch": 0.06672, + "grad_norm": 0.7097746957080621, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 6672 + }, + { + "epoch": 0.06673, + "grad_norm": 0.7112700868221732, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 6673 + }, + { + "epoch": 0.06674, + "grad_norm": 0.6397732093712012, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 6674 + }, + { + "epoch": 0.06675, + "grad_norm": 0.6520340296337761, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 6675 + }, + { + "epoch": 0.06676, + "grad_norm": 0.6333277525295771, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 6676 + }, + { + "epoch": 0.06677, + "grad_norm": 0.5997823871076088, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 6677 + }, + { + "epoch": 0.06678, + "grad_norm": 0.6440753158740498, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 6678 + }, + { + "epoch": 0.06679, + "grad_norm": 0.7199390266397416, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 6679 + }, + { + "epoch": 0.0668, + "grad_norm": 0.7140581191194553, + "learning_rate": 0.003, + "loss": 4.0904, + "step": 6680 + }, + { + "epoch": 0.06681, + "grad_norm": 0.7313344473924521, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 6681 + }, + { + "epoch": 0.06682, + "grad_norm": 0.6701809814974284, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 6682 + }, + { + "epoch": 0.06683, + "grad_norm": 0.5325499780469609, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 6683 + }, + { + "epoch": 0.06684, + "grad_norm": 0.5601897624784417, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 6684 + }, + { + "epoch": 0.06685, + "grad_norm": 0.5319217965913903, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 6685 + }, + { + "epoch": 0.06686, + "grad_norm": 0.5773286353132502, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 6686 + }, + { + "epoch": 0.06687, + "grad_norm": 0.6422691262372372, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 6687 + }, + { + "epoch": 0.06688, + "grad_norm": 0.623736923148395, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 6688 + }, + { + "epoch": 0.06689, + "grad_norm": 0.6508635819889379, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 6689 + }, + { + "epoch": 0.0669, + "grad_norm": 0.7372378563833888, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 6690 + }, + { + "epoch": 0.06691, + "grad_norm": 0.6914168637826256, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 6691 + }, + { + "epoch": 0.06692, + "grad_norm": 0.7653226401535247, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 6692 + }, + { + "epoch": 0.06693, + "grad_norm": 0.9605316227774986, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 6693 + }, + { + "epoch": 0.06694, + "grad_norm": 1.0684149459301475, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6694 + }, + { + "epoch": 0.06695, + "grad_norm": 0.856289433408192, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 6695 + }, + { + "epoch": 0.06696, + "grad_norm": 0.7889378021888084, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 6696 + }, + { + "epoch": 0.06697, + "grad_norm": 0.765457303190137, + "learning_rate": 0.003, + "loss": 4.065, + "step": 6697 + }, + { + "epoch": 0.06698, + "grad_norm": 0.7750756702828132, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 6698 + }, + { + "epoch": 0.06699, + "grad_norm": 0.7850999039846963, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 6699 + }, + { + "epoch": 0.067, + "grad_norm": 0.7333962788652812, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 6700 + }, + { + "epoch": 0.06701, + "grad_norm": 0.6862577636720708, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 6701 + }, + { + "epoch": 0.06702, + "grad_norm": 0.6807236985489369, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 6702 + }, + { + "epoch": 0.06703, + "grad_norm": 0.7186159263376848, + "learning_rate": 0.003, + "loss": 4.0751, + "step": 6703 + }, + { + "epoch": 0.06704, + "grad_norm": 0.7374707118345275, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 6704 + }, + { + "epoch": 0.06705, + "grad_norm": 0.7930615610589977, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 6705 + }, + { + "epoch": 0.06706, + "grad_norm": 0.8815823929188342, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 6706 + }, + { + "epoch": 0.06707, + "grad_norm": 0.7697415849330416, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 6707 + }, + { + "epoch": 0.06708, + "grad_norm": 0.6396516916396368, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 6708 + }, + { + "epoch": 0.06709, + "grad_norm": 0.6380376444830074, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 6709 + }, + { + "epoch": 0.0671, + "grad_norm": 0.6953693168812163, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 6710 + }, + { + "epoch": 0.06711, + "grad_norm": 0.7092953392208627, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 6711 + }, + { + "epoch": 0.06712, + "grad_norm": 0.5810024059869595, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 6712 + }, + { + "epoch": 0.06713, + "grad_norm": 0.6219309427716201, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 6713 + }, + { + "epoch": 0.06714, + "grad_norm": 0.6984801169644851, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 6714 + }, + { + "epoch": 0.06715, + "grad_norm": 0.651421153392065, + "learning_rate": 0.003, + "loss": 4.048, + "step": 6715 + }, + { + "epoch": 0.06716, + "grad_norm": 0.5902010710832922, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 6716 + }, + { + "epoch": 0.06717, + "grad_norm": 0.5940029619159725, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 6717 + }, + { + "epoch": 0.06718, + "grad_norm": 0.6345717355475652, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 6718 + }, + { + "epoch": 0.06719, + "grad_norm": 0.688612876541075, + "learning_rate": 0.003, + "loss": 4.038, + "step": 6719 + }, + { + "epoch": 0.0672, + "grad_norm": 0.7377684892342624, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 6720 + }, + { + "epoch": 0.06721, + "grad_norm": 0.7587881237352552, + "learning_rate": 0.003, + "loss": 4.07, + "step": 6721 + }, + { + "epoch": 0.06722, + "grad_norm": 0.7632773534842684, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 6722 + }, + { + "epoch": 0.06723, + "grad_norm": 0.6993838051811655, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6723 + }, + { + "epoch": 0.06724, + "grad_norm": 0.8341709187614653, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 6724 + }, + { + "epoch": 0.06725, + "grad_norm": 0.8882088184672379, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 6725 + }, + { + "epoch": 0.06726, + "grad_norm": 0.9568613209981662, + "learning_rate": 0.003, + "loss": 4.1018, + "step": 6726 + }, + { + "epoch": 0.06727, + "grad_norm": 0.9091590261680762, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 6727 + }, + { + "epoch": 0.06728, + "grad_norm": 0.8797035226845034, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 6728 + }, + { + "epoch": 0.06729, + "grad_norm": 0.8123692708209892, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 6729 + }, + { + "epoch": 0.0673, + "grad_norm": 0.8317752037287318, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 6730 + }, + { + "epoch": 0.06731, + "grad_norm": 0.8690051356862507, + "learning_rate": 0.003, + "loss": 4.039, + "step": 6731 + }, + { + "epoch": 0.06732, + "grad_norm": 0.8695267641440078, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 6732 + }, + { + "epoch": 0.06733, + "grad_norm": 0.8488243850405971, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 6733 + }, + { + "epoch": 0.06734, + "grad_norm": 0.8249900273162015, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 6734 + }, + { + "epoch": 0.06735, + "grad_norm": 0.8173339321221517, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 6735 + }, + { + "epoch": 0.06736, + "grad_norm": 0.7617761453997925, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 6736 + }, + { + "epoch": 0.06737, + "grad_norm": 0.6867964686232342, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 6737 + }, + { + "epoch": 0.06738, + "grad_norm": 0.6359596586254495, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 6738 + }, + { + "epoch": 0.06739, + "grad_norm": 0.6510352299047496, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 6739 + }, + { + "epoch": 0.0674, + "grad_norm": 0.669369803613436, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 6740 + }, + { + "epoch": 0.06741, + "grad_norm": 0.5787803151348337, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 6741 + }, + { + "epoch": 0.06742, + "grad_norm": 0.5143261655022072, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 6742 + }, + { + "epoch": 0.06743, + "grad_norm": 0.5239924617566188, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 6743 + }, + { + "epoch": 0.06744, + "grad_norm": 0.5115036071413046, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 6744 + }, + { + "epoch": 0.06745, + "grad_norm": 0.5267271686022948, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 6745 + }, + { + "epoch": 0.06746, + "grad_norm": 0.6018686090259493, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 6746 + }, + { + "epoch": 0.06747, + "grad_norm": 0.6339964272094099, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 6747 + }, + { + "epoch": 0.06748, + "grad_norm": 0.5825546251980036, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 6748 + }, + { + "epoch": 0.06749, + "grad_norm": 0.527797162619652, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 6749 + }, + { + "epoch": 0.0675, + "grad_norm": 0.4994973334174798, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 6750 + }, + { + "epoch": 0.06751, + "grad_norm": 0.5644297867133525, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 6751 + }, + { + "epoch": 0.06752, + "grad_norm": 0.7110595343687313, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 6752 + }, + { + "epoch": 0.06753, + "grad_norm": 0.9530987317139209, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 6753 + }, + { + "epoch": 0.06754, + "grad_norm": 1.2175337086478664, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 6754 + }, + { + "epoch": 0.06755, + "grad_norm": 0.6814156527123848, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 6755 + }, + { + "epoch": 0.06756, + "grad_norm": 0.7346345035789635, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 6756 + }, + { + "epoch": 0.06757, + "grad_norm": 0.8428892048558227, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 6757 + }, + { + "epoch": 0.06758, + "grad_norm": 0.8294169762359505, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 6758 + }, + { + "epoch": 0.06759, + "grad_norm": 0.771369239238409, + "learning_rate": 0.003, + "loss": 4.0838, + "step": 6759 + }, + { + "epoch": 0.0676, + "grad_norm": 0.7501594743908848, + "learning_rate": 0.003, + "loss": 4.068, + "step": 6760 + }, + { + "epoch": 0.06761, + "grad_norm": 0.7462112384846398, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 6761 + }, + { + "epoch": 0.06762, + "grad_norm": 0.8269915154365834, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 6762 + }, + { + "epoch": 0.06763, + "grad_norm": 0.738398212272419, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 6763 + }, + { + "epoch": 0.06764, + "grad_norm": 0.7877446706349976, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 6764 + }, + { + "epoch": 0.06765, + "grad_norm": 0.7077874819125928, + "learning_rate": 0.003, + "loss": 4.058, + "step": 6765 + }, + { + "epoch": 0.06766, + "grad_norm": 0.6692111555018156, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 6766 + }, + { + "epoch": 0.06767, + "grad_norm": 0.6294973638059332, + "learning_rate": 0.003, + "loss": 4.069, + "step": 6767 + }, + { + "epoch": 0.06768, + "grad_norm": 0.5272092801916627, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 6768 + }, + { + "epoch": 0.06769, + "grad_norm": 0.5726724310607844, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 6769 + }, + { + "epoch": 0.0677, + "grad_norm": 0.6158933601953914, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6770 + }, + { + "epoch": 0.06771, + "grad_norm": 0.6144603104500397, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 6771 + }, + { + "epoch": 0.06772, + "grad_norm": 0.5642406503999594, + "learning_rate": 0.003, + "loss": 4.051, + "step": 6772 + }, + { + "epoch": 0.06773, + "grad_norm": 0.4980640705608534, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 6773 + }, + { + "epoch": 0.06774, + "grad_norm": 0.4491782450287272, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 6774 + }, + { + "epoch": 0.06775, + "grad_norm": 0.47818412536483945, + "learning_rate": 0.003, + "loss": 4.047, + "step": 6775 + }, + { + "epoch": 0.06776, + "grad_norm": 0.5377299038962707, + "learning_rate": 0.003, + "loss": 4.0753, + "step": 6776 + }, + { + "epoch": 0.06777, + "grad_norm": 0.5305017990865181, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6777 + }, + { + "epoch": 0.06778, + "grad_norm": 0.553572225558693, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 6778 + }, + { + "epoch": 0.06779, + "grad_norm": 0.6130466018898353, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 6779 + }, + { + "epoch": 0.0678, + "grad_norm": 0.7818804333096511, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 6780 + }, + { + "epoch": 0.06781, + "grad_norm": 0.9983418223313943, + "learning_rate": 0.003, + "loss": 4.0889, + "step": 6781 + }, + { + "epoch": 0.06782, + "grad_norm": 1.0219706338872863, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 6782 + }, + { + "epoch": 0.06783, + "grad_norm": 0.8433187490658968, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 6783 + }, + { + "epoch": 0.06784, + "grad_norm": 0.8082313282764686, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 6784 + }, + { + "epoch": 0.06785, + "grad_norm": 0.7964037924243274, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 6785 + }, + { + "epoch": 0.06786, + "grad_norm": 0.8641671290303337, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 6786 + }, + { + "epoch": 0.06787, + "grad_norm": 0.7904458500579361, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 6787 + }, + { + "epoch": 0.06788, + "grad_norm": 0.7617206933533606, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 6788 + }, + { + "epoch": 0.06789, + "grad_norm": 0.7384432620145864, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 6789 + }, + { + "epoch": 0.0679, + "grad_norm": 0.6547770900819717, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 6790 + }, + { + "epoch": 0.06791, + "grad_norm": 0.642125081555029, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 6791 + }, + { + "epoch": 0.06792, + "grad_norm": 0.592179708475192, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 6792 + }, + { + "epoch": 0.06793, + "grad_norm": 0.6551416303028926, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 6793 + }, + { + "epoch": 0.06794, + "grad_norm": 0.6443852912448402, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 6794 + }, + { + "epoch": 0.06795, + "grad_norm": 0.6374481110969081, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 6795 + }, + { + "epoch": 0.06796, + "grad_norm": 0.6401289208621875, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 6796 + }, + { + "epoch": 0.06797, + "grad_norm": 0.6978026627351833, + "learning_rate": 0.003, + "loss": 4.0809, + "step": 6797 + }, + { + "epoch": 0.06798, + "grad_norm": 0.663755212651515, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 6798 + }, + { + "epoch": 0.06799, + "grad_norm": 0.7018813995434555, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 6799 + }, + { + "epoch": 0.068, + "grad_norm": 0.7423380987015208, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6800 + }, + { + "epoch": 0.06801, + "grad_norm": 0.8129302508106757, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 6801 + }, + { + "epoch": 0.06802, + "grad_norm": 0.8331882541865985, + "learning_rate": 0.003, + "loss": 4.053, + "step": 6802 + }, + { + "epoch": 0.06803, + "grad_norm": 0.8496091843176673, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 6803 + }, + { + "epoch": 0.06804, + "grad_norm": 0.8145291712039313, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 6804 + }, + { + "epoch": 0.06805, + "grad_norm": 1.0070886893253241, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 6805 + }, + { + "epoch": 0.06806, + "grad_norm": 1.1242325842695176, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 6806 + }, + { + "epoch": 0.06807, + "grad_norm": 0.894381675435843, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 6807 + }, + { + "epoch": 0.06808, + "grad_norm": 0.7262105928405844, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 6808 + }, + { + "epoch": 0.06809, + "grad_norm": 0.6485024809248104, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 6809 + }, + { + "epoch": 0.0681, + "grad_norm": 0.7192555533753532, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 6810 + }, + { + "epoch": 0.06811, + "grad_norm": 0.6348153358760394, + "learning_rate": 0.003, + "loss": 4.0864, + "step": 6811 + }, + { + "epoch": 0.06812, + "grad_norm": 0.59655126707365, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 6812 + }, + { + "epoch": 0.06813, + "grad_norm": 0.47298779671993707, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 6813 + }, + { + "epoch": 0.06814, + "grad_norm": 0.4567142709424015, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 6814 + }, + { + "epoch": 0.06815, + "grad_norm": 0.47542321715382296, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 6815 + }, + { + "epoch": 0.06816, + "grad_norm": 0.47474338525818915, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 6816 + }, + { + "epoch": 0.06817, + "grad_norm": 0.4512621027490709, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 6817 + }, + { + "epoch": 0.06818, + "grad_norm": 0.4241061335534772, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 6818 + }, + { + "epoch": 0.06819, + "grad_norm": 0.5241748846859439, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 6819 + }, + { + "epoch": 0.0682, + "grad_norm": 0.598130661648289, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 6820 + }, + { + "epoch": 0.06821, + "grad_norm": 0.6577033273856296, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 6821 + }, + { + "epoch": 0.06822, + "grad_norm": 0.7179344088259381, + "learning_rate": 0.003, + "loss": 4.032, + "step": 6822 + }, + { + "epoch": 0.06823, + "grad_norm": 0.8965286307327501, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 6823 + }, + { + "epoch": 0.06824, + "grad_norm": 1.0412813376409753, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 6824 + }, + { + "epoch": 0.06825, + "grad_norm": 0.8629047161683553, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 6825 + }, + { + "epoch": 0.06826, + "grad_norm": 0.7006576897177896, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 6826 + }, + { + "epoch": 0.06827, + "grad_norm": 0.7479235694056444, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 6827 + }, + { + "epoch": 0.06828, + "grad_norm": 0.6921334319094415, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 6828 + }, + { + "epoch": 0.06829, + "grad_norm": 0.659216249076333, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 6829 + }, + { + "epoch": 0.0683, + "grad_norm": 0.640774302012818, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 6830 + }, + { + "epoch": 0.06831, + "grad_norm": 0.6691191829838372, + "learning_rate": 0.003, + "loss": 4.0764, + "step": 6831 + }, + { + "epoch": 0.06832, + "grad_norm": 0.7271225997191756, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 6832 + }, + { + "epoch": 0.06833, + "grad_norm": 0.8379753431103475, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 6833 + }, + { + "epoch": 0.06834, + "grad_norm": 1.0364003673087832, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 6834 + }, + { + "epoch": 0.06835, + "grad_norm": 0.8498777269946786, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 6835 + }, + { + "epoch": 0.06836, + "grad_norm": 0.6592701430882574, + "learning_rate": 0.003, + "loss": 4.088, + "step": 6836 + }, + { + "epoch": 0.06837, + "grad_norm": 0.8521051837828256, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 6837 + }, + { + "epoch": 0.06838, + "grad_norm": 0.8881766754085176, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 6838 + }, + { + "epoch": 0.06839, + "grad_norm": 0.8608753887830588, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 6839 + }, + { + "epoch": 0.0684, + "grad_norm": 0.8525370672556591, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 6840 + }, + { + "epoch": 0.06841, + "grad_norm": 0.7688532013236681, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 6841 + }, + { + "epoch": 0.06842, + "grad_norm": 0.7975277730841799, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 6842 + }, + { + "epoch": 0.06843, + "grad_norm": 0.7204969622011518, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 6843 + }, + { + "epoch": 0.06844, + "grad_norm": 0.6058386360927623, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 6844 + }, + { + "epoch": 0.06845, + "grad_norm": 0.6069969058059, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 6845 + }, + { + "epoch": 0.06846, + "grad_norm": 0.591567291773583, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 6846 + }, + { + "epoch": 0.06847, + "grad_norm": 0.6453892334888602, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 6847 + }, + { + "epoch": 0.06848, + "grad_norm": 0.6126378526887722, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 6848 + }, + { + "epoch": 0.06849, + "grad_norm": 0.5416820152938913, + "learning_rate": 0.003, + "loss": 4.038, + "step": 6849 + }, + { + "epoch": 0.0685, + "grad_norm": 0.560964266834061, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 6850 + }, + { + "epoch": 0.06851, + "grad_norm": 0.6486698399862927, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 6851 + }, + { + "epoch": 0.06852, + "grad_norm": 0.6801220716892844, + "learning_rate": 0.003, + "loss": 4.033, + "step": 6852 + }, + { + "epoch": 0.06853, + "grad_norm": 0.651059817547728, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 6853 + }, + { + "epoch": 0.06854, + "grad_norm": 0.6733201923939611, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 6854 + }, + { + "epoch": 0.06855, + "grad_norm": 0.599839904082283, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 6855 + }, + { + "epoch": 0.06856, + "grad_norm": 0.5672149525549636, + "learning_rate": 0.003, + "loss": 4.07, + "step": 6856 + }, + { + "epoch": 0.06857, + "grad_norm": 0.6170977658589416, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 6857 + }, + { + "epoch": 0.06858, + "grad_norm": 0.6122519298775182, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 6858 + }, + { + "epoch": 0.06859, + "grad_norm": 0.593554375194474, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 6859 + }, + { + "epoch": 0.0686, + "grad_norm": 0.6617704940277238, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 6860 + }, + { + "epoch": 0.06861, + "grad_norm": 0.7547514317035326, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 6861 + }, + { + "epoch": 0.06862, + "grad_norm": 0.9399868532924999, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 6862 + }, + { + "epoch": 0.06863, + "grad_norm": 0.9766642023103487, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 6863 + }, + { + "epoch": 0.06864, + "grad_norm": 0.8774353629478553, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 6864 + }, + { + "epoch": 0.06865, + "grad_norm": 0.7149976152544412, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 6865 + }, + { + "epoch": 0.06866, + "grad_norm": 0.7472568159463555, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 6866 + }, + { + "epoch": 0.06867, + "grad_norm": 0.7382288193805171, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 6867 + }, + { + "epoch": 0.06868, + "grad_norm": 0.8199642657382533, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 6868 + }, + { + "epoch": 0.06869, + "grad_norm": 0.723532910276911, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 6869 + }, + { + "epoch": 0.0687, + "grad_norm": 0.6345461040740767, + "learning_rate": 0.003, + "loss": 4.006, + "step": 6870 + }, + { + "epoch": 0.06871, + "grad_norm": 0.6024146964713776, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 6871 + }, + { + "epoch": 0.06872, + "grad_norm": 0.7022655137968864, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 6872 + }, + { + "epoch": 0.06873, + "grad_norm": 0.7831786818008365, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 6873 + }, + { + "epoch": 0.06874, + "grad_norm": 0.7068714921428121, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 6874 + }, + { + "epoch": 0.06875, + "grad_norm": 0.6289878162917207, + "learning_rate": 0.003, + "loss": 4.061, + "step": 6875 + }, + { + "epoch": 0.06876, + "grad_norm": 0.668758558402013, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 6876 + }, + { + "epoch": 0.06877, + "grad_norm": 0.6709760905942705, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 6877 + }, + { + "epoch": 0.06878, + "grad_norm": 0.6331638141766467, + "learning_rate": 0.003, + "loss": 4.027, + "step": 6878 + }, + { + "epoch": 0.06879, + "grad_norm": 0.5471488380093832, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 6879 + }, + { + "epoch": 0.0688, + "grad_norm": 0.5822913096369849, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 6880 + }, + { + "epoch": 0.06881, + "grad_norm": 0.5598489171228717, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 6881 + }, + { + "epoch": 0.06882, + "grad_norm": 0.5510502396689, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 6882 + }, + { + "epoch": 0.06883, + "grad_norm": 0.5726499255902116, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 6883 + }, + { + "epoch": 0.06884, + "grad_norm": 0.5727476365861327, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 6884 + }, + { + "epoch": 0.06885, + "grad_norm": 0.653679756632453, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 6885 + }, + { + "epoch": 0.06886, + "grad_norm": 0.7136018444366558, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 6886 + }, + { + "epoch": 0.06887, + "grad_norm": 0.7975619584327837, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 6887 + }, + { + "epoch": 0.06888, + "grad_norm": 0.8412363797777029, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 6888 + }, + { + "epoch": 0.06889, + "grad_norm": 0.7937753854917867, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 6889 + }, + { + "epoch": 0.0689, + "grad_norm": 0.7987866913521997, + "learning_rate": 0.003, + "loss": 4.026, + "step": 6890 + }, + { + "epoch": 0.06891, + "grad_norm": 0.7851604632122734, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 6891 + }, + { + "epoch": 0.06892, + "grad_norm": 0.6420493166306677, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 6892 + }, + { + "epoch": 0.06893, + "grad_norm": 0.548040243780554, + "learning_rate": 0.003, + "loss": 4.047, + "step": 6893 + }, + { + "epoch": 0.06894, + "grad_norm": 0.6673602151834961, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 6894 + }, + { + "epoch": 0.06895, + "grad_norm": 0.6613164312149837, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 6895 + }, + { + "epoch": 0.06896, + "grad_norm": 0.7312960015067101, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 6896 + }, + { + "epoch": 0.06897, + "grad_norm": 0.8103566191322464, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 6897 + }, + { + "epoch": 0.06898, + "grad_norm": 0.9727772309051564, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 6898 + }, + { + "epoch": 0.06899, + "grad_norm": 0.9273537268405467, + "learning_rate": 0.003, + "loss": 4.091, + "step": 6899 + }, + { + "epoch": 0.069, + "grad_norm": 0.9224160138125186, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 6900 + }, + { + "epoch": 0.06901, + "grad_norm": 0.8847833603758801, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 6901 + }, + { + "epoch": 0.06902, + "grad_norm": 0.9509194072354902, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 6902 + }, + { + "epoch": 0.06903, + "grad_norm": 0.9682611084659063, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 6903 + }, + { + "epoch": 0.06904, + "grad_norm": 1.0242135741354703, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 6904 + }, + { + "epoch": 0.06905, + "grad_norm": 0.9239625084904938, + "learning_rate": 0.003, + "loss": 4.0958, + "step": 6905 + }, + { + "epoch": 0.06906, + "grad_norm": 0.8804779837480083, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 6906 + }, + { + "epoch": 0.06907, + "grad_norm": 1.0673832505291045, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 6907 + }, + { + "epoch": 0.06908, + "grad_norm": 1.1834159701006775, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 6908 + }, + { + "epoch": 0.06909, + "grad_norm": 0.9674940540605508, + "learning_rate": 0.003, + "loss": 4.0882, + "step": 6909 + }, + { + "epoch": 0.0691, + "grad_norm": 0.8937976925436587, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 6910 + }, + { + "epoch": 0.06911, + "grad_norm": 0.7506795454350067, + "learning_rate": 0.003, + "loss": 4.0968, + "step": 6911 + }, + { + "epoch": 0.06912, + "grad_norm": 0.6836195934893886, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 6912 + }, + { + "epoch": 0.06913, + "grad_norm": 0.6922072435608072, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 6913 + }, + { + "epoch": 0.06914, + "grad_norm": 0.680201979145026, + "learning_rate": 0.003, + "loss": 4.0948, + "step": 6914 + }, + { + "epoch": 0.06915, + "grad_norm": 0.7809791882950092, + "learning_rate": 0.003, + "loss": 4.0987, + "step": 6915 + }, + { + "epoch": 0.06916, + "grad_norm": 0.7995823778386066, + "learning_rate": 0.003, + "loss": 4.0696, + "step": 6916 + }, + { + "epoch": 0.06917, + "grad_norm": 0.5913568460596942, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 6917 + }, + { + "epoch": 0.06918, + "grad_norm": 0.5352927948183505, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 6918 + }, + { + "epoch": 0.06919, + "grad_norm": 0.5957921172065049, + "learning_rate": 0.003, + "loss": 4.0749, + "step": 6919 + }, + { + "epoch": 0.0692, + "grad_norm": 0.6264360797902435, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 6920 + }, + { + "epoch": 0.06921, + "grad_norm": 0.6985360715673118, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 6921 + }, + { + "epoch": 0.06922, + "grad_norm": 0.8381572883781034, + "learning_rate": 0.003, + "loss": 4.066, + "step": 6922 + }, + { + "epoch": 0.06923, + "grad_norm": 0.8942033022976876, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 6923 + }, + { + "epoch": 0.06924, + "grad_norm": 0.7481720874614454, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 6924 + }, + { + "epoch": 0.06925, + "grad_norm": 0.5666921588543413, + "learning_rate": 0.003, + "loss": 4.075, + "step": 6925 + }, + { + "epoch": 0.06926, + "grad_norm": 0.6007414863547075, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 6926 + }, + { + "epoch": 0.06927, + "grad_norm": 0.6251592887382496, + "learning_rate": 0.003, + "loss": 4.063, + "step": 6927 + }, + { + "epoch": 0.06928, + "grad_norm": 0.5696158672938153, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 6928 + }, + { + "epoch": 0.06929, + "grad_norm": 0.5431762364594065, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 6929 + }, + { + "epoch": 0.0693, + "grad_norm": 0.5374186912006703, + "learning_rate": 0.003, + "loss": 4.092, + "step": 6930 + }, + { + "epoch": 0.06931, + "grad_norm": 0.5009432379218931, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 6931 + }, + { + "epoch": 0.06932, + "grad_norm": 0.3987618900552134, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 6932 + }, + { + "epoch": 0.06933, + "grad_norm": 0.41718203803424736, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 6933 + }, + { + "epoch": 0.06934, + "grad_norm": 0.39975796375727773, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 6934 + }, + { + "epoch": 0.06935, + "grad_norm": 0.3955128144957457, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 6935 + }, + { + "epoch": 0.06936, + "grad_norm": 0.45604481301108724, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 6936 + }, + { + "epoch": 0.06937, + "grad_norm": 0.47828420453249537, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 6937 + }, + { + "epoch": 0.06938, + "grad_norm": 0.5604592150158779, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 6938 + }, + { + "epoch": 0.06939, + "grad_norm": 0.6820714549624833, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 6939 + }, + { + "epoch": 0.0694, + "grad_norm": 0.7536698849472765, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 6940 + }, + { + "epoch": 0.06941, + "grad_norm": 0.877247795213436, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 6941 + }, + { + "epoch": 0.06942, + "grad_norm": 0.8649585710805652, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 6942 + }, + { + "epoch": 0.06943, + "grad_norm": 0.7095040666323474, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 6943 + }, + { + "epoch": 0.06944, + "grad_norm": 0.7257799066772886, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 6944 + }, + { + "epoch": 0.06945, + "grad_norm": 0.7132867833824912, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 6945 + }, + { + "epoch": 0.06946, + "grad_norm": 0.7678455568452669, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 6946 + }, + { + "epoch": 0.06947, + "grad_norm": 0.7592561435744833, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 6947 + }, + { + "epoch": 0.06948, + "grad_norm": 0.8198644797892182, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 6948 + }, + { + "epoch": 0.06949, + "grad_norm": 0.7361831240485406, + "learning_rate": 0.003, + "loss": 4.063, + "step": 6949 + }, + { + "epoch": 0.0695, + "grad_norm": 0.6748090785687593, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 6950 + }, + { + "epoch": 0.06951, + "grad_norm": 0.6388177892598923, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 6951 + }, + { + "epoch": 0.06952, + "grad_norm": 0.588279814551527, + "learning_rate": 0.003, + "loss": 4.069, + "step": 6952 + }, + { + "epoch": 0.06953, + "grad_norm": 0.46770007341861713, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 6953 + }, + { + "epoch": 0.06954, + "grad_norm": 0.48508039218936516, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 6954 + }, + { + "epoch": 0.06955, + "grad_norm": 0.41403945825366295, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 6955 + }, + { + "epoch": 0.06956, + "grad_norm": 0.4165642827465793, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 6956 + }, + { + "epoch": 0.06957, + "grad_norm": 0.458052242288519, + "learning_rate": 0.003, + "loss": 4.033, + "step": 6957 + }, + { + "epoch": 0.06958, + "grad_norm": 0.5298053060996272, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 6958 + }, + { + "epoch": 0.06959, + "grad_norm": 0.6423507927695005, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 6959 + }, + { + "epoch": 0.0696, + "grad_norm": 0.783692777704035, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 6960 + }, + { + "epoch": 0.06961, + "grad_norm": 0.8792118744256132, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 6961 + }, + { + "epoch": 0.06962, + "grad_norm": 1.0124082253720965, + "learning_rate": 0.003, + "loss": 4.068, + "step": 6962 + }, + { + "epoch": 0.06963, + "grad_norm": 0.9921905955972825, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 6963 + }, + { + "epoch": 0.06964, + "grad_norm": 0.8556666954842522, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 6964 + }, + { + "epoch": 0.06965, + "grad_norm": 0.8781937663685095, + "learning_rate": 0.003, + "loss": 4.054, + "step": 6965 + }, + { + "epoch": 0.06966, + "grad_norm": 0.8592363510254366, + "learning_rate": 0.003, + "loss": 4.078, + "step": 6966 + }, + { + "epoch": 0.06967, + "grad_norm": 0.7663353174385432, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 6967 + }, + { + "epoch": 0.06968, + "grad_norm": 0.7244907651252543, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 6968 + }, + { + "epoch": 0.06969, + "grad_norm": 0.6517648108882682, + "learning_rate": 0.003, + "loss": 4.0756, + "step": 6969 + }, + { + "epoch": 0.0697, + "grad_norm": 0.5858626747447735, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 6970 + }, + { + "epoch": 0.06971, + "grad_norm": 0.533102816191858, + "learning_rate": 0.003, + "loss": 4.051, + "step": 6971 + }, + { + "epoch": 0.06972, + "grad_norm": 0.506333645434986, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 6972 + }, + { + "epoch": 0.06973, + "grad_norm": 0.5155962384448307, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 6973 + }, + { + "epoch": 0.06974, + "grad_norm": 0.6834235944453704, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 6974 + }, + { + "epoch": 0.06975, + "grad_norm": 0.9255473764298843, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 6975 + }, + { + "epoch": 0.06976, + "grad_norm": 1.1299642864797026, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 6976 + }, + { + "epoch": 0.06977, + "grad_norm": 0.8226033162775157, + "learning_rate": 0.003, + "loss": 4.0994, + "step": 6977 + }, + { + "epoch": 0.06978, + "grad_norm": 0.7696558174276509, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 6978 + }, + { + "epoch": 0.06979, + "grad_norm": 0.8459221862582573, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 6979 + }, + { + "epoch": 0.0698, + "grad_norm": 0.8821065445904699, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 6980 + }, + { + "epoch": 0.06981, + "grad_norm": 0.752730084898518, + "learning_rate": 0.003, + "loss": 4.037, + "step": 6981 + }, + { + "epoch": 0.06982, + "grad_norm": 0.6959138334651674, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 6982 + }, + { + "epoch": 0.06983, + "grad_norm": 0.6779028895996969, + "learning_rate": 0.003, + "loss": 4.031, + "step": 6983 + }, + { + "epoch": 0.06984, + "grad_norm": 0.6365507561774436, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 6984 + }, + { + "epoch": 0.06985, + "grad_norm": 0.6429208175762545, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 6985 + }, + { + "epoch": 0.06986, + "grad_norm": 0.6895688046214052, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 6986 + }, + { + "epoch": 0.06987, + "grad_norm": 0.7534245586179424, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 6987 + }, + { + "epoch": 0.06988, + "grad_norm": 0.7751454206156404, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 6988 + }, + { + "epoch": 0.06989, + "grad_norm": 0.805735444642867, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 6989 + }, + { + "epoch": 0.0699, + "grad_norm": 0.8269736745720014, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 6990 + }, + { + "epoch": 0.06991, + "grad_norm": 0.8348558762512688, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 6991 + }, + { + "epoch": 0.06992, + "grad_norm": 0.6730062200523551, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 6992 + }, + { + "epoch": 0.06993, + "grad_norm": 0.6469355586497911, + "learning_rate": 0.003, + "loss": 4.035, + "step": 6993 + }, + { + "epoch": 0.06994, + "grad_norm": 0.7162220630284637, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 6994 + }, + { + "epoch": 0.06995, + "grad_norm": 0.6859959996065607, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 6995 + }, + { + "epoch": 0.06996, + "grad_norm": 0.6613753250334239, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 6996 + }, + { + "epoch": 0.06997, + "grad_norm": 0.6372576149532143, + "learning_rate": 0.003, + "loss": 4.029, + "step": 6997 + }, + { + "epoch": 0.06998, + "grad_norm": 0.5987760097921951, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 6998 + }, + { + "epoch": 0.06999, + "grad_norm": 0.5629837438371096, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 6999 + }, + { + "epoch": 0.07, + "grad_norm": 0.5863467634297564, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 7000 + }, + { + "epoch": 0.07001, + "grad_norm": 0.5613204129001492, + "learning_rate": 0.003, + "loss": 4.0905, + "step": 7001 + }, + { + "epoch": 0.07002, + "grad_norm": 0.7355456626731732, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 7002 + }, + { + "epoch": 0.07003, + "grad_norm": 0.7440916389327707, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7003 + }, + { + "epoch": 0.07004, + "grad_norm": 0.6682489981389176, + "learning_rate": 0.003, + "loss": 4.02, + "step": 7004 + }, + { + "epoch": 0.07005, + "grad_norm": 0.6896427903817796, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7005 + }, + { + "epoch": 0.07006, + "grad_norm": 0.8285888319297534, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 7006 + }, + { + "epoch": 0.07007, + "grad_norm": 0.9311342883876439, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 7007 + }, + { + "epoch": 0.07008, + "grad_norm": 0.9033562839732313, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7008 + }, + { + "epoch": 0.07009, + "grad_norm": 0.5663304475102136, + "learning_rate": 0.003, + "loss": 4.058, + "step": 7009 + }, + { + "epoch": 0.0701, + "grad_norm": 0.5853377049591081, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 7010 + }, + { + "epoch": 0.07011, + "grad_norm": 0.5397505791696684, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 7011 + }, + { + "epoch": 0.07012, + "grad_norm": 0.6024976124701904, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 7012 + }, + { + "epoch": 0.07013, + "grad_norm": 0.5578939469014146, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 7013 + }, + { + "epoch": 0.07014, + "grad_norm": 0.5533592911412389, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 7014 + }, + { + "epoch": 0.07015, + "grad_norm": 0.5843232426925851, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 7015 + }, + { + "epoch": 0.07016, + "grad_norm": 0.6042554154481627, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 7016 + }, + { + "epoch": 0.07017, + "grad_norm": 0.5980024144612657, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 7017 + }, + { + "epoch": 0.07018, + "grad_norm": 0.6315576267745769, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 7018 + }, + { + "epoch": 0.07019, + "grad_norm": 0.6054931549093797, + "learning_rate": 0.003, + "loss": 4.039, + "step": 7019 + }, + { + "epoch": 0.0702, + "grad_norm": 0.5196537210669834, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 7020 + }, + { + "epoch": 0.07021, + "grad_norm": 0.6168398672020593, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 7021 + }, + { + "epoch": 0.07022, + "grad_norm": 0.7224591580925077, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 7022 + }, + { + "epoch": 0.07023, + "grad_norm": 0.7366306822781177, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 7023 + }, + { + "epoch": 0.07024, + "grad_norm": 0.7548241162899261, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 7024 + }, + { + "epoch": 0.07025, + "grad_norm": 0.7782203171733063, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 7025 + }, + { + "epoch": 0.07026, + "grad_norm": 0.9335531564113231, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 7026 + }, + { + "epoch": 0.07027, + "grad_norm": 0.9333394111986796, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 7027 + }, + { + "epoch": 0.07028, + "grad_norm": 0.8818252614915849, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 7028 + }, + { + "epoch": 0.07029, + "grad_norm": 0.705489180753897, + "learning_rate": 0.003, + "loss": 4.034, + "step": 7029 + }, + { + "epoch": 0.0703, + "grad_norm": 0.7271813542951504, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 7030 + }, + { + "epoch": 0.07031, + "grad_norm": 0.822208591192176, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 7031 + }, + { + "epoch": 0.07032, + "grad_norm": 0.8119813306588858, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 7032 + }, + { + "epoch": 0.07033, + "grad_norm": 0.769904225277831, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 7033 + }, + { + "epoch": 0.07034, + "grad_norm": 0.7995315471181762, + "learning_rate": 0.003, + "loss": 4.052, + "step": 7034 + }, + { + "epoch": 0.07035, + "grad_norm": 0.8814428529842684, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 7035 + }, + { + "epoch": 0.07036, + "grad_norm": 1.1020955985667489, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7036 + }, + { + "epoch": 0.07037, + "grad_norm": 1.039208745826152, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7037 + }, + { + "epoch": 0.07038, + "grad_norm": 0.9000478913472877, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 7038 + }, + { + "epoch": 0.07039, + "grad_norm": 0.83494262359919, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 7039 + }, + { + "epoch": 0.0704, + "grad_norm": 0.9116550821376965, + "learning_rate": 0.003, + "loss": 4.0855, + "step": 7040 + }, + { + "epoch": 0.07041, + "grad_norm": 0.9053681198025633, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 7041 + }, + { + "epoch": 0.07042, + "grad_norm": 0.9413869585580837, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 7042 + }, + { + "epoch": 0.07043, + "grad_norm": 0.8038117299874746, + "learning_rate": 0.003, + "loss": 4.0866, + "step": 7043 + }, + { + "epoch": 0.07044, + "grad_norm": 0.7162125850004302, + "learning_rate": 0.003, + "loss": 4.082, + "step": 7044 + }, + { + "epoch": 0.07045, + "grad_norm": 0.606614172804327, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 7045 + }, + { + "epoch": 0.07046, + "grad_norm": 0.6960317808484672, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 7046 + }, + { + "epoch": 0.07047, + "grad_norm": 0.6912047233502331, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 7047 + }, + { + "epoch": 0.07048, + "grad_norm": 0.6100170460361597, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 7048 + }, + { + "epoch": 0.07049, + "grad_norm": 0.6084359493749789, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7049 + }, + { + "epoch": 0.0705, + "grad_norm": 0.57788714065014, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 7050 + }, + { + "epoch": 0.07051, + "grad_norm": 0.5780483286625386, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 7051 + }, + { + "epoch": 0.07052, + "grad_norm": 0.5207022211614921, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 7052 + }, + { + "epoch": 0.07053, + "grad_norm": 0.4208194068460513, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 7053 + }, + { + "epoch": 0.07054, + "grad_norm": 0.40247575699996974, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 7054 + }, + { + "epoch": 0.07055, + "grad_norm": 0.4264083039743897, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 7055 + }, + { + "epoch": 0.07056, + "grad_norm": 0.4408117478022399, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 7056 + }, + { + "epoch": 0.07057, + "grad_norm": 0.449029991928975, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 7057 + }, + { + "epoch": 0.07058, + "grad_norm": 0.484030477134895, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 7058 + }, + { + "epoch": 0.07059, + "grad_norm": 0.5784174132886775, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 7059 + }, + { + "epoch": 0.0706, + "grad_norm": 0.7090602739576372, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 7060 + }, + { + "epoch": 0.07061, + "grad_norm": 0.7725981198677542, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 7061 + }, + { + "epoch": 0.07062, + "grad_norm": 0.8085849099381437, + "learning_rate": 0.003, + "loss": 4.077, + "step": 7062 + }, + { + "epoch": 0.07063, + "grad_norm": 0.7187822356641893, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 7063 + }, + { + "epoch": 0.07064, + "grad_norm": 0.6137887021607865, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 7064 + }, + { + "epoch": 0.07065, + "grad_norm": 0.7445165446903537, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 7065 + }, + { + "epoch": 0.07066, + "grad_norm": 0.8903525939066376, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 7066 + }, + { + "epoch": 0.07067, + "grad_norm": 0.782395205691175, + "learning_rate": 0.003, + "loss": 4.0924, + "step": 7067 + }, + { + "epoch": 0.07068, + "grad_norm": 0.7669539968310057, + "learning_rate": 0.003, + "loss": 4.028, + "step": 7068 + }, + { + "epoch": 0.07069, + "grad_norm": 0.7863588474409114, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 7069 + }, + { + "epoch": 0.0707, + "grad_norm": 0.7183453119074922, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 7070 + }, + { + "epoch": 0.07071, + "grad_norm": 0.658341629801751, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 7071 + }, + { + "epoch": 0.07072, + "grad_norm": 0.6881658863418513, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 7072 + }, + { + "epoch": 0.07073, + "grad_norm": 0.6933766291818241, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 7073 + }, + { + "epoch": 0.07074, + "grad_norm": 0.8338424422316789, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7074 + }, + { + "epoch": 0.07075, + "grad_norm": 0.8606220414792101, + "learning_rate": 0.003, + "loss": 4.0883, + "step": 7075 + }, + { + "epoch": 0.07076, + "grad_norm": 0.8487096838408261, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 7076 + }, + { + "epoch": 0.07077, + "grad_norm": 0.9522925345263835, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 7077 + }, + { + "epoch": 0.07078, + "grad_norm": 0.878218026911634, + "learning_rate": 0.003, + "loss": 4.0848, + "step": 7078 + }, + { + "epoch": 0.07079, + "grad_norm": 0.7395796732760429, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 7079 + }, + { + "epoch": 0.0708, + "grad_norm": 0.7348882876983373, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 7080 + }, + { + "epoch": 0.07081, + "grad_norm": 0.6957934221297892, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 7081 + }, + { + "epoch": 0.07082, + "grad_norm": 0.6532284452813655, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7082 + }, + { + "epoch": 0.07083, + "grad_norm": 0.5438132938950468, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 7083 + }, + { + "epoch": 0.07084, + "grad_norm": 0.521200220021269, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 7084 + }, + { + "epoch": 0.07085, + "grad_norm": 0.5014394841135668, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 7085 + }, + { + "epoch": 0.07086, + "grad_norm": 0.5185434498610421, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 7086 + }, + { + "epoch": 0.07087, + "grad_norm": 0.5583457732815869, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 7087 + }, + { + "epoch": 0.07088, + "grad_norm": 0.5608554901432256, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 7088 + }, + { + "epoch": 0.07089, + "grad_norm": 0.5386241971675472, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 7089 + }, + { + "epoch": 0.0709, + "grad_norm": 0.5491589213220635, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 7090 + }, + { + "epoch": 0.07091, + "grad_norm": 0.6399226775129114, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 7091 + }, + { + "epoch": 0.07092, + "grad_norm": 0.8185677428408016, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7092 + }, + { + "epoch": 0.07093, + "grad_norm": 1.0068338910184684, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 7093 + }, + { + "epoch": 0.07094, + "grad_norm": 0.9621662809879228, + "learning_rate": 0.003, + "loss": 4.0928, + "step": 7094 + }, + { + "epoch": 0.07095, + "grad_norm": 0.7046697648480488, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 7095 + }, + { + "epoch": 0.07096, + "grad_norm": 0.830409667817406, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 7096 + }, + { + "epoch": 0.07097, + "grad_norm": 0.7164191492665055, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 7097 + }, + { + "epoch": 0.07098, + "grad_norm": 0.5888908224381079, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 7098 + }, + { + "epoch": 0.07099, + "grad_norm": 0.6763617488330542, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 7099 + }, + { + "epoch": 0.071, + "grad_norm": 0.7506499256437289, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 7100 + }, + { + "epoch": 0.07101, + "grad_norm": 0.811127618123077, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 7101 + }, + { + "epoch": 0.07102, + "grad_norm": 0.8423663797466346, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 7102 + }, + { + "epoch": 0.07103, + "grad_norm": 0.939389962250667, + "learning_rate": 0.003, + "loss": 4.0741, + "step": 7103 + }, + { + "epoch": 0.07104, + "grad_norm": 1.065626674042908, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 7104 + }, + { + "epoch": 0.07105, + "grad_norm": 0.8744474092687, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 7105 + }, + { + "epoch": 0.07106, + "grad_norm": 0.7645541605335416, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 7106 + }, + { + "epoch": 0.07107, + "grad_norm": 0.7620002175355072, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 7107 + }, + { + "epoch": 0.07108, + "grad_norm": 0.6514244880039121, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 7108 + }, + { + "epoch": 0.07109, + "grad_norm": 0.6366705583651942, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 7109 + }, + { + "epoch": 0.0711, + "grad_norm": 0.5787547358913719, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 7110 + }, + { + "epoch": 0.07111, + "grad_norm": 0.5084154765180142, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 7111 + }, + { + "epoch": 0.07112, + "grad_norm": 0.46208300687756654, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7112 + }, + { + "epoch": 0.07113, + "grad_norm": 0.5063075926015548, + "learning_rate": 0.003, + "loss": 4.0774, + "step": 7113 + }, + { + "epoch": 0.07114, + "grad_norm": 0.5243308054830852, + "learning_rate": 0.003, + "loss": 4.054, + "step": 7114 + }, + { + "epoch": 0.07115, + "grad_norm": 0.5199159301633056, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 7115 + }, + { + "epoch": 0.07116, + "grad_norm": 0.5225803627411774, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 7116 + }, + { + "epoch": 0.07117, + "grad_norm": 0.4619628203298842, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 7117 + }, + { + "epoch": 0.07118, + "grad_norm": 0.45878013379535126, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 7118 + }, + { + "epoch": 0.07119, + "grad_norm": 0.5448437044322277, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 7119 + }, + { + "epoch": 0.0712, + "grad_norm": 0.6056688720589296, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 7120 + }, + { + "epoch": 0.07121, + "grad_norm": 0.6745888689516769, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 7121 + }, + { + "epoch": 0.07122, + "grad_norm": 0.802276236622279, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 7122 + }, + { + "epoch": 0.07123, + "grad_norm": 0.7535279634067673, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 7123 + }, + { + "epoch": 0.07124, + "grad_norm": 0.626552874865624, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 7124 + }, + { + "epoch": 0.07125, + "grad_norm": 0.6051695094344289, + "learning_rate": 0.003, + "loss": 4.031, + "step": 7125 + }, + { + "epoch": 0.07126, + "grad_norm": 0.5720241864567985, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 7126 + }, + { + "epoch": 0.07127, + "grad_norm": 0.6105910448399364, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 7127 + }, + { + "epoch": 0.07128, + "grad_norm": 0.688002651712525, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 7128 + }, + { + "epoch": 0.07129, + "grad_norm": 0.6098170552466032, + "learning_rate": 0.003, + "loss": 4.053, + "step": 7129 + }, + { + "epoch": 0.0713, + "grad_norm": 0.6244175761948557, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 7130 + }, + { + "epoch": 0.07131, + "grad_norm": 0.6878314267019149, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 7131 + }, + { + "epoch": 0.07132, + "grad_norm": 0.7099741791172226, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 7132 + }, + { + "epoch": 0.07133, + "grad_norm": 0.796953346873059, + "learning_rate": 0.003, + "loss": 4.0795, + "step": 7133 + }, + { + "epoch": 0.07134, + "grad_norm": 0.8231659239758954, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 7134 + }, + { + "epoch": 0.07135, + "grad_norm": 0.8099410836327584, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 7135 + }, + { + "epoch": 0.07136, + "grad_norm": 0.8457069137186146, + "learning_rate": 0.003, + "loss": 4.088, + "step": 7136 + }, + { + "epoch": 0.07137, + "grad_norm": 0.8244822711812634, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 7137 + }, + { + "epoch": 0.07138, + "grad_norm": 0.9439361257796965, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 7138 + }, + { + "epoch": 0.07139, + "grad_norm": 1.0737970326587571, + "learning_rate": 0.003, + "loss": 4.064, + "step": 7139 + }, + { + "epoch": 0.0714, + "grad_norm": 0.8096486913154792, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 7140 + }, + { + "epoch": 0.07141, + "grad_norm": 0.7468411147201653, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 7141 + }, + { + "epoch": 0.07142, + "grad_norm": 0.6635655583101364, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 7142 + }, + { + "epoch": 0.07143, + "grad_norm": 0.6876383094159536, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 7143 + }, + { + "epoch": 0.07144, + "grad_norm": 0.8057954002305314, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 7144 + }, + { + "epoch": 0.07145, + "grad_norm": 0.9212693904349629, + "learning_rate": 0.003, + "loss": 4.062, + "step": 7145 + }, + { + "epoch": 0.07146, + "grad_norm": 1.091047612561957, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 7146 + }, + { + "epoch": 0.07147, + "grad_norm": 0.8455483243741031, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 7147 + }, + { + "epoch": 0.07148, + "grad_norm": 0.696507973751977, + "learning_rate": 0.003, + "loss": 4.0685, + "step": 7148 + }, + { + "epoch": 0.07149, + "grad_norm": 0.6178647287562519, + "learning_rate": 0.003, + "loss": 4.0746, + "step": 7149 + }, + { + "epoch": 0.0715, + "grad_norm": 0.6641984133184746, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 7150 + }, + { + "epoch": 0.07151, + "grad_norm": 0.8193783414315825, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 7151 + }, + { + "epoch": 0.07152, + "grad_norm": 0.7428468308617348, + "learning_rate": 0.003, + "loss": 4.074, + "step": 7152 + }, + { + "epoch": 0.07153, + "grad_norm": 0.7482140401180368, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 7153 + }, + { + "epoch": 0.07154, + "grad_norm": 0.8122599292302012, + "learning_rate": 0.003, + "loss": 4.0957, + "step": 7154 + }, + { + "epoch": 0.07155, + "grad_norm": 0.9325523806413661, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 7155 + }, + { + "epoch": 0.07156, + "grad_norm": 0.8635102757954148, + "learning_rate": 0.003, + "loss": 4.067, + "step": 7156 + }, + { + "epoch": 0.07157, + "grad_norm": 0.8280182513803493, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 7157 + }, + { + "epoch": 0.07158, + "grad_norm": 0.6992218862069915, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 7158 + }, + { + "epoch": 0.07159, + "grad_norm": 0.6765318577061701, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 7159 + }, + { + "epoch": 0.0716, + "grad_norm": 0.7649121421097587, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 7160 + }, + { + "epoch": 0.07161, + "grad_norm": 0.7087877514026757, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 7161 + }, + { + "epoch": 0.07162, + "grad_norm": 0.6763518983903675, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 7162 + }, + { + "epoch": 0.07163, + "grad_norm": 0.6634212922621017, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 7163 + }, + { + "epoch": 0.07164, + "grad_norm": 0.6874444153993609, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 7164 + }, + { + "epoch": 0.07165, + "grad_norm": 0.665885838794523, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 7165 + }, + { + "epoch": 0.07166, + "grad_norm": 0.696238470007808, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 7166 + }, + { + "epoch": 0.07167, + "grad_norm": 0.8404117499955268, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 7167 + }, + { + "epoch": 0.07168, + "grad_norm": 0.8665409776120457, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 7168 + }, + { + "epoch": 0.07169, + "grad_norm": 0.6854422583769841, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 7169 + }, + { + "epoch": 0.0717, + "grad_norm": 0.592156233901982, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 7170 + }, + { + "epoch": 0.07171, + "grad_norm": 0.6628384735019112, + "learning_rate": 0.003, + "loss": 4.061, + "step": 7171 + }, + { + "epoch": 0.07172, + "grad_norm": 0.6822221432345806, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 7172 + }, + { + "epoch": 0.07173, + "grad_norm": 0.721008398550851, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 7173 + }, + { + "epoch": 0.07174, + "grad_norm": 0.7881108944726012, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 7174 + }, + { + "epoch": 0.07175, + "grad_norm": 0.6904667767065579, + "learning_rate": 0.003, + "loss": 4.0742, + "step": 7175 + }, + { + "epoch": 0.07176, + "grad_norm": 0.6216598604657809, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 7176 + }, + { + "epoch": 0.07177, + "grad_norm": 0.5927485303099118, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 7177 + }, + { + "epoch": 0.07178, + "grad_norm": 0.5745048102202693, + "learning_rate": 0.003, + "loss": 4.011, + "step": 7178 + }, + { + "epoch": 0.07179, + "grad_norm": 0.5768709586116577, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 7179 + }, + { + "epoch": 0.0718, + "grad_norm": 0.6427322580878045, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 7180 + }, + { + "epoch": 0.07181, + "grad_norm": 0.7168945742431179, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 7181 + }, + { + "epoch": 0.07182, + "grad_norm": 0.6797015571995331, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 7182 + }, + { + "epoch": 0.07183, + "grad_norm": 0.5089175882630684, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 7183 + }, + { + "epoch": 0.07184, + "grad_norm": 0.5780363345368148, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 7184 + }, + { + "epoch": 0.07185, + "grad_norm": 0.5232058444016439, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 7185 + }, + { + "epoch": 0.07186, + "grad_norm": 0.5921368391698842, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 7186 + }, + { + "epoch": 0.07187, + "grad_norm": 0.5931668408197696, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 7187 + }, + { + "epoch": 0.07188, + "grad_norm": 0.6725385572135081, + "learning_rate": 0.003, + "loss": 4.006, + "step": 7188 + }, + { + "epoch": 0.07189, + "grad_norm": 0.7268022471833311, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 7189 + }, + { + "epoch": 0.0719, + "grad_norm": 0.7685026420721217, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 7190 + }, + { + "epoch": 0.07191, + "grad_norm": 0.6941989050477196, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 7191 + }, + { + "epoch": 0.07192, + "grad_norm": 0.6218128800564352, + "learning_rate": 0.003, + "loss": 4.025, + "step": 7192 + }, + { + "epoch": 0.07193, + "grad_norm": 0.6361090283035431, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7193 + }, + { + "epoch": 0.07194, + "grad_norm": 0.6962784143123117, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7194 + }, + { + "epoch": 0.07195, + "grad_norm": 0.6494614267702561, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 7195 + }, + { + "epoch": 0.07196, + "grad_norm": 0.5858717640165627, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 7196 + }, + { + "epoch": 0.07197, + "grad_norm": 0.6935199111076826, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 7197 + }, + { + "epoch": 0.07198, + "grad_norm": 0.7677248406196279, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 7198 + }, + { + "epoch": 0.07199, + "grad_norm": 0.8674880831870259, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 7199 + }, + { + "epoch": 0.072, + "grad_norm": 1.010695145447074, + "learning_rate": 0.003, + "loss": 4.084, + "step": 7200 + }, + { + "epoch": 0.07201, + "grad_norm": 1.0553323289783902, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 7201 + }, + { + "epoch": 0.07202, + "grad_norm": 0.9224806043024206, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 7202 + }, + { + "epoch": 0.07203, + "grad_norm": 0.741911277779582, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 7203 + }, + { + "epoch": 0.07204, + "grad_norm": 0.7058011310694724, + "learning_rate": 0.003, + "loss": 4.082, + "step": 7204 + }, + { + "epoch": 0.07205, + "grad_norm": 0.720545645768948, + "learning_rate": 0.003, + "loss": 4.0761, + "step": 7205 + }, + { + "epoch": 0.07206, + "grad_norm": 0.8485964705453234, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 7206 + }, + { + "epoch": 0.07207, + "grad_norm": 0.9349298516736723, + "learning_rate": 0.003, + "loss": 4.062, + "step": 7207 + }, + { + "epoch": 0.07208, + "grad_norm": 0.9072423875271604, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 7208 + }, + { + "epoch": 0.07209, + "grad_norm": 0.7988950699065934, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 7209 + }, + { + "epoch": 0.0721, + "grad_norm": 0.7757745524389528, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 7210 + }, + { + "epoch": 0.07211, + "grad_norm": 0.7357552411244734, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 7211 + }, + { + "epoch": 0.07212, + "grad_norm": 0.7015998574199064, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 7212 + }, + { + "epoch": 0.07213, + "grad_norm": 0.5897232134425919, + "learning_rate": 0.003, + "loss": 4.0762, + "step": 7213 + }, + { + "epoch": 0.07214, + "grad_norm": 0.5844745568773955, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 7214 + }, + { + "epoch": 0.07215, + "grad_norm": 0.5427622446559991, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 7215 + }, + { + "epoch": 0.07216, + "grad_norm": 0.5586052738564158, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 7216 + }, + { + "epoch": 0.07217, + "grad_norm": 0.5344298573973888, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7217 + }, + { + "epoch": 0.07218, + "grad_norm": 0.5248366269417295, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 7218 + }, + { + "epoch": 0.07219, + "grad_norm": 0.5328749833995203, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 7219 + }, + { + "epoch": 0.0722, + "grad_norm": 0.5176910586888837, + "learning_rate": 0.003, + "loss": 4.052, + "step": 7220 + }, + { + "epoch": 0.07221, + "grad_norm": 0.564023830480607, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 7221 + }, + { + "epoch": 0.07222, + "grad_norm": 0.6049620145732001, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 7222 + }, + { + "epoch": 0.07223, + "grad_norm": 0.7467241067015448, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 7223 + }, + { + "epoch": 0.07224, + "grad_norm": 0.7327376928205852, + "learning_rate": 0.003, + "loss": 4.024, + "step": 7224 + }, + { + "epoch": 0.07225, + "grad_norm": 0.589858801031246, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 7225 + }, + { + "epoch": 0.07226, + "grad_norm": 0.5918191302973123, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 7226 + }, + { + "epoch": 0.07227, + "grad_norm": 0.6247831805489723, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 7227 + }, + { + "epoch": 0.07228, + "grad_norm": 0.7050534784327177, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 7228 + }, + { + "epoch": 0.07229, + "grad_norm": 0.6200086399041499, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 7229 + }, + { + "epoch": 0.0723, + "grad_norm": 0.5569228438705434, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 7230 + }, + { + "epoch": 0.07231, + "grad_norm": 0.5792705550273988, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 7231 + }, + { + "epoch": 0.07232, + "grad_norm": 0.6760522855527934, + "learning_rate": 0.003, + "loss": 4.046, + "step": 7232 + }, + { + "epoch": 0.07233, + "grad_norm": 0.7736648637917038, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 7233 + }, + { + "epoch": 0.07234, + "grad_norm": 0.8673050197748207, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 7234 + }, + { + "epoch": 0.07235, + "grad_norm": 1.0325195528938493, + "learning_rate": 0.003, + "loss": 4.0851, + "step": 7235 + }, + { + "epoch": 0.07236, + "grad_norm": 1.1691429344683115, + "learning_rate": 0.003, + "loss": 4.0955, + "step": 7236 + }, + { + "epoch": 0.07237, + "grad_norm": 0.7513131070416232, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 7237 + }, + { + "epoch": 0.07238, + "grad_norm": 0.8610815571079256, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 7238 + }, + { + "epoch": 0.07239, + "grad_norm": 0.8559874010083751, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 7239 + }, + { + "epoch": 0.0724, + "grad_norm": 0.953252445182658, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 7240 + }, + { + "epoch": 0.07241, + "grad_norm": 0.8879424302740987, + "learning_rate": 0.003, + "loss": 4.0796, + "step": 7241 + }, + { + "epoch": 0.07242, + "grad_norm": 0.9017878570456369, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 7242 + }, + { + "epoch": 0.07243, + "grad_norm": 0.8633481852462844, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 7243 + }, + { + "epoch": 0.07244, + "grad_norm": 0.801435250240095, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 7244 + }, + { + "epoch": 0.07245, + "grad_norm": 0.8202668333797751, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 7245 + }, + { + "epoch": 0.07246, + "grad_norm": 0.7282305661681244, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 7246 + }, + { + "epoch": 0.07247, + "grad_norm": 0.7469294195660177, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 7247 + }, + { + "epoch": 0.07248, + "grad_norm": 0.6918184093291221, + "learning_rate": 0.003, + "loss": 4.047, + "step": 7248 + }, + { + "epoch": 0.07249, + "grad_norm": 0.6411727093901283, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 7249 + }, + { + "epoch": 0.0725, + "grad_norm": 0.5653970943371169, + "learning_rate": 0.003, + "loss": 4.038, + "step": 7250 + }, + { + "epoch": 0.07251, + "grad_norm": 0.5404451666784519, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 7251 + }, + { + "epoch": 0.07252, + "grad_norm": 0.4815742896862926, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 7252 + }, + { + "epoch": 0.07253, + "grad_norm": 0.5300596641511565, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 7253 + }, + { + "epoch": 0.07254, + "grad_norm": 0.5295354725455971, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 7254 + }, + { + "epoch": 0.07255, + "grad_norm": 0.4433604860605937, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 7255 + }, + { + "epoch": 0.07256, + "grad_norm": 0.4546908836732582, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 7256 + }, + { + "epoch": 0.07257, + "grad_norm": 0.4798560960024966, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 7257 + }, + { + "epoch": 0.07258, + "grad_norm": 0.549687437567593, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 7258 + }, + { + "epoch": 0.07259, + "grad_norm": 0.6392391559196368, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 7259 + }, + { + "epoch": 0.0726, + "grad_norm": 0.6493153694730507, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 7260 + }, + { + "epoch": 0.07261, + "grad_norm": 0.6636862607119886, + "learning_rate": 0.003, + "loss": 4.05, + "step": 7261 + }, + { + "epoch": 0.07262, + "grad_norm": 0.7052001455055708, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 7262 + }, + { + "epoch": 0.07263, + "grad_norm": 0.6529913106783082, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 7263 + }, + { + "epoch": 0.07264, + "grad_norm": 0.6551022299080768, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 7264 + }, + { + "epoch": 0.07265, + "grad_norm": 0.5693321313444362, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 7265 + }, + { + "epoch": 0.07266, + "grad_norm": 0.5219251633022574, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 7266 + }, + { + "epoch": 0.07267, + "grad_norm": 0.7186638082334078, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 7267 + }, + { + "epoch": 0.07268, + "grad_norm": 1.0105510136509115, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 7268 + }, + { + "epoch": 0.07269, + "grad_norm": 1.1318664877584956, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 7269 + }, + { + "epoch": 0.0727, + "grad_norm": 0.7382245498751608, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 7270 + }, + { + "epoch": 0.07271, + "grad_norm": 0.6263537471514521, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 7271 + }, + { + "epoch": 0.07272, + "grad_norm": 0.7664669076907956, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 7272 + }, + { + "epoch": 0.07273, + "grad_norm": 0.744242445816572, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 7273 + }, + { + "epoch": 0.07274, + "grad_norm": 0.6613681427651242, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 7274 + }, + { + "epoch": 0.07275, + "grad_norm": 0.6114009647539533, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 7275 + }, + { + "epoch": 0.07276, + "grad_norm": 0.5998593408044829, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 7276 + }, + { + "epoch": 0.07277, + "grad_norm": 0.5882586326989973, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 7277 + }, + { + "epoch": 0.07278, + "grad_norm": 0.5323611052308881, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 7278 + }, + { + "epoch": 0.07279, + "grad_norm": 0.5060757814870882, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 7279 + }, + { + "epoch": 0.0728, + "grad_norm": 0.5452480223595009, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 7280 + }, + { + "epoch": 0.07281, + "grad_norm": 0.5523654085672058, + "learning_rate": 0.003, + "loss": 4.032, + "step": 7281 + }, + { + "epoch": 0.07282, + "grad_norm": 0.6476341135880502, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 7282 + }, + { + "epoch": 0.07283, + "grad_norm": 0.7856201489813148, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 7283 + }, + { + "epoch": 0.07284, + "grad_norm": 0.8961481525907361, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 7284 + }, + { + "epoch": 0.07285, + "grad_norm": 0.9292330379134205, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 7285 + }, + { + "epoch": 0.07286, + "grad_norm": 0.925382113563149, + "learning_rate": 0.003, + "loss": 4.042, + "step": 7286 + }, + { + "epoch": 0.07287, + "grad_norm": 0.8848807299534056, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 7287 + }, + { + "epoch": 0.07288, + "grad_norm": 0.9222598486401525, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 7288 + }, + { + "epoch": 0.07289, + "grad_norm": 0.8626335207163163, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 7289 + }, + { + "epoch": 0.0729, + "grad_norm": 0.8586596422701189, + "learning_rate": 0.003, + "loss": 4.0911, + "step": 7290 + }, + { + "epoch": 0.07291, + "grad_norm": 0.862744744183233, + "learning_rate": 0.003, + "loss": 4.053, + "step": 7291 + }, + { + "epoch": 0.07292, + "grad_norm": 0.7947689230003856, + "learning_rate": 0.003, + "loss": 4.064, + "step": 7292 + }, + { + "epoch": 0.07293, + "grad_norm": 0.8200656478343068, + "learning_rate": 0.003, + "loss": 4.053, + "step": 7293 + }, + { + "epoch": 0.07294, + "grad_norm": 0.743384529825569, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 7294 + }, + { + "epoch": 0.07295, + "grad_norm": 0.6611048412844297, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 7295 + }, + { + "epoch": 0.07296, + "grad_norm": 0.6427088108690862, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 7296 + }, + { + "epoch": 0.07297, + "grad_norm": 0.6402244391746081, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 7297 + }, + { + "epoch": 0.07298, + "grad_norm": 0.7239030724334098, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 7298 + }, + { + "epoch": 0.07299, + "grad_norm": 0.9038332444699064, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 7299 + }, + { + "epoch": 0.073, + "grad_norm": 1.0259542664049672, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 7300 + }, + { + "epoch": 0.07301, + "grad_norm": 0.8968160347965933, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 7301 + }, + { + "epoch": 0.07302, + "grad_norm": 0.6673124330277329, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 7302 + }, + { + "epoch": 0.07303, + "grad_norm": 0.78493742593234, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 7303 + }, + { + "epoch": 0.07304, + "grad_norm": 0.7141276150990885, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 7304 + }, + { + "epoch": 0.07305, + "grad_norm": 0.7434865994862239, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 7305 + }, + { + "epoch": 0.07306, + "grad_norm": 0.7854290358831699, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 7306 + }, + { + "epoch": 0.07307, + "grad_norm": 0.865281694104162, + "learning_rate": 0.003, + "loss": 4.0886, + "step": 7307 + }, + { + "epoch": 0.07308, + "grad_norm": 0.7761188211672828, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 7308 + }, + { + "epoch": 0.07309, + "grad_norm": 0.6445130800480298, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 7309 + }, + { + "epoch": 0.0731, + "grad_norm": 0.6043931248245128, + "learning_rate": 0.003, + "loss": 4.011, + "step": 7310 + }, + { + "epoch": 0.07311, + "grad_norm": 0.5292851772113796, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 7311 + }, + { + "epoch": 0.07312, + "grad_norm": 0.5412931473750997, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 7312 + }, + { + "epoch": 0.07313, + "grad_norm": 0.5726999776102948, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 7313 + }, + { + "epoch": 0.07314, + "grad_norm": 0.7013368306413956, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 7314 + }, + { + "epoch": 0.07315, + "grad_norm": 0.7394771876821421, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7315 + }, + { + "epoch": 0.07316, + "grad_norm": 0.661962049138519, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 7316 + }, + { + "epoch": 0.07317, + "grad_norm": 0.6302238152737226, + "learning_rate": 0.003, + "loss": 4.048, + "step": 7317 + }, + { + "epoch": 0.07318, + "grad_norm": 0.5900088622998602, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 7318 + }, + { + "epoch": 0.07319, + "grad_norm": 0.5633304551642094, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 7319 + }, + { + "epoch": 0.0732, + "grad_norm": 0.5910759133540996, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 7320 + }, + { + "epoch": 0.07321, + "grad_norm": 0.608270463122907, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 7321 + }, + { + "epoch": 0.07322, + "grad_norm": 0.5869851521167068, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 7322 + }, + { + "epoch": 0.07323, + "grad_norm": 0.5252010540266933, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 7323 + }, + { + "epoch": 0.07324, + "grad_norm": 0.48121462286801525, + "learning_rate": 0.003, + "loss": 4.035, + "step": 7324 + }, + { + "epoch": 0.07325, + "grad_norm": 0.4396342637424572, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 7325 + }, + { + "epoch": 0.07326, + "grad_norm": 0.4429692710763835, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 7326 + }, + { + "epoch": 0.07327, + "grad_norm": 0.5576454019887566, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 7327 + }, + { + "epoch": 0.07328, + "grad_norm": 0.6639322721461474, + "learning_rate": 0.003, + "loss": 4.013, + "step": 7328 + }, + { + "epoch": 0.07329, + "grad_norm": 0.6261487881704807, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 7329 + }, + { + "epoch": 0.0733, + "grad_norm": 0.6096743637562448, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 7330 + }, + { + "epoch": 0.07331, + "grad_norm": 0.6376976418850646, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 7331 + }, + { + "epoch": 0.07332, + "grad_norm": 0.5141739187200024, + "learning_rate": 0.003, + "loss": 4.046, + "step": 7332 + }, + { + "epoch": 0.07333, + "grad_norm": 0.5544564634234118, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 7333 + }, + { + "epoch": 0.07334, + "grad_norm": 0.5355037407334071, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 7334 + }, + { + "epoch": 0.07335, + "grad_norm": 0.5117344473256494, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 7335 + }, + { + "epoch": 0.07336, + "grad_norm": 0.6235670378982092, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 7336 + }, + { + "epoch": 0.07337, + "grad_norm": 0.8783204345349167, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 7337 + }, + { + "epoch": 0.07338, + "grad_norm": 1.156911493168539, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 7338 + }, + { + "epoch": 0.07339, + "grad_norm": 0.9383434722374471, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 7339 + }, + { + "epoch": 0.0734, + "grad_norm": 0.7412748572527361, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 7340 + }, + { + "epoch": 0.07341, + "grad_norm": 0.7247223909749643, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 7341 + }, + { + "epoch": 0.07342, + "grad_norm": 0.7569405953966587, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 7342 + }, + { + "epoch": 0.07343, + "grad_norm": 0.7220974446690054, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 7343 + }, + { + "epoch": 0.07344, + "grad_norm": 0.6332955931285253, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7344 + }, + { + "epoch": 0.07345, + "grad_norm": 0.7038715006781161, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 7345 + }, + { + "epoch": 0.07346, + "grad_norm": 0.70246912592008, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 7346 + }, + { + "epoch": 0.07347, + "grad_norm": 0.816229181245897, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 7347 + }, + { + "epoch": 0.07348, + "grad_norm": 0.8577428330632486, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 7348 + }, + { + "epoch": 0.07349, + "grad_norm": 0.9239871312555437, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 7349 + }, + { + "epoch": 0.0735, + "grad_norm": 0.9323974440314433, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 7350 + }, + { + "epoch": 0.07351, + "grad_norm": 0.8877926156815443, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 7351 + }, + { + "epoch": 0.07352, + "grad_norm": 0.8513579989754952, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 7352 + }, + { + "epoch": 0.07353, + "grad_norm": 0.7773670290040063, + "learning_rate": 0.003, + "loss": 4.0754, + "step": 7353 + }, + { + "epoch": 0.07354, + "grad_norm": 0.8422345572563882, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 7354 + }, + { + "epoch": 0.07355, + "grad_norm": 0.9137015617965155, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 7355 + }, + { + "epoch": 0.07356, + "grad_norm": 0.8111206444502035, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 7356 + }, + { + "epoch": 0.07357, + "grad_norm": 0.7059162419617638, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 7357 + }, + { + "epoch": 0.07358, + "grad_norm": 0.5789996542079138, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 7358 + }, + { + "epoch": 0.07359, + "grad_norm": 0.6097947156533551, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 7359 + }, + { + "epoch": 0.0736, + "grad_norm": 0.5965908700385388, + "learning_rate": 0.003, + "loss": 4.066, + "step": 7360 + }, + { + "epoch": 0.07361, + "grad_norm": 0.673654335977286, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 7361 + }, + { + "epoch": 0.07362, + "grad_norm": 0.7025287661138682, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 7362 + }, + { + "epoch": 0.07363, + "grad_norm": 0.6976772269155177, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7363 + }, + { + "epoch": 0.07364, + "grad_norm": 0.7662948668478051, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 7364 + }, + { + "epoch": 0.07365, + "grad_norm": 0.6992910450725112, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 7365 + }, + { + "epoch": 0.07366, + "grad_norm": 0.7228204763205468, + "learning_rate": 0.003, + "loss": 4.05, + "step": 7366 + }, + { + "epoch": 0.07367, + "grad_norm": 0.8337083830369219, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 7367 + }, + { + "epoch": 0.07368, + "grad_norm": 0.8285651246118787, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 7368 + }, + { + "epoch": 0.07369, + "grad_norm": 0.8942162276273463, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 7369 + }, + { + "epoch": 0.0737, + "grad_norm": 0.8262103948016015, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 7370 + }, + { + "epoch": 0.07371, + "grad_norm": 0.7858901708196185, + "learning_rate": 0.003, + "loss": 4.0808, + "step": 7371 + }, + { + "epoch": 0.07372, + "grad_norm": 0.7698692828208866, + "learning_rate": 0.003, + "loss": 4.0631, + "step": 7372 + }, + { + "epoch": 0.07373, + "grad_norm": 0.7298732599740968, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 7373 + }, + { + "epoch": 0.07374, + "grad_norm": 0.7006541612460654, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 7374 + }, + { + "epoch": 0.07375, + "grad_norm": 0.690299455748127, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 7375 + }, + { + "epoch": 0.07376, + "grad_norm": 0.7095338839679064, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 7376 + }, + { + "epoch": 0.07377, + "grad_norm": 0.7544552005709612, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 7377 + }, + { + "epoch": 0.07378, + "grad_norm": 0.7489801608030118, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 7378 + }, + { + "epoch": 0.07379, + "grad_norm": 0.7280819452003691, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 7379 + }, + { + "epoch": 0.0738, + "grad_norm": 0.7494509682213668, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 7380 + }, + { + "epoch": 0.07381, + "grad_norm": 0.7628142492276991, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 7381 + }, + { + "epoch": 0.07382, + "grad_norm": 0.7784700660028563, + "learning_rate": 0.003, + "loss": 4.0824, + "step": 7382 + }, + { + "epoch": 0.07383, + "grad_norm": 0.7292026057112635, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 7383 + }, + { + "epoch": 0.07384, + "grad_norm": 0.6834411306427035, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 7384 + }, + { + "epoch": 0.07385, + "grad_norm": 0.5863682821058672, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 7385 + }, + { + "epoch": 0.07386, + "grad_norm": 0.5268457225960965, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 7386 + }, + { + "epoch": 0.07387, + "grad_norm": 0.4598866401565974, + "learning_rate": 0.003, + "loss": 4.024, + "step": 7387 + }, + { + "epoch": 0.07388, + "grad_norm": 0.420374397414649, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 7388 + }, + { + "epoch": 0.07389, + "grad_norm": 0.5058954986863584, + "learning_rate": 0.003, + "loss": 4.033, + "step": 7389 + }, + { + "epoch": 0.0739, + "grad_norm": 0.5483500865730521, + "learning_rate": 0.003, + "loss": 4.025, + "step": 7390 + }, + { + "epoch": 0.07391, + "grad_norm": 0.6507265570849518, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 7391 + }, + { + "epoch": 0.07392, + "grad_norm": 0.6816833472733388, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 7392 + }, + { + "epoch": 0.07393, + "grad_norm": 0.6269639267214473, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 7393 + }, + { + "epoch": 0.07394, + "grad_norm": 0.5736337802970459, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 7394 + }, + { + "epoch": 0.07395, + "grad_norm": 0.5992707411720563, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 7395 + }, + { + "epoch": 0.07396, + "grad_norm": 0.815036015836318, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 7396 + }, + { + "epoch": 0.07397, + "grad_norm": 1.054012093181102, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 7397 + }, + { + "epoch": 0.07398, + "grad_norm": 1.1179990759191294, + "learning_rate": 0.003, + "loss": 4.057, + "step": 7398 + }, + { + "epoch": 0.07399, + "grad_norm": 0.6754306878058735, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 7399 + }, + { + "epoch": 0.074, + "grad_norm": 0.6090506399122712, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 7400 + }, + { + "epoch": 0.07401, + "grad_norm": 0.6499762973034969, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 7401 + }, + { + "epoch": 0.07402, + "grad_norm": 0.5796642363397976, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 7402 + }, + { + "epoch": 0.07403, + "grad_norm": 0.5860337047142725, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7403 + }, + { + "epoch": 0.07404, + "grad_norm": 0.6469963993107319, + "learning_rate": 0.003, + "loss": 4.0854, + "step": 7404 + }, + { + "epoch": 0.07405, + "grad_norm": 0.6107455461474278, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 7405 + }, + { + "epoch": 0.07406, + "grad_norm": 0.5998568849692762, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 7406 + }, + { + "epoch": 0.07407, + "grad_norm": 0.7354350045638043, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 7407 + }, + { + "epoch": 0.07408, + "grad_norm": 0.6997781101937441, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 7408 + }, + { + "epoch": 0.07409, + "grad_norm": 0.6242294446939869, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 7409 + }, + { + "epoch": 0.0741, + "grad_norm": 0.6575686890130875, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 7410 + }, + { + "epoch": 0.07411, + "grad_norm": 0.7412881905924752, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 7411 + }, + { + "epoch": 0.07412, + "grad_norm": 0.7439060047307218, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 7412 + }, + { + "epoch": 0.07413, + "grad_norm": 0.7356409257349356, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 7413 + }, + { + "epoch": 0.07414, + "grad_norm": 0.764154031886002, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 7414 + }, + { + "epoch": 0.07415, + "grad_norm": 0.8274415896394334, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 7415 + }, + { + "epoch": 0.07416, + "grad_norm": 0.8026251146995648, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 7416 + }, + { + "epoch": 0.07417, + "grad_norm": 0.7782453867539443, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 7417 + }, + { + "epoch": 0.07418, + "grad_norm": 0.864843513730434, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 7418 + }, + { + "epoch": 0.07419, + "grad_norm": 0.8256969768671512, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 7419 + }, + { + "epoch": 0.0742, + "grad_norm": 0.7842818411434794, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 7420 + }, + { + "epoch": 0.07421, + "grad_norm": 0.8183120108947622, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 7421 + }, + { + "epoch": 0.07422, + "grad_norm": 0.742618670877204, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 7422 + }, + { + "epoch": 0.07423, + "grad_norm": 0.782478555576758, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 7423 + }, + { + "epoch": 0.07424, + "grad_norm": 0.8757465579607198, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 7424 + }, + { + "epoch": 0.07425, + "grad_norm": 0.7814021232952248, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 7425 + }, + { + "epoch": 0.07426, + "grad_norm": 0.6589814945197152, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 7426 + }, + { + "epoch": 0.07427, + "grad_norm": 0.6878947680840289, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 7427 + }, + { + "epoch": 0.07428, + "grad_norm": 0.6844675152546802, + "learning_rate": 0.003, + "loss": 4.0816, + "step": 7428 + }, + { + "epoch": 0.07429, + "grad_norm": 0.8229301998100853, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 7429 + }, + { + "epoch": 0.0743, + "grad_norm": 0.9460447102100716, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 7430 + }, + { + "epoch": 0.07431, + "grad_norm": 0.9332328649332472, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 7431 + }, + { + "epoch": 0.07432, + "grad_norm": 0.9830727368167865, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 7432 + }, + { + "epoch": 0.07433, + "grad_norm": 0.9991366907192593, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 7433 + }, + { + "epoch": 0.07434, + "grad_norm": 0.866600698284322, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 7434 + }, + { + "epoch": 0.07435, + "grad_norm": 0.8743247826380333, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 7435 + }, + { + "epoch": 0.07436, + "grad_norm": 0.8237770378592496, + "learning_rate": 0.003, + "loss": 4.0885, + "step": 7436 + }, + { + "epoch": 0.07437, + "grad_norm": 0.7522595604561741, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 7437 + }, + { + "epoch": 0.07438, + "grad_norm": 0.6964868247436655, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 7438 + }, + { + "epoch": 0.07439, + "grad_norm": 0.6953824483500841, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7439 + }, + { + "epoch": 0.0744, + "grad_norm": 0.6845671843622471, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 7440 + }, + { + "epoch": 0.07441, + "grad_norm": 0.6553559212732933, + "learning_rate": 0.003, + "loss": 4.0919, + "step": 7441 + }, + { + "epoch": 0.07442, + "grad_norm": 0.5898733616787618, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 7442 + }, + { + "epoch": 0.07443, + "grad_norm": 0.4845806113294189, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 7443 + }, + { + "epoch": 0.07444, + "grad_norm": 0.4689560196428437, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 7444 + }, + { + "epoch": 0.07445, + "grad_norm": 0.4480815113682579, + "learning_rate": 0.003, + "loss": 4.041, + "step": 7445 + }, + { + "epoch": 0.07446, + "grad_norm": 0.486010951268112, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 7446 + }, + { + "epoch": 0.07447, + "grad_norm": 0.4812566676557777, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 7447 + }, + { + "epoch": 0.07448, + "grad_norm": 0.4596667049520752, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 7448 + }, + { + "epoch": 0.07449, + "grad_norm": 0.4427259324604104, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 7449 + }, + { + "epoch": 0.0745, + "grad_norm": 0.4514509244837985, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 7450 + }, + { + "epoch": 0.07451, + "grad_norm": 0.4260145734361279, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 7451 + }, + { + "epoch": 0.07452, + "grad_norm": 0.4401709861164209, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 7452 + }, + { + "epoch": 0.07453, + "grad_norm": 0.4576956792567787, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 7453 + }, + { + "epoch": 0.07454, + "grad_norm": 0.49375953883051676, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 7454 + }, + { + "epoch": 0.07455, + "grad_norm": 0.7584370539035131, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 7455 + }, + { + "epoch": 0.07456, + "grad_norm": 1.1570953358608245, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 7456 + }, + { + "epoch": 0.07457, + "grad_norm": 0.8249537205862102, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 7457 + }, + { + "epoch": 0.07458, + "grad_norm": 0.6870906123549062, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 7458 + }, + { + "epoch": 0.07459, + "grad_norm": 0.7663124507912973, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 7459 + }, + { + "epoch": 0.0746, + "grad_norm": 0.7505150245574198, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 7460 + }, + { + "epoch": 0.07461, + "grad_norm": 0.6290211997162473, + "learning_rate": 0.003, + "loss": 4.053, + "step": 7461 + }, + { + "epoch": 0.07462, + "grad_norm": 0.7041466560755248, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 7462 + }, + { + "epoch": 0.07463, + "grad_norm": 0.7477976530136671, + "learning_rate": 0.003, + "loss": 4.0772, + "step": 7463 + }, + { + "epoch": 0.07464, + "grad_norm": 0.651891829398325, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 7464 + }, + { + "epoch": 0.07465, + "grad_norm": 0.6185416950915306, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 7465 + }, + { + "epoch": 0.07466, + "grad_norm": 0.7315246941932765, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 7466 + }, + { + "epoch": 0.07467, + "grad_norm": 0.7370591837642468, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 7467 + }, + { + "epoch": 0.07468, + "grad_norm": 0.7599489906702022, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 7468 + }, + { + "epoch": 0.07469, + "grad_norm": 0.8436058951910103, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 7469 + }, + { + "epoch": 0.0747, + "grad_norm": 0.9002900858344719, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 7470 + }, + { + "epoch": 0.07471, + "grad_norm": 0.8877148734325953, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 7471 + }, + { + "epoch": 0.07472, + "grad_norm": 0.7267532572570661, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 7472 + }, + { + "epoch": 0.07473, + "grad_norm": 0.675027126487361, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 7473 + }, + { + "epoch": 0.07474, + "grad_norm": 0.5881533545369908, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 7474 + }, + { + "epoch": 0.07475, + "grad_norm": 0.6027263178798289, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 7475 + }, + { + "epoch": 0.07476, + "grad_norm": 0.531600622609373, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 7476 + }, + { + "epoch": 0.07477, + "grad_norm": 0.5857700661084179, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 7477 + }, + { + "epoch": 0.07478, + "grad_norm": 0.550946838836272, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 7478 + }, + { + "epoch": 0.07479, + "grad_norm": 0.4731726309966069, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 7479 + }, + { + "epoch": 0.0748, + "grad_norm": 0.45318385316024584, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 7480 + }, + { + "epoch": 0.07481, + "grad_norm": 0.48623712899720595, + "learning_rate": 0.003, + "loss": 4.039, + "step": 7481 + }, + { + "epoch": 0.07482, + "grad_norm": 0.5668319645658599, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 7482 + }, + { + "epoch": 0.07483, + "grad_norm": 0.655014500993717, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 7483 + }, + { + "epoch": 0.07484, + "grad_norm": 0.9006335161935384, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 7484 + }, + { + "epoch": 0.07485, + "grad_norm": 1.3073488350135698, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 7485 + }, + { + "epoch": 0.07486, + "grad_norm": 0.6241397003925787, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 7486 + }, + { + "epoch": 0.07487, + "grad_norm": 0.5684977002949013, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 7487 + }, + { + "epoch": 0.07488, + "grad_norm": 0.6211527439843805, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 7488 + }, + { + "epoch": 0.07489, + "grad_norm": 0.7811270212292504, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 7489 + }, + { + "epoch": 0.0749, + "grad_norm": 1.0089556173136367, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7490 + }, + { + "epoch": 0.07491, + "grad_norm": 1.106593969661189, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 7491 + }, + { + "epoch": 0.07492, + "grad_norm": 0.7637045933315925, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 7492 + }, + { + "epoch": 0.07493, + "grad_norm": 0.835125437039064, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 7493 + }, + { + "epoch": 0.07494, + "grad_norm": 0.8784573493203266, + "learning_rate": 0.003, + "loss": 4.0829, + "step": 7494 + }, + { + "epoch": 0.07495, + "grad_norm": 0.8772811496444803, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 7495 + }, + { + "epoch": 0.07496, + "grad_norm": 0.8901513637433354, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 7496 + }, + { + "epoch": 0.07497, + "grad_norm": 0.9312904175905978, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 7497 + }, + { + "epoch": 0.07498, + "grad_norm": 1.0204554526696257, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 7498 + }, + { + "epoch": 0.07499, + "grad_norm": 0.8999514985066407, + "learning_rate": 0.003, + "loss": 4.0678, + "step": 7499 + }, + { + "epoch": 0.075, + "grad_norm": 0.8745053805730034, + "learning_rate": 0.003, + "loss": 4.0856, + "step": 7500 + }, + { + "epoch": 0.07501, + "grad_norm": 0.8155167444876013, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 7501 + }, + { + "epoch": 0.07502, + "grad_norm": 0.8372977351090519, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 7502 + }, + { + "epoch": 0.07503, + "grad_norm": 0.8437799463433308, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 7503 + }, + { + "epoch": 0.07504, + "grad_norm": 0.9136312841518225, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 7504 + }, + { + "epoch": 0.07505, + "grad_norm": 0.819546274697563, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 7505 + }, + { + "epoch": 0.07506, + "grad_norm": 0.783639730054514, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 7506 + }, + { + "epoch": 0.07507, + "grad_norm": 0.8143001172517857, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 7507 + }, + { + "epoch": 0.07508, + "grad_norm": 0.7203225306002784, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 7508 + }, + { + "epoch": 0.07509, + "grad_norm": 0.7013375957046832, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 7509 + }, + { + "epoch": 0.0751, + "grad_norm": 0.5396207843543408, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 7510 + }, + { + "epoch": 0.07511, + "grad_norm": 0.532015565083253, + "learning_rate": 0.003, + "loss": 4.04, + "step": 7511 + }, + { + "epoch": 0.07512, + "grad_norm": 0.5479166196829706, + "learning_rate": 0.003, + "loss": 4.078, + "step": 7512 + }, + { + "epoch": 0.07513, + "grad_norm": 0.6143194625604115, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 7513 + }, + { + "epoch": 0.07514, + "grad_norm": 0.6593346593901741, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 7514 + }, + { + "epoch": 0.07515, + "grad_norm": 0.5772232271036999, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 7515 + }, + { + "epoch": 0.07516, + "grad_norm": 0.5080419135116755, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 7516 + }, + { + "epoch": 0.07517, + "grad_norm": 0.5763642673624367, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7517 + }, + { + "epoch": 0.07518, + "grad_norm": 0.6679121838623565, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 7518 + }, + { + "epoch": 0.07519, + "grad_norm": 0.7736365752757945, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 7519 + }, + { + "epoch": 0.0752, + "grad_norm": 0.9306067729683678, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 7520 + }, + { + "epoch": 0.07521, + "grad_norm": 1.0552835576856872, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 7521 + }, + { + "epoch": 0.07522, + "grad_norm": 0.7240335978378571, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 7522 + }, + { + "epoch": 0.07523, + "grad_norm": 0.7020278407669496, + "learning_rate": 0.003, + "loss": 4.035, + "step": 7523 + }, + { + "epoch": 0.07524, + "grad_norm": 0.7377250940375731, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 7524 + }, + { + "epoch": 0.07525, + "grad_norm": 0.7793572497010395, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 7525 + }, + { + "epoch": 0.07526, + "grad_norm": 0.7324556076303901, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 7526 + }, + { + "epoch": 0.07527, + "grad_norm": 0.6286502042898248, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 7527 + }, + { + "epoch": 0.07528, + "grad_norm": 0.6260861775458725, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 7528 + }, + { + "epoch": 0.07529, + "grad_norm": 0.6972672810962912, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 7529 + }, + { + "epoch": 0.0753, + "grad_norm": 0.7978591706621103, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 7530 + }, + { + "epoch": 0.07531, + "grad_norm": 0.7712152882427815, + "learning_rate": 0.003, + "loss": 4.0724, + "step": 7531 + }, + { + "epoch": 0.07532, + "grad_norm": 0.7907260445662537, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 7532 + }, + { + "epoch": 0.07533, + "grad_norm": 0.7758872334909813, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 7533 + }, + { + "epoch": 0.07534, + "grad_norm": 0.767472639640663, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 7534 + }, + { + "epoch": 0.07535, + "grad_norm": 0.766187973077159, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 7535 + }, + { + "epoch": 0.07536, + "grad_norm": 0.7150450390551791, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 7536 + }, + { + "epoch": 0.07537, + "grad_norm": 0.6564289096636778, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 7537 + }, + { + "epoch": 0.07538, + "grad_norm": 0.7729121457246384, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 7538 + }, + { + "epoch": 0.07539, + "grad_norm": 0.8277366996193086, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 7539 + }, + { + "epoch": 0.0754, + "grad_norm": 0.8356891585156353, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 7540 + }, + { + "epoch": 0.07541, + "grad_norm": 0.8475324778770377, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 7541 + }, + { + "epoch": 0.07542, + "grad_norm": 0.933260050409068, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 7542 + }, + { + "epoch": 0.07543, + "grad_norm": 0.7376563429405116, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 7543 + }, + { + "epoch": 0.07544, + "grad_norm": 0.6761208303907776, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 7544 + }, + { + "epoch": 0.07545, + "grad_norm": 0.610223183731481, + "learning_rate": 0.003, + "loss": 4.0794, + "step": 7545 + }, + { + "epoch": 0.07546, + "grad_norm": 0.5296982352068923, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 7546 + }, + { + "epoch": 0.07547, + "grad_norm": 0.472463153342835, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 7547 + }, + { + "epoch": 0.07548, + "grad_norm": 0.4638853675047035, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 7548 + }, + { + "epoch": 0.07549, + "grad_norm": 0.5267220424276784, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 7549 + }, + { + "epoch": 0.0755, + "grad_norm": 0.6440791891609617, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 7550 + }, + { + "epoch": 0.07551, + "grad_norm": 0.7477090391252855, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 7551 + }, + { + "epoch": 0.07552, + "grad_norm": 0.881581151916062, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 7552 + }, + { + "epoch": 0.07553, + "grad_norm": 0.9300288193599606, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 7553 + }, + { + "epoch": 0.07554, + "grad_norm": 0.8205772658554302, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 7554 + }, + { + "epoch": 0.07555, + "grad_norm": 0.7763153378964374, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 7555 + }, + { + "epoch": 0.07556, + "grad_norm": 0.777022825713419, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 7556 + }, + { + "epoch": 0.07557, + "grad_norm": 0.7462601146480812, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 7557 + }, + { + "epoch": 0.07558, + "grad_norm": 0.6000427379839431, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 7558 + }, + { + "epoch": 0.07559, + "grad_norm": 0.652365291304031, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 7559 + }, + { + "epoch": 0.0756, + "grad_norm": 0.6209643030840687, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7560 + }, + { + "epoch": 0.07561, + "grad_norm": 0.6247012074046319, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 7561 + }, + { + "epoch": 0.07562, + "grad_norm": 0.6145852151802866, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 7562 + }, + { + "epoch": 0.07563, + "grad_norm": 0.5762017961132972, + "learning_rate": 0.003, + "loss": 4.043, + "step": 7563 + }, + { + "epoch": 0.07564, + "grad_norm": 0.44100691209366116, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 7564 + }, + { + "epoch": 0.07565, + "grad_norm": 0.4120591869310952, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 7565 + }, + { + "epoch": 0.07566, + "grad_norm": 0.43429244439528253, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 7566 + }, + { + "epoch": 0.07567, + "grad_norm": 0.3986723533229915, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 7567 + }, + { + "epoch": 0.07568, + "grad_norm": 0.4419281681197685, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 7568 + }, + { + "epoch": 0.07569, + "grad_norm": 0.5373312792210869, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 7569 + }, + { + "epoch": 0.0757, + "grad_norm": 0.5482490947901065, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 7570 + }, + { + "epoch": 0.07571, + "grad_norm": 0.5807554883143237, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 7571 + }, + { + "epoch": 0.07572, + "grad_norm": 0.539482558876458, + "learning_rate": 0.003, + "loss": 4.013, + "step": 7572 + }, + { + "epoch": 0.07573, + "grad_norm": 0.5578970110644322, + "learning_rate": 0.003, + "loss": 4.031, + "step": 7573 + }, + { + "epoch": 0.07574, + "grad_norm": 0.6321593340495187, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 7574 + }, + { + "epoch": 0.07575, + "grad_norm": 0.6350259997008573, + "learning_rate": 0.003, + "loss": 4.03, + "step": 7575 + }, + { + "epoch": 0.07576, + "grad_norm": 0.605385484218955, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 7576 + }, + { + "epoch": 0.07577, + "grad_norm": 0.6896503401466064, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 7577 + }, + { + "epoch": 0.07578, + "grad_norm": 0.9497296196058909, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 7578 + }, + { + "epoch": 0.07579, + "grad_norm": 1.1362903807211815, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 7579 + }, + { + "epoch": 0.0758, + "grad_norm": 0.715087526196801, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 7580 + }, + { + "epoch": 0.07581, + "grad_norm": 0.7200552056297321, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 7581 + }, + { + "epoch": 0.07582, + "grad_norm": 0.8492796246196268, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 7582 + }, + { + "epoch": 0.07583, + "grad_norm": 0.796839787507485, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 7583 + }, + { + "epoch": 0.07584, + "grad_norm": 0.7500362083340513, + "learning_rate": 0.003, + "loss": 4.0893, + "step": 7584 + }, + { + "epoch": 0.07585, + "grad_norm": 0.7574069537911439, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 7585 + }, + { + "epoch": 0.07586, + "grad_norm": 0.7939087088309319, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 7586 + }, + { + "epoch": 0.07587, + "grad_norm": 0.7876202697324444, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 7587 + }, + { + "epoch": 0.07588, + "grad_norm": 0.816298643717335, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 7588 + }, + { + "epoch": 0.07589, + "grad_norm": 0.8483320696318782, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 7589 + }, + { + "epoch": 0.0759, + "grad_norm": 0.8911579129468967, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 7590 + }, + { + "epoch": 0.07591, + "grad_norm": 0.8329062274704158, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 7591 + }, + { + "epoch": 0.07592, + "grad_norm": 0.7828430360351878, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 7592 + }, + { + "epoch": 0.07593, + "grad_norm": 0.7319086658314898, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 7593 + }, + { + "epoch": 0.07594, + "grad_norm": 0.6551633943241083, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 7594 + }, + { + "epoch": 0.07595, + "grad_norm": 0.6876270730097411, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 7595 + }, + { + "epoch": 0.07596, + "grad_norm": 0.6664009022583618, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 7596 + }, + { + "epoch": 0.07597, + "grad_norm": 0.7675263669667872, + "learning_rate": 0.003, + "loss": 4.074, + "step": 7597 + }, + { + "epoch": 0.07598, + "grad_norm": 0.8570590675925124, + "learning_rate": 0.003, + "loss": 4.069, + "step": 7598 + }, + { + "epoch": 0.07599, + "grad_norm": 0.8247508859755924, + "learning_rate": 0.003, + "loss": 4.0714, + "step": 7599 + }, + { + "epoch": 0.076, + "grad_norm": 0.8026493936878719, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 7600 + }, + { + "epoch": 0.07601, + "grad_norm": 0.776045935797423, + "learning_rate": 0.003, + "loss": 4.0763, + "step": 7601 + }, + { + "epoch": 0.07602, + "grad_norm": 0.7181902085438208, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 7602 + }, + { + "epoch": 0.07603, + "grad_norm": 0.7657077438188911, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7603 + }, + { + "epoch": 0.07604, + "grad_norm": 0.7941977075124232, + "learning_rate": 0.003, + "loss": 4.0671, + "step": 7604 + }, + { + "epoch": 0.07605, + "grad_norm": 0.6428049367649177, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 7605 + }, + { + "epoch": 0.07606, + "grad_norm": 0.6134743132020162, + "learning_rate": 0.003, + "loss": 4.1062, + "step": 7606 + }, + { + "epoch": 0.07607, + "grad_norm": 0.6687305764242613, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 7607 + }, + { + "epoch": 0.07608, + "grad_norm": 0.6743645961373843, + "learning_rate": 0.003, + "loss": 4.0771, + "step": 7608 + }, + { + "epoch": 0.07609, + "grad_norm": 0.735377963859144, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 7609 + }, + { + "epoch": 0.0761, + "grad_norm": 0.7661087762742179, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 7610 + }, + { + "epoch": 0.07611, + "grad_norm": 0.6938653598743548, + "learning_rate": 0.003, + "loss": 4.0748, + "step": 7611 + }, + { + "epoch": 0.07612, + "grad_norm": 0.6759237837598605, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 7612 + }, + { + "epoch": 0.07613, + "grad_norm": 0.6022667403048313, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 7613 + }, + { + "epoch": 0.07614, + "grad_norm": 0.5835862421508177, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 7614 + }, + { + "epoch": 0.07615, + "grad_norm": 0.57220929899597, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 7615 + }, + { + "epoch": 0.07616, + "grad_norm": 0.5447237056099387, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 7616 + }, + { + "epoch": 0.07617, + "grad_norm": 0.4812162438815339, + "learning_rate": 0.003, + "loss": 4.006, + "step": 7617 + }, + { + "epoch": 0.07618, + "grad_norm": 0.42972280290033293, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 7618 + }, + { + "epoch": 0.07619, + "grad_norm": 0.40704236755017975, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 7619 + }, + { + "epoch": 0.0762, + "grad_norm": 0.4561133530071867, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 7620 + }, + { + "epoch": 0.07621, + "grad_norm": 0.4857082553264001, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 7621 + }, + { + "epoch": 0.07622, + "grad_norm": 0.5901433336622326, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 7622 + }, + { + "epoch": 0.07623, + "grad_norm": 0.6561607388709625, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 7623 + }, + { + "epoch": 0.07624, + "grad_norm": 0.7168166328134451, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 7624 + }, + { + "epoch": 0.07625, + "grad_norm": 0.6753061379457685, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 7625 + }, + { + "epoch": 0.07626, + "grad_norm": 0.6009842945781215, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 7626 + }, + { + "epoch": 0.07627, + "grad_norm": 0.605895956937568, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 7627 + }, + { + "epoch": 0.07628, + "grad_norm": 0.6755998831606779, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 7628 + }, + { + "epoch": 0.07629, + "grad_norm": 0.7949403118983037, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 7629 + }, + { + "epoch": 0.0763, + "grad_norm": 0.7835077479923791, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 7630 + }, + { + "epoch": 0.07631, + "grad_norm": 0.8426919379342346, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 7631 + }, + { + "epoch": 0.07632, + "grad_norm": 0.9997742385919999, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 7632 + }, + { + "epoch": 0.07633, + "grad_norm": 1.0805464047135134, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 7633 + }, + { + "epoch": 0.07634, + "grad_norm": 0.8735358301676082, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 7634 + }, + { + "epoch": 0.07635, + "grad_norm": 0.8050555515500526, + "learning_rate": 0.003, + "loss": 4.055, + "step": 7635 + }, + { + "epoch": 0.07636, + "grad_norm": 0.7014926481401261, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 7636 + }, + { + "epoch": 0.07637, + "grad_norm": 0.7115733773755004, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 7637 + }, + { + "epoch": 0.07638, + "grad_norm": 0.8250768935974896, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7638 + }, + { + "epoch": 0.07639, + "grad_norm": 0.8529222211154474, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 7639 + }, + { + "epoch": 0.0764, + "grad_norm": 0.7760093756150678, + "learning_rate": 0.003, + "loss": 4.0873, + "step": 7640 + }, + { + "epoch": 0.07641, + "grad_norm": 0.7969295579882023, + "learning_rate": 0.003, + "loss": 4.08, + "step": 7641 + }, + { + "epoch": 0.07642, + "grad_norm": 0.731176833189907, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 7642 + }, + { + "epoch": 0.07643, + "grad_norm": 0.633837319158038, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 7643 + }, + { + "epoch": 0.07644, + "grad_norm": 0.6082374339779484, + "learning_rate": 0.003, + "loss": 4.027, + "step": 7644 + }, + { + "epoch": 0.07645, + "grad_norm": 0.6354610005957227, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 7645 + }, + { + "epoch": 0.07646, + "grad_norm": 0.656910891742267, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 7646 + }, + { + "epoch": 0.07647, + "grad_norm": 0.6527667241982296, + "learning_rate": 0.003, + "loss": 4.025, + "step": 7647 + }, + { + "epoch": 0.07648, + "grad_norm": 0.5765455587385564, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 7648 + }, + { + "epoch": 0.07649, + "grad_norm": 0.5469679700793885, + "learning_rate": 0.003, + "loss": 4.04, + "step": 7649 + }, + { + "epoch": 0.0765, + "grad_norm": 0.5573395422134555, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 7650 + }, + { + "epoch": 0.07651, + "grad_norm": 0.5529177207107702, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 7651 + }, + { + "epoch": 0.07652, + "grad_norm": 0.49451590777240895, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 7652 + }, + { + "epoch": 0.07653, + "grad_norm": 0.5326949833114069, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 7653 + }, + { + "epoch": 0.07654, + "grad_norm": 0.6052512497561293, + "learning_rate": 0.003, + "loss": 4.041, + "step": 7654 + }, + { + "epoch": 0.07655, + "grad_norm": 0.6284191357020593, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 7655 + }, + { + "epoch": 0.07656, + "grad_norm": 0.6486770470060506, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 7656 + }, + { + "epoch": 0.07657, + "grad_norm": 0.8347726656621142, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 7657 + }, + { + "epoch": 0.07658, + "grad_norm": 0.9303171631678393, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 7658 + }, + { + "epoch": 0.07659, + "grad_norm": 0.9842092966678585, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 7659 + }, + { + "epoch": 0.0766, + "grad_norm": 0.9838097074862379, + "learning_rate": 0.003, + "loss": 4.0881, + "step": 7660 + }, + { + "epoch": 0.07661, + "grad_norm": 0.9864470329466856, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 7661 + }, + { + "epoch": 0.07662, + "grad_norm": 0.9238413121998281, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 7662 + }, + { + "epoch": 0.07663, + "grad_norm": 0.8759981953066126, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 7663 + }, + { + "epoch": 0.07664, + "grad_norm": 0.7364522936236557, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 7664 + }, + { + "epoch": 0.07665, + "grad_norm": 0.6773762415907468, + "learning_rate": 0.003, + "loss": 4.033, + "step": 7665 + }, + { + "epoch": 0.07666, + "grad_norm": 0.6791538751941775, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 7666 + }, + { + "epoch": 0.07667, + "grad_norm": 0.6813007808360633, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 7667 + }, + { + "epoch": 0.07668, + "grad_norm": 0.7219382254041767, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 7668 + }, + { + "epoch": 0.07669, + "grad_norm": 0.7472048500462506, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 7669 + }, + { + "epoch": 0.0767, + "grad_norm": 0.8154744894616025, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 7670 + }, + { + "epoch": 0.07671, + "grad_norm": 0.8000664977278562, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 7671 + }, + { + "epoch": 0.07672, + "grad_norm": 0.6802144710436505, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 7672 + }, + { + "epoch": 0.07673, + "grad_norm": 0.5808732362274317, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 7673 + }, + { + "epoch": 0.07674, + "grad_norm": 0.5104137432330133, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 7674 + }, + { + "epoch": 0.07675, + "grad_norm": 0.6151997358818722, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 7675 + }, + { + "epoch": 0.07676, + "grad_norm": 0.6407140715368373, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 7676 + }, + { + "epoch": 0.07677, + "grad_norm": 0.5969328903585241, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 7677 + }, + { + "epoch": 0.07678, + "grad_norm": 0.6636028970005442, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 7678 + }, + { + "epoch": 0.07679, + "grad_norm": 0.8629227367574791, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 7679 + }, + { + "epoch": 0.0768, + "grad_norm": 0.9057061464304386, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 7680 + }, + { + "epoch": 0.07681, + "grad_norm": 0.7852580126331745, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 7681 + }, + { + "epoch": 0.07682, + "grad_norm": 0.6963928491491511, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 7682 + }, + { + "epoch": 0.07683, + "grad_norm": 0.6392760489865914, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 7683 + }, + { + "epoch": 0.07684, + "grad_norm": 0.6636426909877271, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 7684 + }, + { + "epoch": 0.07685, + "grad_norm": 0.7747904264119995, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 7685 + }, + { + "epoch": 0.07686, + "grad_norm": 0.7373351048383946, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 7686 + }, + { + "epoch": 0.07687, + "grad_norm": 0.6776655947531807, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 7687 + }, + { + "epoch": 0.07688, + "grad_norm": 0.6664927622632959, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 7688 + }, + { + "epoch": 0.07689, + "grad_norm": 0.7036630866262804, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 7689 + }, + { + "epoch": 0.0769, + "grad_norm": 0.8423736962511771, + "learning_rate": 0.003, + "loss": 4.062, + "step": 7690 + }, + { + "epoch": 0.07691, + "grad_norm": 1.0737155954652369, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 7691 + }, + { + "epoch": 0.07692, + "grad_norm": 0.7649285268493068, + "learning_rate": 0.003, + "loss": 4.0654, + "step": 7692 + }, + { + "epoch": 0.07693, + "grad_norm": 0.6542968488963911, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 7693 + }, + { + "epoch": 0.07694, + "grad_norm": 0.7310943790115392, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 7694 + }, + { + "epoch": 0.07695, + "grad_norm": 0.8219474178357663, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 7695 + }, + { + "epoch": 0.07696, + "grad_norm": 0.7937239831250903, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 7696 + }, + { + "epoch": 0.07697, + "grad_norm": 0.7206238136545898, + "learning_rate": 0.003, + "loss": 4.0834, + "step": 7697 + }, + { + "epoch": 0.07698, + "grad_norm": 0.6788497096743122, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 7698 + }, + { + "epoch": 0.07699, + "grad_norm": 0.6779971341102946, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 7699 + }, + { + "epoch": 0.077, + "grad_norm": 0.6030167361153842, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 7700 + }, + { + "epoch": 0.07701, + "grad_norm": 0.6251132015438084, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 7701 + }, + { + "epoch": 0.07702, + "grad_norm": 0.7183860982451377, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 7702 + }, + { + "epoch": 0.07703, + "grad_norm": 0.6711554934408176, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 7703 + }, + { + "epoch": 0.07704, + "grad_norm": 0.6406311128440741, + "learning_rate": 0.003, + "loss": 4.052, + "step": 7704 + }, + { + "epoch": 0.07705, + "grad_norm": 0.6001008412078687, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 7705 + }, + { + "epoch": 0.07706, + "grad_norm": 0.6098995429359666, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 7706 + }, + { + "epoch": 0.07707, + "grad_norm": 0.580799386097219, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 7707 + }, + { + "epoch": 0.07708, + "grad_norm": 0.5526293275379175, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 7708 + }, + { + "epoch": 0.07709, + "grad_norm": 0.5847646665032754, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 7709 + }, + { + "epoch": 0.0771, + "grad_norm": 0.6491186329509964, + "learning_rate": 0.003, + "loss": 4.021, + "step": 7710 + }, + { + "epoch": 0.07711, + "grad_norm": 0.6961727926634954, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 7711 + }, + { + "epoch": 0.07712, + "grad_norm": 0.8366650521294742, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 7712 + }, + { + "epoch": 0.07713, + "grad_norm": 0.8981896944098439, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 7713 + }, + { + "epoch": 0.07714, + "grad_norm": 0.8273405919493078, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 7714 + }, + { + "epoch": 0.07715, + "grad_norm": 0.7373899495810222, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 7715 + }, + { + "epoch": 0.07716, + "grad_norm": 0.7586949421566886, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 7716 + }, + { + "epoch": 0.07717, + "grad_norm": 0.7911829380141604, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 7717 + }, + { + "epoch": 0.07718, + "grad_norm": 0.7188004597874216, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 7718 + }, + { + "epoch": 0.07719, + "grad_norm": 0.6415372632411966, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 7719 + }, + { + "epoch": 0.0772, + "grad_norm": 0.7006654447031044, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 7720 + }, + { + "epoch": 0.07721, + "grad_norm": 0.7193836743695385, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 7721 + }, + { + "epoch": 0.07722, + "grad_norm": 0.676416949388731, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 7722 + }, + { + "epoch": 0.07723, + "grad_norm": 0.6867682077988717, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 7723 + }, + { + "epoch": 0.07724, + "grad_norm": 0.6303547197209717, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 7724 + }, + { + "epoch": 0.07725, + "grad_norm": 0.561537011454582, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 7725 + }, + { + "epoch": 0.07726, + "grad_norm": 0.6094284732398672, + "learning_rate": 0.003, + "loss": 4.062, + "step": 7726 + }, + { + "epoch": 0.07727, + "grad_norm": 0.5977745977827622, + "learning_rate": 0.003, + "loss": 4.02, + "step": 7727 + }, + { + "epoch": 0.07728, + "grad_norm": 0.5974264434681982, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 7728 + }, + { + "epoch": 0.07729, + "grad_norm": 0.6385926728134873, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 7729 + }, + { + "epoch": 0.0773, + "grad_norm": 0.7007420323637349, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 7730 + }, + { + "epoch": 0.07731, + "grad_norm": 0.7023455937867197, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 7731 + }, + { + "epoch": 0.07732, + "grad_norm": 0.7552706430554972, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 7732 + }, + { + "epoch": 0.07733, + "grad_norm": 0.889009980024093, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 7733 + }, + { + "epoch": 0.07734, + "grad_norm": 0.9038778059893959, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 7734 + }, + { + "epoch": 0.07735, + "grad_norm": 0.7387198970535299, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 7735 + }, + { + "epoch": 0.07736, + "grad_norm": 0.5943541307171945, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 7736 + }, + { + "epoch": 0.07737, + "grad_norm": 0.5659082549422041, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 7737 + }, + { + "epoch": 0.07738, + "grad_norm": 0.5785233084939277, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 7738 + }, + { + "epoch": 0.07739, + "grad_norm": 0.5317368388592326, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 7739 + }, + { + "epoch": 0.0774, + "grad_norm": 0.5505549803396771, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 7740 + }, + { + "epoch": 0.07741, + "grad_norm": 0.6042973868730916, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 7741 + }, + { + "epoch": 0.07742, + "grad_norm": 0.7106469283153454, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 7742 + }, + { + "epoch": 0.07743, + "grad_norm": 0.7381244413530601, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 7743 + }, + { + "epoch": 0.07744, + "grad_norm": 0.7449857545859564, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 7744 + }, + { + "epoch": 0.07745, + "grad_norm": 0.7487149832008485, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 7745 + }, + { + "epoch": 0.07746, + "grad_norm": 0.7076426197730348, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 7746 + }, + { + "epoch": 0.07747, + "grad_norm": 0.7117362298174772, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 7747 + }, + { + "epoch": 0.07748, + "grad_norm": 0.7752115492234533, + "learning_rate": 0.003, + "loss": 4.0725, + "step": 7748 + }, + { + "epoch": 0.07749, + "grad_norm": 0.8853015186328781, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 7749 + }, + { + "epoch": 0.0775, + "grad_norm": 0.7574071104773505, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 7750 + }, + { + "epoch": 0.07751, + "grad_norm": 0.8372535177533414, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 7751 + }, + { + "epoch": 0.07752, + "grad_norm": 0.9211587372348612, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 7752 + }, + { + "epoch": 0.07753, + "grad_norm": 0.8973771658760825, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 7753 + }, + { + "epoch": 0.07754, + "grad_norm": 0.7389715432369666, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 7754 + }, + { + "epoch": 0.07755, + "grad_norm": 0.6695677066163175, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 7755 + }, + { + "epoch": 0.07756, + "grad_norm": 0.6942425577327686, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 7756 + }, + { + "epoch": 0.07757, + "grad_norm": 0.7934531618932724, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 7757 + }, + { + "epoch": 0.07758, + "grad_norm": 0.8396215907446974, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 7758 + }, + { + "epoch": 0.07759, + "grad_norm": 0.7362786353417305, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 7759 + }, + { + "epoch": 0.0776, + "grad_norm": 0.6730594548200651, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 7760 + }, + { + "epoch": 0.07761, + "grad_norm": 0.608370890912831, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 7761 + }, + { + "epoch": 0.07762, + "grad_norm": 0.7678796597920887, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 7762 + }, + { + "epoch": 0.07763, + "grad_norm": 0.9034275299946557, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 7763 + }, + { + "epoch": 0.07764, + "grad_norm": 0.9729504349655794, + "learning_rate": 0.003, + "loss": 4.0789, + "step": 7764 + }, + { + "epoch": 0.07765, + "grad_norm": 0.9384405384021188, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 7765 + }, + { + "epoch": 0.07766, + "grad_norm": 1.0381424865947604, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 7766 + }, + { + "epoch": 0.07767, + "grad_norm": 0.8193679135494465, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 7767 + }, + { + "epoch": 0.07768, + "grad_norm": 0.8022689042136126, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 7768 + }, + { + "epoch": 0.07769, + "grad_norm": 0.7142478835068574, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 7769 + }, + { + "epoch": 0.0777, + "grad_norm": 0.7822671319039305, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 7770 + }, + { + "epoch": 0.07771, + "grad_norm": 0.6926353494668392, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 7771 + }, + { + "epoch": 0.07772, + "grad_norm": 0.6595775156027098, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 7772 + }, + { + "epoch": 0.07773, + "grad_norm": 0.6695052598345946, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 7773 + }, + { + "epoch": 0.07774, + "grad_norm": 0.6338916352476932, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 7774 + }, + { + "epoch": 0.07775, + "grad_norm": 0.6660274502055001, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 7775 + }, + { + "epoch": 0.07776, + "grad_norm": 0.653471518012234, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 7776 + }, + { + "epoch": 0.07777, + "grad_norm": 0.7230110562093495, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 7777 + }, + { + "epoch": 0.07778, + "grad_norm": 0.7435494662967048, + "learning_rate": 0.003, + "loss": 4.071, + "step": 7778 + }, + { + "epoch": 0.07779, + "grad_norm": 0.6614597720007421, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 7779 + }, + { + "epoch": 0.0778, + "grad_norm": 0.601646005220641, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 7780 + }, + { + "epoch": 0.07781, + "grad_norm": 0.6828240130970312, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 7781 + }, + { + "epoch": 0.07782, + "grad_norm": 0.764226650533448, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 7782 + }, + { + "epoch": 0.07783, + "grad_norm": 0.7049634652261119, + "learning_rate": 0.003, + "loss": 4.041, + "step": 7783 + }, + { + "epoch": 0.07784, + "grad_norm": 0.7581444851784971, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 7784 + }, + { + "epoch": 0.07785, + "grad_norm": 0.7723694525456029, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 7785 + }, + { + "epoch": 0.07786, + "grad_norm": 0.8859962308944258, + "learning_rate": 0.003, + "loss": 4.036, + "step": 7786 + }, + { + "epoch": 0.07787, + "grad_norm": 0.8041963631383058, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 7787 + }, + { + "epoch": 0.07788, + "grad_norm": 0.6200905982170827, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 7788 + }, + { + "epoch": 0.07789, + "grad_norm": 0.5444716347616775, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 7789 + }, + { + "epoch": 0.0779, + "grad_norm": 0.663865169322728, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 7790 + }, + { + "epoch": 0.07791, + "grad_norm": 0.7333581784180571, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 7791 + }, + { + "epoch": 0.07792, + "grad_norm": 0.6702336961709963, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 7792 + }, + { + "epoch": 0.07793, + "grad_norm": 0.5279105313466235, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 7793 + }, + { + "epoch": 0.07794, + "grad_norm": 0.5495141444372565, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 7794 + }, + { + "epoch": 0.07795, + "grad_norm": 0.5138694083868935, + "learning_rate": 0.003, + "loss": 4.005, + "step": 7795 + }, + { + "epoch": 0.07796, + "grad_norm": 0.5037839268997677, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 7796 + }, + { + "epoch": 0.07797, + "grad_norm": 0.46797596724668883, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 7797 + }, + { + "epoch": 0.07798, + "grad_norm": 0.44722857886849937, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 7798 + }, + { + "epoch": 0.07799, + "grad_norm": 0.4539303292831138, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 7799 + }, + { + "epoch": 0.078, + "grad_norm": 0.4218157948926157, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 7800 + }, + { + "epoch": 0.07801, + "grad_norm": 0.4721697933593621, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 7801 + }, + { + "epoch": 0.07802, + "grad_norm": 0.4563856609389985, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 7802 + }, + { + "epoch": 0.07803, + "grad_norm": 0.4962551404028748, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 7803 + }, + { + "epoch": 0.07804, + "grad_norm": 0.5600757286519844, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 7804 + }, + { + "epoch": 0.07805, + "grad_norm": 0.6591788779638021, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 7805 + }, + { + "epoch": 0.07806, + "grad_norm": 0.8080615622803791, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 7806 + }, + { + "epoch": 0.07807, + "grad_norm": 0.9709261221109322, + "learning_rate": 0.003, + "loss": 4.068, + "step": 7807 + }, + { + "epoch": 0.07808, + "grad_norm": 1.1525170975698793, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 7808 + }, + { + "epoch": 0.07809, + "grad_norm": 0.8654895890040383, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 7809 + }, + { + "epoch": 0.0781, + "grad_norm": 0.7788933331709564, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 7810 + }, + { + "epoch": 0.07811, + "grad_norm": 0.8204759669750299, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 7811 + }, + { + "epoch": 0.07812, + "grad_norm": 0.9125745271848327, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 7812 + }, + { + "epoch": 0.07813, + "grad_norm": 1.1911263381442552, + "learning_rate": 0.003, + "loss": 4.0823, + "step": 7813 + }, + { + "epoch": 0.07814, + "grad_norm": 0.7560965820935954, + "learning_rate": 0.003, + "loss": 4.0939, + "step": 7814 + }, + { + "epoch": 0.07815, + "grad_norm": 0.7448154332499729, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 7815 + }, + { + "epoch": 0.07816, + "grad_norm": 0.8855338315708976, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 7816 + }, + { + "epoch": 0.07817, + "grad_norm": 0.9499754938391336, + "learning_rate": 0.003, + "loss": 4.054, + "step": 7817 + }, + { + "epoch": 0.07818, + "grad_norm": 0.8895037635797708, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 7818 + }, + { + "epoch": 0.07819, + "grad_norm": 0.8553384656101726, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 7819 + }, + { + "epoch": 0.0782, + "grad_norm": 0.845505632593502, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 7820 + }, + { + "epoch": 0.07821, + "grad_norm": 0.8386742644875637, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 7821 + }, + { + "epoch": 0.07822, + "grad_norm": 0.86077603554862, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 7822 + }, + { + "epoch": 0.07823, + "grad_norm": 0.8811105758002923, + "learning_rate": 0.003, + "loss": 4.052, + "step": 7823 + }, + { + "epoch": 0.07824, + "grad_norm": 0.9365145265471595, + "learning_rate": 0.003, + "loss": 4.0896, + "step": 7824 + }, + { + "epoch": 0.07825, + "grad_norm": 0.8818311784282632, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 7825 + }, + { + "epoch": 0.07826, + "grad_norm": 0.7983686146896467, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 7826 + }, + { + "epoch": 0.07827, + "grad_norm": 0.7182637358811055, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 7827 + }, + { + "epoch": 0.07828, + "grad_norm": 0.6641399718249245, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 7828 + }, + { + "epoch": 0.07829, + "grad_norm": 0.6479905197528112, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 7829 + }, + { + "epoch": 0.0783, + "grad_norm": 0.6749520582117384, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 7830 + }, + { + "epoch": 0.07831, + "grad_norm": 0.6603029127595206, + "learning_rate": 0.003, + "loss": 4.045, + "step": 7831 + }, + { + "epoch": 0.07832, + "grad_norm": 0.6027085389629416, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 7832 + }, + { + "epoch": 0.07833, + "grad_norm": 0.6402446914110962, + "learning_rate": 0.003, + "loss": 4.047, + "step": 7833 + }, + { + "epoch": 0.07834, + "grad_norm": 0.6683095627686985, + "learning_rate": 0.003, + "loss": 4.0991, + "step": 7834 + }, + { + "epoch": 0.07835, + "grad_norm": 0.5711119126247584, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 7835 + }, + { + "epoch": 0.07836, + "grad_norm": 0.5184737463548346, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 7836 + }, + { + "epoch": 0.07837, + "grad_norm": 0.5219602815610779, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 7837 + }, + { + "epoch": 0.07838, + "grad_norm": 0.6351415270305494, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 7838 + }, + { + "epoch": 0.07839, + "grad_norm": 0.7030121978243518, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 7839 + }, + { + "epoch": 0.0784, + "grad_norm": 0.8027780287141163, + "learning_rate": 0.003, + "loss": 4.064, + "step": 7840 + }, + { + "epoch": 0.07841, + "grad_norm": 0.82149947192184, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 7841 + }, + { + "epoch": 0.07842, + "grad_norm": 0.8016623648227814, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 7842 + }, + { + "epoch": 0.07843, + "grad_norm": 0.7308070133564583, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 7843 + }, + { + "epoch": 0.07844, + "grad_norm": 0.6440320065328795, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 7844 + }, + { + "epoch": 0.07845, + "grad_norm": 0.7178090238555969, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 7845 + }, + { + "epoch": 0.07846, + "grad_norm": 0.7590167032503078, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 7846 + }, + { + "epoch": 0.07847, + "grad_norm": 0.7071667048882848, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 7847 + }, + { + "epoch": 0.07848, + "grad_norm": 0.6698055979830217, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 7848 + }, + { + "epoch": 0.07849, + "grad_norm": 0.6910132822489703, + "learning_rate": 0.003, + "loss": 4.074, + "step": 7849 + }, + { + "epoch": 0.0785, + "grad_norm": 0.7570928736376142, + "learning_rate": 0.003, + "loss": 4.046, + "step": 7850 + }, + { + "epoch": 0.07851, + "grad_norm": 0.8186017797759164, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 7851 + }, + { + "epoch": 0.07852, + "grad_norm": 0.7833296550337739, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 7852 + }, + { + "epoch": 0.07853, + "grad_norm": 0.7444067994281202, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 7853 + }, + { + "epoch": 0.07854, + "grad_norm": 0.7190264527996346, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 7854 + }, + { + "epoch": 0.07855, + "grad_norm": 0.6589593151571598, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 7855 + }, + { + "epoch": 0.07856, + "grad_norm": 0.6119907561486682, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 7856 + }, + { + "epoch": 0.07857, + "grad_norm": 0.6211860686411367, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 7857 + }, + { + "epoch": 0.07858, + "grad_norm": 0.5633081723962255, + "learning_rate": 0.003, + "loss": 4.052, + "step": 7858 + }, + { + "epoch": 0.07859, + "grad_norm": 0.5837543029746843, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 7859 + }, + { + "epoch": 0.0786, + "grad_norm": 0.7545392661840279, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 7860 + }, + { + "epoch": 0.07861, + "grad_norm": 0.8231221110716561, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 7861 + }, + { + "epoch": 0.07862, + "grad_norm": 0.8259073977921338, + "learning_rate": 0.003, + "loss": 4.084, + "step": 7862 + }, + { + "epoch": 0.07863, + "grad_norm": 0.7309689958542195, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 7863 + }, + { + "epoch": 0.07864, + "grad_norm": 0.6986816218613375, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7864 + }, + { + "epoch": 0.07865, + "grad_norm": 0.7055015494596285, + "learning_rate": 0.003, + "loss": 4.029, + "step": 7865 + }, + { + "epoch": 0.07866, + "grad_norm": 0.745554666158877, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 7866 + }, + { + "epoch": 0.07867, + "grad_norm": 0.6441740918105141, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 7867 + }, + { + "epoch": 0.07868, + "grad_norm": 0.6301368120928853, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 7868 + }, + { + "epoch": 0.07869, + "grad_norm": 0.6109297024486525, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 7869 + }, + { + "epoch": 0.0787, + "grad_norm": 0.5816100500284472, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 7870 + }, + { + "epoch": 0.07871, + "grad_norm": 0.6341495292901554, + "learning_rate": 0.003, + "loss": 4.0704, + "step": 7871 + }, + { + "epoch": 0.07872, + "grad_norm": 0.5827690320389015, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 7872 + }, + { + "epoch": 0.07873, + "grad_norm": 0.5211488124135346, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 7873 + }, + { + "epoch": 0.07874, + "grad_norm": 0.6481435824567602, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 7874 + }, + { + "epoch": 0.07875, + "grad_norm": 0.7040367891345678, + "learning_rate": 0.003, + "loss": 4.056, + "step": 7875 + }, + { + "epoch": 0.07876, + "grad_norm": 0.7837911856871046, + "learning_rate": 0.003, + "loss": 4.056, + "step": 7876 + }, + { + "epoch": 0.07877, + "grad_norm": 0.8781980226315174, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 7877 + }, + { + "epoch": 0.07878, + "grad_norm": 1.0008285994186386, + "learning_rate": 0.003, + "loss": 4.064, + "step": 7878 + }, + { + "epoch": 0.07879, + "grad_norm": 1.0569055896042627, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 7879 + }, + { + "epoch": 0.0788, + "grad_norm": 0.8393626387727805, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 7880 + }, + { + "epoch": 0.07881, + "grad_norm": 0.8415177734749038, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 7881 + }, + { + "epoch": 0.07882, + "grad_norm": 0.97310078821296, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 7882 + }, + { + "epoch": 0.07883, + "grad_norm": 0.9280304232510249, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 7883 + }, + { + "epoch": 0.07884, + "grad_norm": 0.9199718053633089, + "learning_rate": 0.003, + "loss": 4.0661, + "step": 7884 + }, + { + "epoch": 0.07885, + "grad_norm": 0.7548804773432991, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 7885 + }, + { + "epoch": 0.07886, + "grad_norm": 0.6885393382696238, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 7886 + }, + { + "epoch": 0.07887, + "grad_norm": 0.6782598568948106, + "learning_rate": 0.003, + "loss": 4.044, + "step": 7887 + }, + { + "epoch": 0.07888, + "grad_norm": 0.7104644623521696, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 7888 + }, + { + "epoch": 0.07889, + "grad_norm": 0.6832431134398846, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 7889 + }, + { + "epoch": 0.0789, + "grad_norm": 0.6853893830500566, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 7890 + }, + { + "epoch": 0.07891, + "grad_norm": 0.5741690604685803, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 7891 + }, + { + "epoch": 0.07892, + "grad_norm": 0.5621493774366465, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 7892 + }, + { + "epoch": 0.07893, + "grad_norm": 0.5769802254237503, + "learning_rate": 0.003, + "loss": 4.057, + "step": 7893 + }, + { + "epoch": 0.07894, + "grad_norm": 0.5368820040289479, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 7894 + }, + { + "epoch": 0.07895, + "grad_norm": 0.514163951059578, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 7895 + }, + { + "epoch": 0.07896, + "grad_norm": 0.47170751768199254, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 7896 + }, + { + "epoch": 0.07897, + "grad_norm": 0.4857893923604179, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 7897 + }, + { + "epoch": 0.07898, + "grad_norm": 0.5130071094704942, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 7898 + }, + { + "epoch": 0.07899, + "grad_norm": 0.5160174668294417, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 7899 + }, + { + "epoch": 0.079, + "grad_norm": 0.44883524363323285, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 7900 + }, + { + "epoch": 0.07901, + "grad_norm": 0.48138855367166256, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 7901 + }, + { + "epoch": 0.07902, + "grad_norm": 0.5017929296305591, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 7902 + }, + { + "epoch": 0.07903, + "grad_norm": 0.47729437326369023, + "learning_rate": 0.003, + "loss": 4.022, + "step": 7903 + }, + { + "epoch": 0.07904, + "grad_norm": 0.4284188268770118, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 7904 + }, + { + "epoch": 0.07905, + "grad_norm": 0.4789079589334836, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 7905 + }, + { + "epoch": 0.07906, + "grad_norm": 0.4835248588165741, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 7906 + }, + { + "epoch": 0.07907, + "grad_norm": 0.5906565332510613, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 7907 + }, + { + "epoch": 0.07908, + "grad_norm": 0.7752121636381054, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 7908 + }, + { + "epoch": 0.07909, + "grad_norm": 0.979690015927044, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 7909 + }, + { + "epoch": 0.0791, + "grad_norm": 0.9508880805050036, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 7910 + }, + { + "epoch": 0.07911, + "grad_norm": 0.774912451606984, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 7911 + }, + { + "epoch": 0.07912, + "grad_norm": 0.8168567927180777, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 7912 + }, + { + "epoch": 0.07913, + "grad_norm": 0.8068817517073694, + "learning_rate": 0.003, + "loss": 4.025, + "step": 7913 + }, + { + "epoch": 0.07914, + "grad_norm": 0.8398598119385825, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 7914 + }, + { + "epoch": 0.07915, + "grad_norm": 0.843102986236797, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 7915 + }, + { + "epoch": 0.07916, + "grad_norm": 0.8622413525460335, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 7916 + }, + { + "epoch": 0.07917, + "grad_norm": 0.7429858186892445, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 7917 + }, + { + "epoch": 0.07918, + "grad_norm": 0.7737783344887638, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 7918 + }, + { + "epoch": 0.07919, + "grad_norm": 0.8290639433398784, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 7919 + }, + { + "epoch": 0.0792, + "grad_norm": 0.9233782966914783, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 7920 + }, + { + "epoch": 0.07921, + "grad_norm": 0.9728247596383766, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 7921 + }, + { + "epoch": 0.07922, + "grad_norm": 1.0890704850876314, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 7922 + }, + { + "epoch": 0.07923, + "grad_norm": 0.9405935891386482, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 7923 + }, + { + "epoch": 0.07924, + "grad_norm": 0.7858600027998561, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 7924 + }, + { + "epoch": 0.07925, + "grad_norm": 0.7154189583420261, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 7925 + }, + { + "epoch": 0.07926, + "grad_norm": 0.6355134069333063, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 7926 + }, + { + "epoch": 0.07927, + "grad_norm": 0.6426308388957044, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 7927 + }, + { + "epoch": 0.07928, + "grad_norm": 0.6595212521535703, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 7928 + }, + { + "epoch": 0.07929, + "grad_norm": 0.711230184514999, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 7929 + }, + { + "epoch": 0.0793, + "grad_norm": 0.6987375473382824, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 7930 + }, + { + "epoch": 0.07931, + "grad_norm": 0.633099080315121, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 7931 + }, + { + "epoch": 0.07932, + "grad_norm": 0.6898172179333318, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 7932 + }, + { + "epoch": 0.07933, + "grad_norm": 0.7579960608058907, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 7933 + }, + { + "epoch": 0.07934, + "grad_norm": 0.7467242096491132, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 7934 + }, + { + "epoch": 0.07935, + "grad_norm": 0.6889758730550589, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 7935 + }, + { + "epoch": 0.07936, + "grad_norm": 0.6820129349166422, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 7936 + }, + { + "epoch": 0.07937, + "grad_norm": 0.7015821067289403, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 7937 + }, + { + "epoch": 0.07938, + "grad_norm": 0.7289613218476352, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 7938 + }, + { + "epoch": 0.07939, + "grad_norm": 0.6670273935505028, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 7939 + }, + { + "epoch": 0.0794, + "grad_norm": 0.6361229828169337, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 7940 + }, + { + "epoch": 0.07941, + "grad_norm": 0.5728976102414903, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 7941 + }, + { + "epoch": 0.07942, + "grad_norm": 0.5809844518563533, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 7942 + }, + { + "epoch": 0.07943, + "grad_norm": 0.6084017271017472, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 7943 + }, + { + "epoch": 0.07944, + "grad_norm": 0.5891038116762525, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 7944 + }, + { + "epoch": 0.07945, + "grad_norm": 0.6763826557721151, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 7945 + }, + { + "epoch": 0.07946, + "grad_norm": 0.8031625150856421, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 7946 + }, + { + "epoch": 0.07947, + "grad_norm": 0.8584434712049819, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 7947 + }, + { + "epoch": 0.07948, + "grad_norm": 0.8838742237781925, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 7948 + }, + { + "epoch": 0.07949, + "grad_norm": 0.8077296036229514, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 7949 + }, + { + "epoch": 0.0795, + "grad_norm": 0.6923870245876846, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 7950 + }, + { + "epoch": 0.07951, + "grad_norm": 0.6682538527422214, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 7951 + }, + { + "epoch": 0.07952, + "grad_norm": 0.6685795703322273, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 7952 + }, + { + "epoch": 0.07953, + "grad_norm": 0.7459955426970684, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 7953 + }, + { + "epoch": 0.07954, + "grad_norm": 0.8399053668593275, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 7954 + }, + { + "epoch": 0.07955, + "grad_norm": 0.7827728882872442, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 7955 + }, + { + "epoch": 0.07956, + "grad_norm": 0.6438020691454618, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 7956 + }, + { + "epoch": 0.07957, + "grad_norm": 0.6147719289879409, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 7957 + }, + { + "epoch": 0.07958, + "grad_norm": 0.6181852559288525, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 7958 + }, + { + "epoch": 0.07959, + "grad_norm": 0.5729582952318599, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 7959 + }, + { + "epoch": 0.0796, + "grad_norm": 0.5648532008615327, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 7960 + }, + { + "epoch": 0.07961, + "grad_norm": 0.536963959958475, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 7961 + }, + { + "epoch": 0.07962, + "grad_norm": 0.5415863771072796, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 7962 + }, + { + "epoch": 0.07963, + "grad_norm": 0.4890993572862303, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 7963 + }, + { + "epoch": 0.07964, + "grad_norm": 0.5222578268480008, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 7964 + }, + { + "epoch": 0.07965, + "grad_norm": 0.5144173065866917, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 7965 + }, + { + "epoch": 0.07966, + "grad_norm": 0.5648601359141344, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 7966 + }, + { + "epoch": 0.07967, + "grad_norm": 0.7803950395477834, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 7967 + }, + { + "epoch": 0.07968, + "grad_norm": 0.9614915093177762, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 7968 + }, + { + "epoch": 0.07969, + "grad_norm": 1.0062605993299323, + "learning_rate": 0.003, + "loss": 4.073, + "step": 7969 + }, + { + "epoch": 0.0797, + "grad_norm": 0.7316512319345297, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 7970 + }, + { + "epoch": 0.07971, + "grad_norm": 0.7280426095285986, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 7971 + }, + { + "epoch": 0.07972, + "grad_norm": 0.824931381381015, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 7972 + }, + { + "epoch": 0.07973, + "grad_norm": 0.8727428512688569, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 7973 + }, + { + "epoch": 0.07974, + "grad_norm": 0.8061823316515806, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 7974 + }, + { + "epoch": 0.07975, + "grad_norm": 0.8883017502623766, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 7975 + }, + { + "epoch": 0.07976, + "grad_norm": 0.9327131432667375, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 7976 + }, + { + "epoch": 0.07977, + "grad_norm": 0.887962164044367, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 7977 + }, + { + "epoch": 0.07978, + "grad_norm": 0.8471031234650425, + "learning_rate": 0.003, + "loss": 4.065, + "step": 7978 + }, + { + "epoch": 0.07979, + "grad_norm": 0.712339733400124, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 7979 + }, + { + "epoch": 0.0798, + "grad_norm": 0.7403780032835701, + "learning_rate": 0.003, + "loss": 4.0719, + "step": 7980 + }, + { + "epoch": 0.07981, + "grad_norm": 0.6353492977786231, + "learning_rate": 0.003, + "loss": 4.014, + "step": 7981 + }, + { + "epoch": 0.07982, + "grad_norm": 0.6068885125892998, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 7982 + }, + { + "epoch": 0.07983, + "grad_norm": 0.5305254135322562, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 7983 + }, + { + "epoch": 0.07984, + "grad_norm": 0.49379035675044525, + "learning_rate": 0.003, + "loss": 4.051, + "step": 7984 + }, + { + "epoch": 0.07985, + "grad_norm": 0.5270474800749442, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 7985 + }, + { + "epoch": 0.07986, + "grad_norm": 0.5177791119504694, + "learning_rate": 0.003, + "loss": 4.055, + "step": 7986 + }, + { + "epoch": 0.07987, + "grad_norm": 0.4409471209947055, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 7987 + }, + { + "epoch": 0.07988, + "grad_norm": 0.43126318075266545, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 7988 + }, + { + "epoch": 0.07989, + "grad_norm": 0.4846848072067431, + "learning_rate": 0.003, + "loss": 4.013, + "step": 7989 + }, + { + "epoch": 0.0799, + "grad_norm": 0.5205708982357089, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 7990 + }, + { + "epoch": 0.07991, + "grad_norm": 0.5184656031975156, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 7991 + }, + { + "epoch": 0.07992, + "grad_norm": 0.5656735284675292, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 7992 + }, + { + "epoch": 0.07993, + "grad_norm": 0.7342685652165115, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 7993 + }, + { + "epoch": 0.07994, + "grad_norm": 0.9167656002981955, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 7994 + }, + { + "epoch": 0.07995, + "grad_norm": 1.0581942423291801, + "learning_rate": 0.003, + "loss": 4.059, + "step": 7995 + }, + { + "epoch": 0.07996, + "grad_norm": 0.8773670652759645, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 7996 + }, + { + "epoch": 0.07997, + "grad_norm": 0.7465192546578548, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 7997 + }, + { + "epoch": 0.07998, + "grad_norm": 0.797696198052162, + "learning_rate": 0.003, + "loss": 4.074, + "step": 7998 + }, + { + "epoch": 0.07999, + "grad_norm": 0.7382286154402453, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 7999 + }, + { + "epoch": 0.08, + "grad_norm": 0.697128110442625, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 8000 + }, + { + "epoch": 0.08001, + "grad_norm": 0.737935876644138, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 8001 + }, + { + "epoch": 0.08002, + "grad_norm": 0.7781630554529115, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 8002 + }, + { + "epoch": 0.08003, + "grad_norm": 0.7950898014465319, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 8003 + }, + { + "epoch": 0.08004, + "grad_norm": 0.8006046480271932, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 8004 + }, + { + "epoch": 0.08005, + "grad_norm": 0.6657389684503167, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 8005 + }, + { + "epoch": 0.08006, + "grad_norm": 0.533824928370597, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 8006 + }, + { + "epoch": 0.08007, + "grad_norm": 0.5706315146210506, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 8007 + }, + { + "epoch": 0.08008, + "grad_norm": 0.5317487593925204, + "learning_rate": 0.003, + "loss": 4.013, + "step": 8008 + }, + { + "epoch": 0.08009, + "grad_norm": 0.48918867754601303, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 8009 + }, + { + "epoch": 0.0801, + "grad_norm": 0.41720207094912304, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 8010 + }, + { + "epoch": 0.08011, + "grad_norm": 0.43598989191303295, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 8011 + }, + { + "epoch": 0.08012, + "grad_norm": 0.43905123717343536, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 8012 + }, + { + "epoch": 0.08013, + "grad_norm": 0.5209805115565395, + "learning_rate": 0.003, + "loss": 4.033, + "step": 8013 + }, + { + "epoch": 0.08014, + "grad_norm": 0.6212861709740428, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 8014 + }, + { + "epoch": 0.08015, + "grad_norm": 0.8283332144763556, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 8015 + }, + { + "epoch": 0.08016, + "grad_norm": 0.9466822893079881, + "learning_rate": 0.003, + "loss": 4.016, + "step": 8016 + }, + { + "epoch": 0.08017, + "grad_norm": 0.8689760893959395, + "learning_rate": 0.003, + "loss": 4.024, + "step": 8017 + }, + { + "epoch": 0.08018, + "grad_norm": 0.7945197791659722, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 8018 + }, + { + "epoch": 0.08019, + "grad_norm": 0.8989963861228748, + "learning_rate": 0.003, + "loss": 4.0684, + "step": 8019 + }, + { + "epoch": 0.0802, + "grad_norm": 0.9455847696748121, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 8020 + }, + { + "epoch": 0.08021, + "grad_norm": 0.8288606671554433, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 8021 + }, + { + "epoch": 0.08022, + "grad_norm": 0.7425574886601446, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 8022 + }, + { + "epoch": 0.08023, + "grad_norm": 0.7249827789849455, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 8023 + }, + { + "epoch": 0.08024, + "grad_norm": 0.7783366652846129, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 8024 + }, + { + "epoch": 0.08025, + "grad_norm": 0.7483254753876812, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 8025 + }, + { + "epoch": 0.08026, + "grad_norm": 0.7748169048287166, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 8026 + }, + { + "epoch": 0.08027, + "grad_norm": 0.7860312159840537, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 8027 + }, + { + "epoch": 0.08028, + "grad_norm": 0.8065285365829903, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 8028 + }, + { + "epoch": 0.08029, + "grad_norm": 0.6827701843998871, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 8029 + }, + { + "epoch": 0.0803, + "grad_norm": 0.5764644124812746, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 8030 + }, + { + "epoch": 0.08031, + "grad_norm": 0.5535503964411983, + "learning_rate": 0.003, + "loss": 4.043, + "step": 8031 + }, + { + "epoch": 0.08032, + "grad_norm": 0.6086067425825472, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 8032 + }, + { + "epoch": 0.08033, + "grad_norm": 0.6025688268179953, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 8033 + }, + { + "epoch": 0.08034, + "grad_norm": 0.6377349113519364, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 8034 + }, + { + "epoch": 0.08035, + "grad_norm": 0.6486716939150651, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 8035 + }, + { + "epoch": 0.08036, + "grad_norm": 0.6897968932669338, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 8036 + }, + { + "epoch": 0.08037, + "grad_norm": 0.7611846896256969, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 8037 + }, + { + "epoch": 0.08038, + "grad_norm": 0.791184539232403, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 8038 + }, + { + "epoch": 0.08039, + "grad_norm": 0.8579481276292464, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 8039 + }, + { + "epoch": 0.0804, + "grad_norm": 0.8718065294937941, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 8040 + }, + { + "epoch": 0.08041, + "grad_norm": 0.8268177943187309, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 8041 + }, + { + "epoch": 0.08042, + "grad_norm": 0.8161502579393379, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 8042 + }, + { + "epoch": 0.08043, + "grad_norm": 0.8267495473040645, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 8043 + }, + { + "epoch": 0.08044, + "grad_norm": 0.8425336156432075, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 8044 + }, + { + "epoch": 0.08045, + "grad_norm": 0.9454332203845861, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 8045 + }, + { + "epoch": 0.08046, + "grad_norm": 0.900138412552877, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 8046 + }, + { + "epoch": 0.08047, + "grad_norm": 0.7291113353308969, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 8047 + }, + { + "epoch": 0.08048, + "grad_norm": 0.7039403150505659, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 8048 + }, + { + "epoch": 0.08049, + "grad_norm": 0.6839325148646369, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 8049 + }, + { + "epoch": 0.0805, + "grad_norm": 0.580631045713614, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 8050 + }, + { + "epoch": 0.08051, + "grad_norm": 0.5446758378464536, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 8051 + }, + { + "epoch": 0.08052, + "grad_norm": 0.6239651995637002, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 8052 + }, + { + "epoch": 0.08053, + "grad_norm": 0.6603084137325326, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 8053 + }, + { + "epoch": 0.08054, + "grad_norm": 0.5914577746406867, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 8054 + }, + { + "epoch": 0.08055, + "grad_norm": 0.5941054262581457, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 8055 + }, + { + "epoch": 0.08056, + "grad_norm": 0.6408871270318784, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 8056 + }, + { + "epoch": 0.08057, + "grad_norm": 0.818260039785732, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 8057 + }, + { + "epoch": 0.08058, + "grad_norm": 0.9641016231070247, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 8058 + }, + { + "epoch": 0.08059, + "grad_norm": 0.7716897164188243, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 8059 + }, + { + "epoch": 0.0806, + "grad_norm": 0.682136282542081, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 8060 + }, + { + "epoch": 0.08061, + "grad_norm": 0.7487311316978927, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 8061 + }, + { + "epoch": 0.08062, + "grad_norm": 0.6348746636301646, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 8062 + }, + { + "epoch": 0.08063, + "grad_norm": 0.712877085264771, + "learning_rate": 0.003, + "loss": 4.0765, + "step": 8063 + }, + { + "epoch": 0.08064, + "grad_norm": 0.7376698518694439, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8064 + }, + { + "epoch": 0.08065, + "grad_norm": 0.6937035236511004, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 8065 + }, + { + "epoch": 0.08066, + "grad_norm": 0.7219669319784563, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 8066 + }, + { + "epoch": 0.08067, + "grad_norm": 0.7029771722748186, + "learning_rate": 0.003, + "loss": 4.054, + "step": 8067 + }, + { + "epoch": 0.08068, + "grad_norm": 0.5890844511251652, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 8068 + }, + { + "epoch": 0.08069, + "grad_norm": 0.5304650655444964, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 8069 + }, + { + "epoch": 0.0807, + "grad_norm": 0.4987782548311441, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 8070 + }, + { + "epoch": 0.08071, + "grad_norm": 0.559353580119064, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 8071 + }, + { + "epoch": 0.08072, + "grad_norm": 0.6949601222135615, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 8072 + }, + { + "epoch": 0.08073, + "grad_norm": 0.8882128875705677, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 8073 + }, + { + "epoch": 0.08074, + "grad_norm": 0.9349234812463699, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 8074 + }, + { + "epoch": 0.08075, + "grad_norm": 0.8637953939456313, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 8075 + }, + { + "epoch": 0.08076, + "grad_norm": 0.8487440970232598, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 8076 + }, + { + "epoch": 0.08077, + "grad_norm": 0.8008522772243695, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 8077 + }, + { + "epoch": 0.08078, + "grad_norm": 0.894425435359524, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 8078 + }, + { + "epoch": 0.08079, + "grad_norm": 0.8466577571446373, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 8079 + }, + { + "epoch": 0.0808, + "grad_norm": 0.682900383946627, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 8080 + }, + { + "epoch": 0.08081, + "grad_norm": 0.6080401031933714, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 8081 + }, + { + "epoch": 0.08082, + "grad_norm": 0.6034144010670672, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 8082 + }, + { + "epoch": 0.08083, + "grad_norm": 0.5418871419001842, + "learning_rate": 0.003, + "loss": 4.031, + "step": 8083 + }, + { + "epoch": 0.08084, + "grad_norm": 0.6018693457888313, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 8084 + }, + { + "epoch": 0.08085, + "grad_norm": 0.698502170159121, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 8085 + }, + { + "epoch": 0.08086, + "grad_norm": 0.7519664058672254, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 8086 + }, + { + "epoch": 0.08087, + "grad_norm": 0.7536480517284665, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 8087 + }, + { + "epoch": 0.08088, + "grad_norm": 0.6804066889099789, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 8088 + }, + { + "epoch": 0.08089, + "grad_norm": 0.6310344267387439, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 8089 + }, + { + "epoch": 0.0809, + "grad_norm": 0.6074683075102638, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 8090 + }, + { + "epoch": 0.08091, + "grad_norm": 0.5861433322767835, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 8091 + }, + { + "epoch": 0.08092, + "grad_norm": 0.6538537188137997, + "learning_rate": 0.003, + "loss": 4.058, + "step": 8092 + }, + { + "epoch": 0.08093, + "grad_norm": 0.7079337925909709, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 8093 + }, + { + "epoch": 0.08094, + "grad_norm": 0.7140197264812004, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 8094 + }, + { + "epoch": 0.08095, + "grad_norm": 0.6296062818943632, + "learning_rate": 0.003, + "loss": 4.017, + "step": 8095 + }, + { + "epoch": 0.08096, + "grad_norm": 0.5515026772891211, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 8096 + }, + { + "epoch": 0.08097, + "grad_norm": 0.5257599227388431, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 8097 + }, + { + "epoch": 0.08098, + "grad_norm": 0.6041487493629959, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 8098 + }, + { + "epoch": 0.08099, + "grad_norm": 0.6257606321716116, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 8099 + }, + { + "epoch": 0.081, + "grad_norm": 0.7176449814607896, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 8100 + }, + { + "epoch": 0.08101, + "grad_norm": 0.7769778905487897, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 8101 + }, + { + "epoch": 0.08102, + "grad_norm": 0.7499870229397124, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 8102 + }, + { + "epoch": 0.08103, + "grad_norm": 0.7300324502106752, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8103 + }, + { + "epoch": 0.08104, + "grad_norm": 0.7769599370734824, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 8104 + }, + { + "epoch": 0.08105, + "grad_norm": 0.8048510315729345, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 8105 + }, + { + "epoch": 0.08106, + "grad_norm": 0.842191360314911, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 8106 + }, + { + "epoch": 0.08107, + "grad_norm": 0.865049445765839, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 8107 + }, + { + "epoch": 0.08108, + "grad_norm": 1.0160696005499938, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 8108 + }, + { + "epoch": 0.08109, + "grad_norm": 1.0971121824715153, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 8109 + }, + { + "epoch": 0.0811, + "grad_norm": 0.8423879386571966, + "learning_rate": 0.003, + "loss": 4.0681, + "step": 8110 + }, + { + "epoch": 0.08111, + "grad_norm": 0.7171287831595567, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 8111 + }, + { + "epoch": 0.08112, + "grad_norm": 0.7981058907626594, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 8112 + }, + { + "epoch": 0.08113, + "grad_norm": 0.7974210928578152, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 8113 + }, + { + "epoch": 0.08114, + "grad_norm": 0.891489297833214, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 8114 + }, + { + "epoch": 0.08115, + "grad_norm": 0.9024450131368834, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 8115 + }, + { + "epoch": 0.08116, + "grad_norm": 0.8660936858370433, + "learning_rate": 0.003, + "loss": 4.0768, + "step": 8116 + }, + { + "epoch": 0.08117, + "grad_norm": 0.7711633593452822, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 8117 + }, + { + "epoch": 0.08118, + "grad_norm": 0.7293289306150071, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 8118 + }, + { + "epoch": 0.08119, + "grad_norm": 0.6772791477736853, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 8119 + }, + { + "epoch": 0.0812, + "grad_norm": 0.5944113661608328, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 8120 + }, + { + "epoch": 0.08121, + "grad_norm": 0.5277022095149619, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 8121 + }, + { + "epoch": 0.08122, + "grad_norm": 0.4669561804231516, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 8122 + }, + { + "epoch": 0.08123, + "grad_norm": 0.4940261947658655, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8123 + }, + { + "epoch": 0.08124, + "grad_norm": 0.47826483865624025, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 8124 + }, + { + "epoch": 0.08125, + "grad_norm": 0.4910410062884967, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 8125 + }, + { + "epoch": 0.08126, + "grad_norm": 0.5077745388760384, + "learning_rate": 0.003, + "loss": 4.023, + "step": 8126 + }, + { + "epoch": 0.08127, + "grad_norm": 0.5133210378912718, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 8127 + }, + { + "epoch": 0.08128, + "grad_norm": 0.5030462538158976, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 8128 + }, + { + "epoch": 0.08129, + "grad_norm": 0.4351059858970992, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 8129 + }, + { + "epoch": 0.0813, + "grad_norm": 0.45730893306665427, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 8130 + }, + { + "epoch": 0.08131, + "grad_norm": 0.4597470968905339, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 8131 + }, + { + "epoch": 0.08132, + "grad_norm": 0.49808374940218936, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 8132 + }, + { + "epoch": 0.08133, + "grad_norm": 0.5606012398896881, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 8133 + }, + { + "epoch": 0.08134, + "grad_norm": 0.594670993844876, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 8134 + }, + { + "epoch": 0.08135, + "grad_norm": 0.7279063531989127, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 8135 + }, + { + "epoch": 0.08136, + "grad_norm": 0.895502895825732, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 8136 + }, + { + "epoch": 0.08137, + "grad_norm": 0.8526756720650094, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 8137 + }, + { + "epoch": 0.08138, + "grad_norm": 0.799876944167941, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 8138 + }, + { + "epoch": 0.08139, + "grad_norm": 0.8643587526727834, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 8139 + }, + { + "epoch": 0.0814, + "grad_norm": 0.7978744833811383, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 8140 + }, + { + "epoch": 0.08141, + "grad_norm": 0.7992409810817563, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 8141 + }, + { + "epoch": 0.08142, + "grad_norm": 0.8093971046838764, + "learning_rate": 0.003, + "loss": 4.063, + "step": 8142 + }, + { + "epoch": 0.08143, + "grad_norm": 0.9117481680303504, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 8143 + }, + { + "epoch": 0.08144, + "grad_norm": 1.1019154051561295, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 8144 + }, + { + "epoch": 0.08145, + "grad_norm": 0.8835876816489401, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 8145 + }, + { + "epoch": 0.08146, + "grad_norm": 0.9021426486741848, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 8146 + }, + { + "epoch": 0.08147, + "grad_norm": 0.821814137538181, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 8147 + }, + { + "epoch": 0.08148, + "grad_norm": 0.6829721647769539, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8148 + }, + { + "epoch": 0.08149, + "grad_norm": 0.5366614133360115, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 8149 + }, + { + "epoch": 0.0815, + "grad_norm": 0.6089539813973405, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 8150 + }, + { + "epoch": 0.08151, + "grad_norm": 0.7099166082161043, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 8151 + }, + { + "epoch": 0.08152, + "grad_norm": 0.892063823066484, + "learning_rate": 0.003, + "loss": 4.064, + "step": 8152 + }, + { + "epoch": 0.08153, + "grad_norm": 0.9299793348929113, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 8153 + }, + { + "epoch": 0.08154, + "grad_norm": 0.8610362840174799, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 8154 + }, + { + "epoch": 0.08155, + "grad_norm": 0.9085666319316019, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 8155 + }, + { + "epoch": 0.08156, + "grad_norm": 0.8808935683607264, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 8156 + }, + { + "epoch": 0.08157, + "grad_norm": 0.7182061224597421, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 8157 + }, + { + "epoch": 0.08158, + "grad_norm": 0.7560270793763518, + "learning_rate": 0.003, + "loss": 4.0821, + "step": 8158 + }, + { + "epoch": 0.08159, + "grad_norm": 0.7717228606841453, + "learning_rate": 0.003, + "loss": 4.0814, + "step": 8159 + }, + { + "epoch": 0.0816, + "grad_norm": 0.6938998555165131, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 8160 + }, + { + "epoch": 0.08161, + "grad_norm": 0.6210495778514565, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 8161 + }, + { + "epoch": 0.08162, + "grad_norm": 0.5916287301006917, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 8162 + }, + { + "epoch": 0.08163, + "grad_norm": 0.5146797783861714, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 8163 + }, + { + "epoch": 0.08164, + "grad_norm": 0.542586880062633, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 8164 + }, + { + "epoch": 0.08165, + "grad_norm": 0.5826146514504134, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 8165 + }, + { + "epoch": 0.08166, + "grad_norm": 0.6404662692692923, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 8166 + }, + { + "epoch": 0.08167, + "grad_norm": 0.6482944809119093, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 8167 + }, + { + "epoch": 0.08168, + "grad_norm": 0.6941499140075943, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 8168 + }, + { + "epoch": 0.08169, + "grad_norm": 0.7668421841623635, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 8169 + }, + { + "epoch": 0.0817, + "grad_norm": 0.7815089602657143, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 8170 + }, + { + "epoch": 0.08171, + "grad_norm": 0.7585314699362724, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 8171 + }, + { + "epoch": 0.08172, + "grad_norm": 0.640911038634051, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 8172 + }, + { + "epoch": 0.08173, + "grad_norm": 0.5793694348508043, + "learning_rate": 0.003, + "loss": 4.0743, + "step": 8173 + }, + { + "epoch": 0.08174, + "grad_norm": 0.5791284179175312, + "learning_rate": 0.003, + "loss": 4.068, + "step": 8174 + }, + { + "epoch": 0.08175, + "grad_norm": 0.623162613132913, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 8175 + }, + { + "epoch": 0.08176, + "grad_norm": 0.6333169463942829, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 8176 + }, + { + "epoch": 0.08177, + "grad_norm": 0.6571505504635461, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 8177 + }, + { + "epoch": 0.08178, + "grad_norm": 0.676191028411339, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 8178 + }, + { + "epoch": 0.08179, + "grad_norm": 0.6835275131767474, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 8179 + }, + { + "epoch": 0.0818, + "grad_norm": 0.6390531710835708, + "learning_rate": 0.003, + "loss": 4.033, + "step": 8180 + }, + { + "epoch": 0.08181, + "grad_norm": 0.5820045369253264, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 8181 + }, + { + "epoch": 0.08182, + "grad_norm": 0.5754124665523351, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 8182 + }, + { + "epoch": 0.08183, + "grad_norm": 0.5095803353921973, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 8183 + }, + { + "epoch": 0.08184, + "grad_norm": 0.5230305848368935, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 8184 + }, + { + "epoch": 0.08185, + "grad_norm": 0.5528643636697647, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 8185 + }, + { + "epoch": 0.08186, + "grad_norm": 0.5558588282103597, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 8186 + }, + { + "epoch": 0.08187, + "grad_norm": 0.5915623420522257, + "learning_rate": 0.003, + "loss": 4.004, + "step": 8187 + }, + { + "epoch": 0.08188, + "grad_norm": 0.601374768910701, + "learning_rate": 0.003, + "loss": 4.053, + "step": 8188 + }, + { + "epoch": 0.08189, + "grad_norm": 0.5802677730096782, + "learning_rate": 0.003, + "loss": 4.049, + "step": 8189 + }, + { + "epoch": 0.0819, + "grad_norm": 0.627802788019237, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 8190 + }, + { + "epoch": 0.08191, + "grad_norm": 0.8312526648588364, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 8191 + }, + { + "epoch": 0.08192, + "grad_norm": 0.8586724868707798, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 8192 + }, + { + "epoch": 0.08193, + "grad_norm": 0.8325072675871149, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 8193 + }, + { + "epoch": 0.08194, + "grad_norm": 0.7032081651736415, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 8194 + }, + { + "epoch": 0.08195, + "grad_norm": 0.7370917893277578, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8195 + }, + { + "epoch": 0.08196, + "grad_norm": 0.7791876008294636, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 8196 + }, + { + "epoch": 0.08197, + "grad_norm": 0.7920465284198261, + "learning_rate": 0.003, + "loss": 4.045, + "step": 8197 + }, + { + "epoch": 0.08198, + "grad_norm": 0.680491648517299, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 8198 + }, + { + "epoch": 0.08199, + "grad_norm": 0.6358987392931551, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8199 + }, + { + "epoch": 0.082, + "grad_norm": 0.6806937981678177, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 8200 + }, + { + "epoch": 0.08201, + "grad_norm": 0.7688768775940364, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 8201 + }, + { + "epoch": 0.08202, + "grad_norm": 0.7964750657929169, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 8202 + }, + { + "epoch": 0.08203, + "grad_norm": 0.7957703782432136, + "learning_rate": 0.003, + "loss": 4.051, + "step": 8203 + }, + { + "epoch": 0.08204, + "grad_norm": 0.8088407306522456, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 8204 + }, + { + "epoch": 0.08205, + "grad_norm": 0.7672470105656158, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 8205 + }, + { + "epoch": 0.08206, + "grad_norm": 0.774992196334392, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 8206 + }, + { + "epoch": 0.08207, + "grad_norm": 0.6677245474135516, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 8207 + }, + { + "epoch": 0.08208, + "grad_norm": 0.5919302672755191, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 8208 + }, + { + "epoch": 0.08209, + "grad_norm": 0.6191489368264088, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 8209 + }, + { + "epoch": 0.0821, + "grad_norm": 0.5545691743479098, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8210 + }, + { + "epoch": 0.08211, + "grad_norm": 0.5599543160507289, + "learning_rate": 0.003, + "loss": 4.035, + "step": 8211 + }, + { + "epoch": 0.08212, + "grad_norm": 0.5724710275035697, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 8212 + }, + { + "epoch": 0.08213, + "grad_norm": 0.6619550670961922, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 8213 + }, + { + "epoch": 0.08214, + "grad_norm": 0.7119310163613618, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 8214 + }, + { + "epoch": 0.08215, + "grad_norm": 0.815828915722191, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 8215 + }, + { + "epoch": 0.08216, + "grad_norm": 0.7759822050992788, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 8216 + }, + { + "epoch": 0.08217, + "grad_norm": 0.6526161841550577, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 8217 + }, + { + "epoch": 0.08218, + "grad_norm": 0.628249319795086, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 8218 + }, + { + "epoch": 0.08219, + "grad_norm": 0.7067586264202683, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 8219 + }, + { + "epoch": 0.0822, + "grad_norm": 0.770726702217598, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 8220 + }, + { + "epoch": 0.08221, + "grad_norm": 0.8424774253927828, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 8221 + }, + { + "epoch": 0.08222, + "grad_norm": 0.8432725078619205, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 8222 + }, + { + "epoch": 0.08223, + "grad_norm": 0.8980478649233362, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 8223 + }, + { + "epoch": 0.08224, + "grad_norm": 0.9712775146158149, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 8224 + }, + { + "epoch": 0.08225, + "grad_norm": 0.8314083252989968, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 8225 + }, + { + "epoch": 0.08226, + "grad_norm": 0.7190735494024668, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 8226 + }, + { + "epoch": 0.08227, + "grad_norm": 0.8287007499711808, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 8227 + }, + { + "epoch": 0.08228, + "grad_norm": 0.8840159537726527, + "learning_rate": 0.003, + "loss": 4.07, + "step": 8228 + }, + { + "epoch": 0.08229, + "grad_norm": 0.9096033736155922, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 8229 + }, + { + "epoch": 0.0823, + "grad_norm": 0.8897085248009451, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 8230 + }, + { + "epoch": 0.08231, + "grad_norm": 1.0413792114788305, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 8231 + }, + { + "epoch": 0.08232, + "grad_norm": 0.9663770932428598, + "learning_rate": 0.003, + "loss": 4.048, + "step": 8232 + }, + { + "epoch": 0.08233, + "grad_norm": 0.831761991019891, + "learning_rate": 0.003, + "loss": 4.057, + "step": 8233 + }, + { + "epoch": 0.08234, + "grad_norm": 0.7667114583935627, + "learning_rate": 0.003, + "loss": 4.0669, + "step": 8234 + }, + { + "epoch": 0.08235, + "grad_norm": 0.8790529759954461, + "learning_rate": 0.003, + "loss": 4.0895, + "step": 8235 + }, + { + "epoch": 0.08236, + "grad_norm": 0.9169725460866149, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 8236 + }, + { + "epoch": 0.08237, + "grad_norm": 0.9591052971513858, + "learning_rate": 0.003, + "loss": 4.064, + "step": 8237 + }, + { + "epoch": 0.08238, + "grad_norm": 0.8895371548138776, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 8238 + }, + { + "epoch": 0.08239, + "grad_norm": 0.7286600016519696, + "learning_rate": 0.003, + "loss": 4.037, + "step": 8239 + }, + { + "epoch": 0.0824, + "grad_norm": 0.6947507903072705, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 8240 + }, + { + "epoch": 0.08241, + "grad_norm": 0.6476377279015784, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 8241 + }, + { + "epoch": 0.08242, + "grad_norm": 0.6028037781752859, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 8242 + }, + { + "epoch": 0.08243, + "grad_norm": 0.6057057641292943, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 8243 + }, + { + "epoch": 0.08244, + "grad_norm": 0.59502301914224, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 8244 + }, + { + "epoch": 0.08245, + "grad_norm": 0.6189654023514646, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 8245 + }, + { + "epoch": 0.08246, + "grad_norm": 0.5821151749028172, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 8246 + }, + { + "epoch": 0.08247, + "grad_norm": 0.6206980464860076, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 8247 + }, + { + "epoch": 0.08248, + "grad_norm": 0.6110049413533856, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 8248 + }, + { + "epoch": 0.08249, + "grad_norm": 0.4724710779473591, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 8249 + }, + { + "epoch": 0.0825, + "grad_norm": 0.5131803651809121, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 8250 + }, + { + "epoch": 0.08251, + "grad_norm": 0.5579645638315378, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8251 + }, + { + "epoch": 0.08252, + "grad_norm": 0.5470650095184045, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 8252 + }, + { + "epoch": 0.08253, + "grad_norm": 0.4570862595132034, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8253 + }, + { + "epoch": 0.08254, + "grad_norm": 0.4081390656751753, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 8254 + }, + { + "epoch": 0.08255, + "grad_norm": 0.3586502266617014, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 8255 + }, + { + "epoch": 0.08256, + "grad_norm": 0.4234060063586608, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 8256 + }, + { + "epoch": 0.08257, + "grad_norm": 0.4895866557204517, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 8257 + }, + { + "epoch": 0.08258, + "grad_norm": 0.5783304891413638, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 8258 + }, + { + "epoch": 0.08259, + "grad_norm": 0.7084846273332669, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 8259 + }, + { + "epoch": 0.0826, + "grad_norm": 0.7631679558576686, + "learning_rate": 0.003, + "loss": 4.012, + "step": 8260 + }, + { + "epoch": 0.08261, + "grad_norm": 0.7797802752010387, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 8261 + }, + { + "epoch": 0.08262, + "grad_norm": 0.7432843749458414, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 8262 + }, + { + "epoch": 0.08263, + "grad_norm": 0.6557695949520227, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 8263 + }, + { + "epoch": 0.08264, + "grad_norm": 0.6162667444097379, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 8264 + }, + { + "epoch": 0.08265, + "grad_norm": 0.6508542621873344, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 8265 + }, + { + "epoch": 0.08266, + "grad_norm": 0.7797934900587395, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 8266 + }, + { + "epoch": 0.08267, + "grad_norm": 0.8564496161157986, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 8267 + }, + { + "epoch": 0.08268, + "grad_norm": 0.8665503519862852, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 8268 + }, + { + "epoch": 0.08269, + "grad_norm": 0.72240911972141, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8269 + }, + { + "epoch": 0.0827, + "grad_norm": 0.731782272673544, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 8270 + }, + { + "epoch": 0.08271, + "grad_norm": 0.8862316886920656, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 8271 + }, + { + "epoch": 0.08272, + "grad_norm": 0.8917677653096486, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 8272 + }, + { + "epoch": 0.08273, + "grad_norm": 0.8889168434329313, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8273 + }, + { + "epoch": 0.08274, + "grad_norm": 0.8993500515712378, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 8274 + }, + { + "epoch": 0.08275, + "grad_norm": 0.9482586752275424, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 8275 + }, + { + "epoch": 0.08276, + "grad_norm": 0.9015070447280308, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 8276 + }, + { + "epoch": 0.08277, + "grad_norm": 0.8240640792591647, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 8277 + }, + { + "epoch": 0.08278, + "grad_norm": 0.8802556905906979, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 8278 + }, + { + "epoch": 0.08279, + "grad_norm": 0.807093735277175, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 8279 + }, + { + "epoch": 0.0828, + "grad_norm": 0.8377107504962047, + "learning_rate": 0.003, + "loss": 4.077, + "step": 8280 + }, + { + "epoch": 0.08281, + "grad_norm": 0.8868360279408677, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 8281 + }, + { + "epoch": 0.08282, + "grad_norm": 0.8638275086005968, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 8282 + }, + { + "epoch": 0.08283, + "grad_norm": 0.9451383548048465, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 8283 + }, + { + "epoch": 0.08284, + "grad_norm": 1.03579192082318, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 8284 + }, + { + "epoch": 0.08285, + "grad_norm": 0.8969826830288632, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 8285 + }, + { + "epoch": 0.08286, + "grad_norm": 0.9046331663390617, + "learning_rate": 0.003, + "loss": 4.0847, + "step": 8286 + }, + { + "epoch": 0.08287, + "grad_norm": 0.8268569706713185, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 8287 + }, + { + "epoch": 0.08288, + "grad_norm": 0.7167867700207334, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 8288 + }, + { + "epoch": 0.08289, + "grad_norm": 0.8442998194141604, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 8289 + }, + { + "epoch": 0.0829, + "grad_norm": 0.8083897817525761, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 8290 + }, + { + "epoch": 0.08291, + "grad_norm": 0.6979837767325835, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 8291 + }, + { + "epoch": 0.08292, + "grad_norm": 0.7199973224089803, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 8292 + }, + { + "epoch": 0.08293, + "grad_norm": 0.7288404157997311, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 8293 + }, + { + "epoch": 0.08294, + "grad_norm": 0.7274511478683491, + "learning_rate": 0.003, + "loss": 4.045, + "step": 8294 + }, + { + "epoch": 0.08295, + "grad_norm": 0.6936029024618312, + "learning_rate": 0.003, + "loss": 4.047, + "step": 8295 + }, + { + "epoch": 0.08296, + "grad_norm": 0.525069975025596, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 8296 + }, + { + "epoch": 0.08297, + "grad_norm": 0.5768090204882219, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 8297 + }, + { + "epoch": 0.08298, + "grad_norm": 0.5542443315905439, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 8298 + }, + { + "epoch": 0.08299, + "grad_norm": 0.5165653771469979, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 8299 + }, + { + "epoch": 0.083, + "grad_norm": 0.4829129818952746, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 8300 + }, + { + "epoch": 0.08301, + "grad_norm": 0.5134644006399709, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 8301 + }, + { + "epoch": 0.08302, + "grad_norm": 0.5560902193605586, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 8302 + }, + { + "epoch": 0.08303, + "grad_norm": 0.46230049719080635, + "learning_rate": 0.003, + "loss": 4.026, + "step": 8303 + }, + { + "epoch": 0.08304, + "grad_norm": 0.43905584782061735, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 8304 + }, + { + "epoch": 0.08305, + "grad_norm": 0.486904699733299, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 8305 + }, + { + "epoch": 0.08306, + "grad_norm": 0.5100602430800917, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 8306 + }, + { + "epoch": 0.08307, + "grad_norm": 0.5535578951835018, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 8307 + }, + { + "epoch": 0.08308, + "grad_norm": 0.721110103976674, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 8308 + }, + { + "epoch": 0.08309, + "grad_norm": 0.8548567419126608, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 8309 + }, + { + "epoch": 0.0831, + "grad_norm": 0.8140971961009839, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 8310 + }, + { + "epoch": 0.08311, + "grad_norm": 0.6384891756946579, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 8311 + }, + { + "epoch": 0.08312, + "grad_norm": 0.5342583785059578, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 8312 + }, + { + "epoch": 0.08313, + "grad_norm": 0.6198058929600962, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 8313 + }, + { + "epoch": 0.08314, + "grad_norm": 0.5737712097268263, + "learning_rate": 0.003, + "loss": 4.047, + "step": 8314 + }, + { + "epoch": 0.08315, + "grad_norm": 0.536555755119035, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 8315 + }, + { + "epoch": 0.08316, + "grad_norm": 0.5656181881943217, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 8316 + }, + { + "epoch": 0.08317, + "grad_norm": 0.7091625142358896, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 8317 + }, + { + "epoch": 0.08318, + "grad_norm": 0.8278073640051181, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 8318 + }, + { + "epoch": 0.08319, + "grad_norm": 0.6780132932554127, + "learning_rate": 0.003, + "loss": 4.015, + "step": 8319 + }, + { + "epoch": 0.0832, + "grad_norm": 0.6726687674501164, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 8320 + }, + { + "epoch": 0.08321, + "grad_norm": 0.7107311712803911, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 8321 + }, + { + "epoch": 0.08322, + "grad_norm": 0.7398985811590479, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 8322 + }, + { + "epoch": 0.08323, + "grad_norm": 0.6458510578187311, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 8323 + }, + { + "epoch": 0.08324, + "grad_norm": 0.5767394171335257, + "learning_rate": 0.003, + "loss": 4.034, + "step": 8324 + }, + { + "epoch": 0.08325, + "grad_norm": 0.6443531906947946, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 8325 + }, + { + "epoch": 0.08326, + "grad_norm": 0.7086167015129559, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 8326 + }, + { + "epoch": 0.08327, + "grad_norm": 0.6071152433176673, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 8327 + }, + { + "epoch": 0.08328, + "grad_norm": 0.5942849667001413, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 8328 + }, + { + "epoch": 0.08329, + "grad_norm": 0.6813488710540198, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 8329 + }, + { + "epoch": 0.0833, + "grad_norm": 0.6628878295989707, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 8330 + }, + { + "epoch": 0.08331, + "grad_norm": 0.5567221577673772, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 8331 + }, + { + "epoch": 0.08332, + "grad_norm": 0.6174965096223124, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8332 + }, + { + "epoch": 0.08333, + "grad_norm": 0.6565634177351464, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 8333 + }, + { + "epoch": 0.08334, + "grad_norm": 0.6767045628586238, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 8334 + }, + { + "epoch": 0.08335, + "grad_norm": 0.6166315416927646, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 8335 + }, + { + "epoch": 0.08336, + "grad_norm": 0.6399052024001751, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 8336 + }, + { + "epoch": 0.08337, + "grad_norm": 0.6862186240061386, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 8337 + }, + { + "epoch": 0.08338, + "grad_norm": 0.7906106321225831, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 8338 + }, + { + "epoch": 0.08339, + "grad_norm": 0.812816997533076, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 8339 + }, + { + "epoch": 0.0834, + "grad_norm": 0.8746522656532724, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 8340 + }, + { + "epoch": 0.08341, + "grad_norm": 1.1018514135288406, + "learning_rate": 0.003, + "loss": 4.0758, + "step": 8341 + }, + { + "epoch": 0.08342, + "grad_norm": 0.9674425463869587, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 8342 + }, + { + "epoch": 0.08343, + "grad_norm": 0.7565222457684151, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 8343 + }, + { + "epoch": 0.08344, + "grad_norm": 0.5632153885913382, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 8344 + }, + { + "epoch": 0.08345, + "grad_norm": 0.6374596755804804, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 8345 + }, + { + "epoch": 0.08346, + "grad_norm": 0.7118147025443838, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 8346 + }, + { + "epoch": 0.08347, + "grad_norm": 0.7693789038111093, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 8347 + }, + { + "epoch": 0.08348, + "grad_norm": 0.6753132921081036, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 8348 + }, + { + "epoch": 0.08349, + "grad_norm": 0.612145680116179, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 8349 + }, + { + "epoch": 0.0835, + "grad_norm": 0.6020961308665921, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 8350 + }, + { + "epoch": 0.08351, + "grad_norm": 0.6434069332666404, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 8351 + }, + { + "epoch": 0.08352, + "grad_norm": 0.5658623626229315, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 8352 + }, + { + "epoch": 0.08353, + "grad_norm": 0.5635076046944061, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 8353 + }, + { + "epoch": 0.08354, + "grad_norm": 0.6139246182489428, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8354 + }, + { + "epoch": 0.08355, + "grad_norm": 0.5960513839046943, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 8355 + }, + { + "epoch": 0.08356, + "grad_norm": 0.5581682453939032, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 8356 + }, + { + "epoch": 0.08357, + "grad_norm": 0.5489180706284262, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 8357 + }, + { + "epoch": 0.08358, + "grad_norm": 0.6170712351080099, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 8358 + }, + { + "epoch": 0.08359, + "grad_norm": 0.7066815134265435, + "learning_rate": 0.003, + "loss": 4.01, + "step": 8359 + }, + { + "epoch": 0.0836, + "grad_norm": 0.6752112413362072, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 8360 + }, + { + "epoch": 0.08361, + "grad_norm": 0.6650455488353557, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 8361 + }, + { + "epoch": 0.08362, + "grad_norm": 0.7408178543854733, + "learning_rate": 0.003, + "loss": 4.054, + "step": 8362 + }, + { + "epoch": 0.08363, + "grad_norm": 0.6563696625568977, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 8363 + }, + { + "epoch": 0.08364, + "grad_norm": 0.6486157980539546, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 8364 + }, + { + "epoch": 0.08365, + "grad_norm": 0.6718042263914419, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 8365 + }, + { + "epoch": 0.08366, + "grad_norm": 0.7380526549609618, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 8366 + }, + { + "epoch": 0.08367, + "grad_norm": 0.8471842645126781, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 8367 + }, + { + "epoch": 0.08368, + "grad_norm": 0.8453931496276672, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 8368 + }, + { + "epoch": 0.08369, + "grad_norm": 0.8943922489586409, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 8369 + }, + { + "epoch": 0.0837, + "grad_norm": 0.8768709638516257, + "learning_rate": 0.003, + "loss": 4.048, + "step": 8370 + }, + { + "epoch": 0.08371, + "grad_norm": 0.6960086053197525, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 8371 + }, + { + "epoch": 0.08372, + "grad_norm": 0.7028303102019064, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 8372 + }, + { + "epoch": 0.08373, + "grad_norm": 0.7546881083511715, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 8373 + }, + { + "epoch": 0.08374, + "grad_norm": 0.7628906266865179, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 8374 + }, + { + "epoch": 0.08375, + "grad_norm": 0.7340942744921471, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 8375 + }, + { + "epoch": 0.08376, + "grad_norm": 0.7027792561600928, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 8376 + }, + { + "epoch": 0.08377, + "grad_norm": 0.9071016288143495, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 8377 + }, + { + "epoch": 0.08378, + "grad_norm": 1.0591163469343339, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 8378 + }, + { + "epoch": 0.08379, + "grad_norm": 1.115640942129094, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 8379 + }, + { + "epoch": 0.0838, + "grad_norm": 0.8872843207682162, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 8380 + }, + { + "epoch": 0.08381, + "grad_norm": 0.7357641001998408, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 8381 + }, + { + "epoch": 0.08382, + "grad_norm": 0.6833403105966946, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 8382 + }, + { + "epoch": 0.08383, + "grad_norm": 0.6765296886335385, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 8383 + }, + { + "epoch": 0.08384, + "grad_norm": 0.6619111793814234, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 8384 + }, + { + "epoch": 0.08385, + "grad_norm": 0.6754533976806071, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 8385 + }, + { + "epoch": 0.08386, + "grad_norm": 0.6858652688310308, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 8386 + }, + { + "epoch": 0.08387, + "grad_norm": 0.5858846483164577, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 8387 + }, + { + "epoch": 0.08388, + "grad_norm": 0.6416867677167735, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 8388 + }, + { + "epoch": 0.08389, + "grad_norm": 0.6051084268075266, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 8389 + }, + { + "epoch": 0.0839, + "grad_norm": 0.6030361048063958, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 8390 + }, + { + "epoch": 0.08391, + "grad_norm": 0.6875516466372213, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 8391 + }, + { + "epoch": 0.08392, + "grad_norm": 0.8192907945793478, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 8392 + }, + { + "epoch": 0.08393, + "grad_norm": 0.8174556353043552, + "learning_rate": 0.003, + "loss": 4.058, + "step": 8393 + }, + { + "epoch": 0.08394, + "grad_norm": 0.7376143427741151, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 8394 + }, + { + "epoch": 0.08395, + "grad_norm": 0.6967954807419333, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 8395 + }, + { + "epoch": 0.08396, + "grad_norm": 0.6307590092084665, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 8396 + }, + { + "epoch": 0.08397, + "grad_norm": 0.6951719895317683, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 8397 + }, + { + "epoch": 0.08398, + "grad_norm": 0.6790235803124289, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 8398 + }, + { + "epoch": 0.08399, + "grad_norm": 0.6077741673660355, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 8399 + }, + { + "epoch": 0.084, + "grad_norm": 0.614022136187256, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 8400 + }, + { + "epoch": 0.08401, + "grad_norm": 0.6164156175089536, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 8401 + }, + { + "epoch": 0.08402, + "grad_norm": 0.6304326357570794, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 8402 + }, + { + "epoch": 0.08403, + "grad_norm": 0.6212496094873539, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 8403 + }, + { + "epoch": 0.08404, + "grad_norm": 0.7270032965688896, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 8404 + }, + { + "epoch": 0.08405, + "grad_norm": 0.6629952777465283, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 8405 + }, + { + "epoch": 0.08406, + "grad_norm": 0.6053228957142524, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 8406 + }, + { + "epoch": 0.08407, + "grad_norm": 0.6000638079244677, + "learning_rate": 0.003, + "loss": 4.0806, + "step": 8407 + }, + { + "epoch": 0.08408, + "grad_norm": 0.5979167638243237, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 8408 + }, + { + "epoch": 0.08409, + "grad_norm": 0.65923857553275, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 8409 + }, + { + "epoch": 0.0841, + "grad_norm": 0.7946706348358272, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 8410 + }, + { + "epoch": 0.08411, + "grad_norm": 0.888589383941381, + "learning_rate": 0.003, + "loss": 4.0757, + "step": 8411 + }, + { + "epoch": 0.08412, + "grad_norm": 0.8714657792608406, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 8412 + }, + { + "epoch": 0.08413, + "grad_norm": 0.7683633986228294, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 8413 + }, + { + "epoch": 0.08414, + "grad_norm": 0.7468344552873766, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 8414 + }, + { + "epoch": 0.08415, + "grad_norm": 0.6847140193564525, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 8415 + }, + { + "epoch": 0.08416, + "grad_norm": 0.570149397526475, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 8416 + }, + { + "epoch": 0.08417, + "grad_norm": 0.6711885931583211, + "learning_rate": 0.003, + "loss": 4.043, + "step": 8417 + }, + { + "epoch": 0.08418, + "grad_norm": 0.7805625944689407, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 8418 + }, + { + "epoch": 0.08419, + "grad_norm": 0.8201396352846838, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8419 + }, + { + "epoch": 0.0842, + "grad_norm": 0.8274986319334163, + "learning_rate": 0.003, + "loss": 4.03, + "step": 8420 + }, + { + "epoch": 0.08421, + "grad_norm": 0.8174313186704971, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 8421 + }, + { + "epoch": 0.08422, + "grad_norm": 0.6909136962749025, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 8422 + }, + { + "epoch": 0.08423, + "grad_norm": 0.6109615213794316, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 8423 + }, + { + "epoch": 0.08424, + "grad_norm": 0.5833898277128587, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 8424 + }, + { + "epoch": 0.08425, + "grad_norm": 0.5404031983490655, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 8425 + }, + { + "epoch": 0.08426, + "grad_norm": 0.4740648641101616, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 8426 + }, + { + "epoch": 0.08427, + "grad_norm": 0.536059348893056, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 8427 + }, + { + "epoch": 0.08428, + "grad_norm": 0.6539220460111898, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 8428 + }, + { + "epoch": 0.08429, + "grad_norm": 0.7931258625645213, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 8429 + }, + { + "epoch": 0.0843, + "grad_norm": 0.6952590584912312, + "learning_rate": 0.003, + "loss": 4.005, + "step": 8430 + }, + { + "epoch": 0.08431, + "grad_norm": 0.5587142638563276, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 8431 + }, + { + "epoch": 0.08432, + "grad_norm": 0.5682128119579886, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 8432 + }, + { + "epoch": 0.08433, + "grad_norm": 0.602525484121658, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 8433 + }, + { + "epoch": 0.08434, + "grad_norm": 0.7081362256824555, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 8434 + }, + { + "epoch": 0.08435, + "grad_norm": 0.7781357897424593, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 8435 + }, + { + "epoch": 0.08436, + "grad_norm": 0.7617744407463904, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 8436 + }, + { + "epoch": 0.08437, + "grad_norm": 0.7491267062059953, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 8437 + }, + { + "epoch": 0.08438, + "grad_norm": 0.7350653964091405, + "learning_rate": 0.003, + "loss": 4.033, + "step": 8438 + }, + { + "epoch": 0.08439, + "grad_norm": 0.7340304896016234, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 8439 + }, + { + "epoch": 0.0844, + "grad_norm": 0.7844333901566573, + "learning_rate": 0.003, + "loss": 4.06, + "step": 8440 + }, + { + "epoch": 0.08441, + "grad_norm": 0.7634779210510275, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 8441 + }, + { + "epoch": 0.08442, + "grad_norm": 0.6626132595630899, + "learning_rate": 0.003, + "loss": 4.017, + "step": 8442 + }, + { + "epoch": 0.08443, + "grad_norm": 0.6234224850216683, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8443 + }, + { + "epoch": 0.08444, + "grad_norm": 0.6402456397044954, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 8444 + }, + { + "epoch": 0.08445, + "grad_norm": 0.7464903155788806, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 8445 + }, + { + "epoch": 0.08446, + "grad_norm": 0.737620568501164, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8446 + }, + { + "epoch": 0.08447, + "grad_norm": 0.7596295931020005, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 8447 + }, + { + "epoch": 0.08448, + "grad_norm": 0.7579760516974765, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 8448 + }, + { + "epoch": 0.08449, + "grad_norm": 0.8169564687226634, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 8449 + }, + { + "epoch": 0.0845, + "grad_norm": 0.8778255166477998, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 8450 + }, + { + "epoch": 0.08451, + "grad_norm": 0.9423758334618242, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 8451 + }, + { + "epoch": 0.08452, + "grad_norm": 0.9164908928908999, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 8452 + }, + { + "epoch": 0.08453, + "grad_norm": 0.8282332790995582, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 8453 + }, + { + "epoch": 0.08454, + "grad_norm": 0.765090818418697, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 8454 + }, + { + "epoch": 0.08455, + "grad_norm": 0.6218638016450626, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 8455 + }, + { + "epoch": 0.08456, + "grad_norm": 0.609354445222755, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 8456 + }, + { + "epoch": 0.08457, + "grad_norm": 0.5046822562952197, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 8457 + }, + { + "epoch": 0.08458, + "grad_norm": 0.4469148651069899, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 8458 + }, + { + "epoch": 0.08459, + "grad_norm": 0.5542425796029143, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 8459 + }, + { + "epoch": 0.0846, + "grad_norm": 0.7155834817326897, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 8460 + }, + { + "epoch": 0.08461, + "grad_norm": 0.8148177239432159, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 8461 + }, + { + "epoch": 0.08462, + "grad_norm": 0.7491785945667764, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 8462 + }, + { + "epoch": 0.08463, + "grad_norm": 0.6332541574379483, + "learning_rate": 0.003, + "loss": 4.022, + "step": 8463 + }, + { + "epoch": 0.08464, + "grad_norm": 0.607247436508271, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 8464 + }, + { + "epoch": 0.08465, + "grad_norm": 0.6626502055861434, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 8465 + }, + { + "epoch": 0.08466, + "grad_norm": 0.6592599796266105, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 8466 + }, + { + "epoch": 0.08467, + "grad_norm": 0.5907054963668018, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 8467 + }, + { + "epoch": 0.08468, + "grad_norm": 0.49371827739122953, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 8468 + }, + { + "epoch": 0.08469, + "grad_norm": 0.5465383235250868, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 8469 + }, + { + "epoch": 0.0847, + "grad_norm": 0.5904060643594715, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 8470 + }, + { + "epoch": 0.08471, + "grad_norm": 0.703582393867244, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 8471 + }, + { + "epoch": 0.08472, + "grad_norm": 0.8020810150547015, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 8472 + }, + { + "epoch": 0.08473, + "grad_norm": 0.9184918286149558, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 8473 + }, + { + "epoch": 0.08474, + "grad_norm": 0.8668444565744571, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 8474 + }, + { + "epoch": 0.08475, + "grad_norm": 0.7784785954421997, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 8475 + }, + { + "epoch": 0.08476, + "grad_norm": 0.9792309273493047, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 8476 + }, + { + "epoch": 0.08477, + "grad_norm": 1.1029909207529345, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 8477 + }, + { + "epoch": 0.08478, + "grad_norm": 0.8800682004029164, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 8478 + }, + { + "epoch": 0.08479, + "grad_norm": 0.942210330732621, + "learning_rate": 0.003, + "loss": 4.0707, + "step": 8479 + }, + { + "epoch": 0.0848, + "grad_norm": 1.0308840263033148, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 8480 + }, + { + "epoch": 0.08481, + "grad_norm": 0.921662825673339, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 8481 + }, + { + "epoch": 0.08482, + "grad_norm": 0.9035981430116728, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 8482 + }, + { + "epoch": 0.08483, + "grad_norm": 0.9804176549590359, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 8483 + }, + { + "epoch": 0.08484, + "grad_norm": 1.004778580614907, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 8484 + }, + { + "epoch": 0.08485, + "grad_norm": 0.990163114818611, + "learning_rate": 0.003, + "loss": 4.0755, + "step": 8485 + }, + { + "epoch": 0.08486, + "grad_norm": 0.972923755754475, + "learning_rate": 0.003, + "loss": 4.0932, + "step": 8486 + }, + { + "epoch": 0.08487, + "grad_norm": 1.027323957629767, + "learning_rate": 0.003, + "loss": 4.0983, + "step": 8487 + }, + { + "epoch": 0.08488, + "grad_norm": 0.9265526081973542, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 8488 + }, + { + "epoch": 0.08489, + "grad_norm": 0.9756235008002894, + "learning_rate": 0.003, + "loss": 4.083, + "step": 8489 + }, + { + "epoch": 0.0849, + "grad_norm": 1.0956592914853502, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 8490 + }, + { + "epoch": 0.08491, + "grad_norm": 0.815003429066045, + "learning_rate": 0.003, + "loss": 4.0857, + "step": 8491 + }, + { + "epoch": 0.08492, + "grad_norm": 0.7739204266250854, + "learning_rate": 0.003, + "loss": 4.0702, + "step": 8492 + }, + { + "epoch": 0.08493, + "grad_norm": 0.802272841577117, + "learning_rate": 0.003, + "loss": 4.0747, + "step": 8493 + }, + { + "epoch": 0.08494, + "grad_norm": 0.7982953413911512, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 8494 + }, + { + "epoch": 0.08495, + "grad_norm": 0.6566728066199379, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 8495 + }, + { + "epoch": 0.08496, + "grad_norm": 0.6231387389595457, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 8496 + }, + { + "epoch": 0.08497, + "grad_norm": 0.606468961285592, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 8497 + }, + { + "epoch": 0.08498, + "grad_norm": 0.6950424187430964, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 8498 + }, + { + "epoch": 0.08499, + "grad_norm": 0.7527060096232954, + "learning_rate": 0.003, + "loss": 4.0839, + "step": 8499 + }, + { + "epoch": 0.085, + "grad_norm": 0.7940905991268296, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 8500 + }, + { + "epoch": 0.08501, + "grad_norm": 0.680965991446952, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 8501 + }, + { + "epoch": 0.08502, + "grad_norm": 0.5299779377648792, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 8502 + }, + { + "epoch": 0.08503, + "grad_norm": 0.6019674704239707, + "learning_rate": 0.003, + "loss": 4.048, + "step": 8503 + }, + { + "epoch": 0.08504, + "grad_norm": 0.6714435629303511, + "learning_rate": 0.003, + "loss": 4.0778, + "step": 8504 + }, + { + "epoch": 0.08505, + "grad_norm": 0.6815185103606393, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 8505 + }, + { + "epoch": 0.08506, + "grad_norm": 0.5814186286858484, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 8506 + }, + { + "epoch": 0.08507, + "grad_norm": 0.5337608488632412, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 8507 + }, + { + "epoch": 0.08508, + "grad_norm": 0.5271100967862581, + "learning_rate": 0.003, + "loss": 4.028, + "step": 8508 + }, + { + "epoch": 0.08509, + "grad_norm": 0.6251406598301218, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 8509 + }, + { + "epoch": 0.0851, + "grad_norm": 0.5731578969326969, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 8510 + }, + { + "epoch": 0.08511, + "grad_norm": 0.4648016842505978, + "learning_rate": 0.003, + "loss": 4.043, + "step": 8511 + }, + { + "epoch": 0.08512, + "grad_norm": 0.48871002316541456, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 8512 + }, + { + "epoch": 0.08513, + "grad_norm": 0.58492278463361, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 8513 + }, + { + "epoch": 0.08514, + "grad_norm": 0.5950589311455498, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 8514 + }, + { + "epoch": 0.08515, + "grad_norm": 0.4785635204580544, + "learning_rate": 0.003, + "loss": 4.028, + "step": 8515 + }, + { + "epoch": 0.08516, + "grad_norm": 0.4708767402604776, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 8516 + }, + { + "epoch": 0.08517, + "grad_norm": 0.45545685524661467, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 8517 + }, + { + "epoch": 0.08518, + "grad_norm": 0.4427256360185519, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 8518 + }, + { + "epoch": 0.08519, + "grad_norm": 0.46786215581529456, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 8519 + }, + { + "epoch": 0.0852, + "grad_norm": 0.38682997290816806, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 8520 + }, + { + "epoch": 0.08521, + "grad_norm": 0.37734759945270224, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 8521 + }, + { + "epoch": 0.08522, + "grad_norm": 0.3973712458965192, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 8522 + }, + { + "epoch": 0.08523, + "grad_norm": 0.43937425910613265, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 8523 + }, + { + "epoch": 0.08524, + "grad_norm": 0.5464745311545621, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 8524 + }, + { + "epoch": 0.08525, + "grad_norm": 0.6906445357766109, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 8525 + }, + { + "epoch": 0.08526, + "grad_norm": 0.892579141524371, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 8526 + }, + { + "epoch": 0.08527, + "grad_norm": 1.1361248252360383, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 8527 + }, + { + "epoch": 0.08528, + "grad_norm": 0.8981655515731252, + "learning_rate": 0.003, + "loss": 4.02, + "step": 8528 + }, + { + "epoch": 0.08529, + "grad_norm": 0.7301907585240228, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 8529 + }, + { + "epoch": 0.0853, + "grad_norm": 0.5843812277385572, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 8530 + }, + { + "epoch": 0.08531, + "grad_norm": 0.6771262990968485, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 8531 + }, + { + "epoch": 0.08532, + "grad_norm": 0.6751358881433738, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 8532 + }, + { + "epoch": 0.08533, + "grad_norm": 0.6651157902320582, + "learning_rate": 0.003, + "loss": 4.049, + "step": 8533 + }, + { + "epoch": 0.08534, + "grad_norm": 0.7169229773452893, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 8534 + }, + { + "epoch": 0.08535, + "grad_norm": 0.8118937944924871, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 8535 + }, + { + "epoch": 0.08536, + "grad_norm": 0.8292308391865982, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 8536 + }, + { + "epoch": 0.08537, + "grad_norm": 0.6562159140971349, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 8537 + }, + { + "epoch": 0.08538, + "grad_norm": 0.6066906823134351, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 8538 + }, + { + "epoch": 0.08539, + "grad_norm": 0.56828329270359, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 8539 + }, + { + "epoch": 0.0854, + "grad_norm": 0.5613292917624827, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 8540 + }, + { + "epoch": 0.08541, + "grad_norm": 0.5214458071840425, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 8541 + }, + { + "epoch": 0.08542, + "grad_norm": 0.600342626928443, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 8542 + }, + { + "epoch": 0.08543, + "grad_norm": 0.6734807455597496, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 8543 + }, + { + "epoch": 0.08544, + "grad_norm": 0.683264974640567, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 8544 + }, + { + "epoch": 0.08545, + "grad_norm": 0.7735837217695435, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 8545 + }, + { + "epoch": 0.08546, + "grad_norm": 0.9866278953342396, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 8546 + }, + { + "epoch": 0.08547, + "grad_norm": 1.0498166452732907, + "learning_rate": 0.003, + "loss": 4.073, + "step": 8547 + }, + { + "epoch": 0.08548, + "grad_norm": 1.0679357429558958, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 8548 + }, + { + "epoch": 0.08549, + "grad_norm": 0.9622068158135412, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 8549 + }, + { + "epoch": 0.0855, + "grad_norm": 0.8845823972170755, + "learning_rate": 0.003, + "loss": 4.0777, + "step": 8550 + }, + { + "epoch": 0.08551, + "grad_norm": 0.9023821325798975, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 8551 + }, + { + "epoch": 0.08552, + "grad_norm": 0.8672779869256604, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 8552 + }, + { + "epoch": 0.08553, + "grad_norm": 0.6439012847238149, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 8553 + }, + { + "epoch": 0.08554, + "grad_norm": 0.6078352940529411, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 8554 + }, + { + "epoch": 0.08555, + "grad_norm": 0.5955091997466793, + "learning_rate": 0.003, + "loss": 4.0739, + "step": 8555 + }, + { + "epoch": 0.08556, + "grad_norm": 0.6184599082494526, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 8556 + }, + { + "epoch": 0.08557, + "grad_norm": 0.6314571239494786, + "learning_rate": 0.003, + "loss": 4.025, + "step": 8557 + }, + { + "epoch": 0.08558, + "grad_norm": 0.7322253512724209, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 8558 + }, + { + "epoch": 0.08559, + "grad_norm": 0.8085700826949176, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 8559 + }, + { + "epoch": 0.0856, + "grad_norm": 0.8925965932609011, + "learning_rate": 0.003, + "loss": 4.047, + "step": 8560 + }, + { + "epoch": 0.08561, + "grad_norm": 0.8252586285420619, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 8561 + }, + { + "epoch": 0.08562, + "grad_norm": 0.80816386710577, + "learning_rate": 0.003, + "loss": 4.018, + "step": 8562 + }, + { + "epoch": 0.08563, + "grad_norm": 0.8050841468565558, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 8563 + }, + { + "epoch": 0.08564, + "grad_norm": 0.851315671917811, + "learning_rate": 0.003, + "loss": 4.043, + "step": 8564 + }, + { + "epoch": 0.08565, + "grad_norm": 0.790405796628351, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 8565 + }, + { + "epoch": 0.08566, + "grad_norm": 0.7837360970476384, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 8566 + }, + { + "epoch": 0.08567, + "grad_norm": 0.6561654702664893, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 8567 + }, + { + "epoch": 0.08568, + "grad_norm": 0.5544753509497192, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 8568 + }, + { + "epoch": 0.08569, + "grad_norm": 0.5415833462275254, + "learning_rate": 0.003, + "loss": 4.0658, + "step": 8569 + }, + { + "epoch": 0.0857, + "grad_norm": 0.6131696122834572, + "learning_rate": 0.003, + "loss": 4.0619, + "step": 8570 + }, + { + "epoch": 0.08571, + "grad_norm": 0.5851197626843028, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 8571 + }, + { + "epoch": 0.08572, + "grad_norm": 0.5944212179691615, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 8572 + }, + { + "epoch": 0.08573, + "grad_norm": 0.5877726531919563, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 8573 + }, + { + "epoch": 0.08574, + "grad_norm": 0.591038079825928, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 8574 + }, + { + "epoch": 0.08575, + "grad_norm": 0.5385577439417866, + "learning_rate": 0.003, + "loss": 4.045, + "step": 8575 + }, + { + "epoch": 0.08576, + "grad_norm": 0.5642464757318268, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 8576 + }, + { + "epoch": 0.08577, + "grad_norm": 0.5791985351865473, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 8577 + }, + { + "epoch": 0.08578, + "grad_norm": 0.47308985429361367, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 8578 + }, + { + "epoch": 0.08579, + "grad_norm": 0.4078825166658552, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 8579 + }, + { + "epoch": 0.0858, + "grad_norm": 0.41875954317135383, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 8580 + }, + { + "epoch": 0.08581, + "grad_norm": 0.44058692596375004, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 8581 + }, + { + "epoch": 0.08582, + "grad_norm": 0.484134168390493, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 8582 + }, + { + "epoch": 0.08583, + "grad_norm": 0.608252177699836, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 8583 + }, + { + "epoch": 0.08584, + "grad_norm": 0.7887497167529344, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 8584 + }, + { + "epoch": 0.08585, + "grad_norm": 0.9452694884125372, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 8585 + }, + { + "epoch": 0.08586, + "grad_norm": 0.91037297711997, + "learning_rate": 0.003, + "loss": 4.0709, + "step": 8586 + }, + { + "epoch": 0.08587, + "grad_norm": 0.861992590352416, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 8587 + }, + { + "epoch": 0.08588, + "grad_norm": 0.6607748088790797, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 8588 + }, + { + "epoch": 0.08589, + "grad_norm": 0.5358766208391812, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 8589 + }, + { + "epoch": 0.0859, + "grad_norm": 0.6529606365226626, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 8590 + }, + { + "epoch": 0.08591, + "grad_norm": 0.7115858030201645, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 8591 + }, + { + "epoch": 0.08592, + "grad_norm": 0.7890403157067962, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 8592 + }, + { + "epoch": 0.08593, + "grad_norm": 0.7815197032858606, + "learning_rate": 0.003, + "loss": 4.078, + "step": 8593 + }, + { + "epoch": 0.08594, + "grad_norm": 0.720635567256226, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 8594 + }, + { + "epoch": 0.08595, + "grad_norm": 0.8073472257484936, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 8595 + }, + { + "epoch": 0.08596, + "grad_norm": 0.9837699031928535, + "learning_rate": 0.003, + "loss": 4.065, + "step": 8596 + }, + { + "epoch": 0.08597, + "grad_norm": 1.0438607582378798, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 8597 + }, + { + "epoch": 0.08598, + "grad_norm": 0.8100018496645228, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 8598 + }, + { + "epoch": 0.08599, + "grad_norm": 0.7905023103156219, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 8599 + }, + { + "epoch": 0.086, + "grad_norm": 0.8186430106274071, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 8600 + }, + { + "epoch": 0.08601, + "grad_norm": 0.8362729555510072, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 8601 + }, + { + "epoch": 0.08602, + "grad_norm": 0.7268468318032925, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 8602 + }, + { + "epoch": 0.08603, + "grad_norm": 0.7684326711136098, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 8603 + }, + { + "epoch": 0.08604, + "grad_norm": 0.8435536323944499, + "learning_rate": 0.003, + "loss": 4.048, + "step": 8604 + }, + { + "epoch": 0.08605, + "grad_norm": 0.7897400165828594, + "learning_rate": 0.003, + "loss": 4.0618, + "step": 8605 + }, + { + "epoch": 0.08606, + "grad_norm": 0.7860752398698371, + "learning_rate": 0.003, + "loss": 4.03, + "step": 8606 + }, + { + "epoch": 0.08607, + "grad_norm": 0.7137909202924649, + "learning_rate": 0.003, + "loss": 4.0712, + "step": 8607 + }, + { + "epoch": 0.08608, + "grad_norm": 0.6558574047828304, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 8608 + }, + { + "epoch": 0.08609, + "grad_norm": 0.7038641494041802, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 8609 + }, + { + "epoch": 0.0861, + "grad_norm": 0.8279313706979311, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 8610 + }, + { + "epoch": 0.08611, + "grad_norm": 0.7564937185854375, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 8611 + }, + { + "epoch": 0.08612, + "grad_norm": 0.6545002162462241, + "learning_rate": 0.003, + "loss": 4.0793, + "step": 8612 + }, + { + "epoch": 0.08613, + "grad_norm": 0.6384935585565132, + "learning_rate": 0.003, + "loss": 4.053, + "step": 8613 + }, + { + "epoch": 0.08614, + "grad_norm": 0.6413639954422861, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 8614 + }, + { + "epoch": 0.08615, + "grad_norm": 0.5846554145208936, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8615 + }, + { + "epoch": 0.08616, + "grad_norm": 0.624828860839509, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 8616 + }, + { + "epoch": 0.08617, + "grad_norm": 0.6193186719516374, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 8617 + }, + { + "epoch": 0.08618, + "grad_norm": 0.6070985254352729, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 8618 + }, + { + "epoch": 0.08619, + "grad_norm": 0.5729445632753838, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 8619 + }, + { + "epoch": 0.0862, + "grad_norm": 0.6356472887394397, + "learning_rate": 0.003, + "loss": 4.042, + "step": 8620 + }, + { + "epoch": 0.08621, + "grad_norm": 0.6803911484438744, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 8621 + }, + { + "epoch": 0.08622, + "grad_norm": 0.7415589775458606, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 8622 + }, + { + "epoch": 0.08623, + "grad_norm": 0.8622103936368933, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 8623 + }, + { + "epoch": 0.08624, + "grad_norm": 0.9739457742355512, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 8624 + }, + { + "epoch": 0.08625, + "grad_norm": 1.0323041067556005, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 8625 + }, + { + "epoch": 0.08626, + "grad_norm": 0.7120475138489862, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 8626 + }, + { + "epoch": 0.08627, + "grad_norm": 0.7169134408043424, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 8627 + }, + { + "epoch": 0.08628, + "grad_norm": 0.670327860850868, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 8628 + }, + { + "epoch": 0.08629, + "grad_norm": 0.6069895734542193, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 8629 + }, + { + "epoch": 0.0863, + "grad_norm": 0.5142692071204801, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 8630 + }, + { + "epoch": 0.08631, + "grad_norm": 0.4983206067217951, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 8631 + }, + { + "epoch": 0.08632, + "grad_norm": 0.5541000034997056, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 8632 + }, + { + "epoch": 0.08633, + "grad_norm": 0.6106072446364847, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 8633 + }, + { + "epoch": 0.08634, + "grad_norm": 0.7373095043116881, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 8634 + }, + { + "epoch": 0.08635, + "grad_norm": 0.8891752950034024, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 8635 + }, + { + "epoch": 0.08636, + "grad_norm": 0.9635160753972452, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 8636 + }, + { + "epoch": 0.08637, + "grad_norm": 0.8992510900262488, + "learning_rate": 0.003, + "loss": 4.0731, + "step": 8637 + }, + { + "epoch": 0.08638, + "grad_norm": 0.7401335389118073, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 8638 + }, + { + "epoch": 0.08639, + "grad_norm": 0.6692460787676616, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 8639 + }, + { + "epoch": 0.0864, + "grad_norm": 0.6594246971725337, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 8640 + }, + { + "epoch": 0.08641, + "grad_norm": 0.5939338596432617, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 8641 + }, + { + "epoch": 0.08642, + "grad_norm": 0.6199437252011405, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 8642 + }, + { + "epoch": 0.08643, + "grad_norm": 0.7092196621819203, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 8643 + }, + { + "epoch": 0.08644, + "grad_norm": 0.8303182509451258, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 8644 + }, + { + "epoch": 0.08645, + "grad_norm": 0.7735473030802615, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 8645 + }, + { + "epoch": 0.08646, + "grad_norm": 0.706362389672776, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 8646 + }, + { + "epoch": 0.08647, + "grad_norm": 0.6987284935086118, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 8647 + }, + { + "epoch": 0.08648, + "grad_norm": 0.6071522685941764, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 8648 + }, + { + "epoch": 0.08649, + "grad_norm": 0.6863248744606724, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 8649 + }, + { + "epoch": 0.0865, + "grad_norm": 0.8134049686443081, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 8650 + }, + { + "epoch": 0.08651, + "grad_norm": 0.8568412303850552, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 8651 + }, + { + "epoch": 0.08652, + "grad_norm": 0.9761387149201433, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 8652 + }, + { + "epoch": 0.08653, + "grad_norm": 1.0124956659050082, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 8653 + }, + { + "epoch": 0.08654, + "grad_norm": 0.8795248881529959, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 8654 + }, + { + "epoch": 0.08655, + "grad_norm": 0.7311093014320024, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 8655 + }, + { + "epoch": 0.08656, + "grad_norm": 0.5831820823938508, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 8656 + }, + { + "epoch": 0.08657, + "grad_norm": 0.649052011648633, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 8657 + }, + { + "epoch": 0.08658, + "grad_norm": 0.6207039095641198, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 8658 + }, + { + "epoch": 0.08659, + "grad_norm": 0.6711056298603381, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 8659 + }, + { + "epoch": 0.0866, + "grad_norm": 0.6044847306951937, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 8660 + }, + { + "epoch": 0.08661, + "grad_norm": 0.5741604118623871, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8661 + }, + { + "epoch": 0.08662, + "grad_norm": 0.5390282739441018, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 8662 + }, + { + "epoch": 0.08663, + "grad_norm": 0.5670872899437688, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 8663 + }, + { + "epoch": 0.08664, + "grad_norm": 0.5943341986691961, + "learning_rate": 0.003, + "loss": 4.035, + "step": 8664 + }, + { + "epoch": 0.08665, + "grad_norm": 0.5961979062852375, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 8665 + }, + { + "epoch": 0.08666, + "grad_norm": 0.6115429497629993, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 8666 + }, + { + "epoch": 0.08667, + "grad_norm": 0.6151426438370023, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 8667 + }, + { + "epoch": 0.08668, + "grad_norm": 0.6645205264696532, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 8668 + }, + { + "epoch": 0.08669, + "grad_norm": 0.7588406208120678, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 8669 + }, + { + "epoch": 0.0867, + "grad_norm": 0.7756686250916464, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 8670 + }, + { + "epoch": 0.08671, + "grad_norm": 0.6957703950529082, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 8671 + }, + { + "epoch": 0.08672, + "grad_norm": 0.7619763339361051, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 8672 + }, + { + "epoch": 0.08673, + "grad_norm": 0.8582688693310772, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 8673 + }, + { + "epoch": 0.08674, + "grad_norm": 0.9013449692467891, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 8674 + }, + { + "epoch": 0.08675, + "grad_norm": 0.8694296044844825, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 8675 + }, + { + "epoch": 0.08676, + "grad_norm": 0.7834036094984678, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 8676 + }, + { + "epoch": 0.08677, + "grad_norm": 0.7435090171944096, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 8677 + }, + { + "epoch": 0.08678, + "grad_norm": 0.6631933266955773, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 8678 + }, + { + "epoch": 0.08679, + "grad_norm": 0.6448513180539295, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 8679 + }, + { + "epoch": 0.0868, + "grad_norm": 0.654676322170026, + "learning_rate": 0.003, + "loss": 4.005, + "step": 8680 + }, + { + "epoch": 0.08681, + "grad_norm": 0.6159085300903079, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 8681 + }, + { + "epoch": 0.08682, + "grad_norm": 0.648131702458417, + "learning_rate": 0.003, + "loss": 4.034, + "step": 8682 + }, + { + "epoch": 0.08683, + "grad_norm": 0.6142805511623832, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 8683 + }, + { + "epoch": 0.08684, + "grad_norm": 0.6150584274050082, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 8684 + }, + { + "epoch": 0.08685, + "grad_norm": 0.6034640346572434, + "learning_rate": 0.003, + "loss": 4.042, + "step": 8685 + }, + { + "epoch": 0.08686, + "grad_norm": 0.6745967806809965, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 8686 + }, + { + "epoch": 0.08687, + "grad_norm": 0.7725748627480534, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 8687 + }, + { + "epoch": 0.08688, + "grad_norm": 0.667066748722982, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 8688 + }, + { + "epoch": 0.08689, + "grad_norm": 0.6247677401883848, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 8689 + }, + { + "epoch": 0.0869, + "grad_norm": 0.5523391330192599, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 8690 + }, + { + "epoch": 0.08691, + "grad_norm": 0.6155082593949734, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 8691 + }, + { + "epoch": 0.08692, + "grad_norm": 0.6031359874855519, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 8692 + }, + { + "epoch": 0.08693, + "grad_norm": 0.6765403284481811, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 8693 + }, + { + "epoch": 0.08694, + "grad_norm": 0.7505511653440876, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 8694 + }, + { + "epoch": 0.08695, + "grad_norm": 0.8375729352022898, + "learning_rate": 0.003, + "loss": 4.032, + "step": 8695 + }, + { + "epoch": 0.08696, + "grad_norm": 0.8155172498289832, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 8696 + }, + { + "epoch": 0.08697, + "grad_norm": 0.7557496410046333, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 8697 + }, + { + "epoch": 0.08698, + "grad_norm": 0.633161926057819, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 8698 + }, + { + "epoch": 0.08699, + "grad_norm": 0.6765097963282148, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8699 + }, + { + "epoch": 0.087, + "grad_norm": 0.6453242856689506, + "learning_rate": 0.003, + "loss": 4.064, + "step": 8700 + }, + { + "epoch": 0.08701, + "grad_norm": 0.6266238013427983, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 8701 + }, + { + "epoch": 0.08702, + "grad_norm": 0.6419176654223141, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 8702 + }, + { + "epoch": 0.08703, + "grad_norm": 0.6020317767741878, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 8703 + }, + { + "epoch": 0.08704, + "grad_norm": 0.6425265013784115, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 8704 + }, + { + "epoch": 0.08705, + "grad_norm": 0.6875238064743521, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 8705 + }, + { + "epoch": 0.08706, + "grad_norm": 0.7815233995916582, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 8706 + }, + { + "epoch": 0.08707, + "grad_norm": 0.8977643661038409, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 8707 + }, + { + "epoch": 0.08708, + "grad_norm": 0.8396480258274408, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 8708 + }, + { + "epoch": 0.08709, + "grad_norm": 0.7124229512328322, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 8709 + }, + { + "epoch": 0.0871, + "grad_norm": 0.6852024222407596, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 8710 + }, + { + "epoch": 0.08711, + "grad_norm": 0.7186669065824979, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8711 + }, + { + "epoch": 0.08712, + "grad_norm": 0.7655815180969611, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 8712 + }, + { + "epoch": 0.08713, + "grad_norm": 0.9384551762911357, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 8713 + }, + { + "epoch": 0.08714, + "grad_norm": 0.9180799640929881, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 8714 + }, + { + "epoch": 0.08715, + "grad_norm": 0.8568370338780645, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 8715 + }, + { + "epoch": 0.08716, + "grad_norm": 0.8374704025286653, + "learning_rate": 0.003, + "loss": 4.0613, + "step": 8716 + }, + { + "epoch": 0.08717, + "grad_norm": 0.6472121959168161, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 8717 + }, + { + "epoch": 0.08718, + "grad_norm": 0.7424987963603377, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 8718 + }, + { + "epoch": 0.08719, + "grad_norm": 0.7016121430916806, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 8719 + }, + { + "epoch": 0.0872, + "grad_norm": 0.6759240302489125, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 8720 + }, + { + "epoch": 0.08721, + "grad_norm": 0.627577722625845, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 8721 + }, + { + "epoch": 0.08722, + "grad_norm": 0.6099110571190953, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 8722 + }, + { + "epoch": 0.08723, + "grad_norm": 0.5536970679263835, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 8723 + }, + { + "epoch": 0.08724, + "grad_norm": 0.4697779056756519, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 8724 + }, + { + "epoch": 0.08725, + "grad_norm": 0.47894687690769794, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 8725 + }, + { + "epoch": 0.08726, + "grad_norm": 0.4999718719135512, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 8726 + }, + { + "epoch": 0.08727, + "grad_norm": 0.5498283363002583, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 8727 + }, + { + "epoch": 0.08728, + "grad_norm": 0.5487755202000018, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 8728 + }, + { + "epoch": 0.08729, + "grad_norm": 0.7057397578482671, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 8729 + }, + { + "epoch": 0.0873, + "grad_norm": 0.7807793478450717, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 8730 + }, + { + "epoch": 0.08731, + "grad_norm": 0.8179731408543531, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 8731 + }, + { + "epoch": 0.08732, + "grad_norm": 0.7833801508902276, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 8732 + }, + { + "epoch": 0.08733, + "grad_norm": 0.7108270457510092, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 8733 + }, + { + "epoch": 0.08734, + "grad_norm": 0.5769733380002771, + "learning_rate": 0.003, + "loss": 4.055, + "step": 8734 + }, + { + "epoch": 0.08735, + "grad_norm": 0.6339426862617171, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8735 + }, + { + "epoch": 0.08736, + "grad_norm": 0.5940885138544176, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 8736 + }, + { + "epoch": 0.08737, + "grad_norm": 0.6436983015420159, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 8737 + }, + { + "epoch": 0.08738, + "grad_norm": 0.721208453608969, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 8738 + }, + { + "epoch": 0.08739, + "grad_norm": 0.8229160572388433, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 8739 + }, + { + "epoch": 0.0874, + "grad_norm": 0.861140717749566, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 8740 + }, + { + "epoch": 0.08741, + "grad_norm": 0.8495877966082562, + "learning_rate": 0.003, + "loss": 4.028, + "step": 8741 + }, + { + "epoch": 0.08742, + "grad_norm": 0.8304766840806012, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 8742 + }, + { + "epoch": 0.08743, + "grad_norm": 0.9162572906605098, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 8743 + }, + { + "epoch": 0.08744, + "grad_norm": 0.9780933649318432, + "learning_rate": 0.003, + "loss": 4.073, + "step": 8744 + }, + { + "epoch": 0.08745, + "grad_norm": 0.8265842521998155, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 8745 + }, + { + "epoch": 0.08746, + "grad_norm": 0.650053002040002, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8746 + }, + { + "epoch": 0.08747, + "grad_norm": 0.6852000326362472, + "learning_rate": 0.003, + "loss": 4.053, + "step": 8747 + }, + { + "epoch": 0.08748, + "grad_norm": 0.5097594498941278, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8748 + }, + { + "epoch": 0.08749, + "grad_norm": 0.4769625234312351, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 8749 + }, + { + "epoch": 0.0875, + "grad_norm": 0.43484234634377655, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 8750 + }, + { + "epoch": 0.08751, + "grad_norm": 0.45545567369208634, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 8751 + }, + { + "epoch": 0.08752, + "grad_norm": 0.5344767323512793, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 8752 + }, + { + "epoch": 0.08753, + "grad_norm": 0.6180881817631013, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 8753 + }, + { + "epoch": 0.08754, + "grad_norm": 0.716508306916931, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 8754 + }, + { + "epoch": 0.08755, + "grad_norm": 0.6278993758304847, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 8755 + }, + { + "epoch": 0.08756, + "grad_norm": 0.5456901895617449, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 8756 + }, + { + "epoch": 0.08757, + "grad_norm": 0.5901560544924496, + "learning_rate": 0.003, + "loss": 4.055, + "step": 8757 + }, + { + "epoch": 0.08758, + "grad_norm": 0.6959643294543882, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 8758 + }, + { + "epoch": 0.08759, + "grad_norm": 0.818706312632267, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 8759 + }, + { + "epoch": 0.0876, + "grad_norm": 0.895865009204446, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 8760 + }, + { + "epoch": 0.08761, + "grad_norm": 0.9273166345729585, + "learning_rate": 0.003, + "loss": 4.075, + "step": 8761 + }, + { + "epoch": 0.08762, + "grad_norm": 0.9240543156705644, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 8762 + }, + { + "epoch": 0.08763, + "grad_norm": 0.9738495364091213, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 8763 + }, + { + "epoch": 0.08764, + "grad_norm": 0.9843020410166308, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 8764 + }, + { + "epoch": 0.08765, + "grad_norm": 0.8627748236966266, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 8765 + }, + { + "epoch": 0.08766, + "grad_norm": 0.9352781782217522, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 8766 + }, + { + "epoch": 0.08767, + "grad_norm": 1.107452622413729, + "learning_rate": 0.003, + "loss": 4.0606, + "step": 8767 + }, + { + "epoch": 0.08768, + "grad_norm": 0.981315209557018, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 8768 + }, + { + "epoch": 0.08769, + "grad_norm": 0.9518710129287874, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 8769 + }, + { + "epoch": 0.0877, + "grad_norm": 0.925189869269097, + "learning_rate": 0.003, + "loss": 4.0682, + "step": 8770 + }, + { + "epoch": 0.08771, + "grad_norm": 0.8584718860605614, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 8771 + }, + { + "epoch": 0.08772, + "grad_norm": 0.7178860482692307, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 8772 + }, + { + "epoch": 0.08773, + "grad_norm": 0.8145854042461742, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 8773 + }, + { + "epoch": 0.08774, + "grad_norm": 0.8074296841469706, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 8774 + }, + { + "epoch": 0.08775, + "grad_norm": 0.8546761043591158, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 8775 + }, + { + "epoch": 0.08776, + "grad_norm": 0.7096661271387306, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 8776 + }, + { + "epoch": 0.08777, + "grad_norm": 0.6797770552112118, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 8777 + }, + { + "epoch": 0.08778, + "grad_norm": 0.6259036977094732, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 8778 + }, + { + "epoch": 0.08779, + "grad_norm": 0.63380850802779, + "learning_rate": 0.003, + "loss": 4.062, + "step": 8779 + }, + { + "epoch": 0.0878, + "grad_norm": 0.642634137313799, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 8780 + }, + { + "epoch": 0.08781, + "grad_norm": 0.6638201980272972, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 8781 + }, + { + "epoch": 0.08782, + "grad_norm": 0.598075222548536, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 8782 + }, + { + "epoch": 0.08783, + "grad_norm": 0.5409990987272038, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 8783 + }, + { + "epoch": 0.08784, + "grad_norm": 0.5646866850159241, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8784 + }, + { + "epoch": 0.08785, + "grad_norm": 0.6833472947585566, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 8785 + }, + { + "epoch": 0.08786, + "grad_norm": 0.6709025059792196, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 8786 + }, + { + "epoch": 0.08787, + "grad_norm": 0.6346825740074405, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 8787 + }, + { + "epoch": 0.08788, + "grad_norm": 0.6127427384181404, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 8788 + }, + { + "epoch": 0.08789, + "grad_norm": 0.6124851048765029, + "learning_rate": 0.003, + "loss": 4.025, + "step": 8789 + }, + { + "epoch": 0.0879, + "grad_norm": 0.6221194613959666, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 8790 + }, + { + "epoch": 0.08791, + "grad_norm": 0.6309930617186035, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 8791 + }, + { + "epoch": 0.08792, + "grad_norm": 0.6302733108503931, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 8792 + }, + { + "epoch": 0.08793, + "grad_norm": 0.6163583900171477, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 8793 + }, + { + "epoch": 0.08794, + "grad_norm": 0.5677721751534922, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 8794 + }, + { + "epoch": 0.08795, + "grad_norm": 0.5326943646801332, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 8795 + }, + { + "epoch": 0.08796, + "grad_norm": 0.5259195442905321, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 8796 + }, + { + "epoch": 0.08797, + "grad_norm": 0.47706063367237145, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 8797 + }, + { + "epoch": 0.08798, + "grad_norm": 0.4892206882336819, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 8798 + }, + { + "epoch": 0.08799, + "grad_norm": 0.508882455588903, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 8799 + }, + { + "epoch": 0.088, + "grad_norm": 0.5992742432029163, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 8800 + }, + { + "epoch": 0.08801, + "grad_norm": 0.7960608554050718, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 8801 + }, + { + "epoch": 0.08802, + "grad_norm": 0.9500525541076159, + "learning_rate": 0.003, + "loss": 4.047, + "step": 8802 + }, + { + "epoch": 0.08803, + "grad_norm": 0.9444459643791653, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 8803 + }, + { + "epoch": 0.08804, + "grad_norm": 0.9128333175463631, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 8804 + }, + { + "epoch": 0.08805, + "grad_norm": 0.9019472219447205, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 8805 + }, + { + "epoch": 0.08806, + "grad_norm": 0.8404269487145729, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8806 + }, + { + "epoch": 0.08807, + "grad_norm": 0.7697569852084947, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 8807 + }, + { + "epoch": 0.08808, + "grad_norm": 0.6999310079144502, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 8808 + }, + { + "epoch": 0.08809, + "grad_norm": 0.6295791004949703, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 8809 + }, + { + "epoch": 0.0881, + "grad_norm": 0.6181261924038526, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 8810 + }, + { + "epoch": 0.08811, + "grad_norm": 0.5773052840195603, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 8811 + }, + { + "epoch": 0.08812, + "grad_norm": 0.6063319551437623, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 8812 + }, + { + "epoch": 0.08813, + "grad_norm": 0.6082333770291924, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 8813 + }, + { + "epoch": 0.08814, + "grad_norm": 0.6302200296665903, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 8814 + }, + { + "epoch": 0.08815, + "grad_norm": 0.6674558126898404, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 8815 + }, + { + "epoch": 0.08816, + "grad_norm": 0.7591828612173012, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 8816 + }, + { + "epoch": 0.08817, + "grad_norm": 0.8680702286425839, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 8817 + }, + { + "epoch": 0.08818, + "grad_norm": 0.972466242993863, + "learning_rate": 0.003, + "loss": 4.037, + "step": 8818 + }, + { + "epoch": 0.08819, + "grad_norm": 0.983010855959098, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 8819 + }, + { + "epoch": 0.0882, + "grad_norm": 0.8334649246144867, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8820 + }, + { + "epoch": 0.08821, + "grad_norm": 0.7252155572353793, + "learning_rate": 0.003, + "loss": 4.0695, + "step": 8821 + }, + { + "epoch": 0.08822, + "grad_norm": 0.6956801898107225, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 8822 + }, + { + "epoch": 0.08823, + "grad_norm": 0.712206976660391, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 8823 + }, + { + "epoch": 0.08824, + "grad_norm": 0.737181256665671, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 8824 + }, + { + "epoch": 0.08825, + "grad_norm": 0.7261338232977552, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 8825 + }, + { + "epoch": 0.08826, + "grad_norm": 0.7081058756618135, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 8826 + }, + { + "epoch": 0.08827, + "grad_norm": 0.7650692997869962, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 8827 + }, + { + "epoch": 0.08828, + "grad_norm": 0.7292986299875106, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 8828 + }, + { + "epoch": 0.08829, + "grad_norm": 0.7111442306454846, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 8829 + }, + { + "epoch": 0.0883, + "grad_norm": 0.7430081117444886, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 8830 + }, + { + "epoch": 0.08831, + "grad_norm": 0.7205579657770675, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 8831 + }, + { + "epoch": 0.08832, + "grad_norm": 0.541300771439013, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 8832 + }, + { + "epoch": 0.08833, + "grad_norm": 0.434581468066541, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 8833 + }, + { + "epoch": 0.08834, + "grad_norm": 0.4546788095655341, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 8834 + }, + { + "epoch": 0.08835, + "grad_norm": 0.5007988198757275, + "learning_rate": 0.003, + "loss": 4.052, + "step": 8835 + }, + { + "epoch": 0.08836, + "grad_norm": 0.5623777437225381, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 8836 + }, + { + "epoch": 0.08837, + "grad_norm": 0.7104858386378617, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 8837 + }, + { + "epoch": 0.08838, + "grad_norm": 0.841994537009608, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 8838 + }, + { + "epoch": 0.08839, + "grad_norm": 0.7243051530923598, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 8839 + }, + { + "epoch": 0.0884, + "grad_norm": 0.5304847413483016, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 8840 + }, + { + "epoch": 0.08841, + "grad_norm": 0.516283832183782, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 8841 + }, + { + "epoch": 0.08842, + "grad_norm": 0.6008198315774474, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 8842 + }, + { + "epoch": 0.08843, + "grad_norm": 0.6457796446741437, + "learning_rate": 0.003, + "loss": 4.018, + "step": 8843 + }, + { + "epoch": 0.08844, + "grad_norm": 0.6095317741341931, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 8844 + }, + { + "epoch": 0.08845, + "grad_norm": 0.5854264517927529, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 8845 + }, + { + "epoch": 0.08846, + "grad_norm": 0.5920983948335907, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 8846 + }, + { + "epoch": 0.08847, + "grad_norm": 0.7609313471429864, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 8847 + }, + { + "epoch": 0.08848, + "grad_norm": 0.8144228317307849, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 8848 + }, + { + "epoch": 0.08849, + "grad_norm": 0.7044375788821898, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 8849 + }, + { + "epoch": 0.0885, + "grad_norm": 0.6901819680239579, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 8850 + }, + { + "epoch": 0.08851, + "grad_norm": 0.6284157009048142, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 8851 + }, + { + "epoch": 0.08852, + "grad_norm": 0.7230086981777916, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 8852 + }, + { + "epoch": 0.08853, + "grad_norm": 0.8492194262167785, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 8853 + }, + { + "epoch": 0.08854, + "grad_norm": 0.9273345132240217, + "learning_rate": 0.003, + "loss": 4.0752, + "step": 8854 + }, + { + "epoch": 0.08855, + "grad_norm": 0.7992685556521781, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 8855 + }, + { + "epoch": 0.08856, + "grad_norm": 0.8036467187136315, + "learning_rate": 0.003, + "loss": 4.049, + "step": 8856 + }, + { + "epoch": 0.08857, + "grad_norm": 0.8356020720619949, + "learning_rate": 0.003, + "loss": 4.052, + "step": 8857 + }, + { + "epoch": 0.08858, + "grad_norm": 0.9467056929283951, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8858 + }, + { + "epoch": 0.08859, + "grad_norm": 0.8233844069772315, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 8859 + }, + { + "epoch": 0.0886, + "grad_norm": 0.8046608743231581, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 8860 + }, + { + "epoch": 0.08861, + "grad_norm": 0.8291341563638279, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 8861 + }, + { + "epoch": 0.08862, + "grad_norm": 0.9398149439109985, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 8862 + }, + { + "epoch": 0.08863, + "grad_norm": 0.822386182210482, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 8863 + }, + { + "epoch": 0.08864, + "grad_norm": 0.6809117432997165, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 8864 + }, + { + "epoch": 0.08865, + "grad_norm": 0.6571986543823963, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 8865 + }, + { + "epoch": 0.08866, + "grad_norm": 0.6581262811644706, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 8866 + }, + { + "epoch": 0.08867, + "grad_norm": 0.6572912556511142, + "learning_rate": 0.003, + "loss": 4.05, + "step": 8867 + }, + { + "epoch": 0.08868, + "grad_norm": 0.7346605603726869, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 8868 + }, + { + "epoch": 0.08869, + "grad_norm": 0.8592631830226695, + "learning_rate": 0.003, + "loss": 4.09, + "step": 8869 + }, + { + "epoch": 0.0887, + "grad_norm": 0.9970286408891992, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 8870 + }, + { + "epoch": 0.08871, + "grad_norm": 0.8854294463251334, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 8871 + }, + { + "epoch": 0.08872, + "grad_norm": 0.6595651919992755, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 8872 + }, + { + "epoch": 0.08873, + "grad_norm": 0.631057261557234, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 8873 + }, + { + "epoch": 0.08874, + "grad_norm": 0.6638367945064158, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 8874 + }, + { + "epoch": 0.08875, + "grad_norm": 0.6649935035212935, + "learning_rate": 0.003, + "loss": 4.057, + "step": 8875 + }, + { + "epoch": 0.08876, + "grad_norm": 0.6415966048832616, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 8876 + }, + { + "epoch": 0.08877, + "grad_norm": 0.6213949251369892, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 8877 + }, + { + "epoch": 0.08878, + "grad_norm": 0.6454540518495, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 8878 + }, + { + "epoch": 0.08879, + "grad_norm": 0.6078033590098875, + "learning_rate": 0.003, + "loss": 4.026, + "step": 8879 + }, + { + "epoch": 0.0888, + "grad_norm": 0.5827934224593164, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 8880 + }, + { + "epoch": 0.08881, + "grad_norm": 0.5959797840758749, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 8881 + }, + { + "epoch": 0.08882, + "grad_norm": 0.6610875035711973, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 8882 + }, + { + "epoch": 0.08883, + "grad_norm": 0.7066705602539173, + "learning_rate": 0.003, + "loss": 4.0811, + "step": 8883 + }, + { + "epoch": 0.08884, + "grad_norm": 0.6839508865379839, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 8884 + }, + { + "epoch": 0.08885, + "grad_norm": 0.6768243138408803, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 8885 + }, + { + "epoch": 0.08886, + "grad_norm": 0.6476294725104514, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 8886 + }, + { + "epoch": 0.08887, + "grad_norm": 0.7426502082995441, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 8887 + }, + { + "epoch": 0.08888, + "grad_norm": 0.8218257568834777, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 8888 + }, + { + "epoch": 0.08889, + "grad_norm": 0.768174391071648, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 8889 + }, + { + "epoch": 0.0889, + "grad_norm": 0.7184556927046435, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 8890 + }, + { + "epoch": 0.08891, + "grad_norm": 0.6160031763724609, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 8891 + }, + { + "epoch": 0.08892, + "grad_norm": 0.5960597608130939, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8892 + }, + { + "epoch": 0.08893, + "grad_norm": 0.5232513991838169, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 8893 + }, + { + "epoch": 0.08894, + "grad_norm": 0.5079818190740559, + "learning_rate": 0.003, + "loss": 4.007, + "step": 8894 + }, + { + "epoch": 0.08895, + "grad_norm": 0.5236615514547535, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 8895 + }, + { + "epoch": 0.08896, + "grad_norm": 0.5806948288524113, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 8896 + }, + { + "epoch": 0.08897, + "grad_norm": 0.7316975089282107, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 8897 + }, + { + "epoch": 0.08898, + "grad_norm": 0.8270556200554973, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 8898 + }, + { + "epoch": 0.08899, + "grad_norm": 0.726586147789268, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 8899 + }, + { + "epoch": 0.089, + "grad_norm": 0.536671294389095, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 8900 + }, + { + "epoch": 0.08901, + "grad_norm": 0.5540385787462229, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 8901 + }, + { + "epoch": 0.08902, + "grad_norm": 0.5740470918452585, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 8902 + }, + { + "epoch": 0.08903, + "grad_norm": 0.5852072547168969, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 8903 + }, + { + "epoch": 0.08904, + "grad_norm": 0.636557821348245, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 8904 + }, + { + "epoch": 0.08905, + "grad_norm": 0.6450995006379565, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8905 + }, + { + "epoch": 0.08906, + "grad_norm": 0.6573674695433441, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8906 + }, + { + "epoch": 0.08907, + "grad_norm": 0.6604521639649997, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 8907 + }, + { + "epoch": 0.08908, + "grad_norm": 0.7012924928842268, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 8908 + }, + { + "epoch": 0.08909, + "grad_norm": 0.724013684940573, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 8909 + }, + { + "epoch": 0.0891, + "grad_norm": 0.9078496879027451, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 8910 + }, + { + "epoch": 0.08911, + "grad_norm": 1.1149639724159413, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 8911 + }, + { + "epoch": 0.08912, + "grad_norm": 0.8861397730568337, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 8912 + }, + { + "epoch": 0.08913, + "grad_norm": 0.8778472409494876, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 8913 + }, + { + "epoch": 0.08914, + "grad_norm": 0.9020173104968283, + "learning_rate": 0.003, + "loss": 4.0591, + "step": 8914 + }, + { + "epoch": 0.08915, + "grad_norm": 0.9696876181632237, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 8915 + }, + { + "epoch": 0.08916, + "grad_norm": 0.9365859211923084, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 8916 + }, + { + "epoch": 0.08917, + "grad_norm": 0.7958796789291209, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 8917 + }, + { + "epoch": 0.08918, + "grad_norm": 0.7720199755764814, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 8918 + }, + { + "epoch": 0.08919, + "grad_norm": 0.7274695703866249, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 8919 + }, + { + "epoch": 0.0892, + "grad_norm": 0.6117093481249115, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 8920 + }, + { + "epoch": 0.08921, + "grad_norm": 0.6626899696796903, + "learning_rate": 0.003, + "loss": 4.033, + "step": 8921 + }, + { + "epoch": 0.08922, + "grad_norm": 0.6392065519700515, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 8922 + }, + { + "epoch": 0.08923, + "grad_norm": 0.7087135379000218, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 8923 + }, + { + "epoch": 0.08924, + "grad_norm": 0.763005141547236, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 8924 + }, + { + "epoch": 0.08925, + "grad_norm": 0.7879232606481785, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 8925 + }, + { + "epoch": 0.08926, + "grad_norm": 0.8563244413456195, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 8926 + }, + { + "epoch": 0.08927, + "grad_norm": 0.8467487097022208, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 8927 + }, + { + "epoch": 0.08928, + "grad_norm": 0.6908356873677782, + "learning_rate": 0.003, + "loss": 4.034, + "step": 8928 + }, + { + "epoch": 0.08929, + "grad_norm": 0.6575659250686429, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 8929 + }, + { + "epoch": 0.0893, + "grad_norm": 0.7520756648601362, + "learning_rate": 0.003, + "loss": 4.049, + "step": 8930 + }, + { + "epoch": 0.08931, + "grad_norm": 0.714609922915027, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 8931 + }, + { + "epoch": 0.08932, + "grad_norm": 0.6769041148907706, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 8932 + }, + { + "epoch": 0.08933, + "grad_norm": 0.6571970010465689, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 8933 + }, + { + "epoch": 0.08934, + "grad_norm": 0.6169238127428907, + "learning_rate": 0.003, + "loss": 4.036, + "step": 8934 + }, + { + "epoch": 0.08935, + "grad_norm": 0.6946999692684502, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 8935 + }, + { + "epoch": 0.08936, + "grad_norm": 0.6801046646965984, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 8936 + }, + { + "epoch": 0.08937, + "grad_norm": 0.6739254018359605, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 8937 + }, + { + "epoch": 0.08938, + "grad_norm": 0.7135887714784602, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 8938 + }, + { + "epoch": 0.08939, + "grad_norm": 0.7582126575196507, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 8939 + }, + { + "epoch": 0.0894, + "grad_norm": 0.8670311694862214, + "learning_rate": 0.003, + "loss": 4.037, + "step": 8940 + }, + { + "epoch": 0.08941, + "grad_norm": 0.7938282738081877, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 8941 + }, + { + "epoch": 0.08942, + "grad_norm": 0.7363401854252062, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 8942 + }, + { + "epoch": 0.08943, + "grad_norm": 0.8329325768405801, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 8943 + }, + { + "epoch": 0.08944, + "grad_norm": 0.8074174130478008, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 8944 + }, + { + "epoch": 0.08945, + "grad_norm": 0.711914736285136, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 8945 + }, + { + "epoch": 0.08946, + "grad_norm": 0.5948678500080778, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 8946 + }, + { + "epoch": 0.08947, + "grad_norm": 0.6055798533907456, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 8947 + }, + { + "epoch": 0.08948, + "grad_norm": 0.5654216220767131, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 8948 + }, + { + "epoch": 0.08949, + "grad_norm": 0.5200651864662031, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 8949 + }, + { + "epoch": 0.0895, + "grad_norm": 0.5432576403650339, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 8950 + }, + { + "epoch": 0.08951, + "grad_norm": 0.6519932907475817, + "learning_rate": 0.003, + "loss": 4.028, + "step": 8951 + }, + { + "epoch": 0.08952, + "grad_norm": 0.7364906674913794, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 8952 + }, + { + "epoch": 0.08953, + "grad_norm": 0.8285030063053859, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 8953 + }, + { + "epoch": 0.08954, + "grad_norm": 0.8940609439035224, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 8954 + }, + { + "epoch": 0.08955, + "grad_norm": 0.893918944167374, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 8955 + }, + { + "epoch": 0.08956, + "grad_norm": 0.8408128746735558, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 8956 + }, + { + "epoch": 0.08957, + "grad_norm": 0.7865192409013115, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 8957 + }, + { + "epoch": 0.08958, + "grad_norm": 0.788220360494159, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 8958 + }, + { + "epoch": 0.08959, + "grad_norm": 0.6680924162210355, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 8959 + }, + { + "epoch": 0.0896, + "grad_norm": 0.6391318097225066, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 8960 + }, + { + "epoch": 0.08961, + "grad_norm": 0.6164060504929627, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 8961 + }, + { + "epoch": 0.08962, + "grad_norm": 0.6386265964734995, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 8962 + }, + { + "epoch": 0.08963, + "grad_norm": 0.683609546663842, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 8963 + }, + { + "epoch": 0.08964, + "grad_norm": 0.626718375011794, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 8964 + }, + { + "epoch": 0.08965, + "grad_norm": 0.6265943211060453, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 8965 + }, + { + "epoch": 0.08966, + "grad_norm": 0.5785824933671037, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 8966 + }, + { + "epoch": 0.08967, + "grad_norm": 0.6138114472079763, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 8967 + }, + { + "epoch": 0.08968, + "grad_norm": 0.6944351608739829, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 8968 + }, + { + "epoch": 0.08969, + "grad_norm": 0.6445061431476451, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 8969 + }, + { + "epoch": 0.0897, + "grad_norm": 0.5818205269033787, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 8970 + }, + { + "epoch": 0.08971, + "grad_norm": 0.51655994894885, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 8971 + }, + { + "epoch": 0.08972, + "grad_norm": 0.5342011141304492, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 8972 + }, + { + "epoch": 0.08973, + "grad_norm": 0.5229812246606285, + "learning_rate": 0.003, + "loss": 4.05, + "step": 8973 + }, + { + "epoch": 0.08974, + "grad_norm": 0.5224849954571419, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 8974 + }, + { + "epoch": 0.08975, + "grad_norm": 0.6074476549507635, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 8975 + }, + { + "epoch": 0.08976, + "grad_norm": 0.6049886225090529, + "learning_rate": 0.003, + "loss": 4.039, + "step": 8976 + }, + { + "epoch": 0.08977, + "grad_norm": 0.5900543517039476, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 8977 + }, + { + "epoch": 0.08978, + "grad_norm": 0.6020639261246736, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 8978 + }, + { + "epoch": 0.08979, + "grad_norm": 0.6901889510594259, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 8979 + }, + { + "epoch": 0.0898, + "grad_norm": 0.6939145231640582, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 8980 + }, + { + "epoch": 0.08981, + "grad_norm": 0.7613294206268486, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 8981 + }, + { + "epoch": 0.08982, + "grad_norm": 1.0953891603303096, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 8982 + }, + { + "epoch": 0.08983, + "grad_norm": 1.199422908031157, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 8983 + }, + { + "epoch": 0.08984, + "grad_norm": 0.7442938192047172, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 8984 + }, + { + "epoch": 0.08985, + "grad_norm": 0.7111867917843738, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 8985 + }, + { + "epoch": 0.08986, + "grad_norm": 0.8491675767013918, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 8986 + }, + { + "epoch": 0.08987, + "grad_norm": 0.8884630459977397, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 8987 + }, + { + "epoch": 0.08988, + "grad_norm": 0.8788395337564905, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 8988 + }, + { + "epoch": 0.08989, + "grad_norm": 0.8125293617232412, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 8989 + }, + { + "epoch": 0.0899, + "grad_norm": 0.8172963488164683, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 8990 + }, + { + "epoch": 0.08991, + "grad_norm": 0.8515303493652278, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 8991 + }, + { + "epoch": 0.08992, + "grad_norm": 0.8536686574833295, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 8992 + }, + { + "epoch": 0.08993, + "grad_norm": 0.7569955292071441, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 8993 + }, + { + "epoch": 0.08994, + "grad_norm": 0.7762826222285091, + "learning_rate": 0.003, + "loss": 4.043, + "step": 8994 + }, + { + "epoch": 0.08995, + "grad_norm": 0.7082414812968141, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 8995 + }, + { + "epoch": 0.08996, + "grad_norm": 0.665491785556958, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 8996 + }, + { + "epoch": 0.08997, + "grad_norm": 0.6155453176848069, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 8997 + }, + { + "epoch": 0.08998, + "grad_norm": 0.5503880748294181, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 8998 + }, + { + "epoch": 0.08999, + "grad_norm": 0.5467524177530326, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 8999 + }, + { + "epoch": 0.09, + "grad_norm": 0.5280607549781544, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9000 + }, + { + "epoch": 0.09001, + "grad_norm": 0.5566617543233352, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9001 + }, + { + "epoch": 0.09002, + "grad_norm": 0.6016014139288688, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 9002 + }, + { + "epoch": 0.09003, + "grad_norm": 0.6374495326392571, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 9003 + }, + { + "epoch": 0.09004, + "grad_norm": 0.607893132429663, + "learning_rate": 0.003, + "loss": 4.027, + "step": 9004 + }, + { + "epoch": 0.09005, + "grad_norm": 0.6587021105572599, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 9005 + }, + { + "epoch": 0.09006, + "grad_norm": 0.7813636736470209, + "learning_rate": 0.003, + "loss": 4.03, + "step": 9006 + }, + { + "epoch": 0.09007, + "grad_norm": 0.7879091487692228, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 9007 + }, + { + "epoch": 0.09008, + "grad_norm": 0.7518176558957639, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 9008 + }, + { + "epoch": 0.09009, + "grad_norm": 0.6843645693666937, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 9009 + }, + { + "epoch": 0.0901, + "grad_norm": 0.7600333276144022, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 9010 + }, + { + "epoch": 0.09011, + "grad_norm": 0.7928314222091741, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9011 + }, + { + "epoch": 0.09012, + "grad_norm": 0.7701192754922268, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 9012 + }, + { + "epoch": 0.09013, + "grad_norm": 0.6946995419055305, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 9013 + }, + { + "epoch": 0.09014, + "grad_norm": 0.6860895854511582, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 9014 + }, + { + "epoch": 0.09015, + "grad_norm": 0.6528588956465784, + "learning_rate": 0.003, + "loss": 4.072, + "step": 9015 + }, + { + "epoch": 0.09016, + "grad_norm": 0.7118824258196399, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 9016 + }, + { + "epoch": 0.09017, + "grad_norm": 0.8085863372170363, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 9017 + }, + { + "epoch": 0.09018, + "grad_norm": 0.7589315494377418, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 9018 + }, + { + "epoch": 0.09019, + "grad_norm": 0.6874943485937518, + "learning_rate": 0.003, + "loss": 4.019, + "step": 9019 + }, + { + "epoch": 0.0902, + "grad_norm": 0.738278238495712, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 9020 + }, + { + "epoch": 0.09021, + "grad_norm": 0.7003521377569231, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 9021 + }, + { + "epoch": 0.09022, + "grad_norm": 0.6580308514496681, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 9022 + }, + { + "epoch": 0.09023, + "grad_norm": 0.6466748998588814, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 9023 + }, + { + "epoch": 0.09024, + "grad_norm": 0.630050453243334, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 9024 + }, + { + "epoch": 0.09025, + "grad_norm": 0.5583249769495601, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 9025 + }, + { + "epoch": 0.09026, + "grad_norm": 0.5150032086147774, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 9026 + }, + { + "epoch": 0.09027, + "grad_norm": 0.5350641697362482, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 9027 + }, + { + "epoch": 0.09028, + "grad_norm": 0.5023862009082889, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 9028 + }, + { + "epoch": 0.09029, + "grad_norm": 0.553557426217491, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 9029 + }, + { + "epoch": 0.0903, + "grad_norm": 0.5619955855048371, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 9030 + }, + { + "epoch": 0.09031, + "grad_norm": 0.5694477129435391, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 9031 + }, + { + "epoch": 0.09032, + "grad_norm": 0.7405404973615698, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 9032 + }, + { + "epoch": 0.09033, + "grad_norm": 0.8817277238007037, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 9033 + }, + { + "epoch": 0.09034, + "grad_norm": 1.0329524752697614, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 9034 + }, + { + "epoch": 0.09035, + "grad_norm": 0.9192125879978444, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 9035 + }, + { + "epoch": 0.09036, + "grad_norm": 0.7438625201523132, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 9036 + }, + { + "epoch": 0.09037, + "grad_norm": 0.729358227860075, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 9037 + }, + { + "epoch": 0.09038, + "grad_norm": 0.7422015486936542, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 9038 + }, + { + "epoch": 0.09039, + "grad_norm": 0.6944574812200661, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 9039 + }, + { + "epoch": 0.0904, + "grad_norm": 0.7047670538825522, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 9040 + }, + { + "epoch": 0.09041, + "grad_norm": 0.6750282421557959, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 9041 + }, + { + "epoch": 0.09042, + "grad_norm": 0.575314486191444, + "learning_rate": 0.003, + "loss": 4.014, + "step": 9042 + }, + { + "epoch": 0.09043, + "grad_norm": 0.5673878241537689, + "learning_rate": 0.003, + "loss": 4.018, + "step": 9043 + }, + { + "epoch": 0.09044, + "grad_norm": 0.651542071620226, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 9044 + }, + { + "epoch": 0.09045, + "grad_norm": 0.6354822955533254, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 9045 + }, + { + "epoch": 0.09046, + "grad_norm": 0.729930698095081, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 9046 + }, + { + "epoch": 0.09047, + "grad_norm": 0.7600654037909901, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 9047 + }, + { + "epoch": 0.09048, + "grad_norm": 0.7454751055470238, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 9048 + }, + { + "epoch": 0.09049, + "grad_norm": 0.6500339166176137, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 9049 + }, + { + "epoch": 0.0905, + "grad_norm": 0.5567585414412743, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 9050 + }, + { + "epoch": 0.09051, + "grad_norm": 0.5653371933711281, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 9051 + }, + { + "epoch": 0.09052, + "grad_norm": 0.5053698417959593, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 9052 + }, + { + "epoch": 0.09053, + "grad_norm": 0.5090662232945156, + "learning_rate": 0.003, + "loss": 4.011, + "step": 9053 + }, + { + "epoch": 0.09054, + "grad_norm": 0.5586028301490107, + "learning_rate": 0.003, + "loss": 4.014, + "step": 9054 + }, + { + "epoch": 0.09055, + "grad_norm": 0.6382778085468224, + "learning_rate": 0.003, + "loss": 4.007, + "step": 9055 + }, + { + "epoch": 0.09056, + "grad_norm": 0.8382937393931309, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 9056 + }, + { + "epoch": 0.09057, + "grad_norm": 0.9813004504209255, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 9057 + }, + { + "epoch": 0.09058, + "grad_norm": 0.9416827814626626, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 9058 + }, + { + "epoch": 0.09059, + "grad_norm": 0.823389232363663, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 9059 + }, + { + "epoch": 0.0906, + "grad_norm": 0.7746850052755953, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 9060 + }, + { + "epoch": 0.09061, + "grad_norm": 0.7880638606426179, + "learning_rate": 0.003, + "loss": 4.046, + "step": 9061 + }, + { + "epoch": 0.09062, + "grad_norm": 0.7159682446427506, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 9062 + }, + { + "epoch": 0.09063, + "grad_norm": 0.747103877472631, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 9063 + }, + { + "epoch": 0.09064, + "grad_norm": 0.7445562099095523, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 9064 + }, + { + "epoch": 0.09065, + "grad_norm": 0.6248064218898473, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 9065 + }, + { + "epoch": 0.09066, + "grad_norm": 0.6128203219080848, + "learning_rate": 0.003, + "loss": 4.034, + "step": 9066 + }, + { + "epoch": 0.09067, + "grad_norm": 0.6966333618675381, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 9067 + }, + { + "epoch": 0.09068, + "grad_norm": 0.7652371046202191, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 9068 + }, + { + "epoch": 0.09069, + "grad_norm": 0.8454640972695644, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 9069 + }, + { + "epoch": 0.0907, + "grad_norm": 0.8628019440720281, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 9070 + }, + { + "epoch": 0.09071, + "grad_norm": 0.844209654854552, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 9071 + }, + { + "epoch": 0.09072, + "grad_norm": 0.7942976322947239, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 9072 + }, + { + "epoch": 0.09073, + "grad_norm": 0.700494646849296, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 9073 + }, + { + "epoch": 0.09074, + "grad_norm": 0.6520056688541473, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 9074 + }, + { + "epoch": 0.09075, + "grad_norm": 0.7393971971104093, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 9075 + }, + { + "epoch": 0.09076, + "grad_norm": 0.8161443081532538, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 9076 + }, + { + "epoch": 0.09077, + "grad_norm": 0.8307277716405194, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 9077 + }, + { + "epoch": 0.09078, + "grad_norm": 0.7723914056727142, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 9078 + }, + { + "epoch": 0.09079, + "grad_norm": 0.6562328782600603, + "learning_rate": 0.003, + "loss": 4.046, + "step": 9079 + }, + { + "epoch": 0.0908, + "grad_norm": 0.6534810898047567, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 9080 + }, + { + "epoch": 0.09081, + "grad_norm": 0.6836055393068904, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 9081 + }, + { + "epoch": 0.09082, + "grad_norm": 0.7696641926497785, + "learning_rate": 0.003, + "loss": 4.028, + "step": 9082 + }, + { + "epoch": 0.09083, + "grad_norm": 0.7325355846754362, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 9083 + }, + { + "epoch": 0.09084, + "grad_norm": 0.7675497961449819, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 9084 + }, + { + "epoch": 0.09085, + "grad_norm": 0.7970172070447584, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 9085 + }, + { + "epoch": 0.09086, + "grad_norm": 0.6828274592801021, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 9086 + }, + { + "epoch": 0.09087, + "grad_norm": 0.6330167159273872, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 9087 + }, + { + "epoch": 0.09088, + "grad_norm": 0.7314488098298657, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 9088 + }, + { + "epoch": 0.09089, + "grad_norm": 0.7372851114512272, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 9089 + }, + { + "epoch": 0.0909, + "grad_norm": 0.6574426824581374, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 9090 + }, + { + "epoch": 0.09091, + "grad_norm": 0.6192474890913011, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 9091 + }, + { + "epoch": 0.09092, + "grad_norm": 0.6149123728007423, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 9092 + }, + { + "epoch": 0.09093, + "grad_norm": 0.6372660591070994, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 9093 + }, + { + "epoch": 0.09094, + "grad_norm": 0.5740758806569958, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 9094 + }, + { + "epoch": 0.09095, + "grad_norm": 0.5198291725664834, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 9095 + }, + { + "epoch": 0.09096, + "grad_norm": 0.5768077973883934, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 9096 + }, + { + "epoch": 0.09097, + "grad_norm": 0.5891058739999352, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 9097 + }, + { + "epoch": 0.09098, + "grad_norm": 0.6152208712701561, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 9098 + }, + { + "epoch": 0.09099, + "grad_norm": 0.7643215036585383, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 9099 + }, + { + "epoch": 0.091, + "grad_norm": 1.0103571400427667, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 9100 + }, + { + "epoch": 0.09101, + "grad_norm": 1.1682765426102981, + "learning_rate": 0.003, + "loss": 4.049, + "step": 9101 + }, + { + "epoch": 0.09102, + "grad_norm": 0.6286982857088396, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 9102 + }, + { + "epoch": 0.09103, + "grad_norm": 0.7612261608004138, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 9103 + }, + { + "epoch": 0.09104, + "grad_norm": 0.805018108408597, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9104 + }, + { + "epoch": 0.09105, + "grad_norm": 0.7775769172720655, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 9105 + }, + { + "epoch": 0.09106, + "grad_norm": 0.6110610829077867, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 9106 + }, + { + "epoch": 0.09107, + "grad_norm": 0.6523361707915806, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 9107 + }, + { + "epoch": 0.09108, + "grad_norm": 0.6488355583256343, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 9108 + }, + { + "epoch": 0.09109, + "grad_norm": 0.6405119656507225, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 9109 + }, + { + "epoch": 0.0911, + "grad_norm": 0.5848345926795614, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 9110 + }, + { + "epoch": 0.09111, + "grad_norm": 0.5736726356861014, + "learning_rate": 0.003, + "loss": 4.043, + "step": 9111 + }, + { + "epoch": 0.09112, + "grad_norm": 0.658184334745208, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 9112 + }, + { + "epoch": 0.09113, + "grad_norm": 0.7139549612481556, + "learning_rate": 0.003, + "loss": 4.004, + "step": 9113 + }, + { + "epoch": 0.09114, + "grad_norm": 0.6725790798713978, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 9114 + }, + { + "epoch": 0.09115, + "grad_norm": 0.7224350424299525, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 9115 + }, + { + "epoch": 0.09116, + "grad_norm": 0.7487347598243942, + "learning_rate": 0.003, + "loss": 4.073, + "step": 9116 + }, + { + "epoch": 0.09117, + "grad_norm": 0.8601638973618313, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 9117 + }, + { + "epoch": 0.09118, + "grad_norm": 0.869611899950819, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9118 + }, + { + "epoch": 0.09119, + "grad_norm": 0.7396523943541158, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 9119 + }, + { + "epoch": 0.0912, + "grad_norm": 0.8039767175009789, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 9120 + }, + { + "epoch": 0.09121, + "grad_norm": 0.7238367573259703, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 9121 + }, + { + "epoch": 0.09122, + "grad_norm": 0.6532364633774433, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 9122 + }, + { + "epoch": 0.09123, + "grad_norm": 0.6867983323148129, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 9123 + }, + { + "epoch": 0.09124, + "grad_norm": 0.6541067190463723, + "learning_rate": 0.003, + "loss": 4.03, + "step": 9124 + }, + { + "epoch": 0.09125, + "grad_norm": 0.629549312654314, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9125 + }, + { + "epoch": 0.09126, + "grad_norm": 0.6215390688826673, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 9126 + }, + { + "epoch": 0.09127, + "grad_norm": 0.7182287983106544, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 9127 + }, + { + "epoch": 0.09128, + "grad_norm": 0.7456623657291204, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 9128 + }, + { + "epoch": 0.09129, + "grad_norm": 0.7361787433674145, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 9129 + }, + { + "epoch": 0.0913, + "grad_norm": 0.6436106942064022, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 9130 + }, + { + "epoch": 0.09131, + "grad_norm": 0.7027093648349537, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 9131 + }, + { + "epoch": 0.09132, + "grad_norm": 0.8173234940419386, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 9132 + }, + { + "epoch": 0.09133, + "grad_norm": 0.8604950067373429, + "learning_rate": 0.003, + "loss": 4.036, + "step": 9133 + }, + { + "epoch": 0.09134, + "grad_norm": 0.821106936815988, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 9134 + }, + { + "epoch": 0.09135, + "grad_norm": 0.7380305022354678, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 9135 + }, + { + "epoch": 0.09136, + "grad_norm": 0.7301669242062018, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 9136 + }, + { + "epoch": 0.09137, + "grad_norm": 0.6601099783558254, + "learning_rate": 0.003, + "loss": 4.0636, + "step": 9137 + }, + { + "epoch": 0.09138, + "grad_norm": 0.7091143008205792, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 9138 + }, + { + "epoch": 0.09139, + "grad_norm": 0.6635408640964436, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 9139 + }, + { + "epoch": 0.0914, + "grad_norm": 0.6269357450052224, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 9140 + }, + { + "epoch": 0.09141, + "grad_norm": 0.4939988342925455, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 9141 + }, + { + "epoch": 0.09142, + "grad_norm": 0.47666523854052284, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 9142 + }, + { + "epoch": 0.09143, + "grad_norm": 0.4685826521014284, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 9143 + }, + { + "epoch": 0.09144, + "grad_norm": 0.45472685093755694, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 9144 + }, + { + "epoch": 0.09145, + "grad_norm": 0.5846603638369533, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 9145 + }, + { + "epoch": 0.09146, + "grad_norm": 0.668508106779649, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 9146 + }, + { + "epoch": 0.09147, + "grad_norm": 0.8152269301372043, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 9147 + }, + { + "epoch": 0.09148, + "grad_norm": 0.8342442125816247, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 9148 + }, + { + "epoch": 0.09149, + "grad_norm": 0.7711893300038962, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 9149 + }, + { + "epoch": 0.0915, + "grad_norm": 0.8455075111778317, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 9150 + }, + { + "epoch": 0.09151, + "grad_norm": 1.0596380832648387, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 9151 + }, + { + "epoch": 0.09152, + "grad_norm": 1.1231561697598054, + "learning_rate": 0.003, + "loss": 4.06, + "step": 9152 + }, + { + "epoch": 0.09153, + "grad_norm": 0.8525533475370635, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 9153 + }, + { + "epoch": 0.09154, + "grad_norm": 0.749612915983972, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 9154 + }, + { + "epoch": 0.09155, + "grad_norm": 0.6458707044584489, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9155 + }, + { + "epoch": 0.09156, + "grad_norm": 0.6371406845282189, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 9156 + }, + { + "epoch": 0.09157, + "grad_norm": 0.7573294662207288, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 9157 + }, + { + "epoch": 0.09158, + "grad_norm": 0.9612435458855998, + "learning_rate": 0.003, + "loss": 4.038, + "step": 9158 + }, + { + "epoch": 0.09159, + "grad_norm": 1.0598195835920619, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 9159 + }, + { + "epoch": 0.0916, + "grad_norm": 0.7270100802642413, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 9160 + }, + { + "epoch": 0.09161, + "grad_norm": 0.7490173635410152, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 9161 + }, + { + "epoch": 0.09162, + "grad_norm": 0.8107162388962985, + "learning_rate": 0.003, + "loss": 4.0687, + "step": 9162 + }, + { + "epoch": 0.09163, + "grad_norm": 0.6811918373454804, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 9163 + }, + { + "epoch": 0.09164, + "grad_norm": 0.6251874592784055, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 9164 + }, + { + "epoch": 0.09165, + "grad_norm": 0.6799951720158808, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 9165 + }, + { + "epoch": 0.09166, + "grad_norm": 0.8250092680320936, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 9166 + }, + { + "epoch": 0.09167, + "grad_norm": 0.857786758892273, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 9167 + }, + { + "epoch": 0.09168, + "grad_norm": 0.7678998740803434, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 9168 + }, + { + "epoch": 0.09169, + "grad_norm": 0.626335282566508, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 9169 + }, + { + "epoch": 0.0917, + "grad_norm": 0.6724767008366537, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 9170 + }, + { + "epoch": 0.09171, + "grad_norm": 0.7051850503746883, + "learning_rate": 0.003, + "loss": 4.066, + "step": 9171 + }, + { + "epoch": 0.09172, + "grad_norm": 0.6987754393565627, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 9172 + }, + { + "epoch": 0.09173, + "grad_norm": 0.6017927333691439, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 9173 + }, + { + "epoch": 0.09174, + "grad_norm": 0.5999580785616139, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 9174 + }, + { + "epoch": 0.09175, + "grad_norm": 0.7451487918094334, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 9175 + }, + { + "epoch": 0.09176, + "grad_norm": 0.7949048097411222, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 9176 + }, + { + "epoch": 0.09177, + "grad_norm": 0.8609692318090963, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 9177 + }, + { + "epoch": 0.09178, + "grad_norm": 0.7847869480879119, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 9178 + }, + { + "epoch": 0.09179, + "grad_norm": 0.6208320669211154, + "learning_rate": 0.003, + "loss": 4.037, + "step": 9179 + }, + { + "epoch": 0.0918, + "grad_norm": 0.5655322503679457, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 9180 + }, + { + "epoch": 0.09181, + "grad_norm": 0.4669014382152296, + "learning_rate": 0.003, + "loss": 4.022, + "step": 9181 + }, + { + "epoch": 0.09182, + "grad_norm": 0.46181408236257177, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 9182 + }, + { + "epoch": 0.09183, + "grad_norm": 0.4957633349578605, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 9183 + }, + { + "epoch": 0.09184, + "grad_norm": 0.5171117007966203, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 9184 + }, + { + "epoch": 0.09185, + "grad_norm": 0.5973660101376455, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 9185 + }, + { + "epoch": 0.09186, + "grad_norm": 0.6517128487657147, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 9186 + }, + { + "epoch": 0.09187, + "grad_norm": 0.6060457771524193, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 9187 + }, + { + "epoch": 0.09188, + "grad_norm": 0.7273937218757498, + "learning_rate": 0.003, + "loss": 4.038, + "step": 9188 + }, + { + "epoch": 0.09189, + "grad_norm": 0.856935863872665, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 9189 + }, + { + "epoch": 0.0919, + "grad_norm": 0.9400920167482492, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 9190 + }, + { + "epoch": 0.09191, + "grad_norm": 0.8849195918717125, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 9191 + }, + { + "epoch": 0.09192, + "grad_norm": 0.7307878590861009, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 9192 + }, + { + "epoch": 0.09193, + "grad_norm": 0.6839356425517885, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 9193 + }, + { + "epoch": 0.09194, + "grad_norm": 0.6098269241585833, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 9194 + }, + { + "epoch": 0.09195, + "grad_norm": 0.6440122922891024, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9195 + }, + { + "epoch": 0.09196, + "grad_norm": 0.5706640279690692, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 9196 + }, + { + "epoch": 0.09197, + "grad_norm": 0.6554002746458695, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 9197 + }, + { + "epoch": 0.09198, + "grad_norm": 0.7698716735676372, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 9198 + }, + { + "epoch": 0.09199, + "grad_norm": 0.7573050821373973, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 9199 + }, + { + "epoch": 0.092, + "grad_norm": 0.7206692134319719, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 9200 + }, + { + "epoch": 0.09201, + "grad_norm": 0.8116301922170408, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 9201 + }, + { + "epoch": 0.09202, + "grad_norm": 0.8389534336640375, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 9202 + }, + { + "epoch": 0.09203, + "grad_norm": 0.8383383674297936, + "learning_rate": 0.003, + "loss": 4.047, + "step": 9203 + }, + { + "epoch": 0.09204, + "grad_norm": 0.7918671773523751, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 9204 + }, + { + "epoch": 0.09205, + "grad_norm": 0.7452989297865678, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 9205 + }, + { + "epoch": 0.09206, + "grad_norm": 0.6912525498059005, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 9206 + }, + { + "epoch": 0.09207, + "grad_norm": 0.7376946851724611, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 9207 + }, + { + "epoch": 0.09208, + "grad_norm": 0.7352197120824172, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 9208 + }, + { + "epoch": 0.09209, + "grad_norm": 0.7435684733408947, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 9209 + }, + { + "epoch": 0.0921, + "grad_norm": 0.788426512974374, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 9210 + }, + { + "epoch": 0.09211, + "grad_norm": 0.853263816907912, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 9211 + }, + { + "epoch": 0.09212, + "grad_norm": 0.7437839455936851, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 9212 + }, + { + "epoch": 0.09213, + "grad_norm": 0.7530389082892212, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 9213 + }, + { + "epoch": 0.09214, + "grad_norm": 0.834763036809194, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 9214 + }, + { + "epoch": 0.09215, + "grad_norm": 0.741420863122177, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 9215 + }, + { + "epoch": 0.09216, + "grad_norm": 0.7494819529610638, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 9216 + }, + { + "epoch": 0.09217, + "grad_norm": 0.7716334223815386, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 9217 + }, + { + "epoch": 0.09218, + "grad_norm": 0.6945155086656055, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 9218 + }, + { + "epoch": 0.09219, + "grad_norm": 0.5636797184893982, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 9219 + }, + { + "epoch": 0.0922, + "grad_norm": 0.5220016914898918, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 9220 + }, + { + "epoch": 0.09221, + "grad_norm": 0.579637326980459, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 9221 + }, + { + "epoch": 0.09222, + "grad_norm": 0.668226412365656, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 9222 + }, + { + "epoch": 0.09223, + "grad_norm": 0.6356004998006967, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 9223 + }, + { + "epoch": 0.09224, + "grad_norm": 0.6813816183509216, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 9224 + }, + { + "epoch": 0.09225, + "grad_norm": 0.6815960818240244, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 9225 + }, + { + "epoch": 0.09226, + "grad_norm": 0.6967923347799875, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 9226 + }, + { + "epoch": 0.09227, + "grad_norm": 0.6799833699775996, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 9227 + }, + { + "epoch": 0.09228, + "grad_norm": 0.6779572459378999, + "learning_rate": 0.003, + "loss": 4.032, + "step": 9228 + }, + { + "epoch": 0.09229, + "grad_norm": 0.6386333673919082, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 9229 + }, + { + "epoch": 0.0923, + "grad_norm": 0.8361713378382143, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9230 + }, + { + "epoch": 0.09231, + "grad_norm": 0.94646464729932, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 9231 + }, + { + "epoch": 0.09232, + "grad_norm": 0.9171882811026472, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 9232 + }, + { + "epoch": 0.09233, + "grad_norm": 0.6994027349165354, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 9233 + }, + { + "epoch": 0.09234, + "grad_norm": 0.65192723883741, + "learning_rate": 0.003, + "loss": 4.0598, + "step": 9234 + }, + { + "epoch": 0.09235, + "grad_norm": 0.7197501297297813, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 9235 + }, + { + "epoch": 0.09236, + "grad_norm": 0.7058609202647043, + "learning_rate": 0.003, + "loss": 4.0686, + "step": 9236 + }, + { + "epoch": 0.09237, + "grad_norm": 0.7070637418397738, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 9237 + }, + { + "epoch": 0.09238, + "grad_norm": 0.6931756292417025, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 9238 + }, + { + "epoch": 0.09239, + "grad_norm": 0.5879565067927063, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 9239 + }, + { + "epoch": 0.0924, + "grad_norm": 0.6067123667800877, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 9240 + }, + { + "epoch": 0.09241, + "grad_norm": 0.6847070757737074, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 9241 + }, + { + "epoch": 0.09242, + "grad_norm": 0.6207988431604151, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 9242 + }, + { + "epoch": 0.09243, + "grad_norm": 0.5397755056060818, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 9243 + }, + { + "epoch": 0.09244, + "grad_norm": 0.5543713970370245, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 9244 + }, + { + "epoch": 0.09245, + "grad_norm": 0.5541034820095151, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 9245 + }, + { + "epoch": 0.09246, + "grad_norm": 0.5113424027234124, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 9246 + }, + { + "epoch": 0.09247, + "grad_norm": 0.49013506817195773, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 9247 + }, + { + "epoch": 0.09248, + "grad_norm": 0.4444317523058028, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 9248 + }, + { + "epoch": 0.09249, + "grad_norm": 0.46638806041494046, + "learning_rate": 0.003, + "loss": 4.0706, + "step": 9249 + }, + { + "epoch": 0.0925, + "grad_norm": 0.45569538278682137, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9250 + }, + { + "epoch": 0.09251, + "grad_norm": 0.47165791400542584, + "learning_rate": 0.003, + "loss": 4.039, + "step": 9251 + }, + { + "epoch": 0.09252, + "grad_norm": 0.5244970941561967, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 9252 + }, + { + "epoch": 0.09253, + "grad_norm": 0.6243978565572065, + "learning_rate": 0.003, + "loss": 4.023, + "step": 9253 + }, + { + "epoch": 0.09254, + "grad_norm": 0.7828680526572193, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 9254 + }, + { + "epoch": 0.09255, + "grad_norm": 0.9733433368483589, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 9255 + }, + { + "epoch": 0.09256, + "grad_norm": 1.1354630171824305, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9256 + }, + { + "epoch": 0.09257, + "grad_norm": 0.9016412111846891, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 9257 + }, + { + "epoch": 0.09258, + "grad_norm": 0.8617817514186579, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9258 + }, + { + "epoch": 0.09259, + "grad_norm": 0.8408272197263507, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 9259 + }, + { + "epoch": 0.0926, + "grad_norm": 0.8413085639532046, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 9260 + }, + { + "epoch": 0.09261, + "grad_norm": 0.7442192769597559, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 9261 + }, + { + "epoch": 0.09262, + "grad_norm": 0.7696031136349304, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 9262 + }, + { + "epoch": 0.09263, + "grad_norm": 0.7595924411082675, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 9263 + }, + { + "epoch": 0.09264, + "grad_norm": 0.764478000443183, + "learning_rate": 0.003, + "loss": 4.0734, + "step": 9264 + }, + { + "epoch": 0.09265, + "grad_norm": 0.7287453785372998, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 9265 + }, + { + "epoch": 0.09266, + "grad_norm": 0.6745268134133299, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 9266 + }, + { + "epoch": 0.09267, + "grad_norm": 0.6786835955096816, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 9267 + }, + { + "epoch": 0.09268, + "grad_norm": 0.7099498115196785, + "learning_rate": 0.003, + "loss": 4.026, + "step": 9268 + }, + { + "epoch": 0.09269, + "grad_norm": 0.7038180469670884, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 9269 + }, + { + "epoch": 0.0927, + "grad_norm": 0.5556807479339361, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 9270 + }, + { + "epoch": 0.09271, + "grad_norm": 0.5782159019042916, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 9271 + }, + { + "epoch": 0.09272, + "grad_norm": 0.6207049784348632, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 9272 + }, + { + "epoch": 0.09273, + "grad_norm": 0.617446280808945, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 9273 + }, + { + "epoch": 0.09274, + "grad_norm": 0.5829684564147539, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 9274 + }, + { + "epoch": 0.09275, + "grad_norm": 0.6138219660701727, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 9275 + }, + { + "epoch": 0.09276, + "grad_norm": 0.6768634749129705, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 9276 + }, + { + "epoch": 0.09277, + "grad_norm": 0.7109110581514423, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 9277 + }, + { + "epoch": 0.09278, + "grad_norm": 0.8485484921348332, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 9278 + }, + { + "epoch": 0.09279, + "grad_norm": 0.9787054619451452, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 9279 + }, + { + "epoch": 0.0928, + "grad_norm": 1.010922942407405, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 9280 + }, + { + "epoch": 0.09281, + "grad_norm": 1.0441732167854245, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 9281 + }, + { + "epoch": 0.09282, + "grad_norm": 0.9083057438814995, + "learning_rate": 0.003, + "loss": 4.0803, + "step": 9282 + }, + { + "epoch": 0.09283, + "grad_norm": 0.9479603100933025, + "learning_rate": 0.003, + "loss": 4.0744, + "step": 9283 + }, + { + "epoch": 0.09284, + "grad_norm": 0.9695543742160656, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 9284 + }, + { + "epoch": 0.09285, + "grad_norm": 0.9371692921794235, + "learning_rate": 0.003, + "loss": 4.046, + "step": 9285 + }, + { + "epoch": 0.09286, + "grad_norm": 1.043869367050299, + "learning_rate": 0.003, + "loss": 4.0713, + "step": 9286 + }, + { + "epoch": 0.09287, + "grad_norm": 0.9261111538204875, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 9287 + }, + { + "epoch": 0.09288, + "grad_norm": 0.8448315219147443, + "learning_rate": 0.003, + "loss": 4.0657, + "step": 9288 + }, + { + "epoch": 0.09289, + "grad_norm": 0.7342810924655063, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 9289 + }, + { + "epoch": 0.0929, + "grad_norm": 0.6879084821606024, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 9290 + }, + { + "epoch": 0.09291, + "grad_norm": 0.700991959566623, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 9291 + }, + { + "epoch": 0.09292, + "grad_norm": 0.7007242242787303, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 9292 + }, + { + "epoch": 0.09293, + "grad_norm": 0.5987303106277989, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 9293 + }, + { + "epoch": 0.09294, + "grad_norm": 0.5930324590889179, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 9294 + }, + { + "epoch": 0.09295, + "grad_norm": 0.5616016858285632, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 9295 + }, + { + "epoch": 0.09296, + "grad_norm": 0.526547856823013, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 9296 + }, + { + "epoch": 0.09297, + "grad_norm": 0.6303156475875816, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 9297 + }, + { + "epoch": 0.09298, + "grad_norm": 0.735402426907724, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 9298 + }, + { + "epoch": 0.09299, + "grad_norm": 0.8847970721198556, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 9299 + }, + { + "epoch": 0.093, + "grad_norm": 0.9750748873163463, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 9300 + }, + { + "epoch": 0.09301, + "grad_norm": 0.9109957189136523, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 9301 + }, + { + "epoch": 0.09302, + "grad_norm": 0.8672448417242827, + "learning_rate": 0.003, + "loss": 4.0779, + "step": 9302 + }, + { + "epoch": 0.09303, + "grad_norm": 0.7781276247503056, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 9303 + }, + { + "epoch": 0.09304, + "grad_norm": 0.7091733068002525, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 9304 + }, + { + "epoch": 0.09305, + "grad_norm": 0.6621200656030555, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9305 + }, + { + "epoch": 0.09306, + "grad_norm": 0.5836546423796407, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 9306 + }, + { + "epoch": 0.09307, + "grad_norm": 0.5579899105660596, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9307 + }, + { + "epoch": 0.09308, + "grad_norm": 0.563284521522256, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 9308 + }, + { + "epoch": 0.09309, + "grad_norm": 0.5640699283758076, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 9309 + }, + { + "epoch": 0.0931, + "grad_norm": 0.531701356867229, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 9310 + }, + { + "epoch": 0.09311, + "grad_norm": 0.5241222672316439, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 9311 + }, + { + "epoch": 0.09312, + "grad_norm": 0.5695773410710713, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 9312 + }, + { + "epoch": 0.09313, + "grad_norm": 0.5780378315380226, + "learning_rate": 0.003, + "loss": 4.0674, + "step": 9313 + }, + { + "epoch": 0.09314, + "grad_norm": 0.493026424687374, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 9314 + }, + { + "epoch": 0.09315, + "grad_norm": 0.4024459654092229, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 9315 + }, + { + "epoch": 0.09316, + "grad_norm": 0.454268407284728, + "learning_rate": 0.003, + "loss": 4.032, + "step": 9316 + }, + { + "epoch": 0.09317, + "grad_norm": 0.4065651733192461, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 9317 + }, + { + "epoch": 0.09318, + "grad_norm": 0.4585291022954056, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 9318 + }, + { + "epoch": 0.09319, + "grad_norm": 0.6166570957173472, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 9319 + }, + { + "epoch": 0.0932, + "grad_norm": 0.9526892304522626, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 9320 + }, + { + "epoch": 0.09321, + "grad_norm": 1.158747517662686, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 9321 + }, + { + "epoch": 0.09322, + "grad_norm": 0.8017466063632835, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 9322 + }, + { + "epoch": 0.09323, + "grad_norm": 0.730774037689625, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 9323 + }, + { + "epoch": 0.09324, + "grad_norm": 0.8286980875890013, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 9324 + }, + { + "epoch": 0.09325, + "grad_norm": 0.7786714166803139, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 9325 + }, + { + "epoch": 0.09326, + "grad_norm": 0.7837045858696695, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 9326 + }, + { + "epoch": 0.09327, + "grad_norm": 0.7832736334932466, + "learning_rate": 0.003, + "loss": 4.038, + "step": 9327 + }, + { + "epoch": 0.09328, + "grad_norm": 0.7524825697716142, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 9328 + }, + { + "epoch": 0.09329, + "grad_norm": 0.5780872736920424, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 9329 + }, + { + "epoch": 0.0933, + "grad_norm": 0.5721913513567357, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 9330 + }, + { + "epoch": 0.09331, + "grad_norm": 0.508306364702041, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 9331 + }, + { + "epoch": 0.09332, + "grad_norm": 0.5107456816259918, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 9332 + }, + { + "epoch": 0.09333, + "grad_norm": 0.6064773880958333, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 9333 + }, + { + "epoch": 0.09334, + "grad_norm": 0.7201477121805577, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 9334 + }, + { + "epoch": 0.09335, + "grad_norm": 0.7373815200477077, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 9335 + }, + { + "epoch": 0.09336, + "grad_norm": 0.6390726036349523, + "learning_rate": 0.003, + "loss": 4.019, + "step": 9336 + }, + { + "epoch": 0.09337, + "grad_norm": 0.6296648722494202, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 9337 + }, + { + "epoch": 0.09338, + "grad_norm": 0.5786524826926605, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 9338 + }, + { + "epoch": 0.09339, + "grad_norm": 0.5931096032203457, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 9339 + }, + { + "epoch": 0.0934, + "grad_norm": 0.6057473702468537, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 9340 + }, + { + "epoch": 0.09341, + "grad_norm": 0.5310411247584104, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 9341 + }, + { + "epoch": 0.09342, + "grad_norm": 0.48987836523560163, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 9342 + }, + { + "epoch": 0.09343, + "grad_norm": 0.5689591956852221, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 9343 + }, + { + "epoch": 0.09344, + "grad_norm": 0.6959775488676232, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 9344 + }, + { + "epoch": 0.09345, + "grad_norm": 0.9577054931670086, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 9345 + }, + { + "epoch": 0.09346, + "grad_norm": 1.002746407399198, + "learning_rate": 0.003, + "loss": 4.0564, + "step": 9346 + }, + { + "epoch": 0.09347, + "grad_norm": 0.9404242155072631, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 9347 + }, + { + "epoch": 0.09348, + "grad_norm": 1.0644528985126545, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 9348 + }, + { + "epoch": 0.09349, + "grad_norm": 1.0378006559784825, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 9349 + }, + { + "epoch": 0.0935, + "grad_norm": 1.0477498666070477, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 9350 + }, + { + "epoch": 0.09351, + "grad_norm": 0.8257238536179369, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 9351 + }, + { + "epoch": 0.09352, + "grad_norm": 0.8006766955611013, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 9352 + }, + { + "epoch": 0.09353, + "grad_norm": 0.7774170224055247, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 9353 + }, + { + "epoch": 0.09354, + "grad_norm": 0.7394995720132359, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 9354 + }, + { + "epoch": 0.09355, + "grad_norm": 0.8156946203748874, + "learning_rate": 0.003, + "loss": 4.0694, + "step": 9355 + }, + { + "epoch": 0.09356, + "grad_norm": 0.9299331006449973, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 9356 + }, + { + "epoch": 0.09357, + "grad_norm": 1.0066372383390785, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 9357 + }, + { + "epoch": 0.09358, + "grad_norm": 0.9674997852767191, + "learning_rate": 0.003, + "loss": 4.061, + "step": 9358 + }, + { + "epoch": 0.09359, + "grad_norm": 0.9547612172139744, + "learning_rate": 0.003, + "loss": 4.0736, + "step": 9359 + }, + { + "epoch": 0.0936, + "grad_norm": 0.6996201212989205, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 9360 + }, + { + "epoch": 0.09361, + "grad_norm": 0.669386811650118, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 9361 + }, + { + "epoch": 0.09362, + "grad_norm": 0.60001175582994, + "learning_rate": 0.003, + "loss": 4.0836, + "step": 9362 + }, + { + "epoch": 0.09363, + "grad_norm": 0.6831664625017024, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 9363 + }, + { + "epoch": 0.09364, + "grad_norm": 0.6997468269442456, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 9364 + }, + { + "epoch": 0.09365, + "grad_norm": 0.6291249142394587, + "learning_rate": 0.003, + "loss": 4.0632, + "step": 9365 + }, + { + "epoch": 0.09366, + "grad_norm": 0.5058197798053995, + "learning_rate": 0.003, + "loss": 4.006, + "step": 9366 + }, + { + "epoch": 0.09367, + "grad_norm": 0.5236992326215362, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 9367 + }, + { + "epoch": 0.09368, + "grad_norm": 0.5246791634669609, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 9368 + }, + { + "epoch": 0.09369, + "grad_norm": 0.5059727505380633, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 9369 + }, + { + "epoch": 0.0937, + "grad_norm": 0.49227015879478025, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 9370 + }, + { + "epoch": 0.09371, + "grad_norm": 0.5235419974074004, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 9371 + }, + { + "epoch": 0.09372, + "grad_norm": 0.6375121051374159, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 9372 + }, + { + "epoch": 0.09373, + "grad_norm": 0.6787945149511082, + "learning_rate": 0.003, + "loss": 4.0732, + "step": 9373 + }, + { + "epoch": 0.09374, + "grad_norm": 0.7069199754704323, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 9374 + }, + { + "epoch": 0.09375, + "grad_norm": 0.6839409245028282, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 9375 + }, + { + "epoch": 0.09376, + "grad_norm": 0.6559534152133927, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 9376 + }, + { + "epoch": 0.09377, + "grad_norm": 0.6197190093955443, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 9377 + }, + { + "epoch": 0.09378, + "grad_norm": 0.6249226813835516, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 9378 + }, + { + "epoch": 0.09379, + "grad_norm": 0.6371703728162548, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 9379 + }, + { + "epoch": 0.0938, + "grad_norm": 0.6386849933156105, + "learning_rate": 0.003, + "loss": 4.01, + "step": 9380 + }, + { + "epoch": 0.09381, + "grad_norm": 0.677971753959941, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 9381 + }, + { + "epoch": 0.09382, + "grad_norm": 0.6380929014276362, + "learning_rate": 0.003, + "loss": 4.036, + "step": 9382 + }, + { + "epoch": 0.09383, + "grad_norm": 0.5732861897583412, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 9383 + }, + { + "epoch": 0.09384, + "grad_norm": 0.6080881483970928, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 9384 + }, + { + "epoch": 0.09385, + "grad_norm": 0.6003328286941136, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 9385 + }, + { + "epoch": 0.09386, + "grad_norm": 0.7010943971301368, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 9386 + }, + { + "epoch": 0.09387, + "grad_norm": 0.8200701652840898, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 9387 + }, + { + "epoch": 0.09388, + "grad_norm": 0.8680814165483955, + "learning_rate": 0.003, + "loss": 4.039, + "step": 9388 + }, + { + "epoch": 0.09389, + "grad_norm": 0.9077098353364628, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 9389 + }, + { + "epoch": 0.0939, + "grad_norm": 0.8275162633067968, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 9390 + }, + { + "epoch": 0.09391, + "grad_norm": 0.7126772787678065, + "learning_rate": 0.003, + "loss": 4.025, + "step": 9391 + }, + { + "epoch": 0.09392, + "grad_norm": 0.6804754920666622, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 9392 + }, + { + "epoch": 0.09393, + "grad_norm": 0.7432530890012637, + "learning_rate": 0.003, + "loss": 4.049, + "step": 9393 + }, + { + "epoch": 0.09394, + "grad_norm": 0.6160745128275361, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9394 + }, + { + "epoch": 0.09395, + "grad_norm": 0.5655058997408525, + "learning_rate": 0.003, + "loss": 4.033, + "step": 9395 + }, + { + "epoch": 0.09396, + "grad_norm": 0.6095924345574806, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 9396 + }, + { + "epoch": 0.09397, + "grad_norm": 0.6111772936061662, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 9397 + }, + { + "epoch": 0.09398, + "grad_norm": 0.6184628574042604, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9398 + }, + { + "epoch": 0.09399, + "grad_norm": 0.6474420473793211, + "learning_rate": 0.003, + "loss": 4.04, + "step": 9399 + }, + { + "epoch": 0.094, + "grad_norm": 0.5975657668097811, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 9400 + }, + { + "epoch": 0.09401, + "grad_norm": 0.5787286065843571, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 9401 + }, + { + "epoch": 0.09402, + "grad_norm": 0.6176418786305744, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 9402 + }, + { + "epoch": 0.09403, + "grad_norm": 0.7182313870161714, + "learning_rate": 0.003, + "loss": 4.052, + "step": 9403 + }, + { + "epoch": 0.09404, + "grad_norm": 0.6919919998467069, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 9404 + }, + { + "epoch": 0.09405, + "grad_norm": 0.6407535807287313, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 9405 + }, + { + "epoch": 0.09406, + "grad_norm": 0.6170104694112555, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 9406 + }, + { + "epoch": 0.09407, + "grad_norm": 0.587418396336813, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9407 + }, + { + "epoch": 0.09408, + "grad_norm": 0.5713679945496545, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 9408 + }, + { + "epoch": 0.09409, + "grad_norm": 0.5728848248563441, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 9409 + }, + { + "epoch": 0.0941, + "grad_norm": 0.7813893158200088, + "learning_rate": 0.003, + "loss": 4.051, + "step": 9410 + }, + { + "epoch": 0.09411, + "grad_norm": 0.917300921986231, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 9411 + }, + { + "epoch": 0.09412, + "grad_norm": 1.0155949274682867, + "learning_rate": 0.003, + "loss": 4.063, + "step": 9412 + }, + { + "epoch": 0.09413, + "grad_norm": 0.8985260638005931, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 9413 + }, + { + "epoch": 0.09414, + "grad_norm": 0.7809509025756851, + "learning_rate": 0.003, + "loss": 4.055, + "step": 9414 + }, + { + "epoch": 0.09415, + "grad_norm": 0.71084158786822, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 9415 + }, + { + "epoch": 0.09416, + "grad_norm": 0.7421801541204552, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 9416 + }, + { + "epoch": 0.09417, + "grad_norm": 0.7307972736796093, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 9417 + }, + { + "epoch": 0.09418, + "grad_norm": 0.658874877034197, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 9418 + }, + { + "epoch": 0.09419, + "grad_norm": 0.6167332837117512, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 9419 + }, + { + "epoch": 0.0942, + "grad_norm": 0.6014356633192184, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 9420 + }, + { + "epoch": 0.09421, + "grad_norm": 0.6268575285270098, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 9421 + }, + { + "epoch": 0.09422, + "grad_norm": 0.6978954629952625, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 9422 + }, + { + "epoch": 0.09423, + "grad_norm": 0.7182191384822959, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 9423 + }, + { + "epoch": 0.09424, + "grad_norm": 0.7436552893722601, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9424 + }, + { + "epoch": 0.09425, + "grad_norm": 0.8224067476201944, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 9425 + }, + { + "epoch": 0.09426, + "grad_norm": 0.9485703599702464, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 9426 + }, + { + "epoch": 0.09427, + "grad_norm": 1.1672303893220712, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 9427 + }, + { + "epoch": 0.09428, + "grad_norm": 0.8097027213223613, + "learning_rate": 0.003, + "loss": 4.064, + "step": 9428 + }, + { + "epoch": 0.09429, + "grad_norm": 0.7993295612172283, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 9429 + }, + { + "epoch": 0.0943, + "grad_norm": 0.7605014417821012, + "learning_rate": 0.003, + "loss": 4.044, + "step": 9430 + }, + { + "epoch": 0.09431, + "grad_norm": 0.7868273158782408, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 9431 + }, + { + "epoch": 0.09432, + "grad_norm": 0.8680003986125082, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 9432 + }, + { + "epoch": 0.09433, + "grad_norm": 0.8145043515352026, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 9433 + }, + { + "epoch": 0.09434, + "grad_norm": 0.6811428567164926, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 9434 + }, + { + "epoch": 0.09435, + "grad_norm": 0.6677796627524553, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 9435 + }, + { + "epoch": 0.09436, + "grad_norm": 0.6375947384417209, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 9436 + }, + { + "epoch": 0.09437, + "grad_norm": 0.6109438462376267, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 9437 + }, + { + "epoch": 0.09438, + "grad_norm": 0.6880542431470272, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 9438 + }, + { + "epoch": 0.09439, + "grad_norm": 0.713758422842352, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 9439 + }, + { + "epoch": 0.0944, + "grad_norm": 0.7358325326163485, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9440 + }, + { + "epoch": 0.09441, + "grad_norm": 0.6654118419025815, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 9441 + }, + { + "epoch": 0.09442, + "grad_norm": 0.5875294196651112, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 9442 + }, + { + "epoch": 0.09443, + "grad_norm": 0.4971994429668099, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 9443 + }, + { + "epoch": 0.09444, + "grad_norm": 0.46935549741201515, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9444 + }, + { + "epoch": 0.09445, + "grad_norm": 0.5105930505230285, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 9445 + }, + { + "epoch": 0.09446, + "grad_norm": 0.5758051638582016, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 9446 + }, + { + "epoch": 0.09447, + "grad_norm": 0.5721119651604664, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 9447 + }, + { + "epoch": 0.09448, + "grad_norm": 0.5954308875117721, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 9448 + }, + { + "epoch": 0.09449, + "grad_norm": 0.5392728529737144, + "learning_rate": 0.003, + "loss": 4.025, + "step": 9449 + }, + { + "epoch": 0.0945, + "grad_norm": 0.6355454732528022, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 9450 + }, + { + "epoch": 0.09451, + "grad_norm": 0.7211427147613977, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 9451 + }, + { + "epoch": 0.09452, + "grad_norm": 0.6986769261708674, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 9452 + }, + { + "epoch": 0.09453, + "grad_norm": 0.8078309341671657, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 9453 + }, + { + "epoch": 0.09454, + "grad_norm": 0.8348429102500077, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 9454 + }, + { + "epoch": 0.09455, + "grad_norm": 0.9028525221761462, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 9455 + }, + { + "epoch": 0.09456, + "grad_norm": 1.0174412287814851, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 9456 + }, + { + "epoch": 0.09457, + "grad_norm": 0.9584945663976725, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 9457 + }, + { + "epoch": 0.09458, + "grad_norm": 0.8084486402634247, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 9458 + }, + { + "epoch": 0.09459, + "grad_norm": 0.722980760509963, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 9459 + }, + { + "epoch": 0.0946, + "grad_norm": 0.6188966143893488, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 9460 + }, + { + "epoch": 0.09461, + "grad_norm": 0.7221170426729282, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 9461 + }, + { + "epoch": 0.09462, + "grad_norm": 0.8179321951763419, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 9462 + }, + { + "epoch": 0.09463, + "grad_norm": 0.8930048335312827, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9463 + }, + { + "epoch": 0.09464, + "grad_norm": 0.9163698726896933, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 9464 + }, + { + "epoch": 0.09465, + "grad_norm": 0.9881649235098358, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 9465 + }, + { + "epoch": 0.09466, + "grad_norm": 0.9384800273715657, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 9466 + }, + { + "epoch": 0.09467, + "grad_norm": 0.9563512188499613, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9467 + }, + { + "epoch": 0.09468, + "grad_norm": 0.9015485341644711, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 9468 + }, + { + "epoch": 0.09469, + "grad_norm": 0.7899410796741955, + "learning_rate": 0.003, + "loss": 4.025, + "step": 9469 + }, + { + "epoch": 0.0947, + "grad_norm": 0.7088641113545276, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 9470 + }, + { + "epoch": 0.09471, + "grad_norm": 0.5542385838762932, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 9471 + }, + { + "epoch": 0.09472, + "grad_norm": 0.5860312536598133, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 9472 + }, + { + "epoch": 0.09473, + "grad_norm": 0.5348692761892904, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9473 + }, + { + "epoch": 0.09474, + "grad_norm": 0.5066129941741104, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 9474 + }, + { + "epoch": 0.09475, + "grad_norm": 0.5443141204886921, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 9475 + }, + { + "epoch": 0.09476, + "grad_norm": 0.7097989824018621, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 9476 + }, + { + "epoch": 0.09477, + "grad_norm": 0.7123407967652701, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 9477 + }, + { + "epoch": 0.09478, + "grad_norm": 0.6627895469726679, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 9478 + }, + { + "epoch": 0.09479, + "grad_norm": 0.7561049732566557, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 9479 + }, + { + "epoch": 0.0948, + "grad_norm": 0.8990374556074312, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 9480 + }, + { + "epoch": 0.09481, + "grad_norm": 0.8988011061354663, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 9481 + }, + { + "epoch": 0.09482, + "grad_norm": 0.8307532528736723, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9482 + }, + { + "epoch": 0.09483, + "grad_norm": 0.7465778969717352, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 9483 + }, + { + "epoch": 0.09484, + "grad_norm": 0.670689159113372, + "learning_rate": 0.003, + "loss": 4.056, + "step": 9484 + }, + { + "epoch": 0.09485, + "grad_norm": 0.5730023814224957, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9485 + }, + { + "epoch": 0.09486, + "grad_norm": 0.6325123792025878, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 9486 + }, + { + "epoch": 0.09487, + "grad_norm": 0.669529146313543, + "learning_rate": 0.003, + "loss": 4.035, + "step": 9487 + }, + { + "epoch": 0.09488, + "grad_norm": 0.642769915551367, + "learning_rate": 0.003, + "loss": 4.0792, + "step": 9488 + }, + { + "epoch": 0.09489, + "grad_norm": 0.5156590728225835, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 9489 + }, + { + "epoch": 0.0949, + "grad_norm": 0.5402949914448095, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 9490 + }, + { + "epoch": 0.09491, + "grad_norm": 0.521455090533923, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 9491 + }, + { + "epoch": 0.09492, + "grad_norm": 0.5516108795111017, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 9492 + }, + { + "epoch": 0.09493, + "grad_norm": 0.5698287997797269, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 9493 + }, + { + "epoch": 0.09494, + "grad_norm": 0.6398370779733041, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9494 + }, + { + "epoch": 0.09495, + "grad_norm": 0.8380737997363238, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 9495 + }, + { + "epoch": 0.09496, + "grad_norm": 0.9479843549188433, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 9496 + }, + { + "epoch": 0.09497, + "grad_norm": 0.8021862408219196, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 9497 + }, + { + "epoch": 0.09498, + "grad_norm": 0.6503266533281803, + "learning_rate": 0.003, + "loss": 4.011, + "step": 9498 + }, + { + "epoch": 0.09499, + "grad_norm": 0.7650357611935428, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 9499 + }, + { + "epoch": 0.095, + "grad_norm": 0.7955907217881945, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 9500 + }, + { + "epoch": 0.09501, + "grad_norm": 0.6936871366883636, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 9501 + }, + { + "epoch": 0.09502, + "grad_norm": 0.6704711790108537, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 9502 + }, + { + "epoch": 0.09503, + "grad_norm": 0.6304424505370851, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 9503 + }, + { + "epoch": 0.09504, + "grad_norm": 0.717751567790892, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 9504 + }, + { + "epoch": 0.09505, + "grad_norm": 0.7588861814061514, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 9505 + }, + { + "epoch": 0.09506, + "grad_norm": 0.6353085438386181, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 9506 + }, + { + "epoch": 0.09507, + "grad_norm": 0.6594813256379105, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 9507 + }, + { + "epoch": 0.09508, + "grad_norm": 0.6075484416271231, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 9508 + }, + { + "epoch": 0.09509, + "grad_norm": 0.694939836973809, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 9509 + }, + { + "epoch": 0.0951, + "grad_norm": 0.6665450612797287, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 9510 + }, + { + "epoch": 0.09511, + "grad_norm": 0.6892394128230958, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 9511 + }, + { + "epoch": 0.09512, + "grad_norm": 0.6220794415709733, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 9512 + }, + { + "epoch": 0.09513, + "grad_norm": 0.6230216784823006, + "learning_rate": 0.003, + "loss": 4.015, + "step": 9513 + }, + { + "epoch": 0.09514, + "grad_norm": 0.6336509751434405, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 9514 + }, + { + "epoch": 0.09515, + "grad_norm": 0.6548021738288409, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 9515 + }, + { + "epoch": 0.09516, + "grad_norm": 0.6472732625148736, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9516 + }, + { + "epoch": 0.09517, + "grad_norm": 0.7046216510152215, + "learning_rate": 0.003, + "loss": 4.028, + "step": 9517 + }, + { + "epoch": 0.09518, + "grad_norm": 0.7416622584126377, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 9518 + }, + { + "epoch": 0.09519, + "grad_norm": 0.7725268305050113, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 9519 + }, + { + "epoch": 0.0952, + "grad_norm": 0.7400523050974745, + "learning_rate": 0.003, + "loss": 4.003, + "step": 9520 + }, + { + "epoch": 0.09521, + "grad_norm": 0.928024747149792, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 9521 + }, + { + "epoch": 0.09522, + "grad_norm": 1.0883977839395278, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 9522 + }, + { + "epoch": 0.09523, + "grad_norm": 0.7479722169208735, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 9523 + }, + { + "epoch": 0.09524, + "grad_norm": 0.7472200844195505, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 9524 + }, + { + "epoch": 0.09525, + "grad_norm": 0.7890443517847859, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 9525 + }, + { + "epoch": 0.09526, + "grad_norm": 0.9280185115921652, + "learning_rate": 0.003, + "loss": 4.04, + "step": 9526 + }, + { + "epoch": 0.09527, + "grad_norm": 1.044170856660612, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 9527 + }, + { + "epoch": 0.09528, + "grad_norm": 0.8306066657532915, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 9528 + }, + { + "epoch": 0.09529, + "grad_norm": 0.6667166131603599, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 9529 + }, + { + "epoch": 0.0953, + "grad_norm": 0.7096071140872328, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 9530 + }, + { + "epoch": 0.09531, + "grad_norm": 0.6817095157044997, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 9531 + }, + { + "epoch": 0.09532, + "grad_norm": 0.7439295037791601, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 9532 + }, + { + "epoch": 0.09533, + "grad_norm": 0.7030688354825372, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 9533 + }, + { + "epoch": 0.09534, + "grad_norm": 0.5701287437882973, + "learning_rate": 0.003, + "loss": 4.016, + "step": 9534 + }, + { + "epoch": 0.09535, + "grad_norm": 0.5078381305946676, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 9535 + }, + { + "epoch": 0.09536, + "grad_norm": 0.63897221682341, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 9536 + }, + { + "epoch": 0.09537, + "grad_norm": 0.6920302198955521, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 9537 + }, + { + "epoch": 0.09538, + "grad_norm": 0.681879186271492, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 9538 + }, + { + "epoch": 0.09539, + "grad_norm": 0.6488629663058754, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 9539 + }, + { + "epoch": 0.0954, + "grad_norm": 0.6419467576131561, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 9540 + }, + { + "epoch": 0.09541, + "grad_norm": 0.6876843275708705, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9541 + }, + { + "epoch": 0.09542, + "grad_norm": 0.7200145077236769, + "learning_rate": 0.003, + "loss": 4.011, + "step": 9542 + }, + { + "epoch": 0.09543, + "grad_norm": 0.7578411190872801, + "learning_rate": 0.003, + "loss": 4.025, + "step": 9543 + }, + { + "epoch": 0.09544, + "grad_norm": 0.7010598192659876, + "learning_rate": 0.003, + "loss": 4.018, + "step": 9544 + }, + { + "epoch": 0.09545, + "grad_norm": 0.6311272581429666, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 9545 + }, + { + "epoch": 0.09546, + "grad_norm": 0.6238038854495016, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 9546 + }, + { + "epoch": 0.09547, + "grad_norm": 0.7388749137262556, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 9547 + }, + { + "epoch": 0.09548, + "grad_norm": 0.763530094181748, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 9548 + }, + { + "epoch": 0.09549, + "grad_norm": 0.7594914119792266, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 9549 + }, + { + "epoch": 0.0955, + "grad_norm": 0.7378672227547327, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 9550 + }, + { + "epoch": 0.09551, + "grad_norm": 0.6897256697283077, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 9551 + }, + { + "epoch": 0.09552, + "grad_norm": 0.7654921276346474, + "learning_rate": 0.003, + "loss": 4.0844, + "step": 9552 + }, + { + "epoch": 0.09553, + "grad_norm": 0.7590450624156653, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 9553 + }, + { + "epoch": 0.09554, + "grad_norm": 0.668632724946176, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 9554 + }, + { + "epoch": 0.09555, + "grad_norm": 0.5912983608580621, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 9555 + }, + { + "epoch": 0.09556, + "grad_norm": 0.5565293591910623, + "learning_rate": 0.003, + "loss": 4.025, + "step": 9556 + }, + { + "epoch": 0.09557, + "grad_norm": 0.6270158844877468, + "learning_rate": 0.003, + "loss": 4.046, + "step": 9557 + }, + { + "epoch": 0.09558, + "grad_norm": 0.689653418642019, + "learning_rate": 0.003, + "loss": 4.035, + "step": 9558 + }, + { + "epoch": 0.09559, + "grad_norm": 0.8395636779689373, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 9559 + }, + { + "epoch": 0.0956, + "grad_norm": 0.8782146074783973, + "learning_rate": 0.003, + "loss": 4.031, + "step": 9560 + }, + { + "epoch": 0.09561, + "grad_norm": 0.7265691455320403, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 9561 + }, + { + "epoch": 0.09562, + "grad_norm": 0.781051313243115, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 9562 + }, + { + "epoch": 0.09563, + "grad_norm": 0.8326369784417603, + "learning_rate": 0.003, + "loss": 4.0804, + "step": 9563 + }, + { + "epoch": 0.09564, + "grad_norm": 0.7305150419795909, + "learning_rate": 0.003, + "loss": 4.0608, + "step": 9564 + }, + { + "epoch": 0.09565, + "grad_norm": 0.6688292218195512, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 9565 + }, + { + "epoch": 0.09566, + "grad_norm": 0.6361084169360314, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 9566 + }, + { + "epoch": 0.09567, + "grad_norm": 0.6714713399143089, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 9567 + }, + { + "epoch": 0.09568, + "grad_norm": 0.7435190836794898, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 9568 + }, + { + "epoch": 0.09569, + "grad_norm": 0.7959510274747182, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 9569 + }, + { + "epoch": 0.0957, + "grad_norm": 0.8571395730696499, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 9570 + }, + { + "epoch": 0.09571, + "grad_norm": 0.8283879837410419, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 9571 + }, + { + "epoch": 0.09572, + "grad_norm": 0.7595065389539266, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9572 + }, + { + "epoch": 0.09573, + "grad_norm": 0.6384941028440521, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 9573 + }, + { + "epoch": 0.09574, + "grad_norm": 0.6518267167937323, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 9574 + }, + { + "epoch": 0.09575, + "grad_norm": 0.5910084190134272, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 9575 + }, + { + "epoch": 0.09576, + "grad_norm": 0.5537587690836792, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 9576 + }, + { + "epoch": 0.09577, + "grad_norm": 0.5364490944964171, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 9577 + }, + { + "epoch": 0.09578, + "grad_norm": 0.4537361101955507, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 9578 + }, + { + "epoch": 0.09579, + "grad_norm": 0.5166636711443451, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 9579 + }, + { + "epoch": 0.0958, + "grad_norm": 0.56518260916355, + "learning_rate": 0.003, + "loss": 4.04, + "step": 9580 + }, + { + "epoch": 0.09581, + "grad_norm": 0.5409029878604577, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 9581 + }, + { + "epoch": 0.09582, + "grad_norm": 0.4798374655326553, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 9582 + }, + { + "epoch": 0.09583, + "grad_norm": 0.5238056459118106, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 9583 + }, + { + "epoch": 0.09584, + "grad_norm": 0.5833796622341467, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 9584 + }, + { + "epoch": 0.09585, + "grad_norm": 0.6775449740147607, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 9585 + }, + { + "epoch": 0.09586, + "grad_norm": 0.777451514297023, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 9586 + }, + { + "epoch": 0.09587, + "grad_norm": 0.8515770782378858, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 9587 + }, + { + "epoch": 0.09588, + "grad_norm": 0.9342293934794202, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 9588 + }, + { + "epoch": 0.09589, + "grad_norm": 0.9852117294459376, + "learning_rate": 0.003, + "loss": 4.044, + "step": 9589 + }, + { + "epoch": 0.0959, + "grad_norm": 0.8302497491372647, + "learning_rate": 0.003, + "loss": 4.055, + "step": 9590 + }, + { + "epoch": 0.09591, + "grad_norm": 0.8184349524479022, + "learning_rate": 0.003, + "loss": 4.0817, + "step": 9591 + }, + { + "epoch": 0.09592, + "grad_norm": 0.8688949911902178, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 9592 + }, + { + "epoch": 0.09593, + "grad_norm": 0.9187065457776056, + "learning_rate": 0.003, + "loss": 4.067, + "step": 9593 + }, + { + "epoch": 0.09594, + "grad_norm": 0.9509469250312006, + "learning_rate": 0.003, + "loss": 4.035, + "step": 9594 + }, + { + "epoch": 0.09595, + "grad_norm": 1.0162993701888292, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 9595 + }, + { + "epoch": 0.09596, + "grad_norm": 0.8833845700778189, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 9596 + }, + { + "epoch": 0.09597, + "grad_norm": 0.9212045056353114, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 9597 + }, + { + "epoch": 0.09598, + "grad_norm": 0.8366683747836731, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 9598 + }, + { + "epoch": 0.09599, + "grad_norm": 0.7659760978619349, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 9599 + }, + { + "epoch": 0.096, + "grad_norm": 0.6178237409029183, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 9600 + }, + { + "epoch": 0.09601, + "grad_norm": 0.7257885555103538, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 9601 + }, + { + "epoch": 0.09602, + "grad_norm": 0.6503724839865827, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 9602 + }, + { + "epoch": 0.09603, + "grad_norm": 0.6472192449005016, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 9603 + }, + { + "epoch": 0.09604, + "grad_norm": 0.6717321285668933, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 9604 + }, + { + "epoch": 0.09605, + "grad_norm": 0.6122965815507444, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 9605 + }, + { + "epoch": 0.09606, + "grad_norm": 0.5490219424111816, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 9606 + }, + { + "epoch": 0.09607, + "grad_norm": 0.5151599816682494, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9607 + }, + { + "epoch": 0.09608, + "grad_norm": 0.44672321756837635, + "learning_rate": 0.003, + "loss": 4.0623, + "step": 9608 + }, + { + "epoch": 0.09609, + "grad_norm": 0.42605596297492704, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 9609 + }, + { + "epoch": 0.0961, + "grad_norm": 0.49901086547730017, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 9610 + }, + { + "epoch": 0.09611, + "grad_norm": 0.5677334924584306, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 9611 + }, + { + "epoch": 0.09612, + "grad_norm": 0.6822686899444607, + "learning_rate": 0.003, + "loss": 4.033, + "step": 9612 + }, + { + "epoch": 0.09613, + "grad_norm": 0.7470925393326813, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 9613 + }, + { + "epoch": 0.09614, + "grad_norm": 0.7348075008305133, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 9614 + }, + { + "epoch": 0.09615, + "grad_norm": 0.5999504558142992, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 9615 + }, + { + "epoch": 0.09616, + "grad_norm": 0.5776261743151357, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 9616 + }, + { + "epoch": 0.09617, + "grad_norm": 0.5854993026034835, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 9617 + }, + { + "epoch": 0.09618, + "grad_norm": 0.669422425468719, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 9618 + }, + { + "epoch": 0.09619, + "grad_norm": 0.6367194226068549, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 9619 + }, + { + "epoch": 0.0962, + "grad_norm": 0.6988063184950772, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 9620 + }, + { + "epoch": 0.09621, + "grad_norm": 0.8199478228173989, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 9621 + }, + { + "epoch": 0.09622, + "grad_norm": 0.8450174656850685, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 9622 + }, + { + "epoch": 0.09623, + "grad_norm": 0.7507551464729676, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 9623 + }, + { + "epoch": 0.09624, + "grad_norm": 0.8078698640414419, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 9624 + }, + { + "epoch": 0.09625, + "grad_norm": 0.7735268587359604, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 9625 + }, + { + "epoch": 0.09626, + "grad_norm": 0.6364418468413616, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 9626 + }, + { + "epoch": 0.09627, + "grad_norm": 0.5746022924928109, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 9627 + }, + { + "epoch": 0.09628, + "grad_norm": 0.5200092611856018, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 9628 + }, + { + "epoch": 0.09629, + "grad_norm": 0.5288611800091972, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 9629 + }, + { + "epoch": 0.0963, + "grad_norm": 0.5796174181987767, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 9630 + }, + { + "epoch": 0.09631, + "grad_norm": 0.7099375233532926, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 9631 + }, + { + "epoch": 0.09632, + "grad_norm": 0.7244803813572703, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 9632 + }, + { + "epoch": 0.09633, + "grad_norm": 0.7143244247400538, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 9633 + }, + { + "epoch": 0.09634, + "grad_norm": 0.6978514334940835, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 9634 + }, + { + "epoch": 0.09635, + "grad_norm": 0.6417002536555552, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 9635 + }, + { + "epoch": 0.09636, + "grad_norm": 0.6822454727745676, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 9636 + }, + { + "epoch": 0.09637, + "grad_norm": 0.6728070850801018, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 9637 + }, + { + "epoch": 0.09638, + "grad_norm": 0.6558336070863201, + "learning_rate": 0.003, + "loss": 4.011, + "step": 9638 + }, + { + "epoch": 0.09639, + "grad_norm": 0.7099105991789862, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 9639 + }, + { + "epoch": 0.0964, + "grad_norm": 0.7719199945485586, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 9640 + }, + { + "epoch": 0.09641, + "grad_norm": 0.8721903793545053, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 9641 + }, + { + "epoch": 0.09642, + "grad_norm": 0.870181642268105, + "learning_rate": 0.003, + "loss": 4.061, + "step": 9642 + }, + { + "epoch": 0.09643, + "grad_norm": 0.8371822210663833, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 9643 + }, + { + "epoch": 0.09644, + "grad_norm": 0.8164969830431094, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 9644 + }, + { + "epoch": 0.09645, + "grad_norm": 0.802341770041663, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9645 + }, + { + "epoch": 0.09646, + "grad_norm": 0.8436986154931037, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 9646 + }, + { + "epoch": 0.09647, + "grad_norm": 0.8864902589125572, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 9647 + }, + { + "epoch": 0.09648, + "grad_norm": 0.7148335098667227, + "learning_rate": 0.003, + "loss": 4.075, + "step": 9648 + }, + { + "epoch": 0.09649, + "grad_norm": 0.7620890745233596, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 9649 + }, + { + "epoch": 0.0965, + "grad_norm": 0.8225665812577014, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 9650 + }, + { + "epoch": 0.09651, + "grad_norm": 1.0252079607973874, + "learning_rate": 0.003, + "loss": 4.054, + "step": 9651 + }, + { + "epoch": 0.09652, + "grad_norm": 0.9512889234354192, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 9652 + }, + { + "epoch": 0.09653, + "grad_norm": 0.8616913946830312, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 9653 + }, + { + "epoch": 0.09654, + "grad_norm": 0.7802088383084742, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 9654 + }, + { + "epoch": 0.09655, + "grad_norm": 0.7004638902756576, + "learning_rate": 0.003, + "loss": 4.0536, + "step": 9655 + }, + { + "epoch": 0.09656, + "grad_norm": 0.7816255676024597, + "learning_rate": 0.003, + "loss": 4.0653, + "step": 9656 + }, + { + "epoch": 0.09657, + "grad_norm": 0.8059211257395651, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 9657 + }, + { + "epoch": 0.09658, + "grad_norm": 0.8728386697752617, + "learning_rate": 0.003, + "loss": 4.042, + "step": 9658 + }, + { + "epoch": 0.09659, + "grad_norm": 0.8144236544189702, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 9659 + }, + { + "epoch": 0.0966, + "grad_norm": 0.6570080797657473, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 9660 + }, + { + "epoch": 0.09661, + "grad_norm": 0.5979958700311733, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 9661 + }, + { + "epoch": 0.09662, + "grad_norm": 0.5378238908288007, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 9662 + }, + { + "epoch": 0.09663, + "grad_norm": 0.5256661537158236, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 9663 + }, + { + "epoch": 0.09664, + "grad_norm": 0.5438453363683964, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9664 + }, + { + "epoch": 0.09665, + "grad_norm": 0.5269396608941532, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 9665 + }, + { + "epoch": 0.09666, + "grad_norm": 0.45677428379922197, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 9666 + }, + { + "epoch": 0.09667, + "grad_norm": 0.49030095620204905, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 9667 + }, + { + "epoch": 0.09668, + "grad_norm": 0.508766527949961, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 9668 + }, + { + "epoch": 0.09669, + "grad_norm": 0.445555391890408, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 9669 + }, + { + "epoch": 0.0967, + "grad_norm": 0.47313746401744633, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 9670 + }, + { + "epoch": 0.09671, + "grad_norm": 0.47117445930209506, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 9671 + }, + { + "epoch": 0.09672, + "grad_norm": 0.46118051025212853, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 9672 + }, + { + "epoch": 0.09673, + "grad_norm": 0.44906908843181786, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 9673 + }, + { + "epoch": 0.09674, + "grad_norm": 0.48949700746077207, + "learning_rate": 0.003, + "loss": 4.0716, + "step": 9674 + }, + { + "epoch": 0.09675, + "grad_norm": 0.6013049097145232, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 9675 + }, + { + "epoch": 0.09676, + "grad_norm": 0.6593613481604037, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9676 + }, + { + "epoch": 0.09677, + "grad_norm": 0.7363080034277738, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 9677 + }, + { + "epoch": 0.09678, + "grad_norm": 0.7898279469555168, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 9678 + }, + { + "epoch": 0.09679, + "grad_norm": 0.7757788968404743, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 9679 + }, + { + "epoch": 0.0968, + "grad_norm": 0.8466793764913422, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9680 + }, + { + "epoch": 0.09681, + "grad_norm": 0.8773344376147645, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9681 + }, + { + "epoch": 0.09682, + "grad_norm": 0.8558727895083553, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 9682 + }, + { + "epoch": 0.09683, + "grad_norm": 0.9914715152618384, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 9683 + }, + { + "epoch": 0.09684, + "grad_norm": 0.9528350472002182, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 9684 + }, + { + "epoch": 0.09685, + "grad_norm": 0.9807832185081737, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 9685 + }, + { + "epoch": 0.09686, + "grad_norm": 1.07752935303985, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 9686 + }, + { + "epoch": 0.09687, + "grad_norm": 1.0225115586497027, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 9687 + }, + { + "epoch": 0.09688, + "grad_norm": 1.015818671816652, + "learning_rate": 0.003, + "loss": 4.0982, + "step": 9688 + }, + { + "epoch": 0.09689, + "grad_norm": 0.9929526998528431, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 9689 + }, + { + "epoch": 0.0969, + "grad_norm": 0.7813261255117088, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 9690 + }, + { + "epoch": 0.09691, + "grad_norm": 0.7291110156126227, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 9691 + }, + { + "epoch": 0.09692, + "grad_norm": 0.732104481693697, + "learning_rate": 0.003, + "loss": 4.0672, + "step": 9692 + }, + { + "epoch": 0.09693, + "grad_norm": 0.7348264571578732, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 9693 + }, + { + "epoch": 0.09694, + "grad_norm": 0.8154956455402665, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 9694 + }, + { + "epoch": 0.09695, + "grad_norm": 0.868580129936653, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 9695 + }, + { + "epoch": 0.09696, + "grad_norm": 0.8615507807760837, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 9696 + }, + { + "epoch": 0.09697, + "grad_norm": 0.7858346747316877, + "learning_rate": 0.003, + "loss": 4.049, + "step": 9697 + }, + { + "epoch": 0.09698, + "grad_norm": 0.7079498348983355, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 9698 + }, + { + "epoch": 0.09699, + "grad_norm": 0.6397150310600791, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 9699 + }, + { + "epoch": 0.097, + "grad_norm": 0.6908815613830682, + "learning_rate": 0.003, + "loss": 4.0651, + "step": 9700 + }, + { + "epoch": 0.09701, + "grad_norm": 0.6639440710921456, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9701 + }, + { + "epoch": 0.09702, + "grad_norm": 0.5834212258360982, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 9702 + }, + { + "epoch": 0.09703, + "grad_norm": 0.5687363197652693, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 9703 + }, + { + "epoch": 0.09704, + "grad_norm": 0.5713355386279885, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 9704 + }, + { + "epoch": 0.09705, + "grad_norm": 0.5413947910447899, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 9705 + }, + { + "epoch": 0.09706, + "grad_norm": 0.5148815559025326, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 9706 + }, + { + "epoch": 0.09707, + "grad_norm": 0.5519580518466485, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 9707 + }, + { + "epoch": 0.09708, + "grad_norm": 0.5620808717545713, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 9708 + }, + { + "epoch": 0.09709, + "grad_norm": 0.48460690682480734, + "learning_rate": 0.003, + "loss": 4.032, + "step": 9709 + }, + { + "epoch": 0.0971, + "grad_norm": 0.43093259770898656, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 9710 + }, + { + "epoch": 0.09711, + "grad_norm": 0.39333941750174756, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 9711 + }, + { + "epoch": 0.09712, + "grad_norm": 0.38249681987121276, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 9712 + }, + { + "epoch": 0.09713, + "grad_norm": 0.3812967839661506, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 9713 + }, + { + "epoch": 0.09714, + "grad_norm": 0.39709505805804846, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 9714 + }, + { + "epoch": 0.09715, + "grad_norm": 0.3982397783800207, + "learning_rate": 0.003, + "loss": 4.026, + "step": 9715 + }, + { + "epoch": 0.09716, + "grad_norm": 0.4541471098834424, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 9716 + }, + { + "epoch": 0.09717, + "grad_norm": 0.500556402101347, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 9717 + }, + { + "epoch": 0.09718, + "grad_norm": 0.6304849425702689, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 9718 + }, + { + "epoch": 0.09719, + "grad_norm": 0.8218949792771653, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 9719 + }, + { + "epoch": 0.0972, + "grad_norm": 0.9402650036757573, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 9720 + }, + { + "epoch": 0.09721, + "grad_norm": 0.912105280275023, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 9721 + }, + { + "epoch": 0.09722, + "grad_norm": 0.8161984155736592, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 9722 + }, + { + "epoch": 0.09723, + "grad_norm": 0.7588493871897334, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 9723 + }, + { + "epoch": 0.09724, + "grad_norm": 0.7021440809083722, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 9724 + }, + { + "epoch": 0.09725, + "grad_norm": 0.673982202504525, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 9725 + }, + { + "epoch": 0.09726, + "grad_norm": 0.5901967525594085, + "learning_rate": 0.003, + "loss": 4.034, + "step": 9726 + }, + { + "epoch": 0.09727, + "grad_norm": 0.6341340337073639, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 9727 + }, + { + "epoch": 0.09728, + "grad_norm": 0.6877396024850011, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 9728 + }, + { + "epoch": 0.09729, + "grad_norm": 0.8074359056158719, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 9729 + }, + { + "epoch": 0.0973, + "grad_norm": 0.9847830626929746, + "learning_rate": 0.003, + "loss": 4.045, + "step": 9730 + }, + { + "epoch": 0.09731, + "grad_norm": 1.0794364509012435, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 9731 + }, + { + "epoch": 0.09732, + "grad_norm": 0.7946605824665086, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 9732 + }, + { + "epoch": 0.09733, + "grad_norm": 0.8053333083787163, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 9733 + }, + { + "epoch": 0.09734, + "grad_norm": 0.8419973000000773, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 9734 + }, + { + "epoch": 0.09735, + "grad_norm": 0.912588824483311, + "learning_rate": 0.003, + "loss": 4.061, + "step": 9735 + }, + { + "epoch": 0.09736, + "grad_norm": 1.0077884646515058, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 9736 + }, + { + "epoch": 0.09737, + "grad_norm": 0.8553742026342334, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 9737 + }, + { + "epoch": 0.09738, + "grad_norm": 0.6663335025191279, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 9738 + }, + { + "epoch": 0.09739, + "grad_norm": 0.6148567771081777, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 9739 + }, + { + "epoch": 0.0974, + "grad_norm": 0.6518956470592588, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 9740 + }, + { + "epoch": 0.09741, + "grad_norm": 0.640738036711578, + "learning_rate": 0.003, + "loss": 4.0832, + "step": 9741 + }, + { + "epoch": 0.09742, + "grad_norm": 0.6569222468905644, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 9742 + }, + { + "epoch": 0.09743, + "grad_norm": 0.8135918559010041, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 9743 + }, + { + "epoch": 0.09744, + "grad_norm": 0.8542738238984536, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 9744 + }, + { + "epoch": 0.09745, + "grad_norm": 0.827825997101658, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 9745 + }, + { + "epoch": 0.09746, + "grad_norm": 0.8676728635409363, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 9746 + }, + { + "epoch": 0.09747, + "grad_norm": 0.7759686751828946, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 9747 + }, + { + "epoch": 0.09748, + "grad_norm": 0.6977991477507604, + "learning_rate": 0.003, + "loss": 4.0735, + "step": 9748 + }, + { + "epoch": 0.09749, + "grad_norm": 0.7267507352248281, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 9749 + }, + { + "epoch": 0.0975, + "grad_norm": 0.6337194831331616, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 9750 + }, + { + "epoch": 0.09751, + "grad_norm": 0.6890427932396798, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 9751 + }, + { + "epoch": 0.09752, + "grad_norm": 0.6546541997966022, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 9752 + }, + { + "epoch": 0.09753, + "grad_norm": 0.6050186080118325, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 9753 + }, + { + "epoch": 0.09754, + "grad_norm": 0.5587658461041117, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 9754 + }, + { + "epoch": 0.09755, + "grad_norm": 0.6202093109536967, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 9755 + }, + { + "epoch": 0.09756, + "grad_norm": 0.6566930811394645, + "learning_rate": 0.003, + "loss": 4.016, + "step": 9756 + }, + { + "epoch": 0.09757, + "grad_norm": 0.7649988606696811, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 9757 + }, + { + "epoch": 0.09758, + "grad_norm": 0.8267165034574794, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 9758 + }, + { + "epoch": 0.09759, + "grad_norm": 0.8520498192733337, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 9759 + }, + { + "epoch": 0.0976, + "grad_norm": 0.7741759988469258, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 9760 + }, + { + "epoch": 0.09761, + "grad_norm": 0.7013242007051717, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 9761 + }, + { + "epoch": 0.09762, + "grad_norm": 0.6091165402866249, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 9762 + }, + { + "epoch": 0.09763, + "grad_norm": 0.5157642643214935, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 9763 + }, + { + "epoch": 0.09764, + "grad_norm": 0.4822725456380591, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 9764 + }, + { + "epoch": 0.09765, + "grad_norm": 0.5313070114880102, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 9765 + }, + { + "epoch": 0.09766, + "grad_norm": 0.5308808507313943, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 9766 + }, + { + "epoch": 0.09767, + "grad_norm": 0.5145277965619717, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 9767 + }, + { + "epoch": 0.09768, + "grad_norm": 0.593899913891455, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 9768 + }, + { + "epoch": 0.09769, + "grad_norm": 0.630170447805223, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 9769 + }, + { + "epoch": 0.0977, + "grad_norm": 0.6394116986587077, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 9770 + }, + { + "epoch": 0.09771, + "grad_norm": 0.5848704637769577, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 9771 + }, + { + "epoch": 0.09772, + "grad_norm": 0.5332618692294218, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 9772 + }, + { + "epoch": 0.09773, + "grad_norm": 0.4953917549085509, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 9773 + }, + { + "epoch": 0.09774, + "grad_norm": 0.45501803271781743, + "learning_rate": 0.003, + "loss": 4.043, + "step": 9774 + }, + { + "epoch": 0.09775, + "grad_norm": 0.48412150823989486, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 9775 + }, + { + "epoch": 0.09776, + "grad_norm": 0.5239934994168983, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 9776 + }, + { + "epoch": 0.09777, + "grad_norm": 0.6027904838194806, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 9777 + }, + { + "epoch": 0.09778, + "grad_norm": 0.774390323379049, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 9778 + }, + { + "epoch": 0.09779, + "grad_norm": 0.9491060985496452, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 9779 + }, + { + "epoch": 0.0978, + "grad_norm": 0.9191489433732237, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 9780 + }, + { + "epoch": 0.09781, + "grad_norm": 0.7827291983171121, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 9781 + }, + { + "epoch": 0.09782, + "grad_norm": 0.7859688722311206, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 9782 + }, + { + "epoch": 0.09783, + "grad_norm": 0.7101887576858492, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 9783 + }, + { + "epoch": 0.09784, + "grad_norm": 0.6912308016830487, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 9784 + }, + { + "epoch": 0.09785, + "grad_norm": 0.7576531986654672, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9785 + }, + { + "epoch": 0.09786, + "grad_norm": 0.8574544400020419, + "learning_rate": 0.003, + "loss": 4.026, + "step": 9786 + }, + { + "epoch": 0.09787, + "grad_norm": 0.9577180720307842, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 9787 + }, + { + "epoch": 0.09788, + "grad_norm": 0.9461362811535105, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 9788 + }, + { + "epoch": 0.09789, + "grad_norm": 0.9314330660686861, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 9789 + }, + { + "epoch": 0.0979, + "grad_norm": 0.923358376983677, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 9790 + }, + { + "epoch": 0.09791, + "grad_norm": 0.9038877363239036, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 9791 + }, + { + "epoch": 0.09792, + "grad_norm": 0.9990794191698911, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 9792 + }, + { + "epoch": 0.09793, + "grad_norm": 0.9445211894088562, + "learning_rate": 0.003, + "loss": 4.046, + "step": 9793 + }, + { + "epoch": 0.09794, + "grad_norm": 0.8956275058636283, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 9794 + }, + { + "epoch": 0.09795, + "grad_norm": 0.8711859193999932, + "learning_rate": 0.003, + "loss": 4.0693, + "step": 9795 + }, + { + "epoch": 0.09796, + "grad_norm": 0.7806261157146592, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 9796 + }, + { + "epoch": 0.09797, + "grad_norm": 0.7925985037796391, + "learning_rate": 0.003, + "loss": 4.047, + "step": 9797 + }, + { + "epoch": 0.09798, + "grad_norm": 0.7691034440219884, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 9798 + }, + { + "epoch": 0.09799, + "grad_norm": 0.6881987683069155, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 9799 + }, + { + "epoch": 0.098, + "grad_norm": 0.6316113522615392, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 9800 + }, + { + "epoch": 0.09801, + "grad_norm": 0.6558510695079839, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 9801 + }, + { + "epoch": 0.09802, + "grad_norm": 0.6260894568430416, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 9802 + }, + { + "epoch": 0.09803, + "grad_norm": 0.5700793448259298, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 9803 + }, + { + "epoch": 0.09804, + "grad_norm": 0.5940632470486125, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 9804 + }, + { + "epoch": 0.09805, + "grad_norm": 0.5617832910023823, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 9805 + }, + { + "epoch": 0.09806, + "grad_norm": 0.5468180714263916, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 9806 + }, + { + "epoch": 0.09807, + "grad_norm": 0.4843260402813189, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 9807 + }, + { + "epoch": 0.09808, + "grad_norm": 0.4284909120219946, + "learning_rate": 0.003, + "loss": 4.055, + "step": 9808 + }, + { + "epoch": 0.09809, + "grad_norm": 0.41494301501945763, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 9809 + }, + { + "epoch": 0.0981, + "grad_norm": 0.4543658157070918, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 9810 + }, + { + "epoch": 0.09811, + "grad_norm": 0.5451181542933197, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 9811 + }, + { + "epoch": 0.09812, + "grad_norm": 0.7426047362847888, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 9812 + }, + { + "epoch": 0.09813, + "grad_norm": 0.8916597370905406, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 9813 + }, + { + "epoch": 0.09814, + "grad_norm": 0.9340821558195145, + "learning_rate": 0.003, + "loss": 4.024, + "step": 9814 + }, + { + "epoch": 0.09815, + "grad_norm": 0.840005976354376, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 9815 + }, + { + "epoch": 0.09816, + "grad_norm": 0.7253377500860728, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 9816 + }, + { + "epoch": 0.09817, + "grad_norm": 0.7042480510048211, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 9817 + }, + { + "epoch": 0.09818, + "grad_norm": 0.7013161768391724, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 9818 + }, + { + "epoch": 0.09819, + "grad_norm": 0.7303064336493744, + "learning_rate": 0.003, + "loss": 4.048, + "step": 9819 + }, + { + "epoch": 0.0982, + "grad_norm": 0.779803313717684, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 9820 + }, + { + "epoch": 0.09821, + "grad_norm": 0.7158019745532164, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 9821 + }, + { + "epoch": 0.09822, + "grad_norm": 0.6723392450218468, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 9822 + }, + { + "epoch": 0.09823, + "grad_norm": 0.6653162238491962, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 9823 + }, + { + "epoch": 0.09824, + "grad_norm": 0.6571025983723008, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 9824 + }, + { + "epoch": 0.09825, + "grad_norm": 0.7567749206437057, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 9825 + }, + { + "epoch": 0.09826, + "grad_norm": 0.8563331474921774, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 9826 + }, + { + "epoch": 0.09827, + "grad_norm": 0.823440749278152, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 9827 + }, + { + "epoch": 0.09828, + "grad_norm": 0.7539761800770246, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 9828 + }, + { + "epoch": 0.09829, + "grad_norm": 0.5730277298635529, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 9829 + }, + { + "epoch": 0.0983, + "grad_norm": 0.5181768084659204, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 9830 + }, + { + "epoch": 0.09831, + "grad_norm": 0.6143034111628997, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 9831 + }, + { + "epoch": 0.09832, + "grad_norm": 0.6775858998336227, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 9832 + }, + { + "epoch": 0.09833, + "grad_norm": 0.7476121864433801, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 9833 + }, + { + "epoch": 0.09834, + "grad_norm": 0.7607806973191736, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 9834 + }, + { + "epoch": 0.09835, + "grad_norm": 0.7188449373605055, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 9835 + }, + { + "epoch": 0.09836, + "grad_norm": 0.7259906232492397, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 9836 + }, + { + "epoch": 0.09837, + "grad_norm": 0.541540571002883, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 9837 + }, + { + "epoch": 0.09838, + "grad_norm": 0.6324952794927997, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 9838 + }, + { + "epoch": 0.09839, + "grad_norm": 0.6913162330431121, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 9839 + }, + { + "epoch": 0.0984, + "grad_norm": 0.6818187272947319, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 9840 + }, + { + "epoch": 0.09841, + "grad_norm": 0.5824070426308308, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 9841 + }, + { + "epoch": 0.09842, + "grad_norm": 0.5611466674211786, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 9842 + }, + { + "epoch": 0.09843, + "grad_norm": 0.4932559885296401, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 9843 + }, + { + "epoch": 0.09844, + "grad_norm": 0.4725725195369003, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 9844 + }, + { + "epoch": 0.09845, + "grad_norm": 0.6071022710939762, + "learning_rate": 0.003, + "loss": 4.017, + "step": 9845 + }, + { + "epoch": 0.09846, + "grad_norm": 0.6819618405769559, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 9846 + }, + { + "epoch": 0.09847, + "grad_norm": 0.779719526263489, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 9847 + }, + { + "epoch": 0.09848, + "grad_norm": 0.8221349440077464, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 9848 + }, + { + "epoch": 0.09849, + "grad_norm": 0.7383403993800184, + "learning_rate": 0.003, + "loss": 4.021, + "step": 9849 + }, + { + "epoch": 0.0985, + "grad_norm": 0.762340407626035, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 9850 + }, + { + "epoch": 0.09851, + "grad_norm": 0.8002898469418368, + "learning_rate": 0.003, + "loss": 4.056, + "step": 9851 + }, + { + "epoch": 0.09852, + "grad_norm": 0.8638015919975774, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 9852 + }, + { + "epoch": 0.09853, + "grad_norm": 0.9020457680731377, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 9853 + }, + { + "epoch": 0.09854, + "grad_norm": 0.9230165852233206, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 9854 + }, + { + "epoch": 0.09855, + "grad_norm": 0.8571092870371995, + "learning_rate": 0.003, + "loss": 4.003, + "step": 9855 + }, + { + "epoch": 0.09856, + "grad_norm": 0.8089728437147037, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 9856 + }, + { + "epoch": 0.09857, + "grad_norm": 0.7045894622553129, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 9857 + }, + { + "epoch": 0.09858, + "grad_norm": 0.7266777803540795, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 9858 + }, + { + "epoch": 0.09859, + "grad_norm": 0.6490891932312824, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 9859 + }, + { + "epoch": 0.0986, + "grad_norm": 0.7235943604919849, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 9860 + }, + { + "epoch": 0.09861, + "grad_norm": 0.7897878690601106, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 9861 + }, + { + "epoch": 0.09862, + "grad_norm": 0.8899392095065458, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 9862 + }, + { + "epoch": 0.09863, + "grad_norm": 1.0139728221191706, + "learning_rate": 0.003, + "loss": 4.0723, + "step": 9863 + }, + { + "epoch": 0.09864, + "grad_norm": 0.9603265632661463, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 9864 + }, + { + "epoch": 0.09865, + "grad_norm": 0.9307279476663425, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 9865 + }, + { + "epoch": 0.09866, + "grad_norm": 0.9832183157702538, + "learning_rate": 0.003, + "loss": 4.044, + "step": 9866 + }, + { + "epoch": 0.09867, + "grad_norm": 1.0252151517485693, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 9867 + }, + { + "epoch": 0.09868, + "grad_norm": 0.9443178686110089, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 9868 + }, + { + "epoch": 0.09869, + "grad_norm": 0.8876857845956557, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 9869 + }, + { + "epoch": 0.0987, + "grad_norm": 0.8434366172387391, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 9870 + }, + { + "epoch": 0.09871, + "grad_norm": 0.7914496669404404, + "learning_rate": 0.003, + "loss": 4.1063, + "step": 9871 + }, + { + "epoch": 0.09872, + "grad_norm": 0.8176766440271231, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 9872 + }, + { + "epoch": 0.09873, + "grad_norm": 0.7932892425742387, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 9873 + }, + { + "epoch": 0.09874, + "grad_norm": 0.7666745939958074, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 9874 + }, + { + "epoch": 0.09875, + "grad_norm": 0.8658521697296111, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 9875 + }, + { + "epoch": 0.09876, + "grad_norm": 0.8231475090965542, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 9876 + }, + { + "epoch": 0.09877, + "grad_norm": 0.7264922619174438, + "learning_rate": 0.003, + "loss": 4.037, + "step": 9877 + }, + { + "epoch": 0.09878, + "grad_norm": 0.6663778944614389, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 9878 + }, + { + "epoch": 0.09879, + "grad_norm": 0.6699835031593454, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 9879 + }, + { + "epoch": 0.0988, + "grad_norm": 0.6302235680491419, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 9880 + }, + { + "epoch": 0.09881, + "grad_norm": 0.5123711970372388, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 9881 + }, + { + "epoch": 0.09882, + "grad_norm": 0.4553784089829774, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 9882 + }, + { + "epoch": 0.09883, + "grad_norm": 0.4490494248520515, + "learning_rate": 0.003, + "loss": 4.026, + "step": 9883 + }, + { + "epoch": 0.09884, + "grad_norm": 0.46931259410627707, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 9884 + }, + { + "epoch": 0.09885, + "grad_norm": 0.47000862322251247, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 9885 + }, + { + "epoch": 0.09886, + "grad_norm": 0.4356687168935574, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 9886 + }, + { + "epoch": 0.09887, + "grad_norm": 0.4482121755470332, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 9887 + }, + { + "epoch": 0.09888, + "grad_norm": 0.4468090117573381, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 9888 + }, + { + "epoch": 0.09889, + "grad_norm": 0.4037218387554367, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 9889 + }, + { + "epoch": 0.0989, + "grad_norm": 0.41400152426310555, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 9890 + }, + { + "epoch": 0.09891, + "grad_norm": 0.44016298246066426, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 9891 + }, + { + "epoch": 0.09892, + "grad_norm": 0.4903810053685349, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 9892 + }, + { + "epoch": 0.09893, + "grad_norm": 0.5334674766442788, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 9893 + }, + { + "epoch": 0.09894, + "grad_norm": 0.574298737842145, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 9894 + }, + { + "epoch": 0.09895, + "grad_norm": 0.6890874968168291, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 9895 + }, + { + "epoch": 0.09896, + "grad_norm": 0.810421045908322, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 9896 + }, + { + "epoch": 0.09897, + "grad_norm": 0.8189208817871967, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 9897 + }, + { + "epoch": 0.09898, + "grad_norm": 0.6899334701896801, + "learning_rate": 0.003, + "loss": 4.019, + "step": 9898 + }, + { + "epoch": 0.09899, + "grad_norm": 0.5957224442194552, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 9899 + }, + { + "epoch": 0.099, + "grad_norm": 0.587701663289095, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 9900 + }, + { + "epoch": 0.09901, + "grad_norm": 0.7061980938429728, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 9901 + }, + { + "epoch": 0.09902, + "grad_norm": 0.8242698280149364, + "learning_rate": 0.003, + "loss": 4.034, + "step": 9902 + }, + { + "epoch": 0.09903, + "grad_norm": 0.9245033362862743, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 9903 + }, + { + "epoch": 0.09904, + "grad_norm": 0.8213358947877748, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 9904 + }, + { + "epoch": 0.09905, + "grad_norm": 0.8103471342284361, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 9905 + }, + { + "epoch": 0.09906, + "grad_norm": 0.7560871159658464, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 9906 + }, + { + "epoch": 0.09907, + "grad_norm": 0.6641055019660936, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 9907 + }, + { + "epoch": 0.09908, + "grad_norm": 0.7552595109751858, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 9908 + }, + { + "epoch": 0.09909, + "grad_norm": 0.7214935184980603, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 9909 + }, + { + "epoch": 0.0991, + "grad_norm": 0.6756774006864223, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 9910 + }, + { + "epoch": 0.09911, + "grad_norm": 0.7726785146992515, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 9911 + }, + { + "epoch": 0.09912, + "grad_norm": 0.7900365997501597, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 9912 + }, + { + "epoch": 0.09913, + "grad_norm": 0.7569036036541276, + "learning_rate": 0.003, + "loss": 4.0628, + "step": 9913 + }, + { + "epoch": 0.09914, + "grad_norm": 0.7915170995929005, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 9914 + }, + { + "epoch": 0.09915, + "grad_norm": 0.767939597593537, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 9915 + }, + { + "epoch": 0.09916, + "grad_norm": 0.8399818217777323, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 9916 + }, + { + "epoch": 0.09917, + "grad_norm": 0.9895296595346701, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 9917 + }, + { + "epoch": 0.09918, + "grad_norm": 1.2040398002564898, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 9918 + }, + { + "epoch": 0.09919, + "grad_norm": 0.9411487706180013, + "learning_rate": 0.003, + "loss": 4.0715, + "step": 9919 + }, + { + "epoch": 0.0992, + "grad_norm": 0.8965799670097959, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 9920 + }, + { + "epoch": 0.09921, + "grad_norm": 0.811235314205704, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 9921 + }, + { + "epoch": 0.09922, + "grad_norm": 0.819626499264921, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 9922 + }, + { + "epoch": 0.09923, + "grad_norm": 0.9069342193582761, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 9923 + }, + { + "epoch": 0.09924, + "grad_norm": 0.8107621946542576, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 9924 + }, + { + "epoch": 0.09925, + "grad_norm": 0.7362058814388313, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 9925 + }, + { + "epoch": 0.09926, + "grad_norm": 0.6169379642879168, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 9926 + }, + { + "epoch": 0.09927, + "grad_norm": 0.5976790846178441, + "learning_rate": 0.003, + "loss": 4.07, + "step": 9927 + }, + { + "epoch": 0.09928, + "grad_norm": 0.549626491829479, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 9928 + }, + { + "epoch": 0.09929, + "grad_norm": 0.46802673949829054, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 9929 + }, + { + "epoch": 0.0993, + "grad_norm": 0.4637470831691204, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 9930 + }, + { + "epoch": 0.09931, + "grad_norm": 0.4160691191265149, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 9931 + }, + { + "epoch": 0.09932, + "grad_norm": 0.5124655674333713, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 9932 + }, + { + "epoch": 0.09933, + "grad_norm": 0.7160761495189472, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 9933 + }, + { + "epoch": 0.09934, + "grad_norm": 0.8972062441018436, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 9934 + }, + { + "epoch": 0.09935, + "grad_norm": 0.9846730216104669, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 9935 + }, + { + "epoch": 0.09936, + "grad_norm": 0.7683987219083155, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 9936 + }, + { + "epoch": 0.09937, + "grad_norm": 0.610850891094036, + "learning_rate": 0.003, + "loss": 4.0, + "step": 9937 + }, + { + "epoch": 0.09938, + "grad_norm": 0.6438921082061193, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 9938 + }, + { + "epoch": 0.09939, + "grad_norm": 0.7077933160798726, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 9939 + }, + { + "epoch": 0.0994, + "grad_norm": 0.6910051330916097, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 9940 + }, + { + "epoch": 0.09941, + "grad_norm": 0.7045064602640984, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 9941 + }, + { + "epoch": 0.09942, + "grad_norm": 0.644193044939945, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 9942 + }, + { + "epoch": 0.09943, + "grad_norm": 0.6363516231209548, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 9943 + }, + { + "epoch": 0.09944, + "grad_norm": 0.5804019679381356, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 9944 + }, + { + "epoch": 0.09945, + "grad_norm": 0.538579805758387, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 9945 + }, + { + "epoch": 0.09946, + "grad_norm": 0.591697287934823, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 9946 + }, + { + "epoch": 0.09947, + "grad_norm": 0.7345913581539315, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 9947 + }, + { + "epoch": 0.09948, + "grad_norm": 0.8691886957481184, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 9948 + }, + { + "epoch": 0.09949, + "grad_norm": 0.9186101728747905, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 9949 + }, + { + "epoch": 0.0995, + "grad_norm": 0.8566674045177795, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 9950 + }, + { + "epoch": 0.09951, + "grad_norm": 0.6947636709958402, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 9951 + }, + { + "epoch": 0.09952, + "grad_norm": 0.756470522194281, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 9952 + }, + { + "epoch": 0.09953, + "grad_norm": 0.8319642062247762, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 9953 + }, + { + "epoch": 0.09954, + "grad_norm": 0.8944588085742335, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 9954 + }, + { + "epoch": 0.09955, + "grad_norm": 0.8780609834850799, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 9955 + }, + { + "epoch": 0.09956, + "grad_norm": 0.8966415589782, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 9956 + }, + { + "epoch": 0.09957, + "grad_norm": 0.9343888380316887, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 9957 + }, + { + "epoch": 0.09958, + "grad_norm": 0.8107702987068099, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 9958 + }, + { + "epoch": 0.09959, + "grad_norm": 0.7173964027454013, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 9959 + }, + { + "epoch": 0.0996, + "grad_norm": 0.7028801341508425, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 9960 + }, + { + "epoch": 0.09961, + "grad_norm": 0.7085300608233286, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 9961 + }, + { + "epoch": 0.09962, + "grad_norm": 0.701062036044723, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9962 + }, + { + "epoch": 0.09963, + "grad_norm": 0.7091697736024817, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 9963 + }, + { + "epoch": 0.09964, + "grad_norm": 0.7536040901138893, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 9964 + }, + { + "epoch": 0.09965, + "grad_norm": 0.7369827420479316, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 9965 + }, + { + "epoch": 0.09966, + "grad_norm": 0.7333635286577631, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 9966 + }, + { + "epoch": 0.09967, + "grad_norm": 0.7071651375949283, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 9967 + }, + { + "epoch": 0.09968, + "grad_norm": 0.6262764099639417, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 9968 + }, + { + "epoch": 0.09969, + "grad_norm": 0.5750366258555206, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 9969 + }, + { + "epoch": 0.0997, + "grad_norm": 0.6120225511137344, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 9970 + }, + { + "epoch": 0.09971, + "grad_norm": 0.5531541515552928, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 9971 + }, + { + "epoch": 0.09972, + "grad_norm": 0.5749478767024971, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 9972 + }, + { + "epoch": 0.09973, + "grad_norm": 0.5976969689059337, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 9973 + }, + { + "epoch": 0.09974, + "grad_norm": 0.6194515528623985, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 9974 + }, + { + "epoch": 0.09975, + "grad_norm": 0.6322846267378582, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 9975 + }, + { + "epoch": 0.09976, + "grad_norm": 0.6546823198819032, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 9976 + }, + { + "epoch": 0.09977, + "grad_norm": 0.6445131168354787, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 9977 + }, + { + "epoch": 0.09978, + "grad_norm": 0.5038313033053654, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 9978 + }, + { + "epoch": 0.09979, + "grad_norm": 0.5900532519723245, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 9979 + }, + { + "epoch": 0.0998, + "grad_norm": 0.70153854688536, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 9980 + }, + { + "epoch": 0.09981, + "grad_norm": 0.7636922149531054, + "learning_rate": 0.003, + "loss": 4.028, + "step": 9981 + }, + { + "epoch": 0.09982, + "grad_norm": 0.8701071149145536, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 9982 + }, + { + "epoch": 0.09983, + "grad_norm": 0.856295192628826, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 9983 + }, + { + "epoch": 0.09984, + "grad_norm": 0.7512372406931733, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 9984 + }, + { + "epoch": 0.09985, + "grad_norm": 0.6903668941425841, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 9985 + }, + { + "epoch": 0.09986, + "grad_norm": 0.6896663317344449, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 9986 + }, + { + "epoch": 0.09987, + "grad_norm": 0.7192055860207286, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 9987 + }, + { + "epoch": 0.09988, + "grad_norm": 0.7410482407589004, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 9988 + }, + { + "epoch": 0.09989, + "grad_norm": 0.7703203123360397, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 9989 + }, + { + "epoch": 0.0999, + "grad_norm": 0.8068363783340079, + "learning_rate": 0.003, + "loss": 4.032, + "step": 9990 + }, + { + "epoch": 0.09991, + "grad_norm": 0.8199834004107396, + "learning_rate": 0.003, + "loss": 4.052, + "step": 9991 + }, + { + "epoch": 0.09992, + "grad_norm": 0.7483854306979308, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 9992 + }, + { + "epoch": 0.09993, + "grad_norm": 0.7413305723324469, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 9993 + }, + { + "epoch": 0.09994, + "grad_norm": 0.6959641218100445, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 9994 + }, + { + "epoch": 0.09995, + "grad_norm": 0.6380717472714593, + "learning_rate": 0.003, + "loss": 4.042, + "step": 9995 + }, + { + "epoch": 0.09996, + "grad_norm": 0.5416886603215291, + "learning_rate": 0.003, + "loss": 4.017, + "step": 9996 + }, + { + "epoch": 0.09997, + "grad_norm": 0.6766391473081508, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 9997 + }, + { + "epoch": 0.09998, + "grad_norm": 0.674854782292625, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 9998 + }, + { + "epoch": 0.09999, + "grad_norm": 0.6803870156699644, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 9999 + }, + { + "epoch": 0.1, + "grad_norm": 0.738805539612618, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 10000 + }, + { + "epoch": 0.10001, + "grad_norm": 0.7879127897976834, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 10001 + }, + { + "epoch": 0.10002, + "grad_norm": 0.6927797418743683, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 10002 + }, + { + "epoch": 0.10003, + "grad_norm": 0.6829209383892021, + "learning_rate": 0.003, + "loss": 4.019, + "step": 10003 + }, + { + "epoch": 0.10004, + "grad_norm": 0.7237162821161365, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 10004 + }, + { + "epoch": 0.10005, + "grad_norm": 0.690961110200158, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 10005 + }, + { + "epoch": 0.10006, + "grad_norm": 0.6249865003925145, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 10006 + }, + { + "epoch": 0.10007, + "grad_norm": 0.5960963449979667, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 10007 + }, + { + "epoch": 0.10008, + "grad_norm": 0.7311374521787116, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 10008 + }, + { + "epoch": 0.10009, + "grad_norm": 0.7594271413735981, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 10009 + }, + { + "epoch": 0.1001, + "grad_norm": 0.8051339766222374, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 10010 + }, + { + "epoch": 0.10011, + "grad_norm": 0.920012937666569, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 10011 + }, + { + "epoch": 0.10012, + "grad_norm": 0.9165450493331897, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 10012 + }, + { + "epoch": 0.10013, + "grad_norm": 0.7765979697722353, + "learning_rate": 0.003, + "loss": 4.052, + "step": 10013 + }, + { + "epoch": 0.10014, + "grad_norm": 0.7187417304176765, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 10014 + }, + { + "epoch": 0.10015, + "grad_norm": 0.8340316434746166, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 10015 + }, + { + "epoch": 0.10016, + "grad_norm": 0.7198413413360849, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 10016 + }, + { + "epoch": 0.10017, + "grad_norm": 0.6708034739897859, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 10017 + }, + { + "epoch": 0.10018, + "grad_norm": 0.7194708650189626, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 10018 + }, + { + "epoch": 0.10019, + "grad_norm": 0.7307731732212398, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10019 + }, + { + "epoch": 0.1002, + "grad_norm": 0.8096736259729286, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 10020 + }, + { + "epoch": 0.10021, + "grad_norm": 0.7953454944295302, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 10021 + }, + { + "epoch": 0.10022, + "grad_norm": 0.7395603851454843, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 10022 + }, + { + "epoch": 0.10023, + "grad_norm": 0.7518203012555871, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10023 + }, + { + "epoch": 0.10024, + "grad_norm": 0.659882902136577, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 10024 + }, + { + "epoch": 0.10025, + "grad_norm": 0.6247780441375326, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10025 + }, + { + "epoch": 0.10026, + "grad_norm": 0.6296556518050307, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 10026 + }, + { + "epoch": 0.10027, + "grad_norm": 0.6497332870864666, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 10027 + }, + { + "epoch": 0.10028, + "grad_norm": 0.6317133990872036, + "learning_rate": 0.003, + "loss": 4.062, + "step": 10028 + }, + { + "epoch": 0.10029, + "grad_norm": 0.5258528925738564, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 10029 + }, + { + "epoch": 0.1003, + "grad_norm": 0.5031627757627122, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10030 + }, + { + "epoch": 0.10031, + "grad_norm": 0.4784967160553306, + "learning_rate": 0.003, + "loss": 4.06, + "step": 10031 + }, + { + "epoch": 0.10032, + "grad_norm": 0.46377540292156433, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 10032 + }, + { + "epoch": 0.10033, + "grad_norm": 0.44239340394754106, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 10033 + }, + { + "epoch": 0.10034, + "grad_norm": 0.44468223280860447, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 10034 + }, + { + "epoch": 0.10035, + "grad_norm": 0.5570866521142827, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 10035 + }, + { + "epoch": 0.10036, + "grad_norm": 0.7060010758286784, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 10036 + }, + { + "epoch": 0.10037, + "grad_norm": 0.8930944752228595, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 10037 + }, + { + "epoch": 0.10038, + "grad_norm": 0.9797058360339388, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 10038 + }, + { + "epoch": 0.10039, + "grad_norm": 0.975368585046754, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 10039 + }, + { + "epoch": 0.1004, + "grad_norm": 0.8023968776460533, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 10040 + }, + { + "epoch": 0.10041, + "grad_norm": 0.7022984959978487, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10041 + }, + { + "epoch": 0.10042, + "grad_norm": 0.7204181700846608, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 10042 + }, + { + "epoch": 0.10043, + "grad_norm": 0.7295357700440954, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 10043 + }, + { + "epoch": 0.10044, + "grad_norm": 0.7382965716389857, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 10044 + }, + { + "epoch": 0.10045, + "grad_norm": 0.7787964489495389, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 10045 + }, + { + "epoch": 0.10046, + "grad_norm": 0.8761872547205464, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 10046 + }, + { + "epoch": 0.10047, + "grad_norm": 0.8379863004371179, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 10047 + }, + { + "epoch": 0.10048, + "grad_norm": 0.8445258486193494, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 10048 + }, + { + "epoch": 0.10049, + "grad_norm": 0.8196940197771861, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 10049 + }, + { + "epoch": 0.1005, + "grad_norm": 0.7533057743380888, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 10050 + }, + { + "epoch": 0.10051, + "grad_norm": 0.7546428477053554, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 10051 + }, + { + "epoch": 0.10052, + "grad_norm": 0.6802034685877799, + "learning_rate": 0.003, + "loss": 4.025, + "step": 10052 + }, + { + "epoch": 0.10053, + "grad_norm": 0.6938192649625389, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 10053 + }, + { + "epoch": 0.10054, + "grad_norm": 0.6900368974607453, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 10054 + }, + { + "epoch": 0.10055, + "grad_norm": 0.7151877135015803, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 10055 + }, + { + "epoch": 0.10056, + "grad_norm": 0.7120903023477141, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 10056 + }, + { + "epoch": 0.10057, + "grad_norm": 0.704065252107797, + "learning_rate": 0.003, + "loss": 4.043, + "step": 10057 + }, + { + "epoch": 0.10058, + "grad_norm": 0.8208353680859148, + "learning_rate": 0.003, + "loss": 4.043, + "step": 10058 + }, + { + "epoch": 0.10059, + "grad_norm": 0.778153327074, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 10059 + }, + { + "epoch": 0.1006, + "grad_norm": 0.7163158149583583, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 10060 + }, + { + "epoch": 0.10061, + "grad_norm": 0.6699742319817853, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 10061 + }, + { + "epoch": 0.10062, + "grad_norm": 0.6930270461870826, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 10062 + }, + { + "epoch": 0.10063, + "grad_norm": 0.7148092007156384, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 10063 + }, + { + "epoch": 0.10064, + "grad_norm": 0.702595516652481, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 10064 + }, + { + "epoch": 0.10065, + "grad_norm": 0.646036544502377, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 10065 + }, + { + "epoch": 0.10066, + "grad_norm": 0.6453989121415348, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 10066 + }, + { + "epoch": 0.10067, + "grad_norm": 0.6014666409790891, + "learning_rate": 0.003, + "loss": 4.037, + "step": 10067 + }, + { + "epoch": 0.10068, + "grad_norm": 0.5318112659707298, + "learning_rate": 0.003, + "loss": 4.026, + "step": 10068 + }, + { + "epoch": 0.10069, + "grad_norm": 0.5199245414841062, + "learning_rate": 0.003, + "loss": 4.014, + "step": 10069 + }, + { + "epoch": 0.1007, + "grad_norm": 0.5227952923053703, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 10070 + }, + { + "epoch": 0.10071, + "grad_norm": 0.5822239896536773, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 10071 + }, + { + "epoch": 0.10072, + "grad_norm": 0.6617157545130878, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 10072 + }, + { + "epoch": 0.10073, + "grad_norm": 0.6800080296818483, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 10073 + }, + { + "epoch": 0.10074, + "grad_norm": 0.7047949354241044, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 10074 + }, + { + "epoch": 0.10075, + "grad_norm": 0.6661315884332217, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 10075 + }, + { + "epoch": 0.10076, + "grad_norm": 0.6812927860984451, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 10076 + }, + { + "epoch": 0.10077, + "grad_norm": 0.7494216596720659, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 10077 + }, + { + "epoch": 0.10078, + "grad_norm": 0.7273558722363268, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 10078 + }, + { + "epoch": 0.10079, + "grad_norm": 0.7645448845643933, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 10079 + }, + { + "epoch": 0.1008, + "grad_norm": 0.7900490469063955, + "learning_rate": 0.003, + "loss": 4.025, + "step": 10080 + }, + { + "epoch": 0.10081, + "grad_norm": 0.7988779836308102, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 10081 + }, + { + "epoch": 0.10082, + "grad_norm": 1.0085519338299518, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 10082 + }, + { + "epoch": 0.10083, + "grad_norm": 1.1161919450556166, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 10083 + }, + { + "epoch": 0.10084, + "grad_norm": 0.8347983630870844, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 10084 + }, + { + "epoch": 0.10085, + "grad_norm": 0.6627168271283793, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 10085 + }, + { + "epoch": 0.10086, + "grad_norm": 0.8314412865896796, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 10086 + }, + { + "epoch": 0.10087, + "grad_norm": 0.8625547357583004, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 10087 + }, + { + "epoch": 0.10088, + "grad_norm": 0.7824328006751063, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10088 + }, + { + "epoch": 0.10089, + "grad_norm": 0.755306395774321, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 10089 + }, + { + "epoch": 0.1009, + "grad_norm": 0.7002263280228616, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 10090 + }, + { + "epoch": 0.10091, + "grad_norm": 0.6515677930650133, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 10091 + }, + { + "epoch": 0.10092, + "grad_norm": 0.6500520308212595, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 10092 + }, + { + "epoch": 0.10093, + "grad_norm": 0.7224191155205028, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 10093 + }, + { + "epoch": 0.10094, + "grad_norm": 0.7417521980158183, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 10094 + }, + { + "epoch": 0.10095, + "grad_norm": 0.8130893763754354, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 10095 + }, + { + "epoch": 0.10096, + "grad_norm": 0.8042210386234582, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10096 + }, + { + "epoch": 0.10097, + "grad_norm": 0.8411606722066373, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 10097 + }, + { + "epoch": 0.10098, + "grad_norm": 0.7973870764201598, + "learning_rate": 0.003, + "loss": 4.029, + "step": 10098 + }, + { + "epoch": 0.10099, + "grad_norm": 0.6564698064637989, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 10099 + }, + { + "epoch": 0.101, + "grad_norm": 0.5586908354198192, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 10100 + }, + { + "epoch": 0.10101, + "grad_norm": 0.5474002505081143, + "learning_rate": 0.003, + "loss": 4.02, + "step": 10101 + }, + { + "epoch": 0.10102, + "grad_norm": 0.5278787564010663, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 10102 + }, + { + "epoch": 0.10103, + "grad_norm": 0.5543165426526394, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 10103 + }, + { + "epoch": 0.10104, + "grad_norm": 0.5705645524924402, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 10104 + }, + { + "epoch": 0.10105, + "grad_norm": 0.5837930993566786, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 10105 + }, + { + "epoch": 0.10106, + "grad_norm": 0.5574041397129644, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 10106 + }, + { + "epoch": 0.10107, + "grad_norm": 0.6348098374121675, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 10107 + }, + { + "epoch": 0.10108, + "grad_norm": 0.6255439015947829, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 10108 + }, + { + "epoch": 0.10109, + "grad_norm": 0.5951511843098067, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 10109 + }, + { + "epoch": 0.1011, + "grad_norm": 0.6169077478652734, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10110 + }, + { + "epoch": 0.10111, + "grad_norm": 0.6525852604703763, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 10111 + }, + { + "epoch": 0.10112, + "grad_norm": 0.628182392191787, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 10112 + }, + { + "epoch": 0.10113, + "grad_norm": 0.5750727375380823, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 10113 + }, + { + "epoch": 0.10114, + "grad_norm": 0.5521811236590936, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 10114 + }, + { + "epoch": 0.10115, + "grad_norm": 0.56523285320945, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 10115 + }, + { + "epoch": 0.10116, + "grad_norm": 0.6892304407957772, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 10116 + }, + { + "epoch": 0.10117, + "grad_norm": 0.8283629293068506, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 10117 + }, + { + "epoch": 0.10118, + "grad_norm": 1.0472148699881978, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 10118 + }, + { + "epoch": 0.10119, + "grad_norm": 0.83951444083863, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 10119 + }, + { + "epoch": 0.1012, + "grad_norm": 0.7460835475125817, + "learning_rate": 0.003, + "loss": 4.018, + "step": 10120 + }, + { + "epoch": 0.10121, + "grad_norm": 0.8138928001459755, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 10121 + }, + { + "epoch": 0.10122, + "grad_norm": 0.8448183935741884, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 10122 + }, + { + "epoch": 0.10123, + "grad_norm": 0.855151552579211, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 10123 + }, + { + "epoch": 0.10124, + "grad_norm": 1.0013308525017748, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 10124 + }, + { + "epoch": 0.10125, + "grad_norm": 0.8191535808529019, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 10125 + }, + { + "epoch": 0.10126, + "grad_norm": 0.807374585751279, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 10126 + }, + { + "epoch": 0.10127, + "grad_norm": 0.8871827134439824, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 10127 + }, + { + "epoch": 0.10128, + "grad_norm": 1.0584134474077507, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 10128 + }, + { + "epoch": 0.10129, + "grad_norm": 0.8639594049798054, + "learning_rate": 0.003, + "loss": 4.061, + "step": 10129 + }, + { + "epoch": 0.1013, + "grad_norm": 0.9312056321105163, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 10130 + }, + { + "epoch": 0.10131, + "grad_norm": 1.1743861433181717, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 10131 + }, + { + "epoch": 0.10132, + "grad_norm": 0.804980727459406, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 10132 + }, + { + "epoch": 0.10133, + "grad_norm": 0.7532246318973579, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 10133 + }, + { + "epoch": 0.10134, + "grad_norm": 0.7541816660659283, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10134 + }, + { + "epoch": 0.10135, + "grad_norm": 0.6889807174411486, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 10135 + }, + { + "epoch": 0.10136, + "grad_norm": 0.6260230877883078, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 10136 + }, + { + "epoch": 0.10137, + "grad_norm": 0.6881849733109103, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 10137 + }, + { + "epoch": 0.10138, + "grad_norm": 0.6999834769918349, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 10138 + }, + { + "epoch": 0.10139, + "grad_norm": 0.7152695618779359, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 10139 + }, + { + "epoch": 0.1014, + "grad_norm": 0.6497102797024055, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 10140 + }, + { + "epoch": 0.10141, + "grad_norm": 0.6541373901997861, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 10141 + }, + { + "epoch": 0.10142, + "grad_norm": 0.5631288792409277, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 10142 + }, + { + "epoch": 0.10143, + "grad_norm": 0.566586831995206, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 10143 + }, + { + "epoch": 0.10144, + "grad_norm": 0.5652312826709329, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10144 + }, + { + "epoch": 0.10145, + "grad_norm": 0.5688793301912161, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 10145 + }, + { + "epoch": 0.10146, + "grad_norm": 0.6184405002556627, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 10146 + }, + { + "epoch": 0.10147, + "grad_norm": 0.5889445691739414, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 10147 + }, + { + "epoch": 0.10148, + "grad_norm": 0.5751390248489617, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 10148 + }, + { + "epoch": 0.10149, + "grad_norm": 0.5313762723496491, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 10149 + }, + { + "epoch": 0.1015, + "grad_norm": 0.4909011829248259, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 10150 + }, + { + "epoch": 0.10151, + "grad_norm": 0.48593478008885094, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 10151 + }, + { + "epoch": 0.10152, + "grad_norm": 0.4463373006073096, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 10152 + }, + { + "epoch": 0.10153, + "grad_norm": 0.46610317967214226, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 10153 + }, + { + "epoch": 0.10154, + "grad_norm": 0.5054276285147266, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 10154 + }, + { + "epoch": 0.10155, + "grad_norm": 0.5159759849118725, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 10155 + }, + { + "epoch": 0.10156, + "grad_norm": 0.5696070713434969, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 10156 + }, + { + "epoch": 0.10157, + "grad_norm": 0.65443677188871, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 10157 + }, + { + "epoch": 0.10158, + "grad_norm": 0.683457584313693, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 10158 + }, + { + "epoch": 0.10159, + "grad_norm": 0.7311732224993086, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 10159 + }, + { + "epoch": 0.1016, + "grad_norm": 0.7360328449505, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 10160 + }, + { + "epoch": 0.10161, + "grad_norm": 0.7926899669775646, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 10161 + }, + { + "epoch": 0.10162, + "grad_norm": 0.8351466028738757, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 10162 + }, + { + "epoch": 0.10163, + "grad_norm": 0.8908484587405915, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10163 + }, + { + "epoch": 0.10164, + "grad_norm": 0.7971803909553479, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 10164 + }, + { + "epoch": 0.10165, + "grad_norm": 0.6970535882452454, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 10165 + }, + { + "epoch": 0.10166, + "grad_norm": 0.6936132713050579, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 10166 + }, + { + "epoch": 0.10167, + "grad_norm": 0.6432475822464314, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10167 + }, + { + "epoch": 0.10168, + "grad_norm": 0.5880981114382495, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 10168 + }, + { + "epoch": 0.10169, + "grad_norm": 0.5462390118192577, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 10169 + }, + { + "epoch": 0.1017, + "grad_norm": 0.48697287218406543, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 10170 + }, + { + "epoch": 0.10171, + "grad_norm": 0.4712736794548392, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 10171 + }, + { + "epoch": 0.10172, + "grad_norm": 0.5222001684274312, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 10172 + }, + { + "epoch": 0.10173, + "grad_norm": 0.5841599035760559, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 10173 + }, + { + "epoch": 0.10174, + "grad_norm": 0.6913264114322897, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 10174 + }, + { + "epoch": 0.10175, + "grad_norm": 0.8268924724541994, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 10175 + }, + { + "epoch": 0.10176, + "grad_norm": 0.8435710626598261, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 10176 + }, + { + "epoch": 0.10177, + "grad_norm": 0.8095268227697132, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 10177 + }, + { + "epoch": 0.10178, + "grad_norm": 0.8090080160761893, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 10178 + }, + { + "epoch": 0.10179, + "grad_norm": 0.9289264227573335, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 10179 + }, + { + "epoch": 0.1018, + "grad_norm": 0.8575258057917386, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 10180 + }, + { + "epoch": 0.10181, + "grad_norm": 0.9409624332402622, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 10181 + }, + { + "epoch": 0.10182, + "grad_norm": 1.0206378058190897, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 10182 + }, + { + "epoch": 0.10183, + "grad_norm": 0.7900024735758078, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 10183 + }, + { + "epoch": 0.10184, + "grad_norm": 0.7416492574393173, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 10184 + }, + { + "epoch": 0.10185, + "grad_norm": 0.7728598154980276, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 10185 + }, + { + "epoch": 0.10186, + "grad_norm": 0.8289578846931824, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 10186 + }, + { + "epoch": 0.10187, + "grad_norm": 0.8933719111642519, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 10187 + }, + { + "epoch": 0.10188, + "grad_norm": 0.9222139997209723, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 10188 + }, + { + "epoch": 0.10189, + "grad_norm": 0.9108160082636183, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 10189 + }, + { + "epoch": 0.1019, + "grad_norm": 1.0125951978225802, + "learning_rate": 0.003, + "loss": 4.0593, + "step": 10190 + }, + { + "epoch": 0.10191, + "grad_norm": 1.1386507407678024, + "learning_rate": 0.003, + "loss": 4.0726, + "step": 10191 + }, + { + "epoch": 0.10192, + "grad_norm": 0.7436290588259256, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 10192 + }, + { + "epoch": 0.10193, + "grad_norm": 0.7646380006686426, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 10193 + }, + { + "epoch": 0.10194, + "grad_norm": 0.8209583150500758, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 10194 + }, + { + "epoch": 0.10195, + "grad_norm": 0.7704023377407652, + "learning_rate": 0.003, + "loss": 4.0675, + "step": 10195 + }, + { + "epoch": 0.10196, + "grad_norm": 0.7059589629512826, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 10196 + }, + { + "epoch": 0.10197, + "grad_norm": 0.6100815201238473, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 10197 + }, + { + "epoch": 0.10198, + "grad_norm": 0.578515492856178, + "learning_rate": 0.003, + "loss": 4.0689, + "step": 10198 + }, + { + "epoch": 0.10199, + "grad_norm": 0.6152738101855209, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 10199 + }, + { + "epoch": 0.102, + "grad_norm": 0.6467786320923281, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10200 + }, + { + "epoch": 0.10201, + "grad_norm": 0.5045752052817358, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 10201 + }, + { + "epoch": 0.10202, + "grad_norm": 0.5293187153750024, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 10202 + }, + { + "epoch": 0.10203, + "grad_norm": 0.5834911184509722, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 10203 + }, + { + "epoch": 0.10204, + "grad_norm": 0.6225302444237483, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 10204 + }, + { + "epoch": 0.10205, + "grad_norm": 0.7250482711155155, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 10205 + }, + { + "epoch": 0.10206, + "grad_norm": 0.8219363954611312, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 10206 + }, + { + "epoch": 0.10207, + "grad_norm": 0.8414447701054694, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 10207 + }, + { + "epoch": 0.10208, + "grad_norm": 0.7474835133010564, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 10208 + }, + { + "epoch": 0.10209, + "grad_norm": 0.7193400122390704, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 10209 + }, + { + "epoch": 0.1021, + "grad_norm": 0.7232316729482701, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 10210 + }, + { + "epoch": 0.10211, + "grad_norm": 0.7581026741208078, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 10211 + }, + { + "epoch": 0.10212, + "grad_norm": 0.8111722282656838, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 10212 + }, + { + "epoch": 0.10213, + "grad_norm": 0.7876519043307583, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 10213 + }, + { + "epoch": 0.10214, + "grad_norm": 0.7481380136033134, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 10214 + }, + { + "epoch": 0.10215, + "grad_norm": 0.8018740918115793, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 10215 + }, + { + "epoch": 0.10216, + "grad_norm": 0.8125780094887491, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 10216 + }, + { + "epoch": 0.10217, + "grad_norm": 0.6887890154144948, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 10217 + }, + { + "epoch": 0.10218, + "grad_norm": 0.7591141840883562, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 10218 + }, + { + "epoch": 0.10219, + "grad_norm": 0.760293736263083, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 10219 + }, + { + "epoch": 0.1022, + "grad_norm": 0.7293120943583128, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 10220 + }, + { + "epoch": 0.10221, + "grad_norm": 0.6472417034807832, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 10221 + }, + { + "epoch": 0.10222, + "grad_norm": 0.6415087712926342, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 10222 + }, + { + "epoch": 0.10223, + "grad_norm": 0.6045484418376729, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 10223 + }, + { + "epoch": 0.10224, + "grad_norm": 0.5659410721513285, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 10224 + }, + { + "epoch": 0.10225, + "grad_norm": 0.562516793903761, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 10225 + }, + { + "epoch": 0.10226, + "grad_norm": 0.5222596467216768, + "learning_rate": 0.003, + "loss": 4.076, + "step": 10226 + }, + { + "epoch": 0.10227, + "grad_norm": 0.46246456619710413, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10227 + }, + { + "epoch": 0.10228, + "grad_norm": 0.5031287720457579, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 10228 + }, + { + "epoch": 0.10229, + "grad_norm": 0.5693813607581187, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 10229 + }, + { + "epoch": 0.1023, + "grad_norm": 0.5728485235531475, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 10230 + }, + { + "epoch": 0.10231, + "grad_norm": 0.4945081097700335, + "learning_rate": 0.003, + "loss": 4.029, + "step": 10231 + }, + { + "epoch": 0.10232, + "grad_norm": 0.490331254654848, + "learning_rate": 0.003, + "loss": 4.031, + "step": 10232 + }, + { + "epoch": 0.10233, + "grad_norm": 0.5428851585883142, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 10233 + }, + { + "epoch": 0.10234, + "grad_norm": 0.6168227515888373, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 10234 + }, + { + "epoch": 0.10235, + "grad_norm": 0.6541734566249592, + "learning_rate": 0.003, + "loss": 4.022, + "step": 10235 + }, + { + "epoch": 0.10236, + "grad_norm": 0.6770761230825558, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 10236 + }, + { + "epoch": 0.10237, + "grad_norm": 0.6667733151129375, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 10237 + }, + { + "epoch": 0.10238, + "grad_norm": 0.5435852444011133, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 10238 + }, + { + "epoch": 0.10239, + "grad_norm": 0.5392990046890335, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 10239 + }, + { + "epoch": 0.1024, + "grad_norm": 0.6653196275144262, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 10240 + }, + { + "epoch": 0.10241, + "grad_norm": 0.6994424127129869, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 10241 + }, + { + "epoch": 0.10242, + "grad_norm": 0.7072185651119459, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 10242 + }, + { + "epoch": 0.10243, + "grad_norm": 0.7686580020973927, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 10243 + }, + { + "epoch": 0.10244, + "grad_norm": 0.8268106946388039, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 10244 + }, + { + "epoch": 0.10245, + "grad_norm": 0.7524408903015769, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 10245 + }, + { + "epoch": 0.10246, + "grad_norm": 0.8196162929525966, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 10246 + }, + { + "epoch": 0.10247, + "grad_norm": 1.0186044474328348, + "learning_rate": 0.003, + "loss": 4.0662, + "step": 10247 + }, + { + "epoch": 0.10248, + "grad_norm": 1.1097492909390607, + "learning_rate": 0.003, + "loss": 4.0642, + "step": 10248 + }, + { + "epoch": 0.10249, + "grad_norm": 0.7756898544018952, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 10249 + }, + { + "epoch": 0.1025, + "grad_norm": 0.7532825210784639, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 10250 + }, + { + "epoch": 0.10251, + "grad_norm": 0.8505655262410471, + "learning_rate": 0.003, + "loss": 4.042, + "step": 10251 + }, + { + "epoch": 0.10252, + "grad_norm": 0.9181204522292044, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 10252 + }, + { + "epoch": 0.10253, + "grad_norm": 0.8300102636289165, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 10253 + }, + { + "epoch": 0.10254, + "grad_norm": 0.7230477402216509, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 10254 + }, + { + "epoch": 0.10255, + "grad_norm": 0.7101868515739244, + "learning_rate": 0.003, + "loss": 4.028, + "step": 10255 + }, + { + "epoch": 0.10256, + "grad_norm": 0.7936589523378054, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 10256 + }, + { + "epoch": 0.10257, + "grad_norm": 0.8487287246246443, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 10257 + }, + { + "epoch": 0.10258, + "grad_norm": 0.788549770728514, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 10258 + }, + { + "epoch": 0.10259, + "grad_norm": 0.7237944895680625, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 10259 + }, + { + "epoch": 0.1026, + "grad_norm": 0.8263930140436775, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 10260 + }, + { + "epoch": 0.10261, + "grad_norm": 0.8537809490927548, + "learning_rate": 0.003, + "loss": 4.04, + "step": 10261 + }, + { + "epoch": 0.10262, + "grad_norm": 0.8263952461026466, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 10262 + }, + { + "epoch": 0.10263, + "grad_norm": 0.8143986516241929, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 10263 + }, + { + "epoch": 0.10264, + "grad_norm": 0.652857375868988, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10264 + }, + { + "epoch": 0.10265, + "grad_norm": 0.7160848039852288, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 10265 + }, + { + "epoch": 0.10266, + "grad_norm": 0.8234282723454794, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 10266 + }, + { + "epoch": 0.10267, + "grad_norm": 0.9417515761502189, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 10267 + }, + { + "epoch": 0.10268, + "grad_norm": 0.7692117469844199, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 10268 + }, + { + "epoch": 0.10269, + "grad_norm": 0.6760728807490278, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 10269 + }, + { + "epoch": 0.1027, + "grad_norm": 0.5857076643819016, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 10270 + }, + { + "epoch": 0.10271, + "grad_norm": 0.5985288864602891, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 10271 + }, + { + "epoch": 0.10272, + "grad_norm": 0.6727738356983207, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 10272 + }, + { + "epoch": 0.10273, + "grad_norm": 0.6836437967225992, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 10273 + }, + { + "epoch": 0.10274, + "grad_norm": 0.7063857135217846, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 10274 + }, + { + "epoch": 0.10275, + "grad_norm": 0.6718476233375692, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 10275 + }, + { + "epoch": 0.10276, + "grad_norm": 0.571958340122604, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 10276 + }, + { + "epoch": 0.10277, + "grad_norm": 0.5478671884485615, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 10277 + }, + { + "epoch": 0.10278, + "grad_norm": 0.5632395626714399, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 10278 + }, + { + "epoch": 0.10279, + "grad_norm": 0.5398022597295509, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 10279 + }, + { + "epoch": 0.1028, + "grad_norm": 0.574170681167508, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 10280 + }, + { + "epoch": 0.10281, + "grad_norm": 0.5419637276003975, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 10281 + }, + { + "epoch": 0.10282, + "grad_norm": 0.5057606063332976, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 10282 + }, + { + "epoch": 0.10283, + "grad_norm": 0.4765575856507067, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 10283 + }, + { + "epoch": 0.10284, + "grad_norm": 0.5225757517778052, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 10284 + }, + { + "epoch": 0.10285, + "grad_norm": 0.5096927885214138, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 10285 + }, + { + "epoch": 0.10286, + "grad_norm": 0.5394215318857914, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 10286 + }, + { + "epoch": 0.10287, + "grad_norm": 0.6002951322452336, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 10287 + }, + { + "epoch": 0.10288, + "grad_norm": 0.6853280988052667, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 10288 + }, + { + "epoch": 0.10289, + "grad_norm": 0.7651162840557956, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 10289 + }, + { + "epoch": 0.1029, + "grad_norm": 0.8276000581708793, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 10290 + }, + { + "epoch": 0.10291, + "grad_norm": 0.8775186843026329, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 10291 + }, + { + "epoch": 0.10292, + "grad_norm": 0.7678299391845492, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 10292 + }, + { + "epoch": 0.10293, + "grad_norm": 0.7531374230566051, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 10293 + }, + { + "epoch": 0.10294, + "grad_norm": 0.7484717774837976, + "learning_rate": 0.003, + "loss": 4.054, + "step": 10294 + }, + { + "epoch": 0.10295, + "grad_norm": 0.7397927239999857, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 10295 + }, + { + "epoch": 0.10296, + "grad_norm": 0.7838729786359416, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 10296 + }, + { + "epoch": 0.10297, + "grad_norm": 0.7666017371373038, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10297 + }, + { + "epoch": 0.10298, + "grad_norm": 0.8549760800987974, + "learning_rate": 0.003, + "loss": 4.0683, + "step": 10298 + }, + { + "epoch": 0.10299, + "grad_norm": 0.9792365900494505, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 10299 + }, + { + "epoch": 0.103, + "grad_norm": 1.0150749488180213, + "learning_rate": 0.003, + "loss": 4.052, + "step": 10300 + }, + { + "epoch": 0.10301, + "grad_norm": 0.9340391340876893, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 10301 + }, + { + "epoch": 0.10302, + "grad_norm": 0.8374644016968678, + "learning_rate": 0.003, + "loss": 4.0721, + "step": 10302 + }, + { + "epoch": 0.10303, + "grad_norm": 0.8017372805021792, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10303 + }, + { + "epoch": 0.10304, + "grad_norm": 0.916247534205975, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 10304 + }, + { + "epoch": 0.10305, + "grad_norm": 0.8868589441255285, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 10305 + }, + { + "epoch": 0.10306, + "grad_norm": 0.7935717616287131, + "learning_rate": 0.003, + "loss": 4.067, + "step": 10306 + }, + { + "epoch": 0.10307, + "grad_norm": 0.7291838520074945, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 10307 + }, + { + "epoch": 0.10308, + "grad_norm": 0.6368913354729271, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 10308 + }, + { + "epoch": 0.10309, + "grad_norm": 0.6372673155626488, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 10309 + }, + { + "epoch": 0.1031, + "grad_norm": 0.6183739375217865, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 10310 + }, + { + "epoch": 0.10311, + "grad_norm": 0.6206216136183226, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 10311 + }, + { + "epoch": 0.10312, + "grad_norm": 0.6479425006954269, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 10312 + }, + { + "epoch": 0.10313, + "grad_norm": 0.6739550610788139, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 10313 + }, + { + "epoch": 0.10314, + "grad_norm": 0.6236254904162777, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 10314 + }, + { + "epoch": 0.10315, + "grad_norm": 0.6344421932047328, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10315 + }, + { + "epoch": 0.10316, + "grad_norm": 0.7290188234487568, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 10316 + }, + { + "epoch": 0.10317, + "grad_norm": 0.7166478143789822, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10317 + }, + { + "epoch": 0.10318, + "grad_norm": 0.7062794245708208, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 10318 + }, + { + "epoch": 0.10319, + "grad_norm": 0.8156711626991822, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 10319 + }, + { + "epoch": 0.1032, + "grad_norm": 0.6873150916081839, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10320 + }, + { + "epoch": 0.10321, + "grad_norm": 0.5842565687140238, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10321 + }, + { + "epoch": 0.10322, + "grad_norm": 0.6306900476375152, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 10322 + }, + { + "epoch": 0.10323, + "grad_norm": 0.6013181850860404, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 10323 + }, + { + "epoch": 0.10324, + "grad_norm": 0.639286700289023, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 10324 + }, + { + "epoch": 0.10325, + "grad_norm": 0.6375414577089098, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 10325 + }, + { + "epoch": 0.10326, + "grad_norm": 0.6561973618184916, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10326 + }, + { + "epoch": 0.10327, + "grad_norm": 0.6693884103376375, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 10327 + }, + { + "epoch": 0.10328, + "grad_norm": 0.6906622531222646, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 10328 + }, + { + "epoch": 0.10329, + "grad_norm": 0.6571882026248322, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 10329 + }, + { + "epoch": 0.1033, + "grad_norm": 0.7377790186936102, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 10330 + }, + { + "epoch": 0.10331, + "grad_norm": 0.8784358970467926, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 10331 + }, + { + "epoch": 0.10332, + "grad_norm": 0.9315417358159878, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 10332 + }, + { + "epoch": 0.10333, + "grad_norm": 0.8383094952923359, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 10333 + }, + { + "epoch": 0.10334, + "grad_norm": 0.7145014140889367, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 10334 + }, + { + "epoch": 0.10335, + "grad_norm": 0.719625845108547, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 10335 + }, + { + "epoch": 0.10336, + "grad_norm": 0.7022498963102233, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 10336 + }, + { + "epoch": 0.10337, + "grad_norm": 0.631805789601279, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 10337 + }, + { + "epoch": 0.10338, + "grad_norm": 0.7246849470962254, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 10338 + }, + { + "epoch": 0.10339, + "grad_norm": 0.8415921265112231, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 10339 + }, + { + "epoch": 0.1034, + "grad_norm": 0.822515548794032, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 10340 + }, + { + "epoch": 0.10341, + "grad_norm": 0.8842493081693986, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 10341 + }, + { + "epoch": 0.10342, + "grad_norm": 0.8399070430658626, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 10342 + }, + { + "epoch": 0.10343, + "grad_norm": 0.7055243469622291, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 10343 + }, + { + "epoch": 0.10344, + "grad_norm": 0.7908596982305551, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 10344 + }, + { + "epoch": 0.10345, + "grad_norm": 0.7304161556081027, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 10345 + }, + { + "epoch": 0.10346, + "grad_norm": 0.6935605070289242, + "learning_rate": 0.003, + "loss": 4.009, + "step": 10346 + }, + { + "epoch": 0.10347, + "grad_norm": 0.6294536785185108, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 10347 + }, + { + "epoch": 0.10348, + "grad_norm": 0.6465402848172335, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 10348 + }, + { + "epoch": 0.10349, + "grad_norm": 0.700571503916212, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 10349 + }, + { + "epoch": 0.1035, + "grad_norm": 0.6541532241255231, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 10350 + }, + { + "epoch": 0.10351, + "grad_norm": 0.637598532794208, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 10351 + }, + { + "epoch": 0.10352, + "grad_norm": 0.5783913732798877, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 10352 + }, + { + "epoch": 0.10353, + "grad_norm": 0.5947822619014511, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 10353 + }, + { + "epoch": 0.10354, + "grad_norm": 0.5540420150954074, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 10354 + }, + { + "epoch": 0.10355, + "grad_norm": 0.582063922710021, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 10355 + }, + { + "epoch": 0.10356, + "grad_norm": 0.5768040255552972, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 10356 + }, + { + "epoch": 0.10357, + "grad_norm": 0.6411867641320751, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 10357 + }, + { + "epoch": 0.10358, + "grad_norm": 0.744601108422105, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 10358 + }, + { + "epoch": 0.10359, + "grad_norm": 0.7348904777332892, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 10359 + }, + { + "epoch": 0.1036, + "grad_norm": 0.7323866949363924, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 10360 + }, + { + "epoch": 0.10361, + "grad_norm": 0.6180653281877111, + "learning_rate": 0.003, + "loss": 3.985, + "step": 10361 + }, + { + "epoch": 0.10362, + "grad_norm": 0.6391665296507255, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 10362 + }, + { + "epoch": 0.10363, + "grad_norm": 0.7611676231587599, + "learning_rate": 0.003, + "loss": 4.016, + "step": 10363 + }, + { + "epoch": 0.10364, + "grad_norm": 0.8348670952787945, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10364 + }, + { + "epoch": 0.10365, + "grad_norm": 0.8707946593613695, + "learning_rate": 0.003, + "loss": 4.0626, + "step": 10365 + }, + { + "epoch": 0.10366, + "grad_norm": 0.951243459188684, + "learning_rate": 0.003, + "loss": 4.043, + "step": 10366 + }, + { + "epoch": 0.10367, + "grad_norm": 1.120736260247303, + "learning_rate": 0.003, + "loss": 4.0668, + "step": 10367 + }, + { + "epoch": 0.10368, + "grad_norm": 0.8666169570114878, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 10368 + }, + { + "epoch": 0.10369, + "grad_norm": 0.9093901848280879, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 10369 + }, + { + "epoch": 0.1037, + "grad_norm": 0.8413888839441124, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 10370 + }, + { + "epoch": 0.10371, + "grad_norm": 0.7247530316309022, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 10371 + }, + { + "epoch": 0.10372, + "grad_norm": 0.6499749442759747, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 10372 + }, + { + "epoch": 0.10373, + "grad_norm": 0.5962778900987583, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 10373 + }, + { + "epoch": 0.10374, + "grad_norm": 0.537622457260128, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 10374 + }, + { + "epoch": 0.10375, + "grad_norm": 0.5286934774730946, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 10375 + }, + { + "epoch": 0.10376, + "grad_norm": 0.5377168361168837, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 10376 + }, + { + "epoch": 0.10377, + "grad_norm": 0.5241472551950065, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 10377 + }, + { + "epoch": 0.10378, + "grad_norm": 0.5304291146533516, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 10378 + }, + { + "epoch": 0.10379, + "grad_norm": 0.6619683981104225, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 10379 + }, + { + "epoch": 0.1038, + "grad_norm": 0.7800836070738506, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 10380 + }, + { + "epoch": 0.10381, + "grad_norm": 0.7054471909471106, + "learning_rate": 0.003, + "loss": 4.03, + "step": 10381 + }, + { + "epoch": 0.10382, + "grad_norm": 0.6972904342546717, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 10382 + }, + { + "epoch": 0.10383, + "grad_norm": 0.6881482564536944, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10383 + }, + { + "epoch": 0.10384, + "grad_norm": 0.6185731270251982, + "learning_rate": 0.003, + "loss": 4.039, + "step": 10384 + }, + { + "epoch": 0.10385, + "grad_norm": 0.5827527073267142, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 10385 + }, + { + "epoch": 0.10386, + "grad_norm": 0.5597881875379053, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10386 + }, + { + "epoch": 0.10387, + "grad_norm": 0.5589913067180577, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 10387 + }, + { + "epoch": 0.10388, + "grad_norm": 0.6697868888778852, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 10388 + }, + { + "epoch": 0.10389, + "grad_norm": 0.7669911082272918, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 10389 + }, + { + "epoch": 0.1039, + "grad_norm": 0.8283946330051588, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 10390 + }, + { + "epoch": 0.10391, + "grad_norm": 0.836593385144346, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 10391 + }, + { + "epoch": 0.10392, + "grad_norm": 0.7697149566969124, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 10392 + }, + { + "epoch": 0.10393, + "grad_norm": 0.6599361134838084, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 10393 + }, + { + "epoch": 0.10394, + "grad_norm": 0.6894078035226862, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 10394 + }, + { + "epoch": 0.10395, + "grad_norm": 0.786509881013533, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 10395 + }, + { + "epoch": 0.10396, + "grad_norm": 0.8124345324274491, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10396 + }, + { + "epoch": 0.10397, + "grad_norm": 0.7906511614353754, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 10397 + }, + { + "epoch": 0.10398, + "grad_norm": 0.7337753614245606, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 10398 + }, + { + "epoch": 0.10399, + "grad_norm": 0.7375916508933933, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 10399 + }, + { + "epoch": 0.104, + "grad_norm": 0.6797063534956214, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 10400 + }, + { + "epoch": 0.10401, + "grad_norm": 0.720881845880351, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 10401 + }, + { + "epoch": 0.10402, + "grad_norm": 0.59662785995047, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 10402 + }, + { + "epoch": 0.10403, + "grad_norm": 0.5866448801305522, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 10403 + }, + { + "epoch": 0.10404, + "grad_norm": 0.5356662291793891, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 10404 + }, + { + "epoch": 0.10405, + "grad_norm": 0.5042810879884669, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 10405 + }, + { + "epoch": 0.10406, + "grad_norm": 0.5461613874556226, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 10406 + }, + { + "epoch": 0.10407, + "grad_norm": 0.6258528808607489, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 10407 + }, + { + "epoch": 0.10408, + "grad_norm": 0.8468115559722816, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 10408 + }, + { + "epoch": 0.10409, + "grad_norm": 1.014822156744099, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 10409 + }, + { + "epoch": 0.1041, + "grad_norm": 0.7978714526199777, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 10410 + }, + { + "epoch": 0.10411, + "grad_norm": 0.678046449962411, + "learning_rate": 0.003, + "loss": 4.019, + "step": 10411 + }, + { + "epoch": 0.10412, + "grad_norm": 0.7194772669760052, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 10412 + }, + { + "epoch": 0.10413, + "grad_norm": 0.6713145790893477, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10413 + }, + { + "epoch": 0.10414, + "grad_norm": 0.6678797954501012, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 10414 + }, + { + "epoch": 0.10415, + "grad_norm": 0.7436863788969632, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 10415 + }, + { + "epoch": 0.10416, + "grad_norm": 0.8173466147304688, + "learning_rate": 0.003, + "loss": 4.0526, + "step": 10416 + }, + { + "epoch": 0.10417, + "grad_norm": 0.9446403761899012, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 10417 + }, + { + "epoch": 0.10418, + "grad_norm": 0.985790223147594, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 10418 + }, + { + "epoch": 0.10419, + "grad_norm": 0.946586203971559, + "learning_rate": 0.003, + "loss": 4.0872, + "step": 10419 + }, + { + "epoch": 0.1042, + "grad_norm": 0.913089800308823, + "learning_rate": 0.003, + "loss": 4.0652, + "step": 10420 + }, + { + "epoch": 0.10421, + "grad_norm": 0.8343729467399144, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 10421 + }, + { + "epoch": 0.10422, + "grad_norm": 0.7844207732712799, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 10422 + }, + { + "epoch": 0.10423, + "grad_norm": 0.7364577956994709, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 10423 + }, + { + "epoch": 0.10424, + "grad_norm": 0.7471439923826526, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10424 + }, + { + "epoch": 0.10425, + "grad_norm": 0.7676057974102031, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 10425 + }, + { + "epoch": 0.10426, + "grad_norm": 0.7120072418560034, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 10426 + }, + { + "epoch": 0.10427, + "grad_norm": 0.7466538155689676, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 10427 + }, + { + "epoch": 0.10428, + "grad_norm": 0.7516454262660435, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10428 + }, + { + "epoch": 0.10429, + "grad_norm": 0.7290780046405567, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 10429 + }, + { + "epoch": 0.1043, + "grad_norm": 0.599218448386974, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 10430 + }, + { + "epoch": 0.10431, + "grad_norm": 0.5683895987561769, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10431 + }, + { + "epoch": 0.10432, + "grad_norm": 0.6497990546633713, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 10432 + }, + { + "epoch": 0.10433, + "grad_norm": 0.6356558022638671, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 10433 + }, + { + "epoch": 0.10434, + "grad_norm": 0.5793672638695669, + "learning_rate": 0.003, + "loss": 4.017, + "step": 10434 + }, + { + "epoch": 0.10435, + "grad_norm": 0.5649713197243685, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 10435 + }, + { + "epoch": 0.10436, + "grad_norm": 0.5872357230676422, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 10436 + }, + { + "epoch": 0.10437, + "grad_norm": 0.5969357659286831, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 10437 + }, + { + "epoch": 0.10438, + "grad_norm": 0.7067684180972401, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 10438 + }, + { + "epoch": 0.10439, + "grad_norm": 0.8125206642222753, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 10439 + }, + { + "epoch": 0.1044, + "grad_norm": 0.8569503703854483, + "learning_rate": 0.003, + "loss": 4.041, + "step": 10440 + }, + { + "epoch": 0.10441, + "grad_norm": 0.8693331804534691, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 10441 + }, + { + "epoch": 0.10442, + "grad_norm": 0.8619601104608854, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 10442 + }, + { + "epoch": 0.10443, + "grad_norm": 0.6820204975681383, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 10443 + }, + { + "epoch": 0.10444, + "grad_norm": 0.7137591275453445, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 10444 + }, + { + "epoch": 0.10445, + "grad_norm": 0.6707041186003744, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 10445 + }, + { + "epoch": 0.10446, + "grad_norm": 0.7271260956437723, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 10446 + }, + { + "epoch": 0.10447, + "grad_norm": 0.7113479601987269, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 10447 + }, + { + "epoch": 0.10448, + "grad_norm": 0.5479850237341684, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 10448 + }, + { + "epoch": 0.10449, + "grad_norm": 0.5314387548835755, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 10449 + }, + { + "epoch": 0.1045, + "grad_norm": 0.5630899156987479, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 10450 + }, + { + "epoch": 0.10451, + "grad_norm": 0.5216940685474791, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 10451 + }, + { + "epoch": 0.10452, + "grad_norm": 0.4967063298381825, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 10452 + }, + { + "epoch": 0.10453, + "grad_norm": 0.4892711314338148, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 10453 + }, + { + "epoch": 0.10454, + "grad_norm": 0.44235199974682954, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 10454 + }, + { + "epoch": 0.10455, + "grad_norm": 0.4351382572749046, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 10455 + }, + { + "epoch": 0.10456, + "grad_norm": 0.4808614351811253, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10456 + }, + { + "epoch": 0.10457, + "grad_norm": 0.6472685363303916, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 10457 + }, + { + "epoch": 0.10458, + "grad_norm": 0.7487997476109776, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 10458 + }, + { + "epoch": 0.10459, + "grad_norm": 0.7993430864455091, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 10459 + }, + { + "epoch": 0.1046, + "grad_norm": 1.0118653531985184, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 10460 + }, + { + "epoch": 0.10461, + "grad_norm": 1.2328987777030882, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 10461 + }, + { + "epoch": 0.10462, + "grad_norm": 0.751754176680235, + "learning_rate": 0.003, + "loss": 4.043, + "step": 10462 + }, + { + "epoch": 0.10463, + "grad_norm": 0.6492175135096198, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 10463 + }, + { + "epoch": 0.10464, + "grad_norm": 0.6566260389176134, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 10464 + }, + { + "epoch": 0.10465, + "grad_norm": 0.5668787195294556, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 10465 + }, + { + "epoch": 0.10466, + "grad_norm": 0.6151625685344514, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 10466 + }, + { + "epoch": 0.10467, + "grad_norm": 0.6271273354825487, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 10467 + }, + { + "epoch": 0.10468, + "grad_norm": 0.7061920787155498, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 10468 + }, + { + "epoch": 0.10469, + "grad_norm": 0.7172487137802546, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 10469 + }, + { + "epoch": 0.1047, + "grad_norm": 0.7828953683854034, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 10470 + }, + { + "epoch": 0.10471, + "grad_norm": 0.969365379934823, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 10471 + }, + { + "epoch": 0.10472, + "grad_norm": 1.1226329378148765, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 10472 + }, + { + "epoch": 0.10473, + "grad_norm": 0.8785961596671167, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 10473 + }, + { + "epoch": 0.10474, + "grad_norm": 0.7971551382356206, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 10474 + }, + { + "epoch": 0.10475, + "grad_norm": 0.866798672084192, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 10475 + }, + { + "epoch": 0.10476, + "grad_norm": 0.9523202437544412, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 10476 + }, + { + "epoch": 0.10477, + "grad_norm": 0.9462315651776281, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 10477 + }, + { + "epoch": 0.10478, + "grad_norm": 0.8780845671191688, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 10478 + }, + { + "epoch": 0.10479, + "grad_norm": 0.9803892200525625, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 10479 + }, + { + "epoch": 0.1048, + "grad_norm": 0.8808690398914639, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 10480 + }, + { + "epoch": 0.10481, + "grad_norm": 0.6347197576392399, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 10481 + }, + { + "epoch": 0.10482, + "grad_norm": 0.6103801901216314, + "learning_rate": 0.003, + "loss": 4.026, + "step": 10482 + }, + { + "epoch": 0.10483, + "grad_norm": 0.6953586711477222, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 10483 + }, + { + "epoch": 0.10484, + "grad_norm": 0.6962467066160956, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 10484 + }, + { + "epoch": 0.10485, + "grad_norm": 0.6836173243812499, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 10485 + }, + { + "epoch": 0.10486, + "grad_norm": 0.6138142700643281, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 10486 + }, + { + "epoch": 0.10487, + "grad_norm": 0.500347937286851, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 10487 + }, + { + "epoch": 0.10488, + "grad_norm": 0.5213986020089679, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 10488 + }, + { + "epoch": 0.10489, + "grad_norm": 0.5612464273498321, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 10489 + }, + { + "epoch": 0.1049, + "grad_norm": 0.46647882869997837, + "learning_rate": 0.003, + "loss": 4.037, + "step": 10490 + }, + { + "epoch": 0.10491, + "grad_norm": 0.5195358329818492, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10491 + }, + { + "epoch": 0.10492, + "grad_norm": 0.5564348998893403, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 10492 + }, + { + "epoch": 0.10493, + "grad_norm": 0.5987946630820135, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 10493 + }, + { + "epoch": 0.10494, + "grad_norm": 0.5564915823788574, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 10494 + }, + { + "epoch": 0.10495, + "grad_norm": 0.6331257618991059, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 10495 + }, + { + "epoch": 0.10496, + "grad_norm": 0.71292269937087, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 10496 + }, + { + "epoch": 0.10497, + "grad_norm": 0.7468750137721929, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 10497 + }, + { + "epoch": 0.10498, + "grad_norm": 0.6059834020409925, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 10498 + }, + { + "epoch": 0.10499, + "grad_norm": 0.5854358957360962, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 10499 + }, + { + "epoch": 0.105, + "grad_norm": 0.6435805024956625, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 10500 + }, + { + "epoch": 0.10501, + "grad_norm": 0.6393958155654224, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 10501 + }, + { + "epoch": 0.10502, + "grad_norm": 0.7006570769364162, + "learning_rate": 0.003, + "loss": 4.01, + "step": 10502 + }, + { + "epoch": 0.10503, + "grad_norm": 0.7301222733255953, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 10503 + }, + { + "epoch": 0.10504, + "grad_norm": 0.6722021070379833, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 10504 + }, + { + "epoch": 0.10505, + "grad_norm": 0.6632847914521751, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 10505 + }, + { + "epoch": 0.10506, + "grad_norm": 0.6269598734417815, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 10506 + }, + { + "epoch": 0.10507, + "grad_norm": 0.5866713231464757, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 10507 + }, + { + "epoch": 0.10508, + "grad_norm": 0.5540721058919273, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 10508 + }, + { + "epoch": 0.10509, + "grad_norm": 0.5994118820042103, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 10509 + }, + { + "epoch": 0.1051, + "grad_norm": 0.6974341845933946, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 10510 + }, + { + "epoch": 0.10511, + "grad_norm": 0.8295540623538257, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 10511 + }, + { + "epoch": 0.10512, + "grad_norm": 0.9175016045863453, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 10512 + }, + { + "epoch": 0.10513, + "grad_norm": 0.9884149622405866, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 10513 + }, + { + "epoch": 0.10514, + "grad_norm": 0.9478329662420409, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 10514 + }, + { + "epoch": 0.10515, + "grad_norm": 0.83457188138151, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 10515 + }, + { + "epoch": 0.10516, + "grad_norm": 0.7738762251134427, + "learning_rate": 0.003, + "loss": 4.002, + "step": 10516 + }, + { + "epoch": 0.10517, + "grad_norm": 0.7048895445323475, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 10517 + }, + { + "epoch": 0.10518, + "grad_norm": 0.616795198212195, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 10518 + }, + { + "epoch": 0.10519, + "grad_norm": 0.6278000728865825, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 10519 + }, + { + "epoch": 0.1052, + "grad_norm": 0.6580311316069505, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 10520 + }, + { + "epoch": 0.10521, + "grad_norm": 0.6804272990715643, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 10521 + }, + { + "epoch": 0.10522, + "grad_norm": 0.6247612803172558, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 10522 + }, + { + "epoch": 0.10523, + "grad_norm": 0.680173678721832, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 10523 + }, + { + "epoch": 0.10524, + "grad_norm": 0.6725031289483909, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 10524 + }, + { + "epoch": 0.10525, + "grad_norm": 0.6523838766245822, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 10525 + }, + { + "epoch": 0.10526, + "grad_norm": 0.7030667834686546, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 10526 + }, + { + "epoch": 0.10527, + "grad_norm": 0.725253934758613, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 10527 + }, + { + "epoch": 0.10528, + "grad_norm": 0.76964552298143, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 10528 + }, + { + "epoch": 0.10529, + "grad_norm": 0.7283916016054874, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 10529 + }, + { + "epoch": 0.1053, + "grad_norm": 0.6744381741981271, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10530 + }, + { + "epoch": 0.10531, + "grad_norm": 0.6735179950260372, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 10531 + }, + { + "epoch": 0.10532, + "grad_norm": 0.7225402011069861, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 10532 + }, + { + "epoch": 0.10533, + "grad_norm": 0.8102027141580637, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 10533 + }, + { + "epoch": 0.10534, + "grad_norm": 0.8041267683706693, + "learning_rate": 0.003, + "loss": 4.037, + "step": 10534 + }, + { + "epoch": 0.10535, + "grad_norm": 0.7245910407175711, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 10535 + }, + { + "epoch": 0.10536, + "grad_norm": 0.6712353968017497, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 10536 + }, + { + "epoch": 0.10537, + "grad_norm": 0.6420799377737171, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 10537 + }, + { + "epoch": 0.10538, + "grad_norm": 0.748518320082626, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 10538 + }, + { + "epoch": 0.10539, + "grad_norm": 0.6594718685919755, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 10539 + }, + { + "epoch": 0.1054, + "grad_norm": 0.7511236348161502, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 10540 + }, + { + "epoch": 0.10541, + "grad_norm": 0.7218253569362708, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 10541 + }, + { + "epoch": 0.10542, + "grad_norm": 0.772734792367334, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 10542 + }, + { + "epoch": 0.10543, + "grad_norm": 0.8816030597789436, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 10543 + }, + { + "epoch": 0.10544, + "grad_norm": 0.960149644402449, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 10544 + }, + { + "epoch": 0.10545, + "grad_norm": 0.9458718493769596, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 10545 + }, + { + "epoch": 0.10546, + "grad_norm": 0.9453614735228063, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 10546 + }, + { + "epoch": 0.10547, + "grad_norm": 0.8433209188016331, + "learning_rate": 0.003, + "loss": 4.017, + "step": 10547 + }, + { + "epoch": 0.10548, + "grad_norm": 0.7801493921841205, + "learning_rate": 0.003, + "loss": 4.024, + "step": 10548 + }, + { + "epoch": 0.10549, + "grad_norm": 0.7783797443853312, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 10549 + }, + { + "epoch": 0.1055, + "grad_norm": 0.7605321084675686, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 10550 + }, + { + "epoch": 0.10551, + "grad_norm": 0.6940968418077357, + "learning_rate": 0.003, + "loss": 4.0534, + "step": 10551 + }, + { + "epoch": 0.10552, + "grad_norm": 0.7411084999536178, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 10552 + }, + { + "epoch": 0.10553, + "grad_norm": 0.6235881191558008, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 10553 + }, + { + "epoch": 0.10554, + "grad_norm": 0.5931005119395628, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 10554 + }, + { + "epoch": 0.10555, + "grad_norm": 0.5625726018290195, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 10555 + }, + { + "epoch": 0.10556, + "grad_norm": 0.560777120495365, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 10556 + }, + { + "epoch": 0.10557, + "grad_norm": 0.5951045735563233, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 10557 + }, + { + "epoch": 0.10558, + "grad_norm": 0.6538521970108072, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 10558 + }, + { + "epoch": 0.10559, + "grad_norm": 0.777688836876897, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 10559 + }, + { + "epoch": 0.1056, + "grad_norm": 0.8112294642780566, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 10560 + }, + { + "epoch": 0.10561, + "grad_norm": 0.8104770742501418, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 10561 + }, + { + "epoch": 0.10562, + "grad_norm": 0.7701915595714043, + "learning_rate": 0.003, + "loss": 4.008, + "step": 10562 + }, + { + "epoch": 0.10563, + "grad_norm": 0.6718368826889366, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10563 + }, + { + "epoch": 0.10564, + "grad_norm": 0.6067773570449294, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10564 + }, + { + "epoch": 0.10565, + "grad_norm": 0.6513823547890253, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 10565 + }, + { + "epoch": 0.10566, + "grad_norm": 0.6549235115330371, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10566 + }, + { + "epoch": 0.10567, + "grad_norm": 0.6993874175134468, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 10567 + }, + { + "epoch": 0.10568, + "grad_norm": 0.7011947750219327, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 10568 + }, + { + "epoch": 0.10569, + "grad_norm": 0.7947499388336584, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 10569 + }, + { + "epoch": 0.1057, + "grad_norm": 0.7891330954416276, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 10570 + }, + { + "epoch": 0.10571, + "grad_norm": 0.6876701797466066, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 10571 + }, + { + "epoch": 0.10572, + "grad_norm": 0.6634442618863962, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 10572 + }, + { + "epoch": 0.10573, + "grad_norm": 0.6222221539350392, + "learning_rate": 0.003, + "loss": 4.015, + "step": 10573 + }, + { + "epoch": 0.10574, + "grad_norm": 0.5482370783515882, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 10574 + }, + { + "epoch": 0.10575, + "grad_norm": 0.5950338451070626, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 10575 + }, + { + "epoch": 0.10576, + "grad_norm": 0.6931367916258468, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 10576 + }, + { + "epoch": 0.10577, + "grad_norm": 0.6425503123692055, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 10577 + }, + { + "epoch": 0.10578, + "grad_norm": 0.6803775184069868, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 10578 + }, + { + "epoch": 0.10579, + "grad_norm": 0.7540506248939414, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 10579 + }, + { + "epoch": 0.1058, + "grad_norm": 0.7672080218568589, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 10580 + }, + { + "epoch": 0.10581, + "grad_norm": 0.8465921934375964, + "learning_rate": 0.003, + "loss": 4.05, + "step": 10581 + }, + { + "epoch": 0.10582, + "grad_norm": 0.9742077110100167, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 10582 + }, + { + "epoch": 0.10583, + "grad_norm": 1.0902169807693054, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 10583 + }, + { + "epoch": 0.10584, + "grad_norm": 0.8854154762722425, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 10584 + }, + { + "epoch": 0.10585, + "grad_norm": 0.7600459753864369, + "learning_rate": 0.003, + "loss": 4.0718, + "step": 10585 + }, + { + "epoch": 0.10586, + "grad_norm": 0.708509971781235, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 10586 + }, + { + "epoch": 0.10587, + "grad_norm": 0.6719334680910937, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 10587 + }, + { + "epoch": 0.10588, + "grad_norm": 0.6965547473490514, + "learning_rate": 0.003, + "loss": 4.015, + "step": 10588 + }, + { + "epoch": 0.10589, + "grad_norm": 0.6651449666886762, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10589 + }, + { + "epoch": 0.1059, + "grad_norm": 0.6105097507868902, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 10590 + }, + { + "epoch": 0.10591, + "grad_norm": 0.6121278347006777, + "learning_rate": 0.003, + "loss": 4.014, + "step": 10591 + }, + { + "epoch": 0.10592, + "grad_norm": 0.6421516685642247, + "learning_rate": 0.003, + "loss": 4.0656, + "step": 10592 + }, + { + "epoch": 0.10593, + "grad_norm": 0.6369616376429176, + "learning_rate": 0.003, + "loss": 4.002, + "step": 10593 + }, + { + "epoch": 0.10594, + "grad_norm": 0.6514527959403165, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 10594 + }, + { + "epoch": 0.10595, + "grad_norm": 0.7279987506095597, + "learning_rate": 0.003, + "loss": 4.027, + "step": 10595 + }, + { + "epoch": 0.10596, + "grad_norm": 0.776942270591287, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 10596 + }, + { + "epoch": 0.10597, + "grad_norm": 0.8873284856004168, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 10597 + }, + { + "epoch": 0.10598, + "grad_norm": 0.9000427598092046, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 10598 + }, + { + "epoch": 0.10599, + "grad_norm": 0.7361975766151362, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 10599 + }, + { + "epoch": 0.106, + "grad_norm": 0.6389562329936472, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10600 + }, + { + "epoch": 0.10601, + "grad_norm": 0.5779176313391765, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 10601 + }, + { + "epoch": 0.10602, + "grad_norm": 0.5349832379855675, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 10602 + }, + { + "epoch": 0.10603, + "grad_norm": 0.5546351158206553, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 10603 + }, + { + "epoch": 0.10604, + "grad_norm": 0.5604223353757734, + "learning_rate": 0.003, + "loss": 4.055, + "step": 10604 + }, + { + "epoch": 0.10605, + "grad_norm": 0.6270891613265825, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 10605 + }, + { + "epoch": 0.10606, + "grad_norm": 0.661034613350579, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 10606 + }, + { + "epoch": 0.10607, + "grad_norm": 0.6543493952646533, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 10607 + }, + { + "epoch": 0.10608, + "grad_norm": 0.6068288530685426, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 10608 + }, + { + "epoch": 0.10609, + "grad_norm": 0.5376865469663556, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 10609 + }, + { + "epoch": 0.1061, + "grad_norm": 0.5288806469158538, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 10610 + }, + { + "epoch": 0.10611, + "grad_norm": 0.5812357506444282, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 10611 + }, + { + "epoch": 0.10612, + "grad_norm": 0.6565538906624412, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 10612 + }, + { + "epoch": 0.10613, + "grad_norm": 0.7816912831589194, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 10613 + }, + { + "epoch": 0.10614, + "grad_norm": 0.9680389810045579, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 10614 + }, + { + "epoch": 0.10615, + "grad_norm": 1.0000904565520423, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 10615 + }, + { + "epoch": 0.10616, + "grad_norm": 0.8600405330189036, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 10616 + }, + { + "epoch": 0.10617, + "grad_norm": 0.9069726248011081, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 10617 + }, + { + "epoch": 0.10618, + "grad_norm": 1.0087328229713661, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 10618 + }, + { + "epoch": 0.10619, + "grad_norm": 1.22182936578869, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 10619 + }, + { + "epoch": 0.1062, + "grad_norm": 0.788307637140012, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 10620 + }, + { + "epoch": 0.10621, + "grad_norm": 0.9289592406449169, + "learning_rate": 0.003, + "loss": 4.028, + "step": 10621 + }, + { + "epoch": 0.10622, + "grad_norm": 0.8992022861156989, + "learning_rate": 0.003, + "loss": 4.066, + "step": 10622 + }, + { + "epoch": 0.10623, + "grad_norm": 0.8902101265524911, + "learning_rate": 0.003, + "loss": 4.064, + "step": 10623 + }, + { + "epoch": 0.10624, + "grad_norm": 0.7882205960361172, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 10624 + }, + { + "epoch": 0.10625, + "grad_norm": 0.8707287016871034, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 10625 + }, + { + "epoch": 0.10626, + "grad_norm": 0.8951289490200566, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 10626 + }, + { + "epoch": 0.10627, + "grad_norm": 0.7600872522695225, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 10627 + }, + { + "epoch": 0.10628, + "grad_norm": 0.7193646351508711, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 10628 + }, + { + "epoch": 0.10629, + "grad_norm": 0.7892103153306007, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10629 + }, + { + "epoch": 0.1063, + "grad_norm": 0.8120721487774452, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 10630 + }, + { + "epoch": 0.10631, + "grad_norm": 0.6671699958890799, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 10631 + }, + { + "epoch": 0.10632, + "grad_norm": 0.5512601694612315, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 10632 + }, + { + "epoch": 0.10633, + "grad_norm": 0.5388990826108597, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 10633 + }, + { + "epoch": 0.10634, + "grad_norm": 0.44971937275098867, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 10634 + }, + { + "epoch": 0.10635, + "grad_norm": 0.46516595547022704, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 10635 + }, + { + "epoch": 0.10636, + "grad_norm": 0.43527893405745693, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 10636 + }, + { + "epoch": 0.10637, + "grad_norm": 0.4968207719654524, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 10637 + }, + { + "epoch": 0.10638, + "grad_norm": 0.5396626866052947, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 10638 + }, + { + "epoch": 0.10639, + "grad_norm": 0.6260407727867879, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 10639 + }, + { + "epoch": 0.1064, + "grad_norm": 0.7145785118352608, + "learning_rate": 0.003, + "loss": 4.012, + "step": 10640 + }, + { + "epoch": 0.10641, + "grad_norm": 0.8428111085095397, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 10641 + }, + { + "epoch": 0.10642, + "grad_norm": 0.9658764742345358, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10642 + }, + { + "epoch": 0.10643, + "grad_norm": 0.8552048950951512, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 10643 + }, + { + "epoch": 0.10644, + "grad_norm": 0.6682499271575251, + "learning_rate": 0.003, + "loss": 4.043, + "step": 10644 + }, + { + "epoch": 0.10645, + "grad_norm": 0.6869427458069487, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 10645 + }, + { + "epoch": 0.10646, + "grad_norm": 0.7248520615396757, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 10646 + }, + { + "epoch": 0.10647, + "grad_norm": 0.7404455053707706, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 10647 + }, + { + "epoch": 0.10648, + "grad_norm": 0.6704226485517216, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 10648 + }, + { + "epoch": 0.10649, + "grad_norm": 0.681416436414596, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 10649 + }, + { + "epoch": 0.1065, + "grad_norm": 0.7085662642822146, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 10650 + }, + { + "epoch": 0.10651, + "grad_norm": 0.5778309069615387, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 10651 + }, + { + "epoch": 0.10652, + "grad_norm": 0.5592635326587433, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 10652 + }, + { + "epoch": 0.10653, + "grad_norm": 0.583077742876273, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 10653 + }, + { + "epoch": 0.10654, + "grad_norm": 0.4782505748999694, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 10654 + }, + { + "epoch": 0.10655, + "grad_norm": 0.4524054764511718, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 10655 + }, + { + "epoch": 0.10656, + "grad_norm": 0.3948046071703574, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 10656 + }, + { + "epoch": 0.10657, + "grad_norm": 0.4319446532011087, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 10657 + }, + { + "epoch": 0.10658, + "grad_norm": 0.4490222662317716, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 10658 + }, + { + "epoch": 0.10659, + "grad_norm": 0.4540021657763898, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 10659 + }, + { + "epoch": 0.1066, + "grad_norm": 0.5054086933255201, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 10660 + }, + { + "epoch": 0.10661, + "grad_norm": 0.6268656186551538, + "learning_rate": 0.003, + "loss": 4.026, + "step": 10661 + }, + { + "epoch": 0.10662, + "grad_norm": 0.661285393000586, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10662 + }, + { + "epoch": 0.10663, + "grad_norm": 0.6276377103374631, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 10663 + }, + { + "epoch": 0.10664, + "grad_norm": 0.6746051602605098, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 10664 + }, + { + "epoch": 0.10665, + "grad_norm": 0.8393463911828015, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 10665 + }, + { + "epoch": 0.10666, + "grad_norm": 0.8353218321704137, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 10666 + }, + { + "epoch": 0.10667, + "grad_norm": 0.9949931468825084, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 10667 + }, + { + "epoch": 0.10668, + "grad_norm": 1.1998166009057374, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 10668 + }, + { + "epoch": 0.10669, + "grad_norm": 0.7833293652842174, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 10669 + }, + { + "epoch": 0.1067, + "grad_norm": 0.6484601759866336, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 10670 + }, + { + "epoch": 0.10671, + "grad_norm": 0.6242611438443517, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 10671 + }, + { + "epoch": 0.10672, + "grad_norm": 0.53605708759406, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 10672 + }, + { + "epoch": 0.10673, + "grad_norm": 0.5929020702794648, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 10673 + }, + { + "epoch": 0.10674, + "grad_norm": 0.6700610302257654, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 10674 + }, + { + "epoch": 0.10675, + "grad_norm": 0.7537543602832254, + "learning_rate": 0.003, + "loss": 4.048, + "step": 10675 + }, + { + "epoch": 0.10676, + "grad_norm": 0.8606049759763481, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 10676 + }, + { + "epoch": 0.10677, + "grad_norm": 0.7147744351602302, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 10677 + }, + { + "epoch": 0.10678, + "grad_norm": 0.5767235753347171, + "learning_rate": 0.003, + "loss": 4.015, + "step": 10678 + }, + { + "epoch": 0.10679, + "grad_norm": 0.5314826986710607, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 10679 + }, + { + "epoch": 0.1068, + "grad_norm": 0.5760954611535145, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 10680 + }, + { + "epoch": 0.10681, + "grad_norm": 0.6038922095267006, + "learning_rate": 0.003, + "loss": 4.005, + "step": 10681 + }, + { + "epoch": 0.10682, + "grad_norm": 0.6471515825769856, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 10682 + }, + { + "epoch": 0.10683, + "grad_norm": 0.6829983992042876, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 10683 + }, + { + "epoch": 0.10684, + "grad_norm": 0.6789626920130911, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 10684 + }, + { + "epoch": 0.10685, + "grad_norm": 0.7602606692955415, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 10685 + }, + { + "epoch": 0.10686, + "grad_norm": 0.8204036322486187, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 10686 + }, + { + "epoch": 0.10687, + "grad_norm": 0.9079419764301824, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 10687 + }, + { + "epoch": 0.10688, + "grad_norm": 0.9389116492420558, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 10688 + }, + { + "epoch": 0.10689, + "grad_norm": 0.8672414532404001, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 10689 + }, + { + "epoch": 0.1069, + "grad_norm": 0.8305515759840345, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 10690 + }, + { + "epoch": 0.10691, + "grad_norm": 0.9266092080293422, + "learning_rate": 0.003, + "loss": 4.0727, + "step": 10691 + }, + { + "epoch": 0.10692, + "grad_norm": 0.9433846488736128, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 10692 + }, + { + "epoch": 0.10693, + "grad_norm": 0.7943417220999712, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 10693 + }, + { + "epoch": 0.10694, + "grad_norm": 0.6964001755405893, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10694 + }, + { + "epoch": 0.10695, + "grad_norm": 0.7045959002933229, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 10695 + }, + { + "epoch": 0.10696, + "grad_norm": 0.7482583803201165, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 10696 + }, + { + "epoch": 0.10697, + "grad_norm": 0.8582663734978103, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 10697 + }, + { + "epoch": 0.10698, + "grad_norm": 0.8442842596817903, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 10698 + }, + { + "epoch": 0.10699, + "grad_norm": 0.7061724639310707, + "learning_rate": 0.003, + "loss": 4.0692, + "step": 10699 + }, + { + "epoch": 0.107, + "grad_norm": 0.8093322029503804, + "learning_rate": 0.003, + "loss": 4.0614, + "step": 10700 + }, + { + "epoch": 0.10701, + "grad_norm": 0.9537628830710113, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 10701 + }, + { + "epoch": 0.10702, + "grad_norm": 0.8352693529148184, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 10702 + }, + { + "epoch": 0.10703, + "grad_norm": 0.5998843821307622, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 10703 + }, + { + "epoch": 0.10704, + "grad_norm": 0.6112625957390407, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10704 + }, + { + "epoch": 0.10705, + "grad_norm": 0.5747363817219219, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 10705 + }, + { + "epoch": 0.10706, + "grad_norm": 0.5945578142960753, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 10706 + }, + { + "epoch": 0.10707, + "grad_norm": 0.635515444637423, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 10707 + }, + { + "epoch": 0.10708, + "grad_norm": 0.6102183783846504, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 10708 + }, + { + "epoch": 0.10709, + "grad_norm": 0.47782044342549823, + "learning_rate": 0.003, + "loss": 4.038, + "step": 10709 + }, + { + "epoch": 0.1071, + "grad_norm": 0.4706549726807725, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 10710 + }, + { + "epoch": 0.10711, + "grad_norm": 0.5086260216089872, + "learning_rate": 0.003, + "loss": 4.03, + "step": 10711 + }, + { + "epoch": 0.10712, + "grad_norm": 0.6465333963793909, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 10712 + }, + { + "epoch": 0.10713, + "grad_norm": 0.8632409091972613, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 10713 + }, + { + "epoch": 0.10714, + "grad_norm": 1.0164273018830037, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 10714 + }, + { + "epoch": 0.10715, + "grad_norm": 0.8206780353880289, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 10715 + }, + { + "epoch": 0.10716, + "grad_norm": 0.6524801759119113, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 10716 + }, + { + "epoch": 0.10717, + "grad_norm": 0.6996695559335923, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 10717 + }, + { + "epoch": 0.10718, + "grad_norm": 0.6720604808519992, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 10718 + }, + { + "epoch": 0.10719, + "grad_norm": 0.6239003691859747, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 10719 + }, + { + "epoch": 0.1072, + "grad_norm": 0.5994064509163558, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 10720 + }, + { + "epoch": 0.10721, + "grad_norm": 0.6624288149401307, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 10721 + }, + { + "epoch": 0.10722, + "grad_norm": 0.773693029585937, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 10722 + }, + { + "epoch": 0.10723, + "grad_norm": 0.727046012719663, + "learning_rate": 0.003, + "loss": 4.03, + "step": 10723 + }, + { + "epoch": 0.10724, + "grad_norm": 0.6710956875145458, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 10724 + }, + { + "epoch": 0.10725, + "grad_norm": 0.6818132171009754, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 10725 + }, + { + "epoch": 0.10726, + "grad_norm": 0.6452170191903657, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 10726 + }, + { + "epoch": 0.10727, + "grad_norm": 0.6935317338943925, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 10727 + }, + { + "epoch": 0.10728, + "grad_norm": 0.7244522136107102, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 10728 + }, + { + "epoch": 0.10729, + "grad_norm": 0.7915338836551747, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 10729 + }, + { + "epoch": 0.1073, + "grad_norm": 0.7714458714190993, + "learning_rate": 0.003, + "loss": 4.017, + "step": 10730 + }, + { + "epoch": 0.10731, + "grad_norm": 0.7646461993722945, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 10731 + }, + { + "epoch": 0.10732, + "grad_norm": 0.7578502078557265, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 10732 + }, + { + "epoch": 0.10733, + "grad_norm": 0.7691296559778578, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 10733 + }, + { + "epoch": 0.10734, + "grad_norm": 0.7514909971960825, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 10734 + }, + { + "epoch": 0.10735, + "grad_norm": 0.7806798851747229, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 10735 + }, + { + "epoch": 0.10736, + "grad_norm": 0.7542026967688327, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 10736 + }, + { + "epoch": 0.10737, + "grad_norm": 0.6981285034345556, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 10737 + }, + { + "epoch": 0.10738, + "grad_norm": 0.6465333320543774, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 10738 + }, + { + "epoch": 0.10739, + "grad_norm": 0.6299238900719762, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 10739 + }, + { + "epoch": 0.1074, + "grad_norm": 0.653727126689422, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 10740 + }, + { + "epoch": 0.10741, + "grad_norm": 0.7159903384813265, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 10741 + }, + { + "epoch": 0.10742, + "grad_norm": 0.7406194341750261, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 10742 + }, + { + "epoch": 0.10743, + "grad_norm": 0.6032078970312981, + "learning_rate": 0.003, + "loss": 4.045, + "step": 10743 + }, + { + "epoch": 0.10744, + "grad_norm": 0.48030272035462396, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 10744 + }, + { + "epoch": 0.10745, + "grad_norm": 0.5940539246293907, + "learning_rate": 0.003, + "loss": 4.017, + "step": 10745 + }, + { + "epoch": 0.10746, + "grad_norm": 0.6115863841071971, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 10746 + }, + { + "epoch": 0.10747, + "grad_norm": 0.671592142834876, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 10747 + }, + { + "epoch": 0.10748, + "grad_norm": 0.7207706502011647, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 10748 + }, + { + "epoch": 0.10749, + "grad_norm": 0.8860657361804246, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 10749 + }, + { + "epoch": 0.1075, + "grad_norm": 0.9529889501125546, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 10750 + }, + { + "epoch": 0.10751, + "grad_norm": 0.8516677566516427, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 10751 + }, + { + "epoch": 0.10752, + "grad_norm": 0.7212982850263704, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 10752 + }, + { + "epoch": 0.10753, + "grad_norm": 0.8237359837607393, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 10753 + }, + { + "epoch": 0.10754, + "grad_norm": 0.7916808942117067, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10754 + }, + { + "epoch": 0.10755, + "grad_norm": 0.8226600172765114, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 10755 + }, + { + "epoch": 0.10756, + "grad_norm": 0.8873686201575736, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 10756 + }, + { + "epoch": 0.10757, + "grad_norm": 0.8926360847396212, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 10757 + }, + { + "epoch": 0.10758, + "grad_norm": 1.029793651859423, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 10758 + }, + { + "epoch": 0.10759, + "grad_norm": 0.8809458040107688, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 10759 + }, + { + "epoch": 0.1076, + "grad_norm": 0.9306086492589638, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 10760 + }, + { + "epoch": 0.10761, + "grad_norm": 0.9554535338112706, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 10761 + }, + { + "epoch": 0.10762, + "grad_norm": 0.9363989948435605, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 10762 + }, + { + "epoch": 0.10763, + "grad_norm": 0.7929958332970587, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 10763 + }, + { + "epoch": 0.10764, + "grad_norm": 1.000494902499705, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 10764 + }, + { + "epoch": 0.10765, + "grad_norm": 1.0064075888317625, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 10765 + }, + { + "epoch": 0.10766, + "grad_norm": 0.9386767107855049, + "learning_rate": 0.003, + "loss": 4.045, + "step": 10766 + }, + { + "epoch": 0.10767, + "grad_norm": 0.8860687908879845, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 10767 + }, + { + "epoch": 0.10768, + "grad_norm": 0.8175660303893406, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 10768 + }, + { + "epoch": 0.10769, + "grad_norm": 0.7439016563221261, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 10769 + }, + { + "epoch": 0.1077, + "grad_norm": 0.7076004999646608, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 10770 + }, + { + "epoch": 0.10771, + "grad_norm": 0.5790166564152505, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 10771 + }, + { + "epoch": 0.10772, + "grad_norm": 0.5548860074819257, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10772 + }, + { + "epoch": 0.10773, + "grad_norm": 0.49286150439877247, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 10773 + }, + { + "epoch": 0.10774, + "grad_norm": 0.45155523336210845, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 10774 + }, + { + "epoch": 0.10775, + "grad_norm": 0.47307450565317843, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 10775 + }, + { + "epoch": 0.10776, + "grad_norm": 0.4839567568448366, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 10776 + }, + { + "epoch": 0.10777, + "grad_norm": 0.5963580895585321, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 10777 + }, + { + "epoch": 0.10778, + "grad_norm": 0.7654052649432523, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 10778 + }, + { + "epoch": 0.10779, + "grad_norm": 0.9847919605613891, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 10779 + }, + { + "epoch": 0.1078, + "grad_norm": 1.1059416982781307, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 10780 + }, + { + "epoch": 0.10781, + "grad_norm": 0.606026247978062, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 10781 + }, + { + "epoch": 0.10782, + "grad_norm": 0.649599805389313, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 10782 + }, + { + "epoch": 0.10783, + "grad_norm": 0.6741672078878379, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 10783 + }, + { + "epoch": 0.10784, + "grad_norm": 0.6000977883561384, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 10784 + }, + { + "epoch": 0.10785, + "grad_norm": 0.5641734367623169, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 10785 + }, + { + "epoch": 0.10786, + "grad_norm": 0.6710198811786872, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 10786 + }, + { + "epoch": 0.10787, + "grad_norm": 0.6030852600871224, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 10787 + }, + { + "epoch": 0.10788, + "grad_norm": 0.513568822685513, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 10788 + }, + { + "epoch": 0.10789, + "grad_norm": 0.5197928207161628, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 10789 + }, + { + "epoch": 0.1079, + "grad_norm": 0.5094203793510536, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 10790 + }, + { + "epoch": 0.10791, + "grad_norm": 0.47564982989456545, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 10791 + }, + { + "epoch": 0.10792, + "grad_norm": 0.48878128162681117, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 10792 + }, + { + "epoch": 0.10793, + "grad_norm": 0.4618163213943937, + "learning_rate": 0.003, + "loss": 4.008, + "step": 10793 + }, + { + "epoch": 0.10794, + "grad_norm": 0.4947487900421549, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 10794 + }, + { + "epoch": 0.10795, + "grad_norm": 0.5236105009665775, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 10795 + }, + { + "epoch": 0.10796, + "grad_norm": 0.5900983898315401, + "learning_rate": 0.003, + "loss": 4.015, + "step": 10796 + }, + { + "epoch": 0.10797, + "grad_norm": 0.6465244339065456, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 10797 + }, + { + "epoch": 0.10798, + "grad_norm": 0.6335087886659654, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 10798 + }, + { + "epoch": 0.10799, + "grad_norm": 0.5373907135671206, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 10799 + }, + { + "epoch": 0.108, + "grad_norm": 0.579684430276672, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 10800 + }, + { + "epoch": 0.10801, + "grad_norm": 0.6001153839845709, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 10801 + }, + { + "epoch": 0.10802, + "grad_norm": 0.5988714676560685, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 10802 + }, + { + "epoch": 0.10803, + "grad_norm": 0.6917751887978937, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 10803 + }, + { + "epoch": 0.10804, + "grad_norm": 0.8713415856202766, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 10804 + }, + { + "epoch": 0.10805, + "grad_norm": 0.8962304879840934, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 10805 + }, + { + "epoch": 0.10806, + "grad_norm": 0.7685441001192913, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 10806 + }, + { + "epoch": 0.10807, + "grad_norm": 0.6672287102832734, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 10807 + }, + { + "epoch": 0.10808, + "grad_norm": 0.6704094068507815, + "learning_rate": 0.003, + "loss": 4.1035, + "step": 10808 + }, + { + "epoch": 0.10809, + "grad_norm": 0.6941404333586272, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 10809 + }, + { + "epoch": 0.1081, + "grad_norm": 0.8203657609563492, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 10810 + }, + { + "epoch": 0.10811, + "grad_norm": 0.8184911183972927, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 10811 + }, + { + "epoch": 0.10812, + "grad_norm": 0.7823081500664267, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 10812 + }, + { + "epoch": 0.10813, + "grad_norm": 0.7305244628505301, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 10813 + }, + { + "epoch": 0.10814, + "grad_norm": 0.7827703819159203, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 10814 + }, + { + "epoch": 0.10815, + "grad_norm": 0.7761613596045656, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10815 + }, + { + "epoch": 0.10816, + "grad_norm": 0.8212347159717724, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 10816 + }, + { + "epoch": 0.10817, + "grad_norm": 0.8925955197574863, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 10817 + }, + { + "epoch": 0.10818, + "grad_norm": 0.9315969366472022, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 10818 + }, + { + "epoch": 0.10819, + "grad_norm": 0.9412774938374964, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 10819 + }, + { + "epoch": 0.1082, + "grad_norm": 0.8190237015352827, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 10820 + }, + { + "epoch": 0.10821, + "grad_norm": 0.6686909322251393, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 10821 + }, + { + "epoch": 0.10822, + "grad_norm": 0.5995191160597949, + "learning_rate": 0.003, + "loss": 4.014, + "step": 10822 + }, + { + "epoch": 0.10823, + "grad_norm": 0.6412626910288645, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 10823 + }, + { + "epoch": 0.10824, + "grad_norm": 0.5657042313920181, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 10824 + }, + { + "epoch": 0.10825, + "grad_norm": 0.5204176239792738, + "learning_rate": 0.003, + "loss": 4.007, + "step": 10825 + }, + { + "epoch": 0.10826, + "grad_norm": 0.47993980169347344, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 10826 + }, + { + "epoch": 0.10827, + "grad_norm": 0.545802034255981, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 10827 + }, + { + "epoch": 0.10828, + "grad_norm": 0.7103567990189308, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 10828 + }, + { + "epoch": 0.10829, + "grad_norm": 0.8641134269792967, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 10829 + }, + { + "epoch": 0.1083, + "grad_norm": 0.9597725019181991, + "learning_rate": 0.003, + "loss": 4.0538, + "step": 10830 + }, + { + "epoch": 0.10831, + "grad_norm": 0.8905608158235813, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 10831 + }, + { + "epoch": 0.10832, + "grad_norm": 0.7381585107233722, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 10832 + }, + { + "epoch": 0.10833, + "grad_norm": 0.7437115862161257, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 10833 + }, + { + "epoch": 0.10834, + "grad_norm": 0.7105773431472947, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 10834 + }, + { + "epoch": 0.10835, + "grad_norm": 0.7217597346707454, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 10835 + }, + { + "epoch": 0.10836, + "grad_norm": 0.7063572927921363, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 10836 + }, + { + "epoch": 0.10837, + "grad_norm": 0.677047804334423, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 10837 + }, + { + "epoch": 0.10838, + "grad_norm": 0.575646034263939, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 10838 + }, + { + "epoch": 0.10839, + "grad_norm": 0.60973160397982, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 10839 + }, + { + "epoch": 0.1084, + "grad_norm": 0.6692594038811125, + "learning_rate": 0.003, + "loss": 4.045, + "step": 10840 + }, + { + "epoch": 0.10841, + "grad_norm": 0.7566094510985311, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 10841 + }, + { + "epoch": 0.10842, + "grad_norm": 0.7922613054066925, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 10842 + }, + { + "epoch": 0.10843, + "grad_norm": 0.679977566199715, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 10843 + }, + { + "epoch": 0.10844, + "grad_norm": 0.7222213237531506, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 10844 + }, + { + "epoch": 0.10845, + "grad_norm": 0.7107286782881448, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 10845 + }, + { + "epoch": 0.10846, + "grad_norm": 0.7098737301150135, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 10846 + }, + { + "epoch": 0.10847, + "grad_norm": 0.7513147111031714, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 10847 + }, + { + "epoch": 0.10848, + "grad_norm": 0.6808638576033755, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 10848 + }, + { + "epoch": 0.10849, + "grad_norm": 0.6298346894254325, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 10849 + }, + { + "epoch": 0.1085, + "grad_norm": 0.6236313036053704, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 10850 + }, + { + "epoch": 0.10851, + "grad_norm": 0.6578925801526583, + "learning_rate": 0.003, + "loss": 4.029, + "step": 10851 + }, + { + "epoch": 0.10852, + "grad_norm": 0.6272932511439222, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 10852 + }, + { + "epoch": 0.10853, + "grad_norm": 0.6390622883420599, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 10853 + }, + { + "epoch": 0.10854, + "grad_norm": 0.6355846119313069, + "learning_rate": 0.003, + "loss": 4.0, + "step": 10854 + }, + { + "epoch": 0.10855, + "grad_norm": 0.6093512809739514, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 10855 + }, + { + "epoch": 0.10856, + "grad_norm": 0.641818824933501, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 10856 + }, + { + "epoch": 0.10857, + "grad_norm": 0.7233923379490161, + "learning_rate": 0.003, + "loss": 4.017, + "step": 10857 + }, + { + "epoch": 0.10858, + "grad_norm": 0.767923198502708, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 10858 + }, + { + "epoch": 0.10859, + "grad_norm": 0.8173561745596454, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 10859 + }, + { + "epoch": 0.1086, + "grad_norm": 0.8200352985112688, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10860 + }, + { + "epoch": 0.10861, + "grad_norm": 0.8412326407609138, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 10861 + }, + { + "epoch": 0.10862, + "grad_norm": 0.7509351525292228, + "learning_rate": 0.003, + "loss": 4.0728, + "step": 10862 + }, + { + "epoch": 0.10863, + "grad_norm": 0.690097084478729, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 10863 + }, + { + "epoch": 0.10864, + "grad_norm": 0.7025429283324257, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 10864 + }, + { + "epoch": 0.10865, + "grad_norm": 0.7051286149299024, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 10865 + }, + { + "epoch": 0.10866, + "grad_norm": 0.8429043955571983, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 10866 + }, + { + "epoch": 0.10867, + "grad_norm": 0.8025450532394673, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 10867 + }, + { + "epoch": 0.10868, + "grad_norm": 0.7436538751821394, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 10868 + }, + { + "epoch": 0.10869, + "grad_norm": 0.7794642689103424, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 10869 + }, + { + "epoch": 0.1087, + "grad_norm": 0.7615429504130531, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 10870 + }, + { + "epoch": 0.10871, + "grad_norm": 0.8113424988007164, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 10871 + }, + { + "epoch": 0.10872, + "grad_norm": 0.7489049875728313, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 10872 + }, + { + "epoch": 0.10873, + "grad_norm": 0.7730505777239347, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 10873 + }, + { + "epoch": 0.10874, + "grad_norm": 0.7760088755590885, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 10874 + }, + { + "epoch": 0.10875, + "grad_norm": 0.6742911527401453, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 10875 + }, + { + "epoch": 0.10876, + "grad_norm": 0.5998489995570341, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 10876 + }, + { + "epoch": 0.10877, + "grad_norm": 0.6461272254980165, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 10877 + }, + { + "epoch": 0.10878, + "grad_norm": 0.6575742639863154, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 10878 + }, + { + "epoch": 0.10879, + "grad_norm": 0.7727107995696623, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 10879 + }, + { + "epoch": 0.1088, + "grad_norm": 0.7520538756088208, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 10880 + }, + { + "epoch": 0.10881, + "grad_norm": 0.695550850849655, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 10881 + }, + { + "epoch": 0.10882, + "grad_norm": 0.6337334723056178, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 10882 + }, + { + "epoch": 0.10883, + "grad_norm": 0.5898017371462143, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 10883 + }, + { + "epoch": 0.10884, + "grad_norm": 0.4572409904673296, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 10884 + }, + { + "epoch": 0.10885, + "grad_norm": 0.4640820368557108, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 10885 + }, + { + "epoch": 0.10886, + "grad_norm": 0.44025384343150414, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 10886 + }, + { + "epoch": 0.10887, + "grad_norm": 0.48621220588516917, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 10887 + }, + { + "epoch": 0.10888, + "grad_norm": 0.49371747303448743, + "learning_rate": 0.003, + "loss": 3.986, + "step": 10888 + }, + { + "epoch": 0.10889, + "grad_norm": 0.5194297767434859, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 10889 + }, + { + "epoch": 0.1089, + "grad_norm": 0.5137548427265785, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 10890 + }, + { + "epoch": 0.10891, + "grad_norm": 0.5753686878818542, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 10891 + }, + { + "epoch": 0.10892, + "grad_norm": 0.586809883604676, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 10892 + }, + { + "epoch": 0.10893, + "grad_norm": 0.6825301524444486, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 10893 + }, + { + "epoch": 0.10894, + "grad_norm": 0.8044196447669095, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 10894 + }, + { + "epoch": 0.10895, + "grad_norm": 0.8292181108054248, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 10895 + }, + { + "epoch": 0.10896, + "grad_norm": 0.9131477963016381, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 10896 + }, + { + "epoch": 0.10897, + "grad_norm": 1.0043913448316857, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 10897 + }, + { + "epoch": 0.10898, + "grad_norm": 1.0144737496808343, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 10898 + }, + { + "epoch": 0.10899, + "grad_norm": 0.9495069743837166, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 10899 + }, + { + "epoch": 0.109, + "grad_norm": 0.8672586773583812, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 10900 + }, + { + "epoch": 0.10901, + "grad_norm": 0.9791355154324539, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 10901 + }, + { + "epoch": 0.10902, + "grad_norm": 1.3451502665753963, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 10902 + }, + { + "epoch": 0.10903, + "grad_norm": 0.7685214308422036, + "learning_rate": 0.003, + "loss": 4.044, + "step": 10903 + }, + { + "epoch": 0.10904, + "grad_norm": 0.8813275225248429, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 10904 + }, + { + "epoch": 0.10905, + "grad_norm": 0.9540385033492641, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 10905 + }, + { + "epoch": 0.10906, + "grad_norm": 0.8440474463461428, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 10906 + }, + { + "epoch": 0.10907, + "grad_norm": 0.8688238202849223, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 10907 + }, + { + "epoch": 0.10908, + "grad_norm": 0.7502594347498712, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 10908 + }, + { + "epoch": 0.10909, + "grad_norm": 0.8215275066951322, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 10909 + }, + { + "epoch": 0.1091, + "grad_norm": 0.8770147930626795, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 10910 + }, + { + "epoch": 0.10911, + "grad_norm": 0.867289238244349, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 10911 + }, + { + "epoch": 0.10912, + "grad_norm": 0.7398664128729519, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 10912 + }, + { + "epoch": 0.10913, + "grad_norm": 0.6034578202146061, + "learning_rate": 0.003, + "loss": 4.055, + "step": 10913 + }, + { + "epoch": 0.10914, + "grad_norm": 0.6637827476954253, + "learning_rate": 0.003, + "loss": 4.033, + "step": 10914 + }, + { + "epoch": 0.10915, + "grad_norm": 0.710287754054119, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 10915 + }, + { + "epoch": 0.10916, + "grad_norm": 0.8063225297063262, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 10916 + }, + { + "epoch": 0.10917, + "grad_norm": 0.6682129623230212, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 10917 + }, + { + "epoch": 0.10918, + "grad_norm": 0.6339543671662302, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 10918 + }, + { + "epoch": 0.10919, + "grad_norm": 0.6765622754404547, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 10919 + }, + { + "epoch": 0.1092, + "grad_norm": 0.6262111256060062, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 10920 + }, + { + "epoch": 0.10921, + "grad_norm": 0.542103079598935, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 10921 + }, + { + "epoch": 0.10922, + "grad_norm": 0.5137115928240171, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 10922 + }, + { + "epoch": 0.10923, + "grad_norm": 0.4489514527392057, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 10923 + }, + { + "epoch": 0.10924, + "grad_norm": 0.46962905942868366, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 10924 + }, + { + "epoch": 0.10925, + "grad_norm": 0.5402667891031823, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 10925 + }, + { + "epoch": 0.10926, + "grad_norm": 0.6695232660281696, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 10926 + }, + { + "epoch": 0.10927, + "grad_norm": 0.7471489558244931, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 10927 + }, + { + "epoch": 0.10928, + "grad_norm": 0.8023304798349268, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 10928 + }, + { + "epoch": 0.10929, + "grad_norm": 0.7411592112394811, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 10929 + }, + { + "epoch": 0.1093, + "grad_norm": 0.6111593525814102, + "learning_rate": 0.003, + "loss": 4.035, + "step": 10930 + }, + { + "epoch": 0.10931, + "grad_norm": 0.5813022347416936, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 10931 + }, + { + "epoch": 0.10932, + "grad_norm": 0.6822526556184744, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 10932 + }, + { + "epoch": 0.10933, + "grad_norm": 0.6330454400829806, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 10933 + }, + { + "epoch": 0.10934, + "grad_norm": 0.6744256730428067, + "learning_rate": 0.003, + "loss": 4.0638, + "step": 10934 + }, + { + "epoch": 0.10935, + "grad_norm": 0.7183646535422697, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 10935 + }, + { + "epoch": 0.10936, + "grad_norm": 0.7385539785606909, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 10936 + }, + { + "epoch": 0.10937, + "grad_norm": 0.7227381255785887, + "learning_rate": 0.003, + "loss": 4.046, + "step": 10937 + }, + { + "epoch": 0.10938, + "grad_norm": 0.624993333200877, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 10938 + }, + { + "epoch": 0.10939, + "grad_norm": 0.4830989844022265, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 10939 + }, + { + "epoch": 0.1094, + "grad_norm": 0.5242529415152747, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 10940 + }, + { + "epoch": 0.10941, + "grad_norm": 0.569678088764479, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 10941 + }, + { + "epoch": 0.10942, + "grad_norm": 0.6300685802896531, + "learning_rate": 0.003, + "loss": 4.031, + "step": 10942 + }, + { + "epoch": 0.10943, + "grad_norm": 0.6607824274128933, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 10943 + }, + { + "epoch": 0.10944, + "grad_norm": 0.7092241736161636, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 10944 + }, + { + "epoch": 0.10945, + "grad_norm": 0.7595307885911903, + "learning_rate": 0.003, + "loss": 4.028, + "step": 10945 + }, + { + "epoch": 0.10946, + "grad_norm": 0.662113719470435, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 10946 + }, + { + "epoch": 0.10947, + "grad_norm": 0.707324478023083, + "learning_rate": 0.003, + "loss": 4.008, + "step": 10947 + }, + { + "epoch": 0.10948, + "grad_norm": 0.7782194717518239, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 10948 + }, + { + "epoch": 0.10949, + "grad_norm": 0.8847890145885523, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 10949 + }, + { + "epoch": 0.1095, + "grad_norm": 0.8833049862166856, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 10950 + }, + { + "epoch": 0.10951, + "grad_norm": 0.7892810420440493, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 10951 + }, + { + "epoch": 0.10952, + "grad_norm": 0.7569542367947412, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 10952 + }, + { + "epoch": 0.10953, + "grad_norm": 0.8578678966857722, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 10953 + }, + { + "epoch": 0.10954, + "grad_norm": 0.8027300705981797, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 10954 + }, + { + "epoch": 0.10955, + "grad_norm": 0.7622473550768762, + "learning_rate": 0.003, + "loss": 4.017, + "step": 10955 + }, + { + "epoch": 0.10956, + "grad_norm": 0.7139210041298515, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 10956 + }, + { + "epoch": 0.10957, + "grad_norm": 0.7637245172127608, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 10957 + }, + { + "epoch": 0.10958, + "grad_norm": 0.8373548508456686, + "learning_rate": 0.003, + "loss": 4.026, + "step": 10958 + }, + { + "epoch": 0.10959, + "grad_norm": 0.8203016394863843, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 10959 + }, + { + "epoch": 0.1096, + "grad_norm": 0.710250793402643, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 10960 + }, + { + "epoch": 0.10961, + "grad_norm": 0.7473900527185722, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 10961 + }, + { + "epoch": 0.10962, + "grad_norm": 0.8534709486577279, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 10962 + }, + { + "epoch": 0.10963, + "grad_norm": 0.7898606487393048, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 10963 + }, + { + "epoch": 0.10964, + "grad_norm": 0.636361898331718, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 10964 + }, + { + "epoch": 0.10965, + "grad_norm": 0.5787159152178174, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 10965 + }, + { + "epoch": 0.10966, + "grad_norm": 0.5766289261402006, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 10966 + }, + { + "epoch": 0.10967, + "grad_norm": 0.5812427621804647, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 10967 + }, + { + "epoch": 0.10968, + "grad_norm": 0.5719282443131701, + "learning_rate": 0.003, + "loss": 4.045, + "step": 10968 + }, + { + "epoch": 0.10969, + "grad_norm": 0.4832090177288875, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 10969 + }, + { + "epoch": 0.1097, + "grad_norm": 0.47144286426122267, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 10970 + }, + { + "epoch": 0.10971, + "grad_norm": 0.5447471075172171, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 10971 + }, + { + "epoch": 0.10972, + "grad_norm": 0.6702751455722818, + "learning_rate": 0.003, + "loss": 4.066, + "step": 10972 + }, + { + "epoch": 0.10973, + "grad_norm": 0.752485004260628, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 10973 + }, + { + "epoch": 0.10974, + "grad_norm": 0.8006823655192781, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 10974 + }, + { + "epoch": 0.10975, + "grad_norm": 0.8892176030724411, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 10975 + }, + { + "epoch": 0.10976, + "grad_norm": 0.8665798277833652, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 10976 + }, + { + "epoch": 0.10977, + "grad_norm": 0.8430860963539034, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 10977 + }, + { + "epoch": 0.10978, + "grad_norm": 0.8017043899836472, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 10978 + }, + { + "epoch": 0.10979, + "grad_norm": 0.8107140605409325, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 10979 + }, + { + "epoch": 0.1098, + "grad_norm": 0.7972098532960646, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 10980 + }, + { + "epoch": 0.10981, + "grad_norm": 0.748305665492647, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 10981 + }, + { + "epoch": 0.10982, + "grad_norm": 0.7426025281889597, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 10982 + }, + { + "epoch": 0.10983, + "grad_norm": 0.732155482991091, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 10983 + }, + { + "epoch": 0.10984, + "grad_norm": 0.753081561678069, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 10984 + }, + { + "epoch": 0.10985, + "grad_norm": 0.8528703225466834, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 10985 + }, + { + "epoch": 0.10986, + "grad_norm": 1.0197051329750417, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 10986 + }, + { + "epoch": 0.10987, + "grad_norm": 0.9299300128042302, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 10987 + }, + { + "epoch": 0.10988, + "grad_norm": 0.9163538011546353, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 10988 + }, + { + "epoch": 0.10989, + "grad_norm": 0.8902886648729984, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 10989 + }, + { + "epoch": 0.1099, + "grad_norm": 0.8268577628120112, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 10990 + }, + { + "epoch": 0.10991, + "grad_norm": 0.9286024569670189, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 10991 + }, + { + "epoch": 0.10992, + "grad_norm": 0.8786393955128015, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 10992 + }, + { + "epoch": 0.10993, + "grad_norm": 0.8750923594596626, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 10993 + }, + { + "epoch": 0.10994, + "grad_norm": 0.8634420733834784, + "learning_rate": 0.003, + "loss": 4.083, + "step": 10994 + }, + { + "epoch": 0.10995, + "grad_norm": 0.78640686229583, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 10995 + }, + { + "epoch": 0.10996, + "grad_norm": 0.7765970287419944, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 10996 + }, + { + "epoch": 0.10997, + "grad_norm": 0.6514345496567573, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 10997 + }, + { + "epoch": 0.10998, + "grad_norm": 0.6354642631690817, + "learning_rate": 0.003, + "loss": 4.0698, + "step": 10998 + }, + { + "epoch": 0.10999, + "grad_norm": 0.6655306231556521, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 10999 + }, + { + "epoch": 0.11, + "grad_norm": 0.7634796334809951, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 11000 + }, + { + "epoch": 0.11001, + "grad_norm": 0.7411085785538175, + "learning_rate": 0.003, + "loss": 4.0641, + "step": 11001 + }, + { + "epoch": 0.11002, + "grad_norm": 0.6453008946113928, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 11002 + }, + { + "epoch": 0.11003, + "grad_norm": 0.5111267344185738, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 11003 + }, + { + "epoch": 0.11004, + "grad_norm": 0.5426471840405502, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 11004 + }, + { + "epoch": 0.11005, + "grad_norm": 0.5300665637776402, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 11005 + }, + { + "epoch": 0.11006, + "grad_norm": 0.5950506746804459, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11006 + }, + { + "epoch": 0.11007, + "grad_norm": 0.7215379250469829, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 11007 + }, + { + "epoch": 0.11008, + "grad_norm": 0.8955700687126673, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 11008 + }, + { + "epoch": 0.11009, + "grad_norm": 0.9083623297085025, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11009 + }, + { + "epoch": 0.1101, + "grad_norm": 0.7951941645443059, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 11010 + }, + { + "epoch": 0.11011, + "grad_norm": 0.6390419771091258, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 11011 + }, + { + "epoch": 0.11012, + "grad_norm": 0.6203414364621666, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 11012 + }, + { + "epoch": 0.11013, + "grad_norm": 0.7293310829868095, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11013 + }, + { + "epoch": 0.11014, + "grad_norm": 0.7591309536930931, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 11014 + }, + { + "epoch": 0.11015, + "grad_norm": 0.5884424509749211, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 11015 + }, + { + "epoch": 0.11016, + "grad_norm": 0.49416585625390475, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 11016 + }, + { + "epoch": 0.11017, + "grad_norm": 0.46611849402168215, + "learning_rate": 0.003, + "loss": 4.003, + "step": 11017 + }, + { + "epoch": 0.11018, + "grad_norm": 0.4988777813381707, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 11018 + }, + { + "epoch": 0.11019, + "grad_norm": 0.5414844939496928, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 11019 + }, + { + "epoch": 0.1102, + "grad_norm": 0.5361782245621374, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 11020 + }, + { + "epoch": 0.11021, + "grad_norm": 0.47423695557391965, + "learning_rate": 0.003, + "loss": 4.0, + "step": 11021 + }, + { + "epoch": 0.11022, + "grad_norm": 0.5307226140015306, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 11022 + }, + { + "epoch": 0.11023, + "grad_norm": 0.5948519654892592, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 11023 + }, + { + "epoch": 0.11024, + "grad_norm": 0.5979935815923778, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 11024 + }, + { + "epoch": 0.11025, + "grad_norm": 0.5369342506907274, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 11025 + }, + { + "epoch": 0.11026, + "grad_norm": 0.5274228598151012, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 11026 + }, + { + "epoch": 0.11027, + "grad_norm": 0.5285423481862842, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 11027 + }, + { + "epoch": 0.11028, + "grad_norm": 0.46894896514034873, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 11028 + }, + { + "epoch": 0.11029, + "grad_norm": 0.48174443375468473, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 11029 + }, + { + "epoch": 0.1103, + "grad_norm": 0.48453088283760404, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 11030 + }, + { + "epoch": 0.11031, + "grad_norm": 0.5153845202333797, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 11031 + }, + { + "epoch": 0.11032, + "grad_norm": 0.4771081166911705, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 11032 + }, + { + "epoch": 0.11033, + "grad_norm": 0.5428570806920385, + "learning_rate": 0.003, + "loss": 4.01, + "step": 11033 + }, + { + "epoch": 0.11034, + "grad_norm": 0.8033598809675997, + "learning_rate": 0.003, + "loss": 3.988, + "step": 11034 + }, + { + "epoch": 0.11035, + "grad_norm": 1.1452471288351083, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 11035 + }, + { + "epoch": 0.11036, + "grad_norm": 0.8077523779944322, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 11036 + }, + { + "epoch": 0.11037, + "grad_norm": 0.7336981627306273, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 11037 + }, + { + "epoch": 0.11038, + "grad_norm": 0.7533233226466816, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 11038 + }, + { + "epoch": 0.11039, + "grad_norm": 0.7951100799294972, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 11039 + }, + { + "epoch": 0.1104, + "grad_norm": 0.8657928206626313, + "learning_rate": 0.003, + "loss": 4.062, + "step": 11040 + }, + { + "epoch": 0.11041, + "grad_norm": 0.9857094766839676, + "learning_rate": 0.003, + "loss": 4.05, + "step": 11041 + }, + { + "epoch": 0.11042, + "grad_norm": 0.9886160571889028, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 11042 + }, + { + "epoch": 0.11043, + "grad_norm": 0.9149188364211451, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 11043 + }, + { + "epoch": 0.11044, + "grad_norm": 0.9080388965587024, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11044 + }, + { + "epoch": 0.11045, + "grad_norm": 0.9829031733494492, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11045 + }, + { + "epoch": 0.11046, + "grad_norm": 1.041504961199167, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 11046 + }, + { + "epoch": 0.11047, + "grad_norm": 0.909735555792298, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 11047 + }, + { + "epoch": 0.11048, + "grad_norm": 0.8771338978128816, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 11048 + }, + { + "epoch": 0.11049, + "grad_norm": 0.7197555550653928, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11049 + }, + { + "epoch": 0.1105, + "grad_norm": 0.662798400946808, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 11050 + }, + { + "epoch": 0.11051, + "grad_norm": 0.7128681130733057, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 11051 + }, + { + "epoch": 0.11052, + "grad_norm": 0.6436728592378992, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 11052 + }, + { + "epoch": 0.11053, + "grad_norm": 0.5917971728156214, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 11053 + }, + { + "epoch": 0.11054, + "grad_norm": 0.685024744092882, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 11054 + }, + { + "epoch": 0.11055, + "grad_norm": 0.8052923789800625, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 11055 + }, + { + "epoch": 0.11056, + "grad_norm": 0.9022254851270457, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 11056 + }, + { + "epoch": 0.11057, + "grad_norm": 0.8910819484641791, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 11057 + }, + { + "epoch": 0.11058, + "grad_norm": 0.7246411271904685, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 11058 + }, + { + "epoch": 0.11059, + "grad_norm": 0.6894715539003105, + "learning_rate": 0.003, + "loss": 4.038, + "step": 11059 + }, + { + "epoch": 0.1106, + "grad_norm": 0.6003315156027453, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 11060 + }, + { + "epoch": 0.11061, + "grad_norm": 0.6456486197396432, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 11061 + }, + { + "epoch": 0.11062, + "grad_norm": 0.6858268768610688, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 11062 + }, + { + "epoch": 0.11063, + "grad_norm": 0.638863364368278, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 11063 + }, + { + "epoch": 0.11064, + "grad_norm": 0.6395987062033898, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 11064 + }, + { + "epoch": 0.11065, + "grad_norm": 0.6903097544482584, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 11065 + }, + { + "epoch": 0.11066, + "grad_norm": 0.7514864278676738, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 11066 + }, + { + "epoch": 0.11067, + "grad_norm": 0.8137206367306612, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 11067 + }, + { + "epoch": 0.11068, + "grad_norm": 0.761575070823793, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 11068 + }, + { + "epoch": 0.11069, + "grad_norm": 0.6766497718540119, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 11069 + }, + { + "epoch": 0.1107, + "grad_norm": 0.6510407350806987, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 11070 + }, + { + "epoch": 0.11071, + "grad_norm": 0.7518500682383541, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 11071 + }, + { + "epoch": 0.11072, + "grad_norm": 0.7628448692178664, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 11072 + }, + { + "epoch": 0.11073, + "grad_norm": 0.7732153009058398, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 11073 + }, + { + "epoch": 0.11074, + "grad_norm": 0.7520978215633654, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 11074 + }, + { + "epoch": 0.11075, + "grad_norm": 0.6066383268763839, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 11075 + }, + { + "epoch": 0.11076, + "grad_norm": 0.6251694306065598, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 11076 + }, + { + "epoch": 0.11077, + "grad_norm": 0.5822601666120819, + "learning_rate": 0.003, + "loss": 4.0703, + "step": 11077 + }, + { + "epoch": 0.11078, + "grad_norm": 0.6631197118032727, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11078 + }, + { + "epoch": 0.11079, + "grad_norm": 0.7007590816736342, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 11079 + }, + { + "epoch": 0.1108, + "grad_norm": 0.6772803261850265, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 11080 + }, + { + "epoch": 0.11081, + "grad_norm": 0.6144800503904058, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 11081 + }, + { + "epoch": 0.11082, + "grad_norm": 0.5159523521758294, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 11082 + }, + { + "epoch": 0.11083, + "grad_norm": 0.5618156377255118, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 11083 + }, + { + "epoch": 0.11084, + "grad_norm": 0.5273964867628337, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 11084 + }, + { + "epoch": 0.11085, + "grad_norm": 0.4825928410233014, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 11085 + }, + { + "epoch": 0.11086, + "grad_norm": 0.44661752191057275, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 11086 + }, + { + "epoch": 0.11087, + "grad_norm": 0.449128301801873, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11087 + }, + { + "epoch": 0.11088, + "grad_norm": 0.5656756087637175, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 11088 + }, + { + "epoch": 0.11089, + "grad_norm": 0.7677126171299611, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 11089 + }, + { + "epoch": 0.1109, + "grad_norm": 0.9329942484753805, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 11090 + }, + { + "epoch": 0.11091, + "grad_norm": 1.0469855443173957, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 11091 + }, + { + "epoch": 0.11092, + "grad_norm": 0.8228770878517968, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 11092 + }, + { + "epoch": 0.11093, + "grad_norm": 0.7468821729711381, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 11093 + }, + { + "epoch": 0.11094, + "grad_norm": 0.7336698323751016, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 11094 + }, + { + "epoch": 0.11095, + "grad_norm": 0.7973116882298239, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 11095 + }, + { + "epoch": 0.11096, + "grad_norm": 0.9165738647318571, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 11096 + }, + { + "epoch": 0.11097, + "grad_norm": 0.9695671258661086, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 11097 + }, + { + "epoch": 0.11098, + "grad_norm": 1.1048396716376363, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 11098 + }, + { + "epoch": 0.11099, + "grad_norm": 0.9653487324919338, + "learning_rate": 0.003, + "loss": 4.073, + "step": 11099 + }, + { + "epoch": 0.111, + "grad_norm": 1.0290738514081872, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 11100 + }, + { + "epoch": 0.11101, + "grad_norm": 0.8942114071217795, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 11101 + }, + { + "epoch": 0.11102, + "grad_norm": 0.8713504486214351, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 11102 + }, + { + "epoch": 0.11103, + "grad_norm": 0.847237082180822, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 11103 + }, + { + "epoch": 0.11104, + "grad_norm": 0.820248842940786, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 11104 + }, + { + "epoch": 0.11105, + "grad_norm": 0.8171875000553017, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 11105 + }, + { + "epoch": 0.11106, + "grad_norm": 0.7471828581379234, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 11106 + }, + { + "epoch": 0.11107, + "grad_norm": 0.6224624715446629, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 11107 + }, + { + "epoch": 0.11108, + "grad_norm": 0.5685041351462842, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 11108 + }, + { + "epoch": 0.11109, + "grad_norm": 0.4986885317042733, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 11109 + }, + { + "epoch": 0.1111, + "grad_norm": 0.558738095415561, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 11110 + }, + { + "epoch": 0.11111, + "grad_norm": 0.5414749876430993, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11111 + }, + { + "epoch": 0.11112, + "grad_norm": 0.5762906906393105, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 11112 + }, + { + "epoch": 0.11113, + "grad_norm": 0.6078542013008711, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 11113 + }, + { + "epoch": 0.11114, + "grad_norm": 0.59863639842196, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 11114 + }, + { + "epoch": 0.11115, + "grad_norm": 0.5711797510285278, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 11115 + }, + { + "epoch": 0.11116, + "grad_norm": 0.6124493121010609, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 11116 + }, + { + "epoch": 0.11117, + "grad_norm": 0.6938184460827908, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 11117 + }, + { + "epoch": 0.11118, + "grad_norm": 0.7080968425922171, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11118 + }, + { + "epoch": 0.11119, + "grad_norm": 0.6802060963538915, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 11119 + }, + { + "epoch": 0.1112, + "grad_norm": 0.6218630410093731, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 11120 + }, + { + "epoch": 0.11121, + "grad_norm": 0.5323197000918067, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 11121 + }, + { + "epoch": 0.11122, + "grad_norm": 0.4757924167276682, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 11122 + }, + { + "epoch": 0.11123, + "grad_norm": 0.5053355092515431, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 11123 + }, + { + "epoch": 0.11124, + "grad_norm": 0.5323372963182575, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 11124 + }, + { + "epoch": 0.11125, + "grad_norm": 0.5683321662081594, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 11125 + }, + { + "epoch": 0.11126, + "grad_norm": 0.5889874657122641, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 11126 + }, + { + "epoch": 0.11127, + "grad_norm": 0.6342560419915375, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11127 + }, + { + "epoch": 0.11128, + "grad_norm": 0.6464799378884137, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 11128 + }, + { + "epoch": 0.11129, + "grad_norm": 0.6827854935979535, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 11129 + }, + { + "epoch": 0.1113, + "grad_norm": 0.6717445889843138, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 11130 + }, + { + "epoch": 0.11131, + "grad_norm": 0.6968979928204649, + "learning_rate": 0.003, + "loss": 4.02, + "step": 11131 + }, + { + "epoch": 0.11132, + "grad_norm": 0.7413888249094877, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 11132 + }, + { + "epoch": 0.11133, + "grad_norm": 0.80194171241403, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 11133 + }, + { + "epoch": 0.11134, + "grad_norm": 0.8445547690054482, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 11134 + }, + { + "epoch": 0.11135, + "grad_norm": 0.8504710909047272, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 11135 + }, + { + "epoch": 0.11136, + "grad_norm": 0.7880852673843916, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 11136 + }, + { + "epoch": 0.11137, + "grad_norm": 0.7025127872709857, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 11137 + }, + { + "epoch": 0.11138, + "grad_norm": 0.7271037735166327, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 11138 + }, + { + "epoch": 0.11139, + "grad_norm": 0.7038796706319972, + "learning_rate": 0.003, + "loss": 4.042, + "step": 11139 + }, + { + "epoch": 0.1114, + "grad_norm": 0.6940605905713021, + "learning_rate": 0.003, + "loss": 4.004, + "step": 11140 + }, + { + "epoch": 0.11141, + "grad_norm": 0.7340272937162617, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 11141 + }, + { + "epoch": 0.11142, + "grad_norm": 0.8260161092736557, + "learning_rate": 0.003, + "loss": 4.031, + "step": 11142 + }, + { + "epoch": 0.11143, + "grad_norm": 0.8328044994756679, + "learning_rate": 0.003, + "loss": 4.024, + "step": 11143 + }, + { + "epoch": 0.11144, + "grad_norm": 0.8378646231160203, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 11144 + }, + { + "epoch": 0.11145, + "grad_norm": 0.7651507147152254, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 11145 + }, + { + "epoch": 0.11146, + "grad_norm": 0.8782353882115156, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 11146 + }, + { + "epoch": 0.11147, + "grad_norm": 0.8502662465495379, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 11147 + }, + { + "epoch": 0.11148, + "grad_norm": 0.8111181040349941, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 11148 + }, + { + "epoch": 0.11149, + "grad_norm": 0.7504897612397421, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 11149 + }, + { + "epoch": 0.1115, + "grad_norm": 0.8885818929104614, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 11150 + }, + { + "epoch": 0.11151, + "grad_norm": 0.8389499664722264, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 11151 + }, + { + "epoch": 0.11152, + "grad_norm": 0.7181827545901298, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 11152 + }, + { + "epoch": 0.11153, + "grad_norm": 0.7187990088446289, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 11153 + }, + { + "epoch": 0.11154, + "grad_norm": 0.6969860324972129, + "learning_rate": 0.003, + "loss": 4.037, + "step": 11154 + }, + { + "epoch": 0.11155, + "grad_norm": 0.7776562989624802, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 11155 + }, + { + "epoch": 0.11156, + "grad_norm": 0.8218529488835971, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 11156 + }, + { + "epoch": 0.11157, + "grad_norm": 0.9185748781295247, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 11157 + }, + { + "epoch": 0.11158, + "grad_norm": 0.7909595301583643, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 11158 + }, + { + "epoch": 0.11159, + "grad_norm": 0.7118487108343333, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 11159 + }, + { + "epoch": 0.1116, + "grad_norm": 0.734184564867571, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 11160 + }, + { + "epoch": 0.11161, + "grad_norm": 0.6615727770097112, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 11161 + }, + { + "epoch": 0.11162, + "grad_norm": 0.6611045555265338, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 11162 + }, + { + "epoch": 0.11163, + "grad_norm": 0.627552858503542, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 11163 + }, + { + "epoch": 0.11164, + "grad_norm": 0.7179781735178405, + "learning_rate": 0.003, + "loss": 4.033, + "step": 11164 + }, + { + "epoch": 0.11165, + "grad_norm": 0.7194096521849105, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11165 + }, + { + "epoch": 0.11166, + "grad_norm": 0.7643526920434413, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 11166 + }, + { + "epoch": 0.11167, + "grad_norm": 0.8075709547528581, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 11167 + }, + { + "epoch": 0.11168, + "grad_norm": 0.8497242847469146, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 11168 + }, + { + "epoch": 0.11169, + "grad_norm": 0.8476663900400179, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11169 + }, + { + "epoch": 0.1117, + "grad_norm": 0.8393908389719942, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11170 + }, + { + "epoch": 0.11171, + "grad_norm": 0.7889809555263858, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 11171 + }, + { + "epoch": 0.11172, + "grad_norm": 0.7282441749855356, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11172 + }, + { + "epoch": 0.11173, + "grad_norm": 0.7367511031109865, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11173 + }, + { + "epoch": 0.11174, + "grad_norm": 0.702374527043833, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 11174 + }, + { + "epoch": 0.11175, + "grad_norm": 0.646798495956887, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 11175 + }, + { + "epoch": 0.11176, + "grad_norm": 0.646526374636163, + "learning_rate": 0.003, + "loss": 4.006, + "step": 11176 + }, + { + "epoch": 0.11177, + "grad_norm": 0.6367166145876565, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 11177 + }, + { + "epoch": 0.11178, + "grad_norm": 0.6087248205379019, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 11178 + }, + { + "epoch": 0.11179, + "grad_norm": 0.5629941397891994, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 11179 + }, + { + "epoch": 0.1118, + "grad_norm": 0.49226163675300355, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 11180 + }, + { + "epoch": 0.11181, + "grad_norm": 0.496324678743459, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11181 + }, + { + "epoch": 0.11182, + "grad_norm": 0.5689340650747319, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11182 + }, + { + "epoch": 0.11183, + "grad_norm": 0.708681046210253, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 11183 + }, + { + "epoch": 0.11184, + "grad_norm": 0.8053278161333081, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 11184 + }, + { + "epoch": 0.11185, + "grad_norm": 0.7856228607454754, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11185 + }, + { + "epoch": 0.11186, + "grad_norm": 0.7892050876729113, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 11186 + }, + { + "epoch": 0.11187, + "grad_norm": 0.7936187289251392, + "learning_rate": 0.003, + "loss": 3.996, + "step": 11187 + }, + { + "epoch": 0.11188, + "grad_norm": 0.6702246257895588, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11188 + }, + { + "epoch": 0.11189, + "grad_norm": 0.6452770826018868, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 11189 + }, + { + "epoch": 0.1119, + "grad_norm": 0.6509759310307832, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 11190 + }, + { + "epoch": 0.11191, + "grad_norm": 0.647099225744172, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 11191 + }, + { + "epoch": 0.11192, + "grad_norm": 0.6358559793048039, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 11192 + }, + { + "epoch": 0.11193, + "grad_norm": 0.6048969324853621, + "learning_rate": 0.003, + "loss": 4.023, + "step": 11193 + }, + { + "epoch": 0.11194, + "grad_norm": 0.5867605166585314, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 11194 + }, + { + "epoch": 0.11195, + "grad_norm": 0.6328236818388516, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 11195 + }, + { + "epoch": 0.11196, + "grad_norm": 0.6808722249438569, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 11196 + }, + { + "epoch": 0.11197, + "grad_norm": 0.6855342780836314, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 11197 + }, + { + "epoch": 0.11198, + "grad_norm": 0.6944461258780181, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 11198 + }, + { + "epoch": 0.11199, + "grad_norm": 0.6006250008297997, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11199 + }, + { + "epoch": 0.112, + "grad_norm": 0.6660478312784981, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 11200 + }, + { + "epoch": 0.11201, + "grad_norm": 0.8103401387134324, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 11201 + }, + { + "epoch": 0.11202, + "grad_norm": 0.9087171911020394, + "learning_rate": 0.003, + "loss": 4.032, + "step": 11202 + }, + { + "epoch": 0.11203, + "grad_norm": 0.932627970858704, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 11203 + }, + { + "epoch": 0.11204, + "grad_norm": 0.7461858886887853, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 11204 + }, + { + "epoch": 0.11205, + "grad_norm": 0.8382520188776988, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 11205 + }, + { + "epoch": 0.11206, + "grad_norm": 0.9438354393027933, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 11206 + }, + { + "epoch": 0.11207, + "grad_norm": 0.9344071517339366, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 11207 + }, + { + "epoch": 0.11208, + "grad_norm": 0.683781420199349, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 11208 + }, + { + "epoch": 0.11209, + "grad_norm": 0.7663808592976001, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 11209 + }, + { + "epoch": 0.1121, + "grad_norm": 0.7660003572311983, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 11210 + }, + { + "epoch": 0.11211, + "grad_norm": 0.8689077081251736, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 11211 + }, + { + "epoch": 0.11212, + "grad_norm": 0.7287009994488349, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 11212 + }, + { + "epoch": 0.11213, + "grad_norm": 0.6428681579986865, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 11213 + }, + { + "epoch": 0.11214, + "grad_norm": 0.5592311592973406, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 11214 + }, + { + "epoch": 0.11215, + "grad_norm": 0.5860654421142516, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 11215 + }, + { + "epoch": 0.11216, + "grad_norm": 0.5933053146118373, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 11216 + }, + { + "epoch": 0.11217, + "grad_norm": 0.6276493883998596, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11217 + }, + { + "epoch": 0.11218, + "grad_norm": 0.6242329840466302, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 11218 + }, + { + "epoch": 0.11219, + "grad_norm": 0.7402407976828813, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 11219 + }, + { + "epoch": 0.1122, + "grad_norm": 0.7399958113217744, + "learning_rate": 0.003, + "loss": 3.991, + "step": 11220 + }, + { + "epoch": 0.11221, + "grad_norm": 0.6551045323699699, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 11221 + }, + { + "epoch": 0.11222, + "grad_norm": 0.598123389182672, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 11222 + }, + { + "epoch": 0.11223, + "grad_norm": 0.5492494393258556, + "learning_rate": 0.003, + "loss": 4.009, + "step": 11223 + }, + { + "epoch": 0.11224, + "grad_norm": 0.5696006819464087, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 11224 + }, + { + "epoch": 0.11225, + "grad_norm": 0.5491115099630085, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 11225 + }, + { + "epoch": 0.11226, + "grad_norm": 0.5663917990042383, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 11226 + }, + { + "epoch": 0.11227, + "grad_norm": 0.6213876023916232, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 11227 + }, + { + "epoch": 0.11228, + "grad_norm": 0.7003782896584353, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 11228 + }, + { + "epoch": 0.11229, + "grad_norm": 0.7647615280089896, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 11229 + }, + { + "epoch": 0.1123, + "grad_norm": 0.9415183944969855, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 11230 + }, + { + "epoch": 0.11231, + "grad_norm": 1.076485159204218, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 11231 + }, + { + "epoch": 0.11232, + "grad_norm": 0.9240100207853978, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 11232 + }, + { + "epoch": 0.11233, + "grad_norm": 0.8144293618623736, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 11233 + }, + { + "epoch": 0.11234, + "grad_norm": 0.8186947788379302, + "learning_rate": 0.003, + "loss": 4.028, + "step": 11234 + }, + { + "epoch": 0.11235, + "grad_norm": 0.6864079106123024, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 11235 + }, + { + "epoch": 0.11236, + "grad_norm": 0.7487026094501924, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11236 + }, + { + "epoch": 0.11237, + "grad_norm": 0.7684708913428342, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 11237 + }, + { + "epoch": 0.11238, + "grad_norm": 0.8737833644128175, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 11238 + }, + { + "epoch": 0.11239, + "grad_norm": 0.8827669162982216, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 11239 + }, + { + "epoch": 0.1124, + "grad_norm": 0.769176223914237, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 11240 + }, + { + "epoch": 0.11241, + "grad_norm": 0.7298835570558875, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 11241 + }, + { + "epoch": 0.11242, + "grad_norm": 0.6553532902657966, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 11242 + }, + { + "epoch": 0.11243, + "grad_norm": 0.6304303950255465, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 11243 + }, + { + "epoch": 0.11244, + "grad_norm": 0.6585711882228287, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 11244 + }, + { + "epoch": 0.11245, + "grad_norm": 0.5952585396623512, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 11245 + }, + { + "epoch": 0.11246, + "grad_norm": 0.6409828503214502, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 11246 + }, + { + "epoch": 0.11247, + "grad_norm": 0.7200591265909396, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 11247 + }, + { + "epoch": 0.11248, + "grad_norm": 0.7499006140894048, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11248 + }, + { + "epoch": 0.11249, + "grad_norm": 0.7615993543455494, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11249 + }, + { + "epoch": 0.1125, + "grad_norm": 0.7401415979226063, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 11250 + }, + { + "epoch": 0.11251, + "grad_norm": 0.7228426270736222, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 11251 + }, + { + "epoch": 0.11252, + "grad_norm": 0.7683499411728729, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 11252 + }, + { + "epoch": 0.11253, + "grad_norm": 0.7653726955410174, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 11253 + }, + { + "epoch": 0.11254, + "grad_norm": 0.7392495186385193, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 11254 + }, + { + "epoch": 0.11255, + "grad_norm": 0.698958382936842, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 11255 + }, + { + "epoch": 0.11256, + "grad_norm": 0.6725318057992107, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 11256 + }, + { + "epoch": 0.11257, + "grad_norm": 0.7362903918658591, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 11257 + }, + { + "epoch": 0.11258, + "grad_norm": 0.7260458906504509, + "learning_rate": 0.003, + "loss": 4.024, + "step": 11258 + }, + { + "epoch": 0.11259, + "grad_norm": 0.6934970021277412, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 11259 + }, + { + "epoch": 0.1126, + "grad_norm": 0.6770055900441696, + "learning_rate": 0.003, + "loss": 4.006, + "step": 11260 + }, + { + "epoch": 0.11261, + "grad_norm": 0.6990730922835574, + "learning_rate": 0.003, + "loss": 4.015, + "step": 11261 + }, + { + "epoch": 0.11262, + "grad_norm": 0.8948681893646525, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 11262 + }, + { + "epoch": 0.11263, + "grad_norm": 0.9058162704146794, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 11263 + }, + { + "epoch": 0.11264, + "grad_norm": 0.7718936804511218, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 11264 + }, + { + "epoch": 0.11265, + "grad_norm": 0.6562100528511404, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 11265 + }, + { + "epoch": 0.11266, + "grad_norm": 0.6530045205956541, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 11266 + }, + { + "epoch": 0.11267, + "grad_norm": 0.5883616678534866, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 11267 + }, + { + "epoch": 0.11268, + "grad_norm": 0.6103860067959307, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 11268 + }, + { + "epoch": 0.11269, + "grad_norm": 0.5012179752286267, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 11269 + }, + { + "epoch": 0.1127, + "grad_norm": 0.49087422836203315, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11270 + }, + { + "epoch": 0.11271, + "grad_norm": 0.5088804940891966, + "learning_rate": 0.003, + "loss": 4.023, + "step": 11271 + }, + { + "epoch": 0.11272, + "grad_norm": 0.4770487387038262, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 11272 + }, + { + "epoch": 0.11273, + "grad_norm": 0.4132970061118445, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11273 + }, + { + "epoch": 0.11274, + "grad_norm": 0.44183517773424674, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 11274 + }, + { + "epoch": 0.11275, + "grad_norm": 0.4855415454041357, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 11275 + }, + { + "epoch": 0.11276, + "grad_norm": 0.5511246504086101, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 11276 + }, + { + "epoch": 0.11277, + "grad_norm": 0.6117156669229906, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 11277 + }, + { + "epoch": 0.11278, + "grad_norm": 0.6176581368786838, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 11278 + }, + { + "epoch": 0.11279, + "grad_norm": 0.5575859750846393, + "learning_rate": 0.003, + "loss": 4.007, + "step": 11279 + }, + { + "epoch": 0.1128, + "grad_norm": 0.49854003118402845, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 11280 + }, + { + "epoch": 0.11281, + "grad_norm": 0.5154375623553807, + "learning_rate": 0.003, + "loss": 4.02, + "step": 11281 + }, + { + "epoch": 0.11282, + "grad_norm": 0.5232233709942603, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 11282 + }, + { + "epoch": 0.11283, + "grad_norm": 0.6166468635674542, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 11283 + }, + { + "epoch": 0.11284, + "grad_norm": 0.7706296690693606, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 11284 + }, + { + "epoch": 0.11285, + "grad_norm": 1.0433966148029477, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 11285 + }, + { + "epoch": 0.11286, + "grad_norm": 0.9136464330292834, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 11286 + }, + { + "epoch": 0.11287, + "grad_norm": 0.6742474540243751, + "learning_rate": 0.003, + "loss": 4.004, + "step": 11287 + }, + { + "epoch": 0.11288, + "grad_norm": 0.6776561746687654, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 11288 + }, + { + "epoch": 0.11289, + "grad_norm": 0.7023207730554802, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 11289 + }, + { + "epoch": 0.1129, + "grad_norm": 0.9170065518925117, + "learning_rate": 0.003, + "loss": 4.017, + "step": 11290 + }, + { + "epoch": 0.11291, + "grad_norm": 1.0717484298445887, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 11291 + }, + { + "epoch": 0.11292, + "grad_norm": 0.8871854354881722, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 11292 + }, + { + "epoch": 0.11293, + "grad_norm": 0.8554023113033866, + "learning_rate": 0.003, + "loss": 4.0579, + "step": 11293 + }, + { + "epoch": 0.11294, + "grad_norm": 0.8518847442178202, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 11294 + }, + { + "epoch": 0.11295, + "grad_norm": 0.9746823845296133, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 11295 + }, + { + "epoch": 0.11296, + "grad_norm": 0.9740213724274002, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 11296 + }, + { + "epoch": 0.11297, + "grad_norm": 0.9085557224266088, + "learning_rate": 0.003, + "loss": 4.0722, + "step": 11297 + }, + { + "epoch": 0.11298, + "grad_norm": 0.841123051379851, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 11298 + }, + { + "epoch": 0.11299, + "grad_norm": 0.8552388898158186, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 11299 + }, + { + "epoch": 0.113, + "grad_norm": 0.7987700422456869, + "learning_rate": 0.003, + "loss": 4.0605, + "step": 11300 + }, + { + "epoch": 0.11301, + "grad_norm": 0.8237978460165376, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 11301 + }, + { + "epoch": 0.11302, + "grad_norm": 0.7948945627089461, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 11302 + }, + { + "epoch": 0.11303, + "grad_norm": 0.8036636279987573, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 11303 + }, + { + "epoch": 0.11304, + "grad_norm": 0.7023865942078346, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 11304 + }, + { + "epoch": 0.11305, + "grad_norm": 0.7171317295612506, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 11305 + }, + { + "epoch": 0.11306, + "grad_norm": 0.6209921521908508, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 11306 + }, + { + "epoch": 0.11307, + "grad_norm": 0.6637747400695151, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 11307 + }, + { + "epoch": 0.11308, + "grad_norm": 0.664998104023118, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 11308 + }, + { + "epoch": 0.11309, + "grad_norm": 0.7089661003814816, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 11309 + }, + { + "epoch": 0.1131, + "grad_norm": 0.7730050408440713, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 11310 + }, + { + "epoch": 0.11311, + "grad_norm": 0.8464784913828908, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 11311 + }, + { + "epoch": 0.11312, + "grad_norm": 0.7925284161498007, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 11312 + }, + { + "epoch": 0.11313, + "grad_norm": 0.6699218072682307, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 11313 + }, + { + "epoch": 0.11314, + "grad_norm": 0.6774719830458245, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 11314 + }, + { + "epoch": 0.11315, + "grad_norm": 0.7124045324501616, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 11315 + }, + { + "epoch": 0.11316, + "grad_norm": 0.6874664861491411, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 11316 + }, + { + "epoch": 0.11317, + "grad_norm": 0.7374819344394455, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 11317 + }, + { + "epoch": 0.11318, + "grad_norm": 0.8303897730552688, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 11318 + }, + { + "epoch": 0.11319, + "grad_norm": 0.8295865477585235, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 11319 + }, + { + "epoch": 0.1132, + "grad_norm": 0.6974040854700542, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 11320 + }, + { + "epoch": 0.11321, + "grad_norm": 0.5460555856863145, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 11321 + }, + { + "epoch": 0.11322, + "grad_norm": 0.5609934486270876, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 11322 + }, + { + "epoch": 0.11323, + "grad_norm": 0.5942662901975376, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 11323 + }, + { + "epoch": 0.11324, + "grad_norm": 0.5466588375950594, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 11324 + }, + { + "epoch": 0.11325, + "grad_norm": 0.503264464733193, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11325 + }, + { + "epoch": 0.11326, + "grad_norm": 0.49907816973388436, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 11326 + }, + { + "epoch": 0.11327, + "grad_norm": 0.554554264966638, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 11327 + }, + { + "epoch": 0.11328, + "grad_norm": 0.6701808398498045, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 11328 + }, + { + "epoch": 0.11329, + "grad_norm": 0.6797872241050426, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 11329 + }, + { + "epoch": 0.1133, + "grad_norm": 0.6241288959543494, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 11330 + }, + { + "epoch": 0.11331, + "grad_norm": 0.5974880714010611, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 11331 + }, + { + "epoch": 0.11332, + "grad_norm": 0.5876201759520249, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11332 + }, + { + "epoch": 0.11333, + "grad_norm": 0.6390845966731439, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 11333 + }, + { + "epoch": 0.11334, + "grad_norm": 0.6083509321306598, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 11334 + }, + { + "epoch": 0.11335, + "grad_norm": 0.637663354226681, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 11335 + }, + { + "epoch": 0.11336, + "grad_norm": 0.6371118015917777, + "learning_rate": 0.003, + "loss": 4.0, + "step": 11336 + }, + { + "epoch": 0.11337, + "grad_norm": 0.6064335283726469, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 11337 + }, + { + "epoch": 0.11338, + "grad_norm": 0.5459763941391882, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 11338 + }, + { + "epoch": 0.11339, + "grad_norm": 0.5384898687667063, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 11339 + }, + { + "epoch": 0.1134, + "grad_norm": 0.706333978783952, + "learning_rate": 0.003, + "loss": 4.022, + "step": 11340 + }, + { + "epoch": 0.11341, + "grad_norm": 0.8444592435754241, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 11341 + }, + { + "epoch": 0.11342, + "grad_norm": 0.8252437500456932, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 11342 + }, + { + "epoch": 0.11343, + "grad_norm": 0.8391715740890706, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 11343 + }, + { + "epoch": 0.11344, + "grad_norm": 0.8021043635456919, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 11344 + }, + { + "epoch": 0.11345, + "grad_norm": 0.820843078221304, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 11345 + }, + { + "epoch": 0.11346, + "grad_norm": 0.8990445009438861, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 11346 + }, + { + "epoch": 0.11347, + "grad_norm": 0.8205187855660475, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 11347 + }, + { + "epoch": 0.11348, + "grad_norm": 0.860167922053747, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 11348 + }, + { + "epoch": 0.11349, + "grad_norm": 0.9058895032287795, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 11349 + }, + { + "epoch": 0.1135, + "grad_norm": 0.9329072936968235, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 11350 + }, + { + "epoch": 0.11351, + "grad_norm": 1.0025293232447625, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 11351 + }, + { + "epoch": 0.11352, + "grad_norm": 1.105101588145658, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 11352 + }, + { + "epoch": 0.11353, + "grad_norm": 1.0032143270526348, + "learning_rate": 0.003, + "loss": 4.034, + "step": 11353 + }, + { + "epoch": 0.11354, + "grad_norm": 0.8179848874229991, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 11354 + }, + { + "epoch": 0.11355, + "grad_norm": 0.8482061004243556, + "learning_rate": 0.003, + "loss": 4.0798, + "step": 11355 + }, + { + "epoch": 0.11356, + "grad_norm": 0.7723682744487419, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 11356 + }, + { + "epoch": 0.11357, + "grad_norm": 0.6818379437674638, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 11357 + }, + { + "epoch": 0.11358, + "grad_norm": 0.7014483879995032, + "learning_rate": 0.003, + "loss": 4.04, + "step": 11358 + }, + { + "epoch": 0.11359, + "grad_norm": 0.744279339398951, + "learning_rate": 0.003, + "loss": 4.0621, + "step": 11359 + }, + { + "epoch": 0.1136, + "grad_norm": 0.6938038578591156, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 11360 + }, + { + "epoch": 0.11361, + "grad_norm": 0.613784350567357, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 11361 + }, + { + "epoch": 0.11362, + "grad_norm": 0.6196124422108985, + "learning_rate": 0.003, + "loss": 4.032, + "step": 11362 + }, + { + "epoch": 0.11363, + "grad_norm": 0.6501688456659973, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 11363 + }, + { + "epoch": 0.11364, + "grad_norm": 0.6007701303648963, + "learning_rate": 0.003, + "loss": 4.044, + "step": 11364 + }, + { + "epoch": 0.11365, + "grad_norm": 0.5508839382751611, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11365 + }, + { + "epoch": 0.11366, + "grad_norm": 0.5206555247089146, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 11366 + }, + { + "epoch": 0.11367, + "grad_norm": 0.5290295346344233, + "learning_rate": 0.003, + "loss": 3.996, + "step": 11367 + }, + { + "epoch": 0.11368, + "grad_norm": 0.5439010333171783, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 11368 + }, + { + "epoch": 0.11369, + "grad_norm": 0.5239485836827933, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 11369 + }, + { + "epoch": 0.1137, + "grad_norm": 0.5348728084475604, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 11370 + }, + { + "epoch": 0.11371, + "grad_norm": 0.6230286436704715, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 11371 + }, + { + "epoch": 0.11372, + "grad_norm": 0.6854265652633135, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 11372 + }, + { + "epoch": 0.11373, + "grad_norm": 0.6438634027872718, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 11373 + }, + { + "epoch": 0.11374, + "grad_norm": 0.6905320144399599, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 11374 + }, + { + "epoch": 0.11375, + "grad_norm": 0.8050503795718816, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 11375 + }, + { + "epoch": 0.11376, + "grad_norm": 0.8488958334920207, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 11376 + }, + { + "epoch": 0.11377, + "grad_norm": 0.8930027685694822, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 11377 + }, + { + "epoch": 0.11378, + "grad_norm": 0.8473319019998473, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 11378 + }, + { + "epoch": 0.11379, + "grad_norm": 0.7381194114254623, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 11379 + }, + { + "epoch": 0.1138, + "grad_norm": 0.6202881529870268, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 11380 + }, + { + "epoch": 0.11381, + "grad_norm": 0.6156903338560075, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 11381 + }, + { + "epoch": 0.11382, + "grad_norm": 0.7528440754753146, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 11382 + }, + { + "epoch": 0.11383, + "grad_norm": 0.7447760084445988, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 11383 + }, + { + "epoch": 0.11384, + "grad_norm": 0.6417152108153082, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 11384 + }, + { + "epoch": 0.11385, + "grad_norm": 0.5841160374388131, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 11385 + }, + { + "epoch": 0.11386, + "grad_norm": 0.6237088498817305, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 11386 + }, + { + "epoch": 0.11387, + "grad_norm": 0.6204845558958338, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 11387 + }, + { + "epoch": 0.11388, + "grad_norm": 0.6397881975955572, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 11388 + }, + { + "epoch": 0.11389, + "grad_norm": 0.6825372639035735, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 11389 + }, + { + "epoch": 0.1139, + "grad_norm": 0.7046557903205382, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 11390 + }, + { + "epoch": 0.11391, + "grad_norm": 0.7614558889637134, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 11391 + }, + { + "epoch": 0.11392, + "grad_norm": 0.7858498957148728, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 11392 + }, + { + "epoch": 0.11393, + "grad_norm": 0.8107269500017864, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 11393 + }, + { + "epoch": 0.11394, + "grad_norm": 0.7398978576411472, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 11394 + }, + { + "epoch": 0.11395, + "grad_norm": 0.6695318166812531, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 11395 + }, + { + "epoch": 0.11396, + "grad_norm": 0.682597244969728, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 11396 + }, + { + "epoch": 0.11397, + "grad_norm": 0.6919639524833384, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 11397 + }, + { + "epoch": 0.11398, + "grad_norm": 0.7861630924318441, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 11398 + }, + { + "epoch": 0.11399, + "grad_norm": 0.8825330103376888, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 11399 + }, + { + "epoch": 0.114, + "grad_norm": 0.8590975803690928, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 11400 + }, + { + "epoch": 0.11401, + "grad_norm": 0.6746573093942401, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 11401 + }, + { + "epoch": 0.11402, + "grad_norm": 0.6018701787077438, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 11402 + }, + { + "epoch": 0.11403, + "grad_norm": 0.5865799266420479, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 11403 + }, + { + "epoch": 0.11404, + "grad_norm": 0.5672664304109833, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 11404 + }, + { + "epoch": 0.11405, + "grad_norm": 0.5847722646032557, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 11405 + }, + { + "epoch": 0.11406, + "grad_norm": 0.6241980571238591, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 11406 + }, + { + "epoch": 0.11407, + "grad_norm": 0.6094852251635454, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 11407 + }, + { + "epoch": 0.11408, + "grad_norm": 0.5669879167833364, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 11408 + }, + { + "epoch": 0.11409, + "grad_norm": 0.6701961171143586, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 11409 + }, + { + "epoch": 0.1141, + "grad_norm": 0.7694364844769717, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 11410 + }, + { + "epoch": 0.11411, + "grad_norm": 0.9237203738020683, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 11411 + }, + { + "epoch": 0.11412, + "grad_norm": 0.9469982128554808, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 11412 + }, + { + "epoch": 0.11413, + "grad_norm": 1.0260463729621787, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 11413 + }, + { + "epoch": 0.11414, + "grad_norm": 0.7683022512875309, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 11414 + }, + { + "epoch": 0.11415, + "grad_norm": 0.7631002650767835, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 11415 + }, + { + "epoch": 0.11416, + "grad_norm": 0.6375787522404956, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 11416 + }, + { + "epoch": 0.11417, + "grad_norm": 0.6129093693870531, + "learning_rate": 0.003, + "loss": 4.009, + "step": 11417 + }, + { + "epoch": 0.11418, + "grad_norm": 0.5910794131898679, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 11418 + }, + { + "epoch": 0.11419, + "grad_norm": 0.6277850870119095, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 11419 + }, + { + "epoch": 0.1142, + "grad_norm": 0.6416001710530855, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 11420 + }, + { + "epoch": 0.11421, + "grad_norm": 0.6669101407527673, + "learning_rate": 0.003, + "loss": 3.993, + "step": 11421 + }, + { + "epoch": 0.11422, + "grad_norm": 0.6955711902043289, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 11422 + }, + { + "epoch": 0.11423, + "grad_norm": 0.6652923829227967, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 11423 + }, + { + "epoch": 0.11424, + "grad_norm": 0.6521808774677563, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 11424 + }, + { + "epoch": 0.11425, + "grad_norm": 0.6934143594417089, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 11425 + }, + { + "epoch": 0.11426, + "grad_norm": 0.6433270230025018, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 11426 + }, + { + "epoch": 0.11427, + "grad_norm": 0.6770764021762676, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 11427 + }, + { + "epoch": 0.11428, + "grad_norm": 0.7186376149858043, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 11428 + }, + { + "epoch": 0.11429, + "grad_norm": 0.8072180733985791, + "learning_rate": 0.003, + "loss": 4.056, + "step": 11429 + }, + { + "epoch": 0.1143, + "grad_norm": 0.7949663040689542, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 11430 + }, + { + "epoch": 0.11431, + "grad_norm": 0.7180993500062615, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 11431 + }, + { + "epoch": 0.11432, + "grad_norm": 0.7235519007715454, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 11432 + }, + { + "epoch": 0.11433, + "grad_norm": 0.7866623201952263, + "learning_rate": 0.003, + "loss": 4.064, + "step": 11433 + }, + { + "epoch": 0.11434, + "grad_norm": 0.8317059636777636, + "learning_rate": 0.003, + "loss": 4.059, + "step": 11434 + }, + { + "epoch": 0.11435, + "grad_norm": 0.8000885842923394, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 11435 + }, + { + "epoch": 0.11436, + "grad_norm": 0.6900855137776525, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 11436 + }, + { + "epoch": 0.11437, + "grad_norm": 0.6760506044991164, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 11437 + }, + { + "epoch": 0.11438, + "grad_norm": 0.6666591948165018, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 11438 + }, + { + "epoch": 0.11439, + "grad_norm": 0.7542207802994385, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 11439 + }, + { + "epoch": 0.1144, + "grad_norm": 0.7623625969801058, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 11440 + }, + { + "epoch": 0.11441, + "grad_norm": 0.7736139057005444, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 11441 + }, + { + "epoch": 0.11442, + "grad_norm": 0.7449847530955542, + "learning_rate": 0.003, + "loss": 4.022, + "step": 11442 + }, + { + "epoch": 0.11443, + "grad_norm": 0.868795943771287, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11443 + }, + { + "epoch": 0.11444, + "grad_norm": 1.0466762117547996, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11444 + }, + { + "epoch": 0.11445, + "grad_norm": 1.0399008308567075, + "learning_rate": 0.003, + "loss": 4.048, + "step": 11445 + }, + { + "epoch": 0.11446, + "grad_norm": 0.8550516829935646, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 11446 + }, + { + "epoch": 0.11447, + "grad_norm": 0.7587012150628428, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 11447 + }, + { + "epoch": 0.11448, + "grad_norm": 0.7454505970197419, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 11448 + }, + { + "epoch": 0.11449, + "grad_norm": 0.7624948025983831, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 11449 + }, + { + "epoch": 0.1145, + "grad_norm": 0.7964369275888208, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 11450 + }, + { + "epoch": 0.11451, + "grad_norm": 0.8090753804094553, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11451 + }, + { + "epoch": 0.11452, + "grad_norm": 0.6495355747407816, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 11452 + }, + { + "epoch": 0.11453, + "grad_norm": 0.6369491726738808, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 11453 + }, + { + "epoch": 0.11454, + "grad_norm": 0.701552378386301, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 11454 + }, + { + "epoch": 0.11455, + "grad_norm": 0.7015626382092112, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 11455 + }, + { + "epoch": 0.11456, + "grad_norm": 0.6080236485651209, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 11456 + }, + { + "epoch": 0.11457, + "grad_norm": 0.6385171057709782, + "learning_rate": 0.003, + "loss": 4.011, + "step": 11457 + }, + { + "epoch": 0.11458, + "grad_norm": 0.7319166231567201, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 11458 + }, + { + "epoch": 0.11459, + "grad_norm": 0.8158034770531701, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 11459 + }, + { + "epoch": 0.1146, + "grad_norm": 0.8884956622552105, + "learning_rate": 0.003, + "loss": 4.019, + "step": 11460 + }, + { + "epoch": 0.11461, + "grad_norm": 0.8742461317291716, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 11461 + }, + { + "epoch": 0.11462, + "grad_norm": 0.9274878200600121, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 11462 + }, + { + "epoch": 0.11463, + "grad_norm": 0.8681602502135649, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11463 + }, + { + "epoch": 0.11464, + "grad_norm": 0.8333490198799206, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 11464 + }, + { + "epoch": 0.11465, + "grad_norm": 0.7744924799859181, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 11465 + }, + { + "epoch": 0.11466, + "grad_norm": 0.7677706520045092, + "learning_rate": 0.003, + "loss": 4.027, + "step": 11466 + }, + { + "epoch": 0.11467, + "grad_norm": 0.7479464866157822, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 11467 + }, + { + "epoch": 0.11468, + "grad_norm": 0.7495822372316991, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 11468 + }, + { + "epoch": 0.11469, + "grad_norm": 0.825072983937388, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 11469 + }, + { + "epoch": 0.1147, + "grad_norm": 0.7640559736725657, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 11470 + }, + { + "epoch": 0.11471, + "grad_norm": 0.6155740039622885, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 11471 + }, + { + "epoch": 0.11472, + "grad_norm": 0.49321581019324, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 11472 + }, + { + "epoch": 0.11473, + "grad_norm": 0.48529711873416564, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 11473 + }, + { + "epoch": 0.11474, + "grad_norm": 0.49175351342543316, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 11474 + }, + { + "epoch": 0.11475, + "grad_norm": 0.4447890317931365, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 11475 + }, + { + "epoch": 0.11476, + "grad_norm": 0.45241296622707033, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 11476 + }, + { + "epoch": 0.11477, + "grad_norm": 0.46322940824987113, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 11477 + }, + { + "epoch": 0.11478, + "grad_norm": 0.48862362581554364, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 11478 + }, + { + "epoch": 0.11479, + "grad_norm": 0.45652857582094597, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 11479 + }, + { + "epoch": 0.1148, + "grad_norm": 0.4895281177961268, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 11480 + }, + { + "epoch": 0.11481, + "grad_norm": 0.6000548594907348, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 11481 + }, + { + "epoch": 0.11482, + "grad_norm": 0.6812738834634311, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 11482 + }, + { + "epoch": 0.11483, + "grad_norm": 0.6856211039369874, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 11483 + }, + { + "epoch": 0.11484, + "grad_norm": 0.7055256892075299, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 11484 + }, + { + "epoch": 0.11485, + "grad_norm": 0.6953957975752768, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 11485 + }, + { + "epoch": 0.11486, + "grad_norm": 0.7414118490585749, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 11486 + }, + { + "epoch": 0.11487, + "grad_norm": 0.7502166980674726, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11487 + }, + { + "epoch": 0.11488, + "grad_norm": 0.6504020520836259, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 11488 + }, + { + "epoch": 0.11489, + "grad_norm": 0.6648489127481131, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 11489 + }, + { + "epoch": 0.1149, + "grad_norm": 0.6245019396955147, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 11490 + }, + { + "epoch": 0.11491, + "grad_norm": 0.6354173945744755, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 11491 + }, + { + "epoch": 0.11492, + "grad_norm": 0.6819311236706822, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11492 + }, + { + "epoch": 0.11493, + "grad_norm": 0.6828612621086475, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 11493 + }, + { + "epoch": 0.11494, + "grad_norm": 0.7577761710921894, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 11494 + }, + { + "epoch": 0.11495, + "grad_norm": 0.7354075792008131, + "learning_rate": 0.003, + "loss": 4.025, + "step": 11495 + }, + { + "epoch": 0.11496, + "grad_norm": 0.7759939479798487, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 11496 + }, + { + "epoch": 0.11497, + "grad_norm": 0.8987019653781375, + "learning_rate": 0.003, + "loss": 4.016, + "step": 11497 + }, + { + "epoch": 0.11498, + "grad_norm": 0.9790533965617638, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 11498 + }, + { + "epoch": 0.11499, + "grad_norm": 1.0668999136572301, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 11499 + }, + { + "epoch": 0.115, + "grad_norm": 0.9776410500713499, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 11500 + }, + { + "epoch": 0.11501, + "grad_norm": 1.0403486112166538, + "learning_rate": 0.003, + "loss": 4.0705, + "step": 11501 + }, + { + "epoch": 0.11502, + "grad_norm": 0.8945206632399854, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 11502 + }, + { + "epoch": 0.11503, + "grad_norm": 0.8682989700726157, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11503 + }, + { + "epoch": 0.11504, + "grad_norm": 0.9946773381001365, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 11504 + }, + { + "epoch": 0.11505, + "grad_norm": 0.8977978045736503, + "learning_rate": 0.003, + "loss": 4.042, + "step": 11505 + }, + { + "epoch": 0.11506, + "grad_norm": 0.8517053942267596, + "learning_rate": 0.003, + "loss": 4.0697, + "step": 11506 + }, + { + "epoch": 0.11507, + "grad_norm": 0.7163973234490041, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 11507 + }, + { + "epoch": 0.11508, + "grad_norm": 0.7105543648593496, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 11508 + }, + { + "epoch": 0.11509, + "grad_norm": 0.6854840383128519, + "learning_rate": 0.003, + "loss": 4.022, + "step": 11509 + }, + { + "epoch": 0.1151, + "grad_norm": 0.588990980349762, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 11510 + }, + { + "epoch": 0.11511, + "grad_norm": 0.5824874370234361, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 11511 + }, + { + "epoch": 0.11512, + "grad_norm": 0.46862660203771767, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 11512 + }, + { + "epoch": 0.11513, + "grad_norm": 0.470390151868247, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 11513 + }, + { + "epoch": 0.11514, + "grad_norm": 0.5467663852595315, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 11514 + }, + { + "epoch": 0.11515, + "grad_norm": 0.6130936401355774, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 11515 + }, + { + "epoch": 0.11516, + "grad_norm": 0.6718355097971194, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 11516 + }, + { + "epoch": 0.11517, + "grad_norm": 0.7380303997866261, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 11517 + }, + { + "epoch": 0.11518, + "grad_norm": 0.668778710748661, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 11518 + }, + { + "epoch": 0.11519, + "grad_norm": 0.5203879439475595, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 11519 + }, + { + "epoch": 0.1152, + "grad_norm": 0.5220018335206423, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 11520 + }, + { + "epoch": 0.11521, + "grad_norm": 0.513380162641383, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 11521 + }, + { + "epoch": 0.11522, + "grad_norm": 0.5532249361225984, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 11522 + }, + { + "epoch": 0.11523, + "grad_norm": 0.5725125144160189, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 11523 + }, + { + "epoch": 0.11524, + "grad_norm": 0.5943263687783578, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 11524 + }, + { + "epoch": 0.11525, + "grad_norm": 0.5422037745866911, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 11525 + }, + { + "epoch": 0.11526, + "grad_norm": 0.5505296928806156, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 11526 + }, + { + "epoch": 0.11527, + "grad_norm": 0.5392692569587317, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 11527 + }, + { + "epoch": 0.11528, + "grad_norm": 0.5665463424656342, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 11528 + }, + { + "epoch": 0.11529, + "grad_norm": 0.6174097454946429, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 11529 + }, + { + "epoch": 0.1153, + "grad_norm": 0.7043279986388752, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 11530 + }, + { + "epoch": 0.11531, + "grad_norm": 0.6948193263651169, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 11531 + }, + { + "epoch": 0.11532, + "grad_norm": 0.6167114553212313, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 11532 + }, + { + "epoch": 0.11533, + "grad_norm": 0.5798734948283979, + "learning_rate": 0.003, + "loss": 4.003, + "step": 11533 + }, + { + "epoch": 0.11534, + "grad_norm": 0.5815146567022272, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 11534 + }, + { + "epoch": 0.11535, + "grad_norm": 0.5631359319884381, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 11535 + }, + { + "epoch": 0.11536, + "grad_norm": 0.52718325195169, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 11536 + }, + { + "epoch": 0.11537, + "grad_norm": 0.5452730641349494, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 11537 + }, + { + "epoch": 0.11538, + "grad_norm": 0.6647736777324647, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 11538 + }, + { + "epoch": 0.11539, + "grad_norm": 0.7993453518784995, + "learning_rate": 0.003, + "loss": 4.001, + "step": 11539 + }, + { + "epoch": 0.1154, + "grad_norm": 0.8676583526008426, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 11540 + }, + { + "epoch": 0.11541, + "grad_norm": 0.8467762593631131, + "learning_rate": 0.003, + "loss": 4.043, + "step": 11541 + }, + { + "epoch": 0.11542, + "grad_norm": 0.8195695606971716, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 11542 + }, + { + "epoch": 0.11543, + "grad_norm": 0.7696212382432006, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 11543 + }, + { + "epoch": 0.11544, + "grad_norm": 0.7261240151106989, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 11544 + }, + { + "epoch": 0.11545, + "grad_norm": 0.8142156009129503, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 11545 + }, + { + "epoch": 0.11546, + "grad_norm": 0.8679829883444653, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 11546 + }, + { + "epoch": 0.11547, + "grad_norm": 0.9601855719917695, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 11547 + }, + { + "epoch": 0.11548, + "grad_norm": 1.0528976219935362, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 11548 + }, + { + "epoch": 0.11549, + "grad_norm": 1.195710076998539, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 11549 + }, + { + "epoch": 0.1155, + "grad_norm": 0.8256075275051304, + "learning_rate": 0.003, + "loss": 4.022, + "step": 11550 + }, + { + "epoch": 0.11551, + "grad_norm": 0.7847606526533497, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 11551 + }, + { + "epoch": 0.11552, + "grad_norm": 0.8725398112675128, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 11552 + }, + { + "epoch": 0.11553, + "grad_norm": 0.955185108666202, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 11553 + }, + { + "epoch": 0.11554, + "grad_norm": 1.0885180881430478, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 11554 + }, + { + "epoch": 0.11555, + "grad_norm": 0.9526378945255434, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 11555 + }, + { + "epoch": 0.11556, + "grad_norm": 0.916091153648662, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 11556 + }, + { + "epoch": 0.11557, + "grad_norm": 0.8740467465226177, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 11557 + }, + { + "epoch": 0.11558, + "grad_norm": 0.9305879891272867, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 11558 + }, + { + "epoch": 0.11559, + "grad_norm": 0.8789270784115336, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 11559 + }, + { + "epoch": 0.1156, + "grad_norm": 0.7723177340863612, + "learning_rate": 0.003, + "loss": 4.072, + "step": 11560 + }, + { + "epoch": 0.11561, + "grad_norm": 0.6797165449666542, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 11561 + }, + { + "epoch": 0.11562, + "grad_norm": 0.6258056132676906, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11562 + }, + { + "epoch": 0.11563, + "grad_norm": 0.6334046252594361, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 11563 + }, + { + "epoch": 0.11564, + "grad_norm": 0.733922379195102, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 11564 + }, + { + "epoch": 0.11565, + "grad_norm": 0.8441136812832606, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 11565 + }, + { + "epoch": 0.11566, + "grad_norm": 0.8958490157963556, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 11566 + }, + { + "epoch": 0.11567, + "grad_norm": 0.7154625841434611, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 11567 + }, + { + "epoch": 0.11568, + "grad_norm": 0.5591807257585931, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11568 + }, + { + "epoch": 0.11569, + "grad_norm": 0.5764103557923296, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 11569 + }, + { + "epoch": 0.1157, + "grad_norm": 0.6175755684693234, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 11570 + }, + { + "epoch": 0.11571, + "grad_norm": 0.5832623479145422, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 11571 + }, + { + "epoch": 0.11572, + "grad_norm": 0.5528600827931691, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 11572 + }, + { + "epoch": 0.11573, + "grad_norm": 0.5411069957506757, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 11573 + }, + { + "epoch": 0.11574, + "grad_norm": 0.5380592073893413, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 11574 + }, + { + "epoch": 0.11575, + "grad_norm": 0.44887611887503975, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 11575 + }, + { + "epoch": 0.11576, + "grad_norm": 0.4477901261825951, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 11576 + }, + { + "epoch": 0.11577, + "grad_norm": 0.4797258945321979, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 11577 + }, + { + "epoch": 0.11578, + "grad_norm": 0.5022017668768843, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 11578 + }, + { + "epoch": 0.11579, + "grad_norm": 0.5307638885867693, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 11579 + }, + { + "epoch": 0.1158, + "grad_norm": 0.5138004795233932, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 11580 + }, + { + "epoch": 0.11581, + "grad_norm": 0.5084972919214781, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 11581 + }, + { + "epoch": 0.11582, + "grad_norm": 0.5220176103851334, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 11582 + }, + { + "epoch": 0.11583, + "grad_norm": 0.5887665195546566, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 11583 + }, + { + "epoch": 0.11584, + "grad_norm": 0.6351467685767302, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 11584 + }, + { + "epoch": 0.11585, + "grad_norm": 0.7699218933028461, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 11585 + }, + { + "epoch": 0.11586, + "grad_norm": 1.0991051478244098, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 11586 + }, + { + "epoch": 0.11587, + "grad_norm": 1.2406961115235668, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 11587 + }, + { + "epoch": 0.11588, + "grad_norm": 0.5809998580034214, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 11588 + }, + { + "epoch": 0.11589, + "grad_norm": 0.6621924063369177, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 11589 + }, + { + "epoch": 0.1159, + "grad_norm": 0.823648448705278, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 11590 + }, + { + "epoch": 0.11591, + "grad_norm": 0.8290072195867793, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 11591 + }, + { + "epoch": 0.11592, + "grad_norm": 0.7365830656656727, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 11592 + }, + { + "epoch": 0.11593, + "grad_norm": 0.7968680444272531, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 11593 + }, + { + "epoch": 0.11594, + "grad_norm": 0.7916698297875813, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 11594 + }, + { + "epoch": 0.11595, + "grad_norm": 0.7357734272056956, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 11595 + }, + { + "epoch": 0.11596, + "grad_norm": 0.7015233841856916, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 11596 + }, + { + "epoch": 0.11597, + "grad_norm": 0.5763554274905541, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 11597 + }, + { + "epoch": 0.11598, + "grad_norm": 0.5802965045457192, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 11598 + }, + { + "epoch": 0.11599, + "grad_norm": 0.5483457858777204, + "learning_rate": 0.003, + "loss": 4.009, + "step": 11599 + }, + { + "epoch": 0.116, + "grad_norm": 0.533286793487087, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 11600 + }, + { + "epoch": 0.11601, + "grad_norm": 0.5235915831003445, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 11601 + }, + { + "epoch": 0.11602, + "grad_norm": 0.6005986636452881, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 11602 + }, + { + "epoch": 0.11603, + "grad_norm": 0.6707084162378779, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 11603 + }, + { + "epoch": 0.11604, + "grad_norm": 0.6101669844958316, + "learning_rate": 0.003, + "loss": 4.009, + "step": 11604 + }, + { + "epoch": 0.11605, + "grad_norm": 0.5561475201882816, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 11605 + }, + { + "epoch": 0.11606, + "grad_norm": 0.5935486043372863, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 11606 + }, + { + "epoch": 0.11607, + "grad_norm": 0.6212691295664468, + "learning_rate": 0.003, + "loss": 4.014, + "step": 11607 + }, + { + "epoch": 0.11608, + "grad_norm": 0.6639877852861372, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 11608 + }, + { + "epoch": 0.11609, + "grad_norm": 0.7595348264164212, + "learning_rate": 0.003, + "loss": 4.0556, + "step": 11609 + }, + { + "epoch": 0.1161, + "grad_norm": 0.947913630421829, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 11610 + }, + { + "epoch": 0.11611, + "grad_norm": 1.015717554764375, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 11611 + }, + { + "epoch": 0.11612, + "grad_norm": 0.9762262895587774, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 11612 + }, + { + "epoch": 0.11613, + "grad_norm": 1.0038277815340253, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11613 + }, + { + "epoch": 0.11614, + "grad_norm": 0.8738197350661631, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 11614 + }, + { + "epoch": 0.11615, + "grad_norm": 0.8811821342462653, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 11615 + }, + { + "epoch": 0.11616, + "grad_norm": 0.8593785751218805, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 11616 + }, + { + "epoch": 0.11617, + "grad_norm": 0.866785981657161, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 11617 + }, + { + "epoch": 0.11618, + "grad_norm": 0.8279049316131087, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 11618 + }, + { + "epoch": 0.11619, + "grad_norm": 0.8469728676052809, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 11619 + }, + { + "epoch": 0.1162, + "grad_norm": 0.7542621476680093, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 11620 + }, + { + "epoch": 0.11621, + "grad_norm": 0.6803925608685063, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 11621 + }, + { + "epoch": 0.11622, + "grad_norm": 0.6171740495794559, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 11622 + }, + { + "epoch": 0.11623, + "grad_norm": 0.5554015111527739, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 11623 + }, + { + "epoch": 0.11624, + "grad_norm": 0.5992988120907846, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 11624 + }, + { + "epoch": 0.11625, + "grad_norm": 0.6211098152701376, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 11625 + }, + { + "epoch": 0.11626, + "grad_norm": 0.6385121054761209, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 11626 + }, + { + "epoch": 0.11627, + "grad_norm": 0.6736493506001526, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 11627 + }, + { + "epoch": 0.11628, + "grad_norm": 0.6708655815563072, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 11628 + }, + { + "epoch": 0.11629, + "grad_norm": 0.6384498821133718, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11629 + }, + { + "epoch": 0.1163, + "grad_norm": 0.6237857110742233, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 11630 + }, + { + "epoch": 0.11631, + "grad_norm": 0.6532661188957888, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 11631 + }, + { + "epoch": 0.11632, + "grad_norm": 0.6591899019675099, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 11632 + }, + { + "epoch": 0.11633, + "grad_norm": 0.7852671822319117, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 11633 + }, + { + "epoch": 0.11634, + "grad_norm": 0.9478206187455215, + "learning_rate": 0.003, + "loss": 4.045, + "step": 11634 + }, + { + "epoch": 0.11635, + "grad_norm": 0.9990046669205872, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 11635 + }, + { + "epoch": 0.11636, + "grad_norm": 0.9580112747574446, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 11636 + }, + { + "epoch": 0.11637, + "grad_norm": 1.020686509053307, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11637 + }, + { + "epoch": 0.11638, + "grad_norm": 0.8753297208730062, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11638 + }, + { + "epoch": 0.11639, + "grad_norm": 0.7547843802739779, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 11639 + }, + { + "epoch": 0.1164, + "grad_norm": 0.7821481993675838, + "learning_rate": 0.003, + "loss": 4.061, + "step": 11640 + }, + { + "epoch": 0.11641, + "grad_norm": 0.9576831209205888, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 11641 + }, + { + "epoch": 0.11642, + "grad_norm": 0.846572973140994, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 11642 + }, + { + "epoch": 0.11643, + "grad_norm": 0.7794393903451743, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 11643 + }, + { + "epoch": 0.11644, + "grad_norm": 0.6985685066701429, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 11644 + }, + { + "epoch": 0.11645, + "grad_norm": 0.7263475630254262, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 11645 + }, + { + "epoch": 0.11646, + "grad_norm": 0.6808678068985357, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 11646 + }, + { + "epoch": 0.11647, + "grad_norm": 0.7271283917994703, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 11647 + }, + { + "epoch": 0.11648, + "grad_norm": 0.6804793633810743, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 11648 + }, + { + "epoch": 0.11649, + "grad_norm": 0.5355601545725887, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 11649 + }, + { + "epoch": 0.1165, + "grad_norm": 0.5142680176109897, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 11650 + }, + { + "epoch": 0.11651, + "grad_norm": 0.5033845710516135, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 11651 + }, + { + "epoch": 0.11652, + "grad_norm": 0.5437069191320955, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 11652 + }, + { + "epoch": 0.11653, + "grad_norm": 0.5897444520613784, + "learning_rate": 0.003, + "loss": 4.027, + "step": 11653 + }, + { + "epoch": 0.11654, + "grad_norm": 0.636692130787932, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 11654 + }, + { + "epoch": 0.11655, + "grad_norm": 0.6920710935641994, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11655 + }, + { + "epoch": 0.11656, + "grad_norm": 0.6499076120617068, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 11656 + }, + { + "epoch": 0.11657, + "grad_norm": 0.7827839705512943, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 11657 + }, + { + "epoch": 0.11658, + "grad_norm": 0.8830756945199015, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 11658 + }, + { + "epoch": 0.11659, + "grad_norm": 0.8539286683180792, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 11659 + }, + { + "epoch": 0.1166, + "grad_norm": 0.6096452273575732, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 11660 + }, + { + "epoch": 0.11661, + "grad_norm": 0.6236359050003065, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 11661 + }, + { + "epoch": 0.11662, + "grad_norm": 0.6907507351957158, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 11662 + }, + { + "epoch": 0.11663, + "grad_norm": 0.7203829825634007, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 11663 + }, + { + "epoch": 0.11664, + "grad_norm": 0.6810514381294753, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 11664 + }, + { + "epoch": 0.11665, + "grad_norm": 0.5776773869211755, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 11665 + }, + { + "epoch": 0.11666, + "grad_norm": 0.5206570519439138, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 11666 + }, + { + "epoch": 0.11667, + "grad_norm": 0.5458692605683314, + "learning_rate": 0.003, + "loss": 4.047, + "step": 11667 + }, + { + "epoch": 0.11668, + "grad_norm": 0.585260249637911, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 11668 + }, + { + "epoch": 0.11669, + "grad_norm": 0.5386683649170171, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11669 + }, + { + "epoch": 0.1167, + "grad_norm": 0.546691236978891, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 11670 + }, + { + "epoch": 0.11671, + "grad_norm": 0.6639566994460306, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 11671 + }, + { + "epoch": 0.11672, + "grad_norm": 0.7297525541839405, + "learning_rate": 0.003, + "loss": 4.045, + "step": 11672 + }, + { + "epoch": 0.11673, + "grad_norm": 0.8134692951310698, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 11673 + }, + { + "epoch": 0.11674, + "grad_norm": 0.8546870907072038, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 11674 + }, + { + "epoch": 0.11675, + "grad_norm": 0.7328946575808639, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 11675 + }, + { + "epoch": 0.11676, + "grad_norm": 0.6106483388263626, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 11676 + }, + { + "epoch": 0.11677, + "grad_norm": 0.6027647525233404, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 11677 + }, + { + "epoch": 0.11678, + "grad_norm": 0.6603800573710713, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 11678 + }, + { + "epoch": 0.11679, + "grad_norm": 0.6306040384970805, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 11679 + }, + { + "epoch": 0.1168, + "grad_norm": 0.6022129753521962, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 11680 + }, + { + "epoch": 0.11681, + "grad_norm": 0.6449068434744858, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 11681 + }, + { + "epoch": 0.11682, + "grad_norm": 0.7461008280476099, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 11682 + }, + { + "epoch": 0.11683, + "grad_norm": 0.822885205740269, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 11683 + }, + { + "epoch": 0.11684, + "grad_norm": 0.7988874627184184, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 11684 + }, + { + "epoch": 0.11685, + "grad_norm": 0.8576166234804711, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 11685 + }, + { + "epoch": 0.11686, + "grad_norm": 0.7877268478523088, + "learning_rate": 0.003, + "loss": 4.035, + "step": 11686 + }, + { + "epoch": 0.11687, + "grad_norm": 0.825797210404978, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 11687 + }, + { + "epoch": 0.11688, + "grad_norm": 0.7905496231696543, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 11688 + }, + { + "epoch": 0.11689, + "grad_norm": 0.6736484267946543, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 11689 + }, + { + "epoch": 0.1169, + "grad_norm": 0.6986394424336312, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 11690 + }, + { + "epoch": 0.11691, + "grad_norm": 0.6992031296435066, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 11691 + }, + { + "epoch": 0.11692, + "grad_norm": 0.7158385515138562, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 11692 + }, + { + "epoch": 0.11693, + "grad_norm": 0.7794290866873266, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 11693 + }, + { + "epoch": 0.11694, + "grad_norm": 0.8057783684075511, + "learning_rate": 0.003, + "loss": 4.025, + "step": 11694 + }, + { + "epoch": 0.11695, + "grad_norm": 0.8369251244417676, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11695 + }, + { + "epoch": 0.11696, + "grad_norm": 0.885723541427596, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 11696 + }, + { + "epoch": 0.11697, + "grad_norm": 0.8570006045065341, + "learning_rate": 0.003, + "loss": 4.009, + "step": 11697 + }, + { + "epoch": 0.11698, + "grad_norm": 0.7914030174163976, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 11698 + }, + { + "epoch": 0.11699, + "grad_norm": 0.9027266540152856, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 11699 + }, + { + "epoch": 0.117, + "grad_norm": 0.79954517487066, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 11700 + }, + { + "epoch": 0.11701, + "grad_norm": 0.8302668814257321, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 11701 + }, + { + "epoch": 0.11702, + "grad_norm": 0.7677134592167869, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11702 + }, + { + "epoch": 0.11703, + "grad_norm": 0.6537282374453911, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 11703 + }, + { + "epoch": 0.11704, + "grad_norm": 0.6475747196490685, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 11704 + }, + { + "epoch": 0.11705, + "grad_norm": 0.6202130201338306, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 11705 + }, + { + "epoch": 0.11706, + "grad_norm": 0.6634283127971773, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 11706 + }, + { + "epoch": 0.11707, + "grad_norm": 0.698877647581644, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11707 + }, + { + "epoch": 0.11708, + "grad_norm": 0.6939191637381062, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 11708 + }, + { + "epoch": 0.11709, + "grad_norm": 0.6955985889558415, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 11709 + }, + { + "epoch": 0.1171, + "grad_norm": 0.7907468920992144, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11710 + }, + { + "epoch": 0.11711, + "grad_norm": 0.856955311537213, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 11711 + }, + { + "epoch": 0.11712, + "grad_norm": 0.8745538781634515, + "learning_rate": 0.003, + "loss": 4.0666, + "step": 11712 + }, + { + "epoch": 0.11713, + "grad_norm": 0.8134166726562375, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 11713 + }, + { + "epoch": 0.11714, + "grad_norm": 0.6584717894815153, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 11714 + }, + { + "epoch": 0.11715, + "grad_norm": 0.616344306177011, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 11715 + }, + { + "epoch": 0.11716, + "grad_norm": 0.6232358461387776, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 11716 + }, + { + "epoch": 0.11717, + "grad_norm": 0.608369353004611, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 11717 + }, + { + "epoch": 0.11718, + "grad_norm": 0.5773947784102242, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 11718 + }, + { + "epoch": 0.11719, + "grad_norm": 0.5415060101075149, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 11719 + }, + { + "epoch": 0.1172, + "grad_norm": 0.5282353807552288, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 11720 + }, + { + "epoch": 0.11721, + "grad_norm": 0.5787683735579456, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 11721 + }, + { + "epoch": 0.11722, + "grad_norm": 0.6366210752016446, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 11722 + }, + { + "epoch": 0.11723, + "grad_norm": 0.6807024120222568, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 11723 + }, + { + "epoch": 0.11724, + "grad_norm": 0.6387273720946737, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 11724 + }, + { + "epoch": 0.11725, + "grad_norm": 0.5888707294322582, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 11725 + }, + { + "epoch": 0.11726, + "grad_norm": 0.6164515540342893, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 11726 + }, + { + "epoch": 0.11727, + "grad_norm": 0.6823359808308882, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 11727 + }, + { + "epoch": 0.11728, + "grad_norm": 0.7647589809369874, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 11728 + }, + { + "epoch": 0.11729, + "grad_norm": 0.9385562386961991, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 11729 + }, + { + "epoch": 0.1173, + "grad_norm": 1.0726277540970819, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 11730 + }, + { + "epoch": 0.11731, + "grad_norm": 0.9301080660484357, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 11731 + }, + { + "epoch": 0.11732, + "grad_norm": 0.8833411507779316, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 11732 + }, + { + "epoch": 0.11733, + "grad_norm": 0.885569601250323, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 11733 + }, + { + "epoch": 0.11734, + "grad_norm": 0.8244992560658364, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 11734 + }, + { + "epoch": 0.11735, + "grad_norm": 0.7250380912024432, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 11735 + }, + { + "epoch": 0.11736, + "grad_norm": 0.739902325317191, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 11736 + }, + { + "epoch": 0.11737, + "grad_norm": 0.7328352607256183, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 11737 + }, + { + "epoch": 0.11738, + "grad_norm": 0.7053761126374648, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 11738 + }, + { + "epoch": 0.11739, + "grad_norm": 0.666493443280725, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 11739 + }, + { + "epoch": 0.1174, + "grad_norm": 0.7441229362276008, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 11740 + }, + { + "epoch": 0.11741, + "grad_norm": 0.7638242979681878, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11741 + }, + { + "epoch": 0.11742, + "grad_norm": 0.6756814431906355, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 11742 + }, + { + "epoch": 0.11743, + "grad_norm": 0.6733638982692363, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 11743 + }, + { + "epoch": 0.11744, + "grad_norm": 0.7060651227276081, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 11744 + }, + { + "epoch": 0.11745, + "grad_norm": 0.801288958276668, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 11745 + }, + { + "epoch": 0.11746, + "grad_norm": 0.8060635069363732, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 11746 + }, + { + "epoch": 0.11747, + "grad_norm": 0.8336119193872935, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 11747 + }, + { + "epoch": 0.11748, + "grad_norm": 0.7304250075247144, + "learning_rate": 0.003, + "loss": 4.011, + "step": 11748 + }, + { + "epoch": 0.11749, + "grad_norm": 0.7108677637856038, + "learning_rate": 0.003, + "loss": 4.022, + "step": 11749 + }, + { + "epoch": 0.1175, + "grad_norm": 0.6297903410227579, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 11750 + }, + { + "epoch": 0.11751, + "grad_norm": 0.6206239053520888, + "learning_rate": 0.003, + "loss": 4.004, + "step": 11751 + }, + { + "epoch": 0.11752, + "grad_norm": 0.5768408085831103, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 11752 + }, + { + "epoch": 0.11753, + "grad_norm": 0.5741656040678591, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 11753 + }, + { + "epoch": 0.11754, + "grad_norm": 0.537905035223838, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 11754 + }, + { + "epoch": 0.11755, + "grad_norm": 0.5639391687013475, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 11755 + }, + { + "epoch": 0.11756, + "grad_norm": 0.6290472854882676, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 11756 + }, + { + "epoch": 0.11757, + "grad_norm": 0.669499940230968, + "learning_rate": 0.003, + "loss": 4.015, + "step": 11757 + }, + { + "epoch": 0.11758, + "grad_norm": 0.6980873827267486, + "learning_rate": 0.003, + "loss": 4.047, + "step": 11758 + }, + { + "epoch": 0.11759, + "grad_norm": 0.630708707619196, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 11759 + }, + { + "epoch": 0.1176, + "grad_norm": 0.6114802099260452, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 11760 + }, + { + "epoch": 0.11761, + "grad_norm": 0.6664145620144772, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 11761 + }, + { + "epoch": 0.11762, + "grad_norm": 0.7848994905943532, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 11762 + }, + { + "epoch": 0.11763, + "grad_norm": 0.8310345884812363, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 11763 + }, + { + "epoch": 0.11764, + "grad_norm": 0.8080395778743169, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 11764 + }, + { + "epoch": 0.11765, + "grad_norm": 0.7523601788910337, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 11765 + }, + { + "epoch": 0.11766, + "grad_norm": 0.7046952485235659, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 11766 + }, + { + "epoch": 0.11767, + "grad_norm": 0.7166146088329161, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11767 + }, + { + "epoch": 0.11768, + "grad_norm": 0.7461397758800733, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 11768 + }, + { + "epoch": 0.11769, + "grad_norm": 0.8135043659326636, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 11769 + }, + { + "epoch": 0.1177, + "grad_norm": 0.8822374319064122, + "learning_rate": 0.003, + "loss": 4.029, + "step": 11770 + }, + { + "epoch": 0.11771, + "grad_norm": 0.7887451337416144, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 11771 + }, + { + "epoch": 0.11772, + "grad_norm": 0.7463084197712188, + "learning_rate": 0.003, + "loss": 4.031, + "step": 11772 + }, + { + "epoch": 0.11773, + "grad_norm": 0.7275362219546454, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 11773 + }, + { + "epoch": 0.11774, + "grad_norm": 0.6494477937678065, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 11774 + }, + { + "epoch": 0.11775, + "grad_norm": 0.7664187847567804, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 11775 + }, + { + "epoch": 0.11776, + "grad_norm": 0.9761527211831801, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 11776 + }, + { + "epoch": 0.11777, + "grad_norm": 0.9255870012896777, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 11777 + }, + { + "epoch": 0.11778, + "grad_norm": 0.8025372585843031, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 11778 + }, + { + "epoch": 0.11779, + "grad_norm": 0.7022603790461175, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 11779 + }, + { + "epoch": 0.1178, + "grad_norm": 0.6977940941203047, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 11780 + }, + { + "epoch": 0.11781, + "grad_norm": 0.8054707149432545, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 11781 + }, + { + "epoch": 0.11782, + "grad_norm": 0.748109951367685, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 11782 + }, + { + "epoch": 0.11783, + "grad_norm": 0.6051933416046825, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 11783 + }, + { + "epoch": 0.11784, + "grad_norm": 0.5864780493295264, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 11784 + }, + { + "epoch": 0.11785, + "grad_norm": 0.5659783871191792, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 11785 + }, + { + "epoch": 0.11786, + "grad_norm": 0.5862185447777339, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 11786 + }, + { + "epoch": 0.11787, + "grad_norm": 0.4889303000538718, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 11787 + }, + { + "epoch": 0.11788, + "grad_norm": 0.5045361328285877, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 11788 + }, + { + "epoch": 0.11789, + "grad_norm": 0.5512770357957741, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 11789 + }, + { + "epoch": 0.1179, + "grad_norm": 0.5801790184741503, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 11790 + }, + { + "epoch": 0.11791, + "grad_norm": 0.6523428842357628, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 11791 + }, + { + "epoch": 0.11792, + "grad_norm": 0.7379765739045967, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 11792 + }, + { + "epoch": 0.11793, + "grad_norm": 0.7116558252932835, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 11793 + }, + { + "epoch": 0.11794, + "grad_norm": 0.660367079391319, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 11794 + }, + { + "epoch": 0.11795, + "grad_norm": 0.6473288835613878, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 11795 + }, + { + "epoch": 0.11796, + "grad_norm": 0.7974453879003772, + "learning_rate": 0.003, + "loss": 4.04, + "step": 11796 + }, + { + "epoch": 0.11797, + "grad_norm": 0.9511544456566466, + "learning_rate": 0.003, + "loss": 4.009, + "step": 11797 + }, + { + "epoch": 0.11798, + "grad_norm": 0.8421609281251441, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 11798 + }, + { + "epoch": 0.11799, + "grad_norm": 0.6399357629906068, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 11799 + }, + { + "epoch": 0.118, + "grad_norm": 0.6960660761298735, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 11800 + }, + { + "epoch": 0.11801, + "grad_norm": 0.6718841282215802, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 11801 + }, + { + "epoch": 0.11802, + "grad_norm": 0.7046618837651971, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11802 + }, + { + "epoch": 0.11803, + "grad_norm": 0.7442953711895961, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 11803 + }, + { + "epoch": 0.11804, + "grad_norm": 0.8207168197916488, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 11804 + }, + { + "epoch": 0.11805, + "grad_norm": 0.8535467031633971, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 11805 + }, + { + "epoch": 0.11806, + "grad_norm": 0.8485605558565251, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 11806 + }, + { + "epoch": 0.11807, + "grad_norm": 0.8337824900656992, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11807 + }, + { + "epoch": 0.11808, + "grad_norm": 0.7652619922524897, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 11808 + }, + { + "epoch": 0.11809, + "grad_norm": 0.7452580361970524, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 11809 + }, + { + "epoch": 0.1181, + "grad_norm": 0.7298264085555294, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 11810 + }, + { + "epoch": 0.11811, + "grad_norm": 0.7043678681072829, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 11811 + }, + { + "epoch": 0.11812, + "grad_norm": 0.6118211878701959, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 11812 + }, + { + "epoch": 0.11813, + "grad_norm": 0.5336704458040081, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 11813 + }, + { + "epoch": 0.11814, + "grad_norm": 0.5045286445463982, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 11814 + }, + { + "epoch": 0.11815, + "grad_norm": 0.5647740153648169, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 11815 + }, + { + "epoch": 0.11816, + "grad_norm": 0.6020506588203145, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 11816 + }, + { + "epoch": 0.11817, + "grad_norm": 0.5870125821147595, + "learning_rate": 0.003, + "loss": 4.002, + "step": 11817 + }, + { + "epoch": 0.11818, + "grad_norm": 0.547287674931146, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 11818 + }, + { + "epoch": 0.11819, + "grad_norm": 0.6210789929958523, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 11819 + }, + { + "epoch": 0.1182, + "grad_norm": 0.7055314733616728, + "learning_rate": 0.003, + "loss": 4.03, + "step": 11820 + }, + { + "epoch": 0.11821, + "grad_norm": 0.9067591794390456, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 11821 + }, + { + "epoch": 0.11822, + "grad_norm": 1.0346711138096842, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 11822 + }, + { + "epoch": 0.11823, + "grad_norm": 1.0393776558671386, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 11823 + }, + { + "epoch": 0.11824, + "grad_norm": 0.9639588318613533, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 11824 + }, + { + "epoch": 0.11825, + "grad_norm": 0.8477491108835306, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 11825 + }, + { + "epoch": 0.11826, + "grad_norm": 0.8394086686847315, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 11826 + }, + { + "epoch": 0.11827, + "grad_norm": 0.8083417949938473, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 11827 + }, + { + "epoch": 0.11828, + "grad_norm": 0.7316162427349613, + "learning_rate": 0.003, + "loss": 4.047, + "step": 11828 + }, + { + "epoch": 0.11829, + "grad_norm": 0.6711926476439745, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 11829 + }, + { + "epoch": 0.1183, + "grad_norm": 0.6488541246190482, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 11830 + }, + { + "epoch": 0.11831, + "grad_norm": 0.6438655318252756, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 11831 + }, + { + "epoch": 0.11832, + "grad_norm": 0.7230857281330204, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 11832 + }, + { + "epoch": 0.11833, + "grad_norm": 0.6926974609227671, + "learning_rate": 0.003, + "loss": 4.0559, + "step": 11833 + }, + { + "epoch": 0.11834, + "grad_norm": 0.7206336438127308, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 11834 + }, + { + "epoch": 0.11835, + "grad_norm": 0.7727627987600394, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 11835 + }, + { + "epoch": 0.11836, + "grad_norm": 0.8331834828670103, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 11836 + }, + { + "epoch": 0.11837, + "grad_norm": 0.885481354748773, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 11837 + }, + { + "epoch": 0.11838, + "grad_norm": 0.7252874860887687, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 11838 + }, + { + "epoch": 0.11839, + "grad_norm": 0.6070262198682099, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 11839 + }, + { + "epoch": 0.1184, + "grad_norm": 0.696652388268785, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11840 + }, + { + "epoch": 0.11841, + "grad_norm": 0.717633876906244, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 11841 + }, + { + "epoch": 0.11842, + "grad_norm": 0.7020528825530948, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 11842 + }, + { + "epoch": 0.11843, + "grad_norm": 0.6879440911457572, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 11843 + }, + { + "epoch": 0.11844, + "grad_norm": 0.5881846043332865, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 11844 + }, + { + "epoch": 0.11845, + "grad_norm": 0.6043492826606103, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 11845 + }, + { + "epoch": 0.11846, + "grad_norm": 0.5665112430239451, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 11846 + }, + { + "epoch": 0.11847, + "grad_norm": 0.5141762256914572, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 11847 + }, + { + "epoch": 0.11848, + "grad_norm": 0.536058481628137, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 11848 + }, + { + "epoch": 0.11849, + "grad_norm": 0.5330083385594756, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 11849 + }, + { + "epoch": 0.1185, + "grad_norm": 0.5726652384459687, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 11850 + }, + { + "epoch": 0.11851, + "grad_norm": 0.5618146701301187, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 11851 + }, + { + "epoch": 0.11852, + "grad_norm": 0.6118931317442301, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 11852 + }, + { + "epoch": 0.11853, + "grad_norm": 0.6886954984456732, + "learning_rate": 0.003, + "loss": 4.021, + "step": 11853 + }, + { + "epoch": 0.11854, + "grad_norm": 0.8196650765902581, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 11854 + }, + { + "epoch": 0.11855, + "grad_norm": 0.9348272042037867, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 11855 + }, + { + "epoch": 0.11856, + "grad_norm": 0.8371775841850643, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 11856 + }, + { + "epoch": 0.11857, + "grad_norm": 0.7469797663651593, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 11857 + }, + { + "epoch": 0.11858, + "grad_norm": 0.8277149673767332, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 11858 + }, + { + "epoch": 0.11859, + "grad_norm": 0.8636560126061351, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 11859 + }, + { + "epoch": 0.1186, + "grad_norm": 0.8377445708612422, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 11860 + }, + { + "epoch": 0.11861, + "grad_norm": 0.9441061052970291, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 11861 + }, + { + "epoch": 0.11862, + "grad_norm": 1.0355588746697295, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 11862 + }, + { + "epoch": 0.11863, + "grad_norm": 0.9376848440352863, + "learning_rate": 0.003, + "loss": 4.051, + "step": 11863 + }, + { + "epoch": 0.11864, + "grad_norm": 0.8428012472437676, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 11864 + }, + { + "epoch": 0.11865, + "grad_norm": 0.869444614143773, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 11865 + }, + { + "epoch": 0.11866, + "grad_norm": 0.7844348049457556, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 11866 + }, + { + "epoch": 0.11867, + "grad_norm": 0.78757878806857, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 11867 + }, + { + "epoch": 0.11868, + "grad_norm": 0.7449055458214403, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 11868 + }, + { + "epoch": 0.11869, + "grad_norm": 0.6435875451997652, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 11869 + }, + { + "epoch": 0.1187, + "grad_norm": 0.6360779848779996, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 11870 + }, + { + "epoch": 0.11871, + "grad_norm": 0.6235373977597173, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 11871 + }, + { + "epoch": 0.11872, + "grad_norm": 0.59440778449655, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 11872 + }, + { + "epoch": 0.11873, + "grad_norm": 0.55347464850433, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 11873 + }, + { + "epoch": 0.11874, + "grad_norm": 0.5208128935686388, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 11874 + }, + { + "epoch": 0.11875, + "grad_norm": 0.5576201101963276, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 11875 + }, + { + "epoch": 0.11876, + "grad_norm": 0.5567254238210742, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 11876 + }, + { + "epoch": 0.11877, + "grad_norm": 0.4935826560483898, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 11877 + }, + { + "epoch": 0.11878, + "grad_norm": 0.5370468034124781, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 11878 + }, + { + "epoch": 0.11879, + "grad_norm": 0.6058049881612229, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 11879 + }, + { + "epoch": 0.1188, + "grad_norm": 0.6576890698563241, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11880 + }, + { + "epoch": 0.11881, + "grad_norm": 0.7793116605473348, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 11881 + }, + { + "epoch": 0.11882, + "grad_norm": 0.7637409489913605, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 11882 + }, + { + "epoch": 0.11883, + "grad_norm": 0.6717363135799559, + "learning_rate": 0.003, + "loss": 4.037, + "step": 11883 + }, + { + "epoch": 0.11884, + "grad_norm": 0.7257539374150797, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 11884 + }, + { + "epoch": 0.11885, + "grad_norm": 0.7255615671088532, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 11885 + }, + { + "epoch": 0.11886, + "grad_norm": 0.8528143562033638, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 11886 + }, + { + "epoch": 0.11887, + "grad_norm": 0.8672394776532841, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 11887 + }, + { + "epoch": 0.11888, + "grad_norm": 0.9046749572806995, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 11888 + }, + { + "epoch": 0.11889, + "grad_norm": 0.8982729393644544, + "learning_rate": 0.003, + "loss": 4.041, + "step": 11889 + }, + { + "epoch": 0.1189, + "grad_norm": 0.8369073273680693, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 11890 + }, + { + "epoch": 0.11891, + "grad_norm": 0.8699208418141651, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 11891 + }, + { + "epoch": 0.11892, + "grad_norm": 0.9374468816162076, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 11892 + }, + { + "epoch": 0.11893, + "grad_norm": 0.7824495645363464, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 11893 + }, + { + "epoch": 0.11894, + "grad_norm": 0.6983555419565606, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 11894 + }, + { + "epoch": 0.11895, + "grad_norm": 0.6912376210246582, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 11895 + }, + { + "epoch": 0.11896, + "grad_norm": 0.7418829796103178, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 11896 + }, + { + "epoch": 0.11897, + "grad_norm": 0.7839085847446305, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 11897 + }, + { + "epoch": 0.11898, + "grad_norm": 0.7410655926141283, + "learning_rate": 0.003, + "loss": 4.0497, + "step": 11898 + }, + { + "epoch": 0.11899, + "grad_norm": 0.6168642961102667, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 11899 + }, + { + "epoch": 0.119, + "grad_norm": 0.7223777471766412, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 11900 + }, + { + "epoch": 0.11901, + "grad_norm": 0.7915923897129133, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 11901 + }, + { + "epoch": 0.11902, + "grad_norm": 0.8583831715650115, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 11902 + }, + { + "epoch": 0.11903, + "grad_norm": 0.728322327812673, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 11903 + }, + { + "epoch": 0.11904, + "grad_norm": 0.621875705110738, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 11904 + }, + { + "epoch": 0.11905, + "grad_norm": 0.6096248398319991, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 11905 + }, + { + "epoch": 0.11906, + "grad_norm": 0.6348875199347772, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 11906 + }, + { + "epoch": 0.11907, + "grad_norm": 0.6451711643742117, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11907 + }, + { + "epoch": 0.11908, + "grad_norm": 0.67452598732063, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 11908 + }, + { + "epoch": 0.11909, + "grad_norm": 0.7553514487234914, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 11909 + }, + { + "epoch": 0.1191, + "grad_norm": 0.783317307615553, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 11910 + }, + { + "epoch": 0.11911, + "grad_norm": 0.7563503997670367, + "learning_rate": 0.003, + "loss": 4.026, + "step": 11911 + }, + { + "epoch": 0.11912, + "grad_norm": 0.628165675963684, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 11912 + }, + { + "epoch": 0.11913, + "grad_norm": 0.5701911739204995, + "learning_rate": 0.003, + "loss": 4.041, + "step": 11913 + }, + { + "epoch": 0.11914, + "grad_norm": 0.616337887711915, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 11914 + }, + { + "epoch": 0.11915, + "grad_norm": 0.6099866964234723, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11915 + }, + { + "epoch": 0.11916, + "grad_norm": 0.6658540571315383, + "learning_rate": 0.003, + "loss": 4.041, + "step": 11916 + }, + { + "epoch": 0.11917, + "grad_norm": 0.6619686978387046, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 11917 + }, + { + "epoch": 0.11918, + "grad_norm": 0.7015628004920093, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 11918 + }, + { + "epoch": 0.11919, + "grad_norm": 0.8417227839933116, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 11919 + }, + { + "epoch": 0.1192, + "grad_norm": 0.8502861734933629, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 11920 + }, + { + "epoch": 0.11921, + "grad_norm": 0.8103319104023667, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 11921 + }, + { + "epoch": 0.11922, + "grad_norm": 0.7051649944311005, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 11922 + }, + { + "epoch": 0.11923, + "grad_norm": 0.6910390936286565, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 11923 + }, + { + "epoch": 0.11924, + "grad_norm": 0.681507722599937, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 11924 + }, + { + "epoch": 0.11925, + "grad_norm": 0.662282809193132, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 11925 + }, + { + "epoch": 0.11926, + "grad_norm": 0.6976691175483847, + "learning_rate": 0.003, + "loss": 4.02, + "step": 11926 + }, + { + "epoch": 0.11927, + "grad_norm": 0.6742080266022281, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 11927 + }, + { + "epoch": 0.11928, + "grad_norm": 0.6520605184573993, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 11928 + }, + { + "epoch": 0.11929, + "grad_norm": 0.6146954518974194, + "learning_rate": 0.003, + "loss": 4.016, + "step": 11929 + }, + { + "epoch": 0.1193, + "grad_norm": 0.600913125507246, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 11930 + }, + { + "epoch": 0.11931, + "grad_norm": 0.6268073142335412, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 11931 + }, + { + "epoch": 0.11932, + "grad_norm": 0.6662417893616144, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 11932 + }, + { + "epoch": 0.11933, + "grad_norm": 0.6188411626422793, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 11933 + }, + { + "epoch": 0.11934, + "grad_norm": 0.7232537832827948, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 11934 + }, + { + "epoch": 0.11935, + "grad_norm": 0.8368795067324477, + "learning_rate": 0.003, + "loss": 4.01, + "step": 11935 + }, + { + "epoch": 0.11936, + "grad_norm": 0.744029143826714, + "learning_rate": 0.003, + "loss": 4.018, + "step": 11936 + }, + { + "epoch": 0.11937, + "grad_norm": 0.7042427903886734, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11937 + }, + { + "epoch": 0.11938, + "grad_norm": 0.7151473941810412, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 11938 + }, + { + "epoch": 0.11939, + "grad_norm": 0.7132279402722266, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 11939 + }, + { + "epoch": 0.1194, + "grad_norm": 0.7625792331943054, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 11940 + }, + { + "epoch": 0.11941, + "grad_norm": 0.8871957853360128, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 11941 + }, + { + "epoch": 0.11942, + "grad_norm": 0.8555606463554024, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 11942 + }, + { + "epoch": 0.11943, + "grad_norm": 0.8140763349069451, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 11943 + }, + { + "epoch": 0.11944, + "grad_norm": 0.7445615919773416, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 11944 + }, + { + "epoch": 0.11945, + "grad_norm": 0.7868709980122751, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 11945 + }, + { + "epoch": 0.11946, + "grad_norm": 0.7080759789422179, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 11946 + }, + { + "epoch": 0.11947, + "grad_norm": 0.6045898573205123, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 11947 + }, + { + "epoch": 0.11948, + "grad_norm": 0.6372859395574728, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 11948 + }, + { + "epoch": 0.11949, + "grad_norm": 0.7778441804500063, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 11949 + }, + { + "epoch": 0.1195, + "grad_norm": 0.9111719607950312, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 11950 + }, + { + "epoch": 0.11951, + "grad_norm": 0.9485047402140333, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 11951 + }, + { + "epoch": 0.11952, + "grad_norm": 0.8798045866839522, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 11952 + }, + { + "epoch": 0.11953, + "grad_norm": 0.8576906557706743, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 11953 + }, + { + "epoch": 0.11954, + "grad_norm": 0.9596383507095367, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 11954 + }, + { + "epoch": 0.11955, + "grad_norm": 0.8736563620331866, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 11955 + }, + { + "epoch": 0.11956, + "grad_norm": 0.8317427163762077, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 11956 + }, + { + "epoch": 0.11957, + "grad_norm": 0.9062926977695371, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 11957 + }, + { + "epoch": 0.11958, + "grad_norm": 0.8885848132465142, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 11958 + }, + { + "epoch": 0.11959, + "grad_norm": 0.7309444112797675, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 11959 + }, + { + "epoch": 0.1196, + "grad_norm": 0.6648602312090239, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 11960 + }, + { + "epoch": 0.11961, + "grad_norm": 0.663962224371706, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 11961 + }, + { + "epoch": 0.11962, + "grad_norm": 0.6760926970190428, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 11962 + }, + { + "epoch": 0.11963, + "grad_norm": 0.6619953719380439, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 11963 + }, + { + "epoch": 0.11964, + "grad_norm": 0.7095188915298404, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 11964 + }, + { + "epoch": 0.11965, + "grad_norm": 0.6247672945016234, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 11965 + }, + { + "epoch": 0.11966, + "grad_norm": 0.6964192585663163, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 11966 + }, + { + "epoch": 0.11967, + "grad_norm": 0.7660277333798857, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 11967 + }, + { + "epoch": 0.11968, + "grad_norm": 0.7010842015719172, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 11968 + }, + { + "epoch": 0.11969, + "grad_norm": 0.7756545595226101, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 11969 + }, + { + "epoch": 0.1197, + "grad_norm": 0.7417350170132095, + "learning_rate": 0.003, + "loss": 4.033, + "step": 11970 + }, + { + "epoch": 0.11971, + "grad_norm": 0.8121504846756478, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 11971 + }, + { + "epoch": 0.11972, + "grad_norm": 0.8398312445112348, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 11972 + }, + { + "epoch": 0.11973, + "grad_norm": 0.7573760585475262, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 11973 + }, + { + "epoch": 0.11974, + "grad_norm": 0.7305960324154813, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 11974 + }, + { + "epoch": 0.11975, + "grad_norm": 0.6369004403505815, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 11975 + }, + { + "epoch": 0.11976, + "grad_norm": 0.6328124308454056, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 11976 + }, + { + "epoch": 0.11977, + "grad_norm": 0.563466158606599, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 11977 + }, + { + "epoch": 0.11978, + "grad_norm": 0.6396081380845896, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 11978 + }, + { + "epoch": 0.11979, + "grad_norm": 0.5933144963709056, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 11979 + }, + { + "epoch": 0.1198, + "grad_norm": 0.6224433822427069, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 11980 + }, + { + "epoch": 0.11981, + "grad_norm": 0.6258837119181038, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 11981 + }, + { + "epoch": 0.11982, + "grad_norm": 0.5707047532431749, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 11982 + }, + { + "epoch": 0.11983, + "grad_norm": 0.6659010815469281, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 11983 + }, + { + "epoch": 0.11984, + "grad_norm": 0.662754132460922, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 11984 + }, + { + "epoch": 0.11985, + "grad_norm": 0.5323791832832907, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 11985 + }, + { + "epoch": 0.11986, + "grad_norm": 0.5768517661488387, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 11986 + }, + { + "epoch": 0.11987, + "grad_norm": 0.5765965022463027, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 11987 + }, + { + "epoch": 0.11988, + "grad_norm": 0.6216967116763082, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 11988 + }, + { + "epoch": 0.11989, + "grad_norm": 0.7156971968709224, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 11989 + }, + { + "epoch": 0.1199, + "grad_norm": 0.9051983898737442, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 11990 + }, + { + "epoch": 0.11991, + "grad_norm": 1.0513944569721305, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 11991 + }, + { + "epoch": 0.11992, + "grad_norm": 0.9274946946780174, + "learning_rate": 0.003, + "loss": 4.0525, + "step": 11992 + }, + { + "epoch": 0.11993, + "grad_norm": 0.8427752174469074, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 11993 + }, + { + "epoch": 0.11994, + "grad_norm": 0.7703982309501328, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 11994 + }, + { + "epoch": 0.11995, + "grad_norm": 0.7561883495812756, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 11995 + }, + { + "epoch": 0.11996, + "grad_norm": 0.8494108825642828, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 11996 + }, + { + "epoch": 0.11997, + "grad_norm": 1.029048344311733, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 11997 + }, + { + "epoch": 0.11998, + "grad_norm": 1.03973821102631, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 11998 + }, + { + "epoch": 0.11999, + "grad_norm": 0.9406231931134269, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 11999 + }, + { + "epoch": 0.12, + "grad_norm": 0.811642392667539, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 12000 + }, + { + "epoch": 0.12001, + "grad_norm": 0.7403042659204061, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 12001 + }, + { + "epoch": 0.12002, + "grad_norm": 0.6506576216847788, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 12002 + }, + { + "epoch": 0.12003, + "grad_norm": 0.6736850824349279, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 12003 + }, + { + "epoch": 0.12004, + "grad_norm": 0.6688832819450603, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 12004 + }, + { + "epoch": 0.12005, + "grad_norm": 0.6906791067816925, + "learning_rate": 0.003, + "loss": 4.013, + "step": 12005 + }, + { + "epoch": 0.12006, + "grad_norm": 0.6754619472371577, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 12006 + }, + { + "epoch": 0.12007, + "grad_norm": 0.6118640618236973, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 12007 + }, + { + "epoch": 0.12008, + "grad_norm": 0.5689880303478466, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 12008 + }, + { + "epoch": 0.12009, + "grad_norm": 0.5291877139683215, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 12009 + }, + { + "epoch": 0.1201, + "grad_norm": 0.5138285526018626, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 12010 + }, + { + "epoch": 0.12011, + "grad_norm": 0.4862925310854174, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 12011 + }, + { + "epoch": 0.12012, + "grad_norm": 0.4993037003672302, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 12012 + }, + { + "epoch": 0.12013, + "grad_norm": 0.4640075708914935, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 12013 + }, + { + "epoch": 0.12014, + "grad_norm": 0.434783512355801, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 12014 + }, + { + "epoch": 0.12015, + "grad_norm": 0.4504462905761205, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 12015 + }, + { + "epoch": 0.12016, + "grad_norm": 0.499189332586657, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12016 + }, + { + "epoch": 0.12017, + "grad_norm": 0.5301704480230034, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 12017 + }, + { + "epoch": 0.12018, + "grad_norm": 0.604911368187421, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 12018 + }, + { + "epoch": 0.12019, + "grad_norm": 0.6075374074928546, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 12019 + }, + { + "epoch": 0.1202, + "grad_norm": 0.529216116217747, + "learning_rate": 0.003, + "loss": 4.021, + "step": 12020 + }, + { + "epoch": 0.12021, + "grad_norm": 0.5733756748875284, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 12021 + }, + { + "epoch": 0.12022, + "grad_norm": 0.7303591458755176, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 12022 + }, + { + "epoch": 0.12023, + "grad_norm": 0.8749873107364549, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 12023 + }, + { + "epoch": 0.12024, + "grad_norm": 1.0245342908032087, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 12024 + }, + { + "epoch": 0.12025, + "grad_norm": 1.1344881476739344, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 12025 + }, + { + "epoch": 0.12026, + "grad_norm": 0.8142104224122245, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 12026 + }, + { + "epoch": 0.12027, + "grad_norm": 0.6702033262394588, + "learning_rate": 0.003, + "loss": 4.015, + "step": 12027 + }, + { + "epoch": 0.12028, + "grad_norm": 0.7070221773101605, + "learning_rate": 0.003, + "loss": 4.019, + "step": 12028 + }, + { + "epoch": 0.12029, + "grad_norm": 0.7796586140915465, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 12029 + }, + { + "epoch": 0.1203, + "grad_norm": 0.8656303706565928, + "learning_rate": 0.003, + "loss": 4.02, + "step": 12030 + }, + { + "epoch": 0.12031, + "grad_norm": 0.804502599861992, + "learning_rate": 0.003, + "loss": 4.034, + "step": 12031 + }, + { + "epoch": 0.12032, + "grad_norm": 0.8054406246149151, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 12032 + }, + { + "epoch": 0.12033, + "grad_norm": 0.8034688329985341, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 12033 + }, + { + "epoch": 0.12034, + "grad_norm": 0.8323570047614531, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 12034 + }, + { + "epoch": 0.12035, + "grad_norm": 0.8217740040529699, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 12035 + }, + { + "epoch": 0.12036, + "grad_norm": 0.6547988515295423, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 12036 + }, + { + "epoch": 0.12037, + "grad_norm": 0.6642533159447495, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 12037 + }, + { + "epoch": 0.12038, + "grad_norm": 0.6771411180637877, + "learning_rate": 0.003, + "loss": 4.035, + "step": 12038 + }, + { + "epoch": 0.12039, + "grad_norm": 0.6839420240369137, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 12039 + }, + { + "epoch": 0.1204, + "grad_norm": 0.6447671284054175, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12040 + }, + { + "epoch": 0.12041, + "grad_norm": 0.6596540319702127, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 12041 + }, + { + "epoch": 0.12042, + "grad_norm": 0.7145938630363269, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 12042 + }, + { + "epoch": 0.12043, + "grad_norm": 0.7171071302960172, + "learning_rate": 0.003, + "loss": 4.044, + "step": 12043 + }, + { + "epoch": 0.12044, + "grad_norm": 0.6047128077702999, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 12044 + }, + { + "epoch": 0.12045, + "grad_norm": 0.6319457511456265, + "learning_rate": 0.003, + "loss": 4.028, + "step": 12045 + }, + { + "epoch": 0.12046, + "grad_norm": 0.7051353604156957, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 12046 + }, + { + "epoch": 0.12047, + "grad_norm": 0.6588014154977462, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 12047 + }, + { + "epoch": 0.12048, + "grad_norm": 0.6610988005462909, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 12048 + }, + { + "epoch": 0.12049, + "grad_norm": 0.7864265422012263, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 12049 + }, + { + "epoch": 0.1205, + "grad_norm": 1.001817334786125, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 12050 + }, + { + "epoch": 0.12051, + "grad_norm": 1.0886367080104264, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 12051 + }, + { + "epoch": 0.12052, + "grad_norm": 0.6702578060439546, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 12052 + }, + { + "epoch": 0.12053, + "grad_norm": 0.69933699655572, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 12053 + }, + { + "epoch": 0.12054, + "grad_norm": 0.6334165157299887, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 12054 + }, + { + "epoch": 0.12055, + "grad_norm": 0.643421775331982, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 12055 + }, + { + "epoch": 0.12056, + "grad_norm": 0.6559336962956491, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 12056 + }, + { + "epoch": 0.12057, + "grad_norm": 0.8303178000140117, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 12057 + }, + { + "epoch": 0.12058, + "grad_norm": 0.8454586951455099, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 12058 + }, + { + "epoch": 0.12059, + "grad_norm": 0.7459844727968716, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 12059 + }, + { + "epoch": 0.1206, + "grad_norm": 0.6514041405536811, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 12060 + }, + { + "epoch": 0.12061, + "grad_norm": 0.6434610782503174, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 12061 + }, + { + "epoch": 0.12062, + "grad_norm": 0.6726668834699381, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 12062 + }, + { + "epoch": 0.12063, + "grad_norm": 0.5938842748850162, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 12063 + }, + { + "epoch": 0.12064, + "grad_norm": 0.539585164076628, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 12064 + }, + { + "epoch": 0.12065, + "grad_norm": 0.6082514120354404, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 12065 + }, + { + "epoch": 0.12066, + "grad_norm": 0.6044661882066388, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 12066 + }, + { + "epoch": 0.12067, + "grad_norm": 0.5802873102304844, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 12067 + }, + { + "epoch": 0.12068, + "grad_norm": 0.5396209420260506, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 12068 + }, + { + "epoch": 0.12069, + "grad_norm": 0.5242130634268191, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 12069 + }, + { + "epoch": 0.1207, + "grad_norm": 0.5999263008398198, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 12070 + }, + { + "epoch": 0.12071, + "grad_norm": 0.7456233629291499, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 12071 + }, + { + "epoch": 0.12072, + "grad_norm": 0.7917890033622567, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 12072 + }, + { + "epoch": 0.12073, + "grad_norm": 0.688774412949923, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 12073 + }, + { + "epoch": 0.12074, + "grad_norm": 0.601640234113646, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 12074 + }, + { + "epoch": 0.12075, + "grad_norm": 0.7049158458593933, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 12075 + }, + { + "epoch": 0.12076, + "grad_norm": 0.816397764703132, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 12076 + }, + { + "epoch": 0.12077, + "grad_norm": 0.8900810594032367, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 12077 + }, + { + "epoch": 0.12078, + "grad_norm": 0.8309247460893535, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 12078 + }, + { + "epoch": 0.12079, + "grad_norm": 0.7131869542069469, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 12079 + }, + { + "epoch": 0.1208, + "grad_norm": 0.733454125028105, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 12080 + }, + { + "epoch": 0.12081, + "grad_norm": 0.7739759050886179, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 12081 + }, + { + "epoch": 0.12082, + "grad_norm": 0.7133295425617965, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 12082 + }, + { + "epoch": 0.12083, + "grad_norm": 0.5734420181338293, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 12083 + }, + { + "epoch": 0.12084, + "grad_norm": 0.7138568346434969, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 12084 + }, + { + "epoch": 0.12085, + "grad_norm": 0.8436817184820651, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 12085 + }, + { + "epoch": 0.12086, + "grad_norm": 0.8502506068928323, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 12086 + }, + { + "epoch": 0.12087, + "grad_norm": 0.8608523620829593, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 12087 + }, + { + "epoch": 0.12088, + "grad_norm": 0.7582202192225752, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12088 + }, + { + "epoch": 0.12089, + "grad_norm": 0.6604654485525896, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 12089 + }, + { + "epoch": 0.1209, + "grad_norm": 0.65676539125732, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 12090 + }, + { + "epoch": 0.12091, + "grad_norm": 0.6745913914806698, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12091 + }, + { + "epoch": 0.12092, + "grad_norm": 0.7701607505240651, + "learning_rate": 0.003, + "loss": 4.015, + "step": 12092 + }, + { + "epoch": 0.12093, + "grad_norm": 0.868059900177147, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 12093 + }, + { + "epoch": 0.12094, + "grad_norm": 0.9532755653366511, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 12094 + }, + { + "epoch": 0.12095, + "grad_norm": 0.9426144645038864, + "learning_rate": 0.003, + "loss": 4.045, + "step": 12095 + }, + { + "epoch": 0.12096, + "grad_norm": 0.9437023548975929, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 12096 + }, + { + "epoch": 0.12097, + "grad_norm": 0.9002928976426777, + "learning_rate": 0.003, + "loss": 4.0616, + "step": 12097 + }, + { + "epoch": 0.12098, + "grad_norm": 0.809549176165328, + "learning_rate": 0.003, + "loss": 4.016, + "step": 12098 + }, + { + "epoch": 0.12099, + "grad_norm": 0.7012160271025067, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 12099 + }, + { + "epoch": 0.121, + "grad_norm": 0.7761033410832523, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 12100 + }, + { + "epoch": 0.12101, + "grad_norm": 0.873680836259069, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 12101 + }, + { + "epoch": 0.12102, + "grad_norm": 0.9862505086707579, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 12102 + }, + { + "epoch": 0.12103, + "grad_norm": 0.8823925589073143, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 12103 + }, + { + "epoch": 0.12104, + "grad_norm": 0.7706987174406941, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 12104 + }, + { + "epoch": 0.12105, + "grad_norm": 0.7478599139155664, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12105 + }, + { + "epoch": 0.12106, + "grad_norm": 0.7681197631358119, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 12106 + }, + { + "epoch": 0.12107, + "grad_norm": 0.7387118820067682, + "learning_rate": 0.003, + "loss": 4.0575, + "step": 12107 + }, + { + "epoch": 0.12108, + "grad_norm": 0.6986454943934429, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 12108 + }, + { + "epoch": 0.12109, + "grad_norm": 0.6407428466712115, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 12109 + }, + { + "epoch": 0.1211, + "grad_norm": 0.6487705123357317, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 12110 + }, + { + "epoch": 0.12111, + "grad_norm": 0.6329320193755316, + "learning_rate": 0.003, + "loss": 4.023, + "step": 12111 + }, + { + "epoch": 0.12112, + "grad_norm": 0.689119306546642, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 12112 + }, + { + "epoch": 0.12113, + "grad_norm": 0.7406959924027086, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 12113 + }, + { + "epoch": 0.12114, + "grad_norm": 0.6478450760085939, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 12114 + }, + { + "epoch": 0.12115, + "grad_norm": 0.5970566788725558, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 12115 + }, + { + "epoch": 0.12116, + "grad_norm": 0.5338273742676821, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 12116 + }, + { + "epoch": 0.12117, + "grad_norm": 0.6169617392078441, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 12117 + }, + { + "epoch": 0.12118, + "grad_norm": 0.7410385769206627, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 12118 + }, + { + "epoch": 0.12119, + "grad_norm": 0.7322383368406364, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 12119 + }, + { + "epoch": 0.1212, + "grad_norm": 0.6816586359519928, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 12120 + }, + { + "epoch": 0.12121, + "grad_norm": 0.7998166031838393, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 12121 + }, + { + "epoch": 0.12122, + "grad_norm": 0.7583564503765432, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 12122 + }, + { + "epoch": 0.12123, + "grad_norm": 0.633479477753893, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 12123 + }, + { + "epoch": 0.12124, + "grad_norm": 0.6712173929970848, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12124 + }, + { + "epoch": 0.12125, + "grad_norm": 0.6811471416408752, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 12125 + }, + { + "epoch": 0.12126, + "grad_norm": 0.6691132883690597, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 12126 + }, + { + "epoch": 0.12127, + "grad_norm": 0.6232572542157325, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12127 + }, + { + "epoch": 0.12128, + "grad_norm": 0.6977222766575649, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 12128 + }, + { + "epoch": 0.12129, + "grad_norm": 0.7766809437956769, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 12129 + }, + { + "epoch": 0.1213, + "grad_norm": 0.6596323127923106, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 12130 + }, + { + "epoch": 0.12131, + "grad_norm": 0.6178169170243882, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 12131 + }, + { + "epoch": 0.12132, + "grad_norm": 0.6234459353289795, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 12132 + }, + { + "epoch": 0.12133, + "grad_norm": 0.6882748669502268, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 12133 + }, + { + "epoch": 0.12134, + "grad_norm": 0.6994686741316841, + "learning_rate": 0.003, + "loss": 4.037, + "step": 12134 + }, + { + "epoch": 0.12135, + "grad_norm": 0.6957745265744402, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 12135 + }, + { + "epoch": 0.12136, + "grad_norm": 0.8004246808651364, + "learning_rate": 0.003, + "loss": 4.019, + "step": 12136 + }, + { + "epoch": 0.12137, + "grad_norm": 0.8036410825845225, + "learning_rate": 0.003, + "loss": 4.057, + "step": 12137 + }, + { + "epoch": 0.12138, + "grad_norm": 0.6829140657327091, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 12138 + }, + { + "epoch": 0.12139, + "grad_norm": 0.5705191216610135, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 12139 + }, + { + "epoch": 0.1214, + "grad_norm": 0.642137471283096, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 12140 + }, + { + "epoch": 0.12141, + "grad_norm": 0.677272598089275, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 12141 + }, + { + "epoch": 0.12142, + "grad_norm": 0.6576234495647907, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 12142 + }, + { + "epoch": 0.12143, + "grad_norm": 0.7110350635412036, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 12143 + }, + { + "epoch": 0.12144, + "grad_norm": 0.6507603646978299, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 12144 + }, + { + "epoch": 0.12145, + "grad_norm": 0.6387495895928103, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 12145 + }, + { + "epoch": 0.12146, + "grad_norm": 0.6048234016671572, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 12146 + }, + { + "epoch": 0.12147, + "grad_norm": 0.6365875974864904, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 12147 + }, + { + "epoch": 0.12148, + "grad_norm": 0.6870730750737807, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 12148 + }, + { + "epoch": 0.12149, + "grad_norm": 0.6910250281694469, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 12149 + }, + { + "epoch": 0.1215, + "grad_norm": 0.6867432661511554, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 12150 + }, + { + "epoch": 0.12151, + "grad_norm": 0.6529481104077838, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 12151 + }, + { + "epoch": 0.12152, + "grad_norm": 0.7366371839271125, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 12152 + }, + { + "epoch": 0.12153, + "grad_norm": 0.7872711631528664, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 12153 + }, + { + "epoch": 0.12154, + "grad_norm": 0.8406882389463518, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 12154 + }, + { + "epoch": 0.12155, + "grad_norm": 0.8158892452522399, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 12155 + }, + { + "epoch": 0.12156, + "grad_norm": 0.6926091164050826, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 12156 + }, + { + "epoch": 0.12157, + "grad_norm": 0.566554681063402, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 12157 + }, + { + "epoch": 0.12158, + "grad_norm": 0.5993638373740453, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 12158 + }, + { + "epoch": 0.12159, + "grad_norm": 0.6519132626758449, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 12159 + }, + { + "epoch": 0.1216, + "grad_norm": 0.6878868602708305, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 12160 + }, + { + "epoch": 0.12161, + "grad_norm": 0.6759060524779874, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 12161 + }, + { + "epoch": 0.12162, + "grad_norm": 0.6615904814845578, + "learning_rate": 0.003, + "loss": 4.016, + "step": 12162 + }, + { + "epoch": 0.12163, + "grad_norm": 0.7091514066793031, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 12163 + }, + { + "epoch": 0.12164, + "grad_norm": 0.8097526441599476, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 12164 + }, + { + "epoch": 0.12165, + "grad_norm": 0.7287328533750221, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 12165 + }, + { + "epoch": 0.12166, + "grad_norm": 0.8433992085254097, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 12166 + }, + { + "epoch": 0.12167, + "grad_norm": 1.0142976820548564, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 12167 + }, + { + "epoch": 0.12168, + "grad_norm": 0.969186034017279, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12168 + }, + { + "epoch": 0.12169, + "grad_norm": 1.0254999277696972, + "learning_rate": 0.003, + "loss": 4.0568, + "step": 12169 + }, + { + "epoch": 0.1217, + "grad_norm": 1.107765505239678, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 12170 + }, + { + "epoch": 0.12171, + "grad_norm": 0.9390494013996052, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 12171 + }, + { + "epoch": 0.12172, + "grad_norm": 0.9523151809469707, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 12172 + }, + { + "epoch": 0.12173, + "grad_norm": 1.0029873496229191, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 12173 + }, + { + "epoch": 0.12174, + "grad_norm": 1.096941652030204, + "learning_rate": 0.003, + "loss": 4.0737, + "step": 12174 + }, + { + "epoch": 0.12175, + "grad_norm": 0.9422716966945642, + "learning_rate": 0.003, + "loss": 4.0643, + "step": 12175 + }, + { + "epoch": 0.12176, + "grad_norm": 1.0404017733433406, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 12176 + }, + { + "epoch": 0.12177, + "grad_norm": 1.0005392346678363, + "learning_rate": 0.003, + "loss": 4.0544, + "step": 12177 + }, + { + "epoch": 0.12178, + "grad_norm": 0.9553213430165856, + "learning_rate": 0.003, + "loss": 4.0622, + "step": 12178 + }, + { + "epoch": 0.12179, + "grad_norm": 0.9599814879335268, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 12179 + }, + { + "epoch": 0.1218, + "grad_norm": 0.9240612212459492, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 12180 + }, + { + "epoch": 0.12181, + "grad_norm": 0.8465767440413747, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 12181 + }, + { + "epoch": 0.12182, + "grad_norm": 0.7755530759731928, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 12182 + }, + { + "epoch": 0.12183, + "grad_norm": 0.7666450566458284, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 12183 + }, + { + "epoch": 0.12184, + "grad_norm": 0.6313657774627187, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 12184 + }, + { + "epoch": 0.12185, + "grad_norm": 0.5580902775223618, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 12185 + }, + { + "epoch": 0.12186, + "grad_norm": 0.5822916163750812, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 12186 + }, + { + "epoch": 0.12187, + "grad_norm": 0.5236089138138592, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 12187 + }, + { + "epoch": 0.12188, + "grad_norm": 0.5184396857325714, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 12188 + }, + { + "epoch": 0.12189, + "grad_norm": 0.581734089941018, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 12189 + }, + { + "epoch": 0.1219, + "grad_norm": 0.5763614394674101, + "learning_rate": 0.003, + "loss": 4.033, + "step": 12190 + }, + { + "epoch": 0.12191, + "grad_norm": 0.5376636029170956, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 12191 + }, + { + "epoch": 0.12192, + "grad_norm": 0.5614433922373305, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 12192 + }, + { + "epoch": 0.12193, + "grad_norm": 0.685616549793136, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 12193 + }, + { + "epoch": 0.12194, + "grad_norm": 0.7826533595337244, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 12194 + }, + { + "epoch": 0.12195, + "grad_norm": 0.7122642626947873, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 12195 + }, + { + "epoch": 0.12196, + "grad_norm": 0.5981954582483818, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 12196 + }, + { + "epoch": 0.12197, + "grad_norm": 0.5093086780398463, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 12197 + }, + { + "epoch": 0.12198, + "grad_norm": 0.5608684125209227, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 12198 + }, + { + "epoch": 0.12199, + "grad_norm": 0.5522989962910022, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 12199 + }, + { + "epoch": 0.122, + "grad_norm": 0.4733454621839758, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 12200 + }, + { + "epoch": 0.12201, + "grad_norm": 0.5354688254698354, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 12201 + }, + { + "epoch": 0.12202, + "grad_norm": 0.617347645609809, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 12202 + }, + { + "epoch": 0.12203, + "grad_norm": 0.5709161725562115, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 12203 + }, + { + "epoch": 0.12204, + "grad_norm": 0.5992212374312984, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 12204 + }, + { + "epoch": 0.12205, + "grad_norm": 0.7332851710616025, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 12205 + }, + { + "epoch": 0.12206, + "grad_norm": 0.8380169886222433, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 12206 + }, + { + "epoch": 0.12207, + "grad_norm": 0.9117680146663315, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 12207 + }, + { + "epoch": 0.12208, + "grad_norm": 0.8072370238314731, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 12208 + }, + { + "epoch": 0.12209, + "grad_norm": 0.6601530633732182, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12209 + }, + { + "epoch": 0.1221, + "grad_norm": 0.7828479735983167, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 12210 + }, + { + "epoch": 0.12211, + "grad_norm": 0.8170766876900553, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 12211 + }, + { + "epoch": 0.12212, + "grad_norm": 0.6940794541390121, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 12212 + }, + { + "epoch": 0.12213, + "grad_norm": 0.5802359092193385, + "learning_rate": 0.003, + "loss": 4.029, + "step": 12213 + }, + { + "epoch": 0.12214, + "grad_norm": 0.5632204879509363, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 12214 + }, + { + "epoch": 0.12215, + "grad_norm": 0.4750539555614441, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 12215 + }, + { + "epoch": 0.12216, + "grad_norm": 0.4815212274575964, + "learning_rate": 0.003, + "loss": 4.011, + "step": 12216 + }, + { + "epoch": 0.12217, + "grad_norm": 0.46609536015072917, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 12217 + }, + { + "epoch": 0.12218, + "grad_norm": 0.47835292787726835, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 12218 + }, + { + "epoch": 0.12219, + "grad_norm": 0.5038740589496469, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 12219 + }, + { + "epoch": 0.1222, + "grad_norm": 0.6291203854146187, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 12220 + }, + { + "epoch": 0.12221, + "grad_norm": 0.7661107155838461, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 12221 + }, + { + "epoch": 0.12222, + "grad_norm": 0.7336270065175331, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 12222 + }, + { + "epoch": 0.12223, + "grad_norm": 0.6612368045042317, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 12223 + }, + { + "epoch": 0.12224, + "grad_norm": 0.713620862308351, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 12224 + }, + { + "epoch": 0.12225, + "grad_norm": 0.6833328132701639, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12225 + }, + { + "epoch": 0.12226, + "grad_norm": 0.7616860536056672, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 12226 + }, + { + "epoch": 0.12227, + "grad_norm": 0.7456229749412068, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 12227 + }, + { + "epoch": 0.12228, + "grad_norm": 0.7888412174393651, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 12228 + }, + { + "epoch": 0.12229, + "grad_norm": 0.7323615773187732, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 12229 + }, + { + "epoch": 0.1223, + "grad_norm": 0.700621637110036, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 12230 + }, + { + "epoch": 0.12231, + "grad_norm": 0.6166201626923198, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 12231 + }, + { + "epoch": 0.12232, + "grad_norm": 0.6845813795187442, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 12232 + }, + { + "epoch": 0.12233, + "grad_norm": 0.7859508883410434, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 12233 + }, + { + "epoch": 0.12234, + "grad_norm": 0.8679742113967279, + "learning_rate": 0.003, + "loss": 4.049, + "step": 12234 + }, + { + "epoch": 0.12235, + "grad_norm": 0.8745939562909472, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 12235 + }, + { + "epoch": 0.12236, + "grad_norm": 0.7943786160987274, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 12236 + }, + { + "epoch": 0.12237, + "grad_norm": 0.7462281574712099, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 12237 + }, + { + "epoch": 0.12238, + "grad_norm": 0.776316920983184, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 12238 + }, + { + "epoch": 0.12239, + "grad_norm": 0.817240187687761, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 12239 + }, + { + "epoch": 0.1224, + "grad_norm": 0.8545485543540197, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 12240 + }, + { + "epoch": 0.12241, + "grad_norm": 0.8417365443876423, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 12241 + }, + { + "epoch": 0.12242, + "grad_norm": 0.7309885485012778, + "learning_rate": 0.003, + "loss": 4.026, + "step": 12242 + }, + { + "epoch": 0.12243, + "grad_norm": 0.6297442356666894, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 12243 + }, + { + "epoch": 0.12244, + "grad_norm": 0.7013778362855986, + "learning_rate": 0.003, + "loss": 4.04, + "step": 12244 + }, + { + "epoch": 0.12245, + "grad_norm": 0.6627982259816627, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 12245 + }, + { + "epoch": 0.12246, + "grad_norm": 0.8618879835141277, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 12246 + }, + { + "epoch": 0.12247, + "grad_norm": 1.1023256452841512, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 12247 + }, + { + "epoch": 0.12248, + "grad_norm": 0.9226861717532847, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 12248 + }, + { + "epoch": 0.12249, + "grad_norm": 0.8045398787822502, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 12249 + }, + { + "epoch": 0.1225, + "grad_norm": 0.7768127483722461, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 12250 + }, + { + "epoch": 0.12251, + "grad_norm": 0.8821540898165011, + "learning_rate": 0.003, + "loss": 4.056, + "step": 12251 + }, + { + "epoch": 0.12252, + "grad_norm": 0.8584113417880848, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 12252 + }, + { + "epoch": 0.12253, + "grad_norm": 0.8403673689646527, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 12253 + }, + { + "epoch": 0.12254, + "grad_norm": 0.8420690700211214, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 12254 + }, + { + "epoch": 0.12255, + "grad_norm": 0.8367651914213688, + "learning_rate": 0.003, + "loss": 4.03, + "step": 12255 + }, + { + "epoch": 0.12256, + "grad_norm": 0.8903278171687453, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 12256 + }, + { + "epoch": 0.12257, + "grad_norm": 0.7344398904994207, + "learning_rate": 0.003, + "loss": 4.0578, + "step": 12257 + }, + { + "epoch": 0.12258, + "grad_norm": 0.6746528604452621, + "learning_rate": 0.003, + "loss": 4.017, + "step": 12258 + }, + { + "epoch": 0.12259, + "grad_norm": 0.5873985490901731, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12259 + }, + { + "epoch": 0.1226, + "grad_norm": 0.5123118525327408, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 12260 + }, + { + "epoch": 0.12261, + "grad_norm": 0.5077577363817305, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 12261 + }, + { + "epoch": 0.12262, + "grad_norm": 0.40719165301405646, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 12262 + }, + { + "epoch": 0.12263, + "grad_norm": 0.42090685973062686, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 12263 + }, + { + "epoch": 0.12264, + "grad_norm": 0.4952019305338533, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 12264 + }, + { + "epoch": 0.12265, + "grad_norm": 0.6632679863879422, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 12265 + }, + { + "epoch": 0.12266, + "grad_norm": 0.8340033389899314, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 12266 + }, + { + "epoch": 0.12267, + "grad_norm": 0.9259276899610961, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 12267 + }, + { + "epoch": 0.12268, + "grad_norm": 0.7918753284469815, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 12268 + }, + { + "epoch": 0.12269, + "grad_norm": 0.6281077052903904, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 12269 + }, + { + "epoch": 0.1227, + "grad_norm": 0.7243693412388729, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 12270 + }, + { + "epoch": 0.12271, + "grad_norm": 0.8223355925669122, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 12271 + }, + { + "epoch": 0.12272, + "grad_norm": 0.7639367556504116, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 12272 + }, + { + "epoch": 0.12273, + "grad_norm": 0.72394348558991, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 12273 + }, + { + "epoch": 0.12274, + "grad_norm": 0.7095518438740538, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 12274 + }, + { + "epoch": 0.12275, + "grad_norm": 0.7803305108725178, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 12275 + }, + { + "epoch": 0.12276, + "grad_norm": 0.8227980053317807, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 12276 + }, + { + "epoch": 0.12277, + "grad_norm": 0.7690550649039816, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 12277 + }, + { + "epoch": 0.12278, + "grad_norm": 0.8129071577909464, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 12278 + }, + { + "epoch": 0.12279, + "grad_norm": 0.7430284979609999, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 12279 + }, + { + "epoch": 0.1228, + "grad_norm": 0.6071886581814888, + "learning_rate": 0.003, + "loss": 4.032, + "step": 12280 + }, + { + "epoch": 0.12281, + "grad_norm": 0.4898922161046083, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 12281 + }, + { + "epoch": 0.12282, + "grad_norm": 0.5834883813428733, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 12282 + }, + { + "epoch": 0.12283, + "grad_norm": 0.6308641863671628, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 12283 + }, + { + "epoch": 0.12284, + "grad_norm": 0.6462259198933907, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 12284 + }, + { + "epoch": 0.12285, + "grad_norm": 0.5833586817635454, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 12285 + }, + { + "epoch": 0.12286, + "grad_norm": 0.6410113195018471, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 12286 + }, + { + "epoch": 0.12287, + "grad_norm": 0.7131542916350695, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 12287 + }, + { + "epoch": 0.12288, + "grad_norm": 0.8059560971204641, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 12288 + }, + { + "epoch": 0.12289, + "grad_norm": 0.8980429361274765, + "learning_rate": 0.003, + "loss": 3.997, + "step": 12289 + }, + { + "epoch": 0.1229, + "grad_norm": 0.8066947914082162, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 12290 + }, + { + "epoch": 0.12291, + "grad_norm": 0.7559190015697826, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12291 + }, + { + "epoch": 0.12292, + "grad_norm": 0.8691708819910544, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 12292 + }, + { + "epoch": 0.12293, + "grad_norm": 0.8553669739941994, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 12293 + }, + { + "epoch": 0.12294, + "grad_norm": 0.8456664807927002, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 12294 + }, + { + "epoch": 0.12295, + "grad_norm": 0.9088976337070604, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 12295 + }, + { + "epoch": 0.12296, + "grad_norm": 0.9337460352959691, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 12296 + }, + { + "epoch": 0.12297, + "grad_norm": 0.9109523397554256, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 12297 + }, + { + "epoch": 0.12298, + "grad_norm": 0.9862208585310054, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 12298 + }, + { + "epoch": 0.12299, + "grad_norm": 0.9671371054368184, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 12299 + }, + { + "epoch": 0.123, + "grad_norm": 0.9254818086800719, + "learning_rate": 0.003, + "loss": 4.0581, + "step": 12300 + }, + { + "epoch": 0.12301, + "grad_norm": 0.7664677817410706, + "learning_rate": 0.003, + "loss": 4.0633, + "step": 12301 + }, + { + "epoch": 0.12302, + "grad_norm": 0.6272252394923856, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 12302 + }, + { + "epoch": 0.12303, + "grad_norm": 0.6473190266274695, + "learning_rate": 0.003, + "loss": 4.0985, + "step": 12303 + }, + { + "epoch": 0.12304, + "grad_norm": 0.6799472084647306, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 12304 + }, + { + "epoch": 0.12305, + "grad_norm": 0.7479864701159473, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 12305 + }, + { + "epoch": 0.12306, + "grad_norm": 0.779785774300088, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 12306 + }, + { + "epoch": 0.12307, + "grad_norm": 0.8208676755755425, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 12307 + }, + { + "epoch": 0.12308, + "grad_norm": 0.6604521717462359, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 12308 + }, + { + "epoch": 0.12309, + "grad_norm": 0.4747013636781702, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 12309 + }, + { + "epoch": 0.1231, + "grad_norm": 0.46222212874102686, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 12310 + }, + { + "epoch": 0.12311, + "grad_norm": 0.49393521966568543, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 12311 + }, + { + "epoch": 0.12312, + "grad_norm": 0.4986894838058389, + "learning_rate": 0.003, + "loss": 4.036, + "step": 12312 + }, + { + "epoch": 0.12313, + "grad_norm": 0.4895449242803725, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 12313 + }, + { + "epoch": 0.12314, + "grad_norm": 0.47279351646409473, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 12314 + }, + { + "epoch": 0.12315, + "grad_norm": 0.5097332422213202, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 12315 + }, + { + "epoch": 0.12316, + "grad_norm": 0.5623257079336488, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12316 + }, + { + "epoch": 0.12317, + "grad_norm": 0.5487603543227612, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 12317 + }, + { + "epoch": 0.12318, + "grad_norm": 0.579927406560016, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 12318 + }, + { + "epoch": 0.12319, + "grad_norm": 0.5704596940444074, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 12319 + }, + { + "epoch": 0.1232, + "grad_norm": 0.6323498253219983, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 12320 + }, + { + "epoch": 0.12321, + "grad_norm": 0.7169928286511796, + "learning_rate": 0.003, + "loss": 4.008, + "step": 12321 + }, + { + "epoch": 0.12322, + "grad_norm": 0.7763549225619278, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 12322 + }, + { + "epoch": 0.12323, + "grad_norm": 0.6755304974130839, + "learning_rate": 0.003, + "loss": 4.013, + "step": 12323 + }, + { + "epoch": 0.12324, + "grad_norm": 0.5658809704365689, + "learning_rate": 0.003, + "loss": 4.001, + "step": 12324 + }, + { + "epoch": 0.12325, + "grad_norm": 0.5291041463212703, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 12325 + }, + { + "epoch": 0.12326, + "grad_norm": 0.5474674065607366, + "learning_rate": 0.003, + "loss": 4.007, + "step": 12326 + }, + { + "epoch": 0.12327, + "grad_norm": 0.5656169231105567, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 12327 + }, + { + "epoch": 0.12328, + "grad_norm": 0.501071817134013, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 12328 + }, + { + "epoch": 0.12329, + "grad_norm": 0.48068770594275617, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12329 + }, + { + "epoch": 0.1233, + "grad_norm": 0.45177562052992265, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 12330 + }, + { + "epoch": 0.12331, + "grad_norm": 0.42773610126704115, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 12331 + }, + { + "epoch": 0.12332, + "grad_norm": 0.5012766645332282, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 12332 + }, + { + "epoch": 0.12333, + "grad_norm": 0.6062063963448587, + "learning_rate": 0.003, + "loss": 3.998, + "step": 12333 + }, + { + "epoch": 0.12334, + "grad_norm": 0.7642991013765253, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 12334 + }, + { + "epoch": 0.12335, + "grad_norm": 0.8993947092998722, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 12335 + }, + { + "epoch": 0.12336, + "grad_norm": 1.0224572343575784, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 12336 + }, + { + "epoch": 0.12337, + "grad_norm": 1.0715597189870802, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 12337 + }, + { + "epoch": 0.12338, + "grad_norm": 0.8993550470910423, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 12338 + }, + { + "epoch": 0.12339, + "grad_norm": 0.8176692783782188, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 12339 + }, + { + "epoch": 0.1234, + "grad_norm": 0.7606785950663456, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 12340 + }, + { + "epoch": 0.12341, + "grad_norm": 0.7696808312801278, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 12341 + }, + { + "epoch": 0.12342, + "grad_norm": 0.806554319144266, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12342 + }, + { + "epoch": 0.12343, + "grad_norm": 0.8362708834038468, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 12343 + }, + { + "epoch": 0.12344, + "grad_norm": 0.8266243424850089, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 12344 + }, + { + "epoch": 0.12345, + "grad_norm": 0.8302043093315791, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 12345 + }, + { + "epoch": 0.12346, + "grad_norm": 0.821254906060925, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 12346 + }, + { + "epoch": 0.12347, + "grad_norm": 0.8027621550042261, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 12347 + }, + { + "epoch": 0.12348, + "grad_norm": 0.8709355474373699, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 12348 + }, + { + "epoch": 0.12349, + "grad_norm": 0.9967505279194293, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 12349 + }, + { + "epoch": 0.1235, + "grad_norm": 1.0796805356938282, + "learning_rate": 0.003, + "loss": 4.033, + "step": 12350 + }, + { + "epoch": 0.12351, + "grad_norm": 0.8406993322271702, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 12351 + }, + { + "epoch": 0.12352, + "grad_norm": 1.0021482297962394, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 12352 + }, + { + "epoch": 0.12353, + "grad_norm": 1.0778347425394175, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 12353 + }, + { + "epoch": 0.12354, + "grad_norm": 0.8393633879019791, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 12354 + }, + { + "epoch": 0.12355, + "grad_norm": 0.6917192232892381, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 12355 + }, + { + "epoch": 0.12356, + "grad_norm": 0.6080842813991842, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 12356 + }, + { + "epoch": 0.12357, + "grad_norm": 0.6023400145486159, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 12357 + }, + { + "epoch": 0.12358, + "grad_norm": 0.6193724519184767, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 12358 + }, + { + "epoch": 0.12359, + "grad_norm": 0.5515048436229043, + "learning_rate": 0.003, + "loss": 4.0, + "step": 12359 + }, + { + "epoch": 0.1236, + "grad_norm": 0.5587857949096928, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 12360 + }, + { + "epoch": 0.12361, + "grad_norm": 0.4803038817528164, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 12361 + }, + { + "epoch": 0.12362, + "grad_norm": 0.49944158234184455, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 12362 + }, + { + "epoch": 0.12363, + "grad_norm": 0.5141555549067546, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 12363 + }, + { + "epoch": 0.12364, + "grad_norm": 0.5793774308649, + "learning_rate": 0.003, + "loss": 4.008, + "step": 12364 + }, + { + "epoch": 0.12365, + "grad_norm": 0.6251696237009211, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 12365 + }, + { + "epoch": 0.12366, + "grad_norm": 0.6063515023020518, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 12366 + }, + { + "epoch": 0.12367, + "grad_norm": 0.5799657290304754, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 12367 + }, + { + "epoch": 0.12368, + "grad_norm": 0.5846292193431708, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 12368 + }, + { + "epoch": 0.12369, + "grad_norm": 0.6340900964410617, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 12369 + }, + { + "epoch": 0.1237, + "grad_norm": 0.7051686342791814, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 12370 + }, + { + "epoch": 0.12371, + "grad_norm": 0.7890284925160758, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 12371 + }, + { + "epoch": 0.12372, + "grad_norm": 0.8806190961405924, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 12372 + }, + { + "epoch": 0.12373, + "grad_norm": 0.8939263045149182, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 12373 + }, + { + "epoch": 0.12374, + "grad_norm": 0.940985600666141, + "learning_rate": 0.003, + "loss": 4.031, + "step": 12374 + }, + { + "epoch": 0.12375, + "grad_norm": 0.8367093536876299, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 12375 + }, + { + "epoch": 0.12376, + "grad_norm": 0.8260090436890121, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 12376 + }, + { + "epoch": 0.12377, + "grad_norm": 0.7592757967452624, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 12377 + }, + { + "epoch": 0.12378, + "grad_norm": 0.7713217069461218, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 12378 + }, + { + "epoch": 0.12379, + "grad_norm": 0.7052414521057527, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 12379 + }, + { + "epoch": 0.1238, + "grad_norm": 0.6541912621524845, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 12380 + }, + { + "epoch": 0.12381, + "grad_norm": 0.5122456756962388, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 12381 + }, + { + "epoch": 0.12382, + "grad_norm": 0.5585644147307102, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 12382 + }, + { + "epoch": 0.12383, + "grad_norm": 0.6603531354055283, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 12383 + }, + { + "epoch": 0.12384, + "grad_norm": 0.7693679769312304, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 12384 + }, + { + "epoch": 0.12385, + "grad_norm": 0.7167806672263544, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 12385 + }, + { + "epoch": 0.12386, + "grad_norm": 0.6823679704296719, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 12386 + }, + { + "epoch": 0.12387, + "grad_norm": 0.7148869392814954, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12387 + }, + { + "epoch": 0.12388, + "grad_norm": 0.7519963366339514, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 12388 + }, + { + "epoch": 0.12389, + "grad_norm": 0.8529701276964179, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 12389 + }, + { + "epoch": 0.1239, + "grad_norm": 0.8017054403359029, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 12390 + }, + { + "epoch": 0.12391, + "grad_norm": 0.5994430649079796, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 12391 + }, + { + "epoch": 0.12392, + "grad_norm": 0.6440164469180917, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 12392 + }, + { + "epoch": 0.12393, + "grad_norm": 0.6451669094933188, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 12393 + }, + { + "epoch": 0.12394, + "grad_norm": 0.6715935528310077, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 12394 + }, + { + "epoch": 0.12395, + "grad_norm": 0.620593024934601, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 12395 + }, + { + "epoch": 0.12396, + "grad_norm": 0.6899258629437591, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 12396 + }, + { + "epoch": 0.12397, + "grad_norm": 0.770291572422279, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 12397 + }, + { + "epoch": 0.12398, + "grad_norm": 0.9175635101037003, + "learning_rate": 0.003, + "loss": 4.053, + "step": 12398 + }, + { + "epoch": 0.12399, + "grad_norm": 0.9638996849381228, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 12399 + }, + { + "epoch": 0.124, + "grad_norm": 0.8704995341922315, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 12400 + }, + { + "epoch": 0.12401, + "grad_norm": 0.7137045464671262, + "learning_rate": 0.003, + "loss": 4.003, + "step": 12401 + }, + { + "epoch": 0.12402, + "grad_norm": 0.7616721933609448, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 12402 + }, + { + "epoch": 0.12403, + "grad_norm": 0.7655622738477017, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 12403 + }, + { + "epoch": 0.12404, + "grad_norm": 0.7454115780979849, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 12404 + }, + { + "epoch": 0.12405, + "grad_norm": 0.79504643058947, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 12405 + }, + { + "epoch": 0.12406, + "grad_norm": 0.885383068411899, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 12406 + }, + { + "epoch": 0.12407, + "grad_norm": 0.8330572846362223, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12407 + }, + { + "epoch": 0.12408, + "grad_norm": 0.7207005477102404, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 12408 + }, + { + "epoch": 0.12409, + "grad_norm": 0.6309805893063962, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 12409 + }, + { + "epoch": 0.1241, + "grad_norm": 0.5923499120143008, + "learning_rate": 0.003, + "loss": 4.039, + "step": 12410 + }, + { + "epoch": 0.12411, + "grad_norm": 0.6304360211835824, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 12411 + }, + { + "epoch": 0.12412, + "grad_norm": 0.6731596606139947, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 12412 + }, + { + "epoch": 0.12413, + "grad_norm": 0.6057840224389032, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 12413 + }, + { + "epoch": 0.12414, + "grad_norm": 0.6163760605435324, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 12414 + }, + { + "epoch": 0.12415, + "grad_norm": 0.6037770573875274, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12415 + }, + { + "epoch": 0.12416, + "grad_norm": 0.5607531545824679, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 12416 + }, + { + "epoch": 0.12417, + "grad_norm": 0.5748238109074049, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 12417 + }, + { + "epoch": 0.12418, + "grad_norm": 0.6064861887456456, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 12418 + }, + { + "epoch": 0.12419, + "grad_norm": 0.6249432733870561, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 12419 + }, + { + "epoch": 0.1242, + "grad_norm": 0.6614127996331236, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 12420 + }, + { + "epoch": 0.12421, + "grad_norm": 0.718409864754456, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 12421 + }, + { + "epoch": 0.12422, + "grad_norm": 0.7740941410506029, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12422 + }, + { + "epoch": 0.12423, + "grad_norm": 0.8694386154158571, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 12423 + }, + { + "epoch": 0.12424, + "grad_norm": 1.0939304776182959, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 12424 + }, + { + "epoch": 0.12425, + "grad_norm": 1.015332541463675, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 12425 + }, + { + "epoch": 0.12426, + "grad_norm": 0.7828955810940796, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 12426 + }, + { + "epoch": 0.12427, + "grad_norm": 0.7322463077362746, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 12427 + }, + { + "epoch": 0.12428, + "grad_norm": 0.7195327758268578, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 12428 + }, + { + "epoch": 0.12429, + "grad_norm": 0.8325911394801826, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 12429 + }, + { + "epoch": 0.1243, + "grad_norm": 0.7490875695713716, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 12430 + }, + { + "epoch": 0.12431, + "grad_norm": 0.6829546327052212, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 12431 + }, + { + "epoch": 0.12432, + "grad_norm": 0.6956627105311378, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 12432 + }, + { + "epoch": 0.12433, + "grad_norm": 0.6601996567093849, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 12433 + }, + { + "epoch": 0.12434, + "grad_norm": 0.7076823486628853, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 12434 + }, + { + "epoch": 0.12435, + "grad_norm": 0.8199399571272888, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12435 + }, + { + "epoch": 0.12436, + "grad_norm": 0.8212919355772265, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 12436 + }, + { + "epoch": 0.12437, + "grad_norm": 0.8169338717404583, + "learning_rate": 0.003, + "loss": 4.007, + "step": 12437 + }, + { + "epoch": 0.12438, + "grad_norm": 0.7074443554394689, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 12438 + }, + { + "epoch": 0.12439, + "grad_norm": 0.7754290813667579, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 12439 + }, + { + "epoch": 0.1244, + "grad_norm": 0.7763171694020949, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 12440 + }, + { + "epoch": 0.12441, + "grad_norm": 0.7373750779691393, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 12441 + }, + { + "epoch": 0.12442, + "grad_norm": 0.7158345346471956, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 12442 + }, + { + "epoch": 0.12443, + "grad_norm": 0.6857826624017002, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12443 + }, + { + "epoch": 0.12444, + "grad_norm": 0.6539946315257654, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 12444 + }, + { + "epoch": 0.12445, + "grad_norm": 0.6052258462356226, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12445 + }, + { + "epoch": 0.12446, + "grad_norm": 0.571728791743933, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 12446 + }, + { + "epoch": 0.12447, + "grad_norm": 0.5824477515548783, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 12447 + }, + { + "epoch": 0.12448, + "grad_norm": 0.5621766449061799, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 12448 + }, + { + "epoch": 0.12449, + "grad_norm": 0.5279542142424815, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 12449 + }, + { + "epoch": 0.1245, + "grad_norm": 0.5272876801933257, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 12450 + }, + { + "epoch": 0.12451, + "grad_norm": 0.48956243110978387, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 12451 + }, + { + "epoch": 0.12452, + "grad_norm": 0.4369320274418459, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 12452 + }, + { + "epoch": 0.12453, + "grad_norm": 0.43818662481472104, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 12453 + }, + { + "epoch": 0.12454, + "grad_norm": 0.41350751199578706, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 12454 + }, + { + "epoch": 0.12455, + "grad_norm": 0.4203070656244474, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 12455 + }, + { + "epoch": 0.12456, + "grad_norm": 0.519370049816637, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 12456 + }, + { + "epoch": 0.12457, + "grad_norm": 0.7252122128209719, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 12457 + }, + { + "epoch": 0.12458, + "grad_norm": 0.9970764609817455, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 12458 + }, + { + "epoch": 0.12459, + "grad_norm": 1.0895213234743544, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 12459 + }, + { + "epoch": 0.1246, + "grad_norm": 0.7585607356690333, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 12460 + }, + { + "epoch": 0.12461, + "grad_norm": 0.7183919335589978, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 12461 + }, + { + "epoch": 0.12462, + "grad_norm": 0.7764688437932653, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 12462 + }, + { + "epoch": 0.12463, + "grad_norm": 0.7396678083282455, + "learning_rate": 0.003, + "loss": 4.015, + "step": 12463 + }, + { + "epoch": 0.12464, + "grad_norm": 0.7507537102654952, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 12464 + }, + { + "epoch": 0.12465, + "grad_norm": 0.645177440353463, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 12465 + }, + { + "epoch": 0.12466, + "grad_norm": 0.6825057857392811, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 12466 + }, + { + "epoch": 0.12467, + "grad_norm": 0.762958680503123, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 12467 + }, + { + "epoch": 0.12468, + "grad_norm": 0.9584677094939008, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 12468 + }, + { + "epoch": 0.12469, + "grad_norm": 0.9876287254078742, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 12469 + }, + { + "epoch": 0.1247, + "grad_norm": 0.8695036779081644, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 12470 + }, + { + "epoch": 0.12471, + "grad_norm": 0.7501409090502065, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 12471 + }, + { + "epoch": 0.12472, + "grad_norm": 0.6943305623406335, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 12472 + }, + { + "epoch": 0.12473, + "grad_norm": 0.7453606497977399, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 12473 + }, + { + "epoch": 0.12474, + "grad_norm": 0.87629266169875, + "learning_rate": 0.003, + "loss": 4.04, + "step": 12474 + }, + { + "epoch": 0.12475, + "grad_norm": 0.8514619774297043, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12475 + }, + { + "epoch": 0.12476, + "grad_norm": 0.7622485566593947, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 12476 + }, + { + "epoch": 0.12477, + "grad_norm": 0.6346584310567632, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 12477 + }, + { + "epoch": 0.12478, + "grad_norm": 0.636561316682709, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 12478 + }, + { + "epoch": 0.12479, + "grad_norm": 0.5646242791879021, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 12479 + }, + { + "epoch": 0.1248, + "grad_norm": 0.575958030238272, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 12480 + }, + { + "epoch": 0.12481, + "grad_norm": 0.5751247316860684, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 12481 + }, + { + "epoch": 0.12482, + "grad_norm": 0.5569854687227197, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 12482 + }, + { + "epoch": 0.12483, + "grad_norm": 0.5101451595387215, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 12483 + }, + { + "epoch": 0.12484, + "grad_norm": 0.5084889765146162, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 12484 + }, + { + "epoch": 0.12485, + "grad_norm": 0.5540340512059015, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 12485 + }, + { + "epoch": 0.12486, + "grad_norm": 0.6915524715264099, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 12486 + }, + { + "epoch": 0.12487, + "grad_norm": 0.7554127691436334, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 12487 + }, + { + "epoch": 0.12488, + "grad_norm": 0.7439352147530245, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 12488 + }, + { + "epoch": 0.12489, + "grad_norm": 0.6897342725232187, + "learning_rate": 0.003, + "loss": 4.001, + "step": 12489 + }, + { + "epoch": 0.1249, + "grad_norm": 0.713551249379984, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 12490 + }, + { + "epoch": 0.12491, + "grad_norm": 0.6648342882996953, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 12491 + }, + { + "epoch": 0.12492, + "grad_norm": 0.6308484967904898, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 12492 + }, + { + "epoch": 0.12493, + "grad_norm": 0.7040667286173012, + "learning_rate": 0.003, + "loss": 4.03, + "step": 12493 + }, + { + "epoch": 0.12494, + "grad_norm": 0.8091222767765611, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 12494 + }, + { + "epoch": 0.12495, + "grad_norm": 0.9732881920663594, + "learning_rate": 0.003, + "loss": 4.029, + "step": 12495 + }, + { + "epoch": 0.12496, + "grad_norm": 1.0567199066000927, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 12496 + }, + { + "epoch": 0.12497, + "grad_norm": 0.8450124839856124, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 12497 + }, + { + "epoch": 0.12498, + "grad_norm": 0.7138446081856737, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12498 + }, + { + "epoch": 0.12499, + "grad_norm": 0.9079859599512766, + "learning_rate": 0.003, + "loss": 4.037, + "step": 12499 + }, + { + "epoch": 0.125, + "grad_norm": 0.9800653947813974, + "learning_rate": 0.003, + "loss": 4.0733, + "step": 12500 + }, + { + "epoch": 0.12501, + "grad_norm": 1.0578058930764078, + "learning_rate": 0.003, + "loss": 4.0569, + "step": 12501 + }, + { + "epoch": 0.12502, + "grad_norm": 0.8475156711311451, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 12502 + }, + { + "epoch": 0.12503, + "grad_norm": 0.8824740658019747, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 12503 + }, + { + "epoch": 0.12504, + "grad_norm": 0.9692355075537794, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 12504 + }, + { + "epoch": 0.12505, + "grad_norm": 0.9615498699912345, + "learning_rate": 0.003, + "loss": 4.0555, + "step": 12505 + }, + { + "epoch": 0.12506, + "grad_norm": 1.0498181374113595, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 12506 + }, + { + "epoch": 0.12507, + "grad_norm": 0.8870154230052183, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 12507 + }, + { + "epoch": 0.12508, + "grad_norm": 0.6945658331105243, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 12508 + }, + { + "epoch": 0.12509, + "grad_norm": 0.6361428229244743, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 12509 + }, + { + "epoch": 0.1251, + "grad_norm": 0.6781966115937983, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 12510 + }, + { + "epoch": 0.12511, + "grad_norm": 0.6434049429942361, + "learning_rate": 0.003, + "loss": 4.019, + "step": 12511 + }, + { + "epoch": 0.12512, + "grad_norm": 0.6216335373271412, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 12512 + }, + { + "epoch": 0.12513, + "grad_norm": 0.6175337469469572, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 12513 + }, + { + "epoch": 0.12514, + "grad_norm": 0.6813866360533597, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 12514 + }, + { + "epoch": 0.12515, + "grad_norm": 0.7366604452587351, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 12515 + }, + { + "epoch": 0.12516, + "grad_norm": 0.6301987089909329, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 12516 + }, + { + "epoch": 0.12517, + "grad_norm": 0.6103628941634551, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 12517 + }, + { + "epoch": 0.12518, + "grad_norm": 0.6252040868353, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 12518 + }, + { + "epoch": 0.12519, + "grad_norm": 0.5595075119549755, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12519 + }, + { + "epoch": 0.1252, + "grad_norm": 0.5008743726728848, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 12520 + }, + { + "epoch": 0.12521, + "grad_norm": 0.621070788701964, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 12521 + }, + { + "epoch": 0.12522, + "grad_norm": 0.7996762758651681, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 12522 + }, + { + "epoch": 0.12523, + "grad_norm": 0.9093901732148741, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12523 + }, + { + "epoch": 0.12524, + "grad_norm": 0.8553800443647471, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 12524 + }, + { + "epoch": 0.12525, + "grad_norm": 0.786042541196193, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 12525 + }, + { + "epoch": 0.12526, + "grad_norm": 0.7113829851776844, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 12526 + }, + { + "epoch": 0.12527, + "grad_norm": 0.734037106118266, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 12527 + }, + { + "epoch": 0.12528, + "grad_norm": 0.7353801843638095, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 12528 + }, + { + "epoch": 0.12529, + "grad_norm": 0.6257156826861133, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12529 + }, + { + "epoch": 0.1253, + "grad_norm": 0.5472550626772146, + "learning_rate": 0.003, + "loss": 3.998, + "step": 12530 + }, + { + "epoch": 0.12531, + "grad_norm": 0.5011931380263308, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 12531 + }, + { + "epoch": 0.12532, + "grad_norm": 0.5181004507560574, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 12532 + }, + { + "epoch": 0.12533, + "grad_norm": 0.4989300667717308, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12533 + }, + { + "epoch": 0.12534, + "grad_norm": 0.5451229972899683, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 12534 + }, + { + "epoch": 0.12535, + "grad_norm": 0.5343451680023377, + "learning_rate": 0.003, + "loss": 4.021, + "step": 12535 + }, + { + "epoch": 0.12536, + "grad_norm": 0.5472675471304776, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 12536 + }, + { + "epoch": 0.12537, + "grad_norm": 0.5004378129150562, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 12537 + }, + { + "epoch": 0.12538, + "grad_norm": 0.5480993517031261, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 12538 + }, + { + "epoch": 0.12539, + "grad_norm": 0.6333990339680527, + "learning_rate": 0.003, + "loss": 4.042, + "step": 12539 + }, + { + "epoch": 0.1254, + "grad_norm": 0.7570721919612216, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 12540 + }, + { + "epoch": 0.12541, + "grad_norm": 0.8555047869435546, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 12541 + }, + { + "epoch": 0.12542, + "grad_norm": 0.8594707791058096, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12542 + }, + { + "epoch": 0.12543, + "grad_norm": 0.8048899871219781, + "learning_rate": 0.003, + "loss": 3.98, + "step": 12543 + }, + { + "epoch": 0.12544, + "grad_norm": 0.8044310832604974, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 12544 + }, + { + "epoch": 0.12545, + "grad_norm": 0.757760144208355, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 12545 + }, + { + "epoch": 0.12546, + "grad_norm": 0.708025830965386, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 12546 + }, + { + "epoch": 0.12547, + "grad_norm": 0.7821022673542635, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 12547 + }, + { + "epoch": 0.12548, + "grad_norm": 0.8242120651599885, + "learning_rate": 0.003, + "loss": 4.003, + "step": 12548 + }, + { + "epoch": 0.12549, + "grad_norm": 0.7855258881004425, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 12549 + }, + { + "epoch": 0.1255, + "grad_norm": 0.6995680497935454, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 12550 + }, + { + "epoch": 0.12551, + "grad_norm": 0.8296509443491399, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 12551 + }, + { + "epoch": 0.12552, + "grad_norm": 0.8818198196258639, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 12552 + }, + { + "epoch": 0.12553, + "grad_norm": 0.9798039684460182, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 12553 + }, + { + "epoch": 0.12554, + "grad_norm": 0.9126738144861728, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 12554 + }, + { + "epoch": 0.12555, + "grad_norm": 0.863904003274379, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 12555 + }, + { + "epoch": 0.12556, + "grad_norm": 0.9403148918761077, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 12556 + }, + { + "epoch": 0.12557, + "grad_norm": 1.0633744181343225, + "learning_rate": 0.003, + "loss": 4.0624, + "step": 12557 + }, + { + "epoch": 0.12558, + "grad_norm": 0.7171053332831061, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 12558 + }, + { + "epoch": 0.12559, + "grad_norm": 0.7270031863681754, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 12559 + }, + { + "epoch": 0.1256, + "grad_norm": 0.6451078755533459, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 12560 + }, + { + "epoch": 0.12561, + "grad_norm": 0.7136007373509008, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 12561 + }, + { + "epoch": 0.12562, + "grad_norm": 0.7733686206960838, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 12562 + }, + { + "epoch": 0.12563, + "grad_norm": 0.8605277388575127, + "learning_rate": 0.003, + "loss": 4.035, + "step": 12563 + }, + { + "epoch": 0.12564, + "grad_norm": 0.9472243452708682, + "learning_rate": 0.003, + "loss": 4.042, + "step": 12564 + }, + { + "epoch": 0.12565, + "grad_norm": 0.9806181579933716, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 12565 + }, + { + "epoch": 0.12566, + "grad_norm": 0.9611459820903897, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 12566 + }, + { + "epoch": 0.12567, + "grad_norm": 0.8798905961091766, + "learning_rate": 0.003, + "loss": 4.0802, + "step": 12567 + }, + { + "epoch": 0.12568, + "grad_norm": 0.8822136152271015, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 12568 + }, + { + "epoch": 0.12569, + "grad_norm": 0.8461660477698004, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 12569 + }, + { + "epoch": 0.1257, + "grad_norm": 0.9263903418882089, + "learning_rate": 0.003, + "loss": 4.0843, + "step": 12570 + }, + { + "epoch": 0.12571, + "grad_norm": 0.9157280655035767, + "learning_rate": 0.003, + "loss": 4.053, + "step": 12571 + }, + { + "epoch": 0.12572, + "grad_norm": 0.8135265194486356, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 12572 + }, + { + "epoch": 0.12573, + "grad_norm": 0.6724697828511556, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 12573 + }, + { + "epoch": 0.12574, + "grad_norm": 0.6036771103667091, + "learning_rate": 0.003, + "loss": 4.043, + "step": 12574 + }, + { + "epoch": 0.12575, + "grad_norm": 0.4958203031430189, + "learning_rate": 0.003, + "loss": 4.0663, + "step": 12575 + }, + { + "epoch": 0.12576, + "grad_norm": 0.4975621505196232, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 12576 + }, + { + "epoch": 0.12577, + "grad_norm": 0.5095062702745917, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 12577 + }, + { + "epoch": 0.12578, + "grad_norm": 0.49428856693699763, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 12578 + }, + { + "epoch": 0.12579, + "grad_norm": 0.5510240536273308, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 12579 + }, + { + "epoch": 0.1258, + "grad_norm": 0.503844361316043, + "learning_rate": 0.003, + "loss": 4.029, + "step": 12580 + }, + { + "epoch": 0.12581, + "grad_norm": 0.5245675670340305, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 12581 + }, + { + "epoch": 0.12582, + "grad_norm": 0.5427826842241954, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 12582 + }, + { + "epoch": 0.12583, + "grad_norm": 0.5281550591646336, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 12583 + }, + { + "epoch": 0.12584, + "grad_norm": 0.5009155003371542, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 12584 + }, + { + "epoch": 0.12585, + "grad_norm": 0.46780406733192964, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 12585 + }, + { + "epoch": 0.12586, + "grad_norm": 0.44536742766779547, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12586 + }, + { + "epoch": 0.12587, + "grad_norm": 0.43547794380921895, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 12587 + }, + { + "epoch": 0.12588, + "grad_norm": 0.40870375998696234, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 12588 + }, + { + "epoch": 0.12589, + "grad_norm": 0.4607634928155922, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 12589 + }, + { + "epoch": 0.1259, + "grad_norm": 0.5263411997518783, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 12590 + }, + { + "epoch": 0.12591, + "grad_norm": 0.6584183265349862, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 12591 + }, + { + "epoch": 0.12592, + "grad_norm": 0.836732438438581, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 12592 + }, + { + "epoch": 0.12593, + "grad_norm": 0.9320660994503748, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 12593 + }, + { + "epoch": 0.12594, + "grad_norm": 0.8712549012878829, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 12594 + }, + { + "epoch": 0.12595, + "grad_norm": 0.7386402727519519, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 12595 + }, + { + "epoch": 0.12596, + "grad_norm": 0.7166041780002438, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 12596 + }, + { + "epoch": 0.12597, + "grad_norm": 0.6930760397691004, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 12597 + }, + { + "epoch": 0.12598, + "grad_norm": 0.7149348163088268, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 12598 + }, + { + "epoch": 0.12599, + "grad_norm": 0.8447184389852475, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 12599 + }, + { + "epoch": 0.126, + "grad_norm": 0.8397489608185655, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 12600 + }, + { + "epoch": 0.12601, + "grad_norm": 0.8300874812769428, + "learning_rate": 0.003, + "loss": 4.0634, + "step": 12601 + }, + { + "epoch": 0.12602, + "grad_norm": 0.6765991123412859, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 12602 + }, + { + "epoch": 0.12603, + "grad_norm": 0.6050687035068897, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 12603 + }, + { + "epoch": 0.12604, + "grad_norm": 0.6663817347022107, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 12604 + }, + { + "epoch": 0.12605, + "grad_norm": 0.7675435740178078, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 12605 + }, + { + "epoch": 0.12606, + "grad_norm": 0.7279677848177217, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 12606 + }, + { + "epoch": 0.12607, + "grad_norm": 0.8102693277690036, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 12607 + }, + { + "epoch": 0.12608, + "grad_norm": 0.9252139022822419, + "learning_rate": 0.003, + "loss": 4.023, + "step": 12608 + }, + { + "epoch": 0.12609, + "grad_norm": 0.9772294095914237, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 12609 + }, + { + "epoch": 0.1261, + "grad_norm": 0.900903932435985, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 12610 + }, + { + "epoch": 0.12611, + "grad_norm": 0.887105803615812, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 12611 + }, + { + "epoch": 0.12612, + "grad_norm": 0.8672350986954146, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 12612 + }, + { + "epoch": 0.12613, + "grad_norm": 0.8743637054452182, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 12613 + }, + { + "epoch": 0.12614, + "grad_norm": 0.805613213407976, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 12614 + }, + { + "epoch": 0.12615, + "grad_norm": 0.784837573781296, + "learning_rate": 0.003, + "loss": 4.0592, + "step": 12615 + }, + { + "epoch": 0.12616, + "grad_norm": 0.7438623888753287, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 12616 + }, + { + "epoch": 0.12617, + "grad_norm": 0.7859635803623858, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 12617 + }, + { + "epoch": 0.12618, + "grad_norm": 0.7686008312001422, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 12618 + }, + { + "epoch": 0.12619, + "grad_norm": 0.6669553218402603, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 12619 + }, + { + "epoch": 0.1262, + "grad_norm": 0.8526238826090266, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 12620 + }, + { + "epoch": 0.12621, + "grad_norm": 0.9618260533212021, + "learning_rate": 0.003, + "loss": 4.036, + "step": 12621 + }, + { + "epoch": 0.12622, + "grad_norm": 0.936976560209041, + "learning_rate": 0.003, + "loss": 4.042, + "step": 12622 + }, + { + "epoch": 0.12623, + "grad_norm": 0.9642726214724836, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 12623 + }, + { + "epoch": 0.12624, + "grad_norm": 0.8357645252661476, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 12624 + }, + { + "epoch": 0.12625, + "grad_norm": 0.6237213741572843, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 12625 + }, + { + "epoch": 0.12626, + "grad_norm": 0.6596686858155588, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 12626 + }, + { + "epoch": 0.12627, + "grad_norm": 0.6444835356209108, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 12627 + }, + { + "epoch": 0.12628, + "grad_norm": 0.6791495672314437, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 12628 + }, + { + "epoch": 0.12629, + "grad_norm": 0.6387062173463641, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 12629 + }, + { + "epoch": 0.1263, + "grad_norm": 0.5538925467862449, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 12630 + }, + { + "epoch": 0.12631, + "grad_norm": 0.6121970808287163, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 12631 + }, + { + "epoch": 0.12632, + "grad_norm": 0.6792975489288957, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 12632 + }, + { + "epoch": 0.12633, + "grad_norm": 0.6916454467920624, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 12633 + }, + { + "epoch": 0.12634, + "grad_norm": 0.7462165609696618, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 12634 + }, + { + "epoch": 0.12635, + "grad_norm": 0.6897713728163887, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 12635 + }, + { + "epoch": 0.12636, + "grad_norm": 0.6084702433864402, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 12636 + }, + { + "epoch": 0.12637, + "grad_norm": 0.5670691606696705, + "learning_rate": 0.003, + "loss": 4.023, + "step": 12637 + }, + { + "epoch": 0.12638, + "grad_norm": 0.5117865272795981, + "learning_rate": 0.003, + "loss": 4.02, + "step": 12638 + }, + { + "epoch": 0.12639, + "grad_norm": 0.538078168811481, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 12639 + }, + { + "epoch": 0.1264, + "grad_norm": 0.5689655917261884, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 12640 + }, + { + "epoch": 0.12641, + "grad_norm": 0.5623666727662494, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 12641 + }, + { + "epoch": 0.12642, + "grad_norm": 0.5575819542019951, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 12642 + }, + { + "epoch": 0.12643, + "grad_norm": 0.5520268326068702, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12643 + }, + { + "epoch": 0.12644, + "grad_norm": 0.561996326867302, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 12644 + }, + { + "epoch": 0.12645, + "grad_norm": 0.5791197677603279, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 12645 + }, + { + "epoch": 0.12646, + "grad_norm": 0.5801813409389764, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 12646 + }, + { + "epoch": 0.12647, + "grad_norm": 0.5255495643487403, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 12647 + }, + { + "epoch": 0.12648, + "grad_norm": 0.4994842836905242, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 12648 + }, + { + "epoch": 0.12649, + "grad_norm": 0.4919555449016544, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 12649 + }, + { + "epoch": 0.1265, + "grad_norm": 0.5602674499273467, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 12650 + }, + { + "epoch": 0.12651, + "grad_norm": 0.6131976651451538, + "learning_rate": 0.003, + "loss": 4.016, + "step": 12651 + }, + { + "epoch": 0.12652, + "grad_norm": 0.671614922881488, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 12652 + }, + { + "epoch": 0.12653, + "grad_norm": 0.6652403759927256, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 12653 + }, + { + "epoch": 0.12654, + "grad_norm": 0.5713555565804928, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 12654 + }, + { + "epoch": 0.12655, + "grad_norm": 0.6508393777983524, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 12655 + }, + { + "epoch": 0.12656, + "grad_norm": 0.6812045586708333, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 12656 + }, + { + "epoch": 0.12657, + "grad_norm": 0.6599436782699626, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 12657 + }, + { + "epoch": 0.12658, + "grad_norm": 0.6616565779625634, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 12658 + }, + { + "epoch": 0.12659, + "grad_norm": 0.7522543077697439, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 12659 + }, + { + "epoch": 0.1266, + "grad_norm": 0.8176888449706824, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 12660 + }, + { + "epoch": 0.12661, + "grad_norm": 0.8397657321012029, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 12661 + }, + { + "epoch": 0.12662, + "grad_norm": 0.8415550150536829, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 12662 + }, + { + "epoch": 0.12663, + "grad_norm": 0.8836550239076064, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 12663 + }, + { + "epoch": 0.12664, + "grad_norm": 0.8382451297203382, + "learning_rate": 0.003, + "loss": 4.021, + "step": 12664 + }, + { + "epoch": 0.12665, + "grad_norm": 0.8008017207127601, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 12665 + }, + { + "epoch": 0.12666, + "grad_norm": 0.9308638528396662, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 12666 + }, + { + "epoch": 0.12667, + "grad_norm": 0.9428075287283626, + "learning_rate": 0.003, + "loss": 4.014, + "step": 12667 + }, + { + "epoch": 0.12668, + "grad_norm": 1.0107029401777152, + "learning_rate": 0.003, + "loss": 4.028, + "step": 12668 + }, + { + "epoch": 0.12669, + "grad_norm": 0.9788860789286673, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 12669 + }, + { + "epoch": 0.1267, + "grad_norm": 0.9857956813987275, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 12670 + }, + { + "epoch": 0.12671, + "grad_norm": 1.1023069770228997, + "learning_rate": 0.003, + "loss": 4.0587, + "step": 12671 + }, + { + "epoch": 0.12672, + "grad_norm": 1.0046358406354212, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 12672 + }, + { + "epoch": 0.12673, + "grad_norm": 0.9009204068150752, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 12673 + }, + { + "epoch": 0.12674, + "grad_norm": 0.7941864251673758, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 12674 + }, + { + "epoch": 0.12675, + "grad_norm": 0.7995351150774043, + "learning_rate": 0.003, + "loss": 4.0612, + "step": 12675 + }, + { + "epoch": 0.12676, + "grad_norm": 0.7429221299065081, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 12676 + }, + { + "epoch": 0.12677, + "grad_norm": 0.8115578671987949, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 12677 + }, + { + "epoch": 0.12678, + "grad_norm": 0.887608701350907, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 12678 + }, + { + "epoch": 0.12679, + "grad_norm": 0.8729284399907145, + "learning_rate": 0.003, + "loss": 4.045, + "step": 12679 + }, + { + "epoch": 0.1268, + "grad_norm": 0.7438855961312207, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 12680 + }, + { + "epoch": 0.12681, + "grad_norm": 0.7311190230395612, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 12681 + }, + { + "epoch": 0.12682, + "grad_norm": 0.8156596178248374, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 12682 + }, + { + "epoch": 0.12683, + "grad_norm": 0.8067708747410027, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 12683 + }, + { + "epoch": 0.12684, + "grad_norm": 0.6864492034340791, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 12684 + }, + { + "epoch": 0.12685, + "grad_norm": 0.6393371540821392, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 12685 + }, + { + "epoch": 0.12686, + "grad_norm": 0.578161537749626, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 12686 + }, + { + "epoch": 0.12687, + "grad_norm": 0.5266883817074044, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 12687 + }, + { + "epoch": 0.12688, + "grad_norm": 0.5113717052603445, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 12688 + }, + { + "epoch": 0.12689, + "grad_norm": 0.5323646893655529, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 12689 + }, + { + "epoch": 0.1269, + "grad_norm": 0.6066401787795697, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 12690 + }, + { + "epoch": 0.12691, + "grad_norm": 0.6087632307543199, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 12691 + }, + { + "epoch": 0.12692, + "grad_norm": 0.6295334817523162, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 12692 + }, + { + "epoch": 0.12693, + "grad_norm": 0.607223564136249, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12693 + }, + { + "epoch": 0.12694, + "grad_norm": 0.5630346061003605, + "learning_rate": 0.003, + "loss": 3.989, + "step": 12694 + }, + { + "epoch": 0.12695, + "grad_norm": 0.5162362652245667, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 12695 + }, + { + "epoch": 0.12696, + "grad_norm": 0.46362331808447826, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 12696 + }, + { + "epoch": 0.12697, + "grad_norm": 0.4669323975993397, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 12697 + }, + { + "epoch": 0.12698, + "grad_norm": 0.4898514894057004, + "learning_rate": 0.003, + "loss": 3.989, + "step": 12698 + }, + { + "epoch": 0.12699, + "grad_norm": 0.4414935736782532, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 12699 + }, + { + "epoch": 0.127, + "grad_norm": 0.5169725547797842, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 12700 + }, + { + "epoch": 0.12701, + "grad_norm": 0.5900515197892178, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 12701 + }, + { + "epoch": 0.12702, + "grad_norm": 0.8453778857581035, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 12702 + }, + { + "epoch": 0.12703, + "grad_norm": 1.0886568749879246, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12703 + }, + { + "epoch": 0.12704, + "grad_norm": 0.8812296530731106, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 12704 + }, + { + "epoch": 0.12705, + "grad_norm": 0.7981966097669712, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12705 + }, + { + "epoch": 0.12706, + "grad_norm": 0.8791804171501783, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 12706 + }, + { + "epoch": 0.12707, + "grad_norm": 0.8751843147896976, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 12707 + }, + { + "epoch": 0.12708, + "grad_norm": 0.7357704648139376, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 12708 + }, + { + "epoch": 0.12709, + "grad_norm": 0.6951578598771816, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 12709 + }, + { + "epoch": 0.1271, + "grad_norm": 0.613246637185972, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 12710 + }, + { + "epoch": 0.12711, + "grad_norm": 0.59015611007869, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 12711 + }, + { + "epoch": 0.12712, + "grad_norm": 0.5852118964376581, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 12712 + }, + { + "epoch": 0.12713, + "grad_norm": 0.663170739181278, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 12713 + }, + { + "epoch": 0.12714, + "grad_norm": 0.6353079740380254, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 12714 + }, + { + "epoch": 0.12715, + "grad_norm": 0.6271102341426709, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 12715 + }, + { + "epoch": 0.12716, + "grad_norm": 0.6433296318540345, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 12716 + }, + { + "epoch": 0.12717, + "grad_norm": 0.6303506632088267, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12717 + }, + { + "epoch": 0.12718, + "grad_norm": 0.6211508260318281, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 12718 + }, + { + "epoch": 0.12719, + "grad_norm": 0.6528343828026612, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 12719 + }, + { + "epoch": 0.1272, + "grad_norm": 0.639082757581153, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 12720 + }, + { + "epoch": 0.12721, + "grad_norm": 0.669651264570942, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 12721 + }, + { + "epoch": 0.12722, + "grad_norm": 0.6622040289619314, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12722 + }, + { + "epoch": 0.12723, + "grad_norm": 0.642878142188778, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 12723 + }, + { + "epoch": 0.12724, + "grad_norm": 0.6145604007878287, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 12724 + }, + { + "epoch": 0.12725, + "grad_norm": 0.5440693624516679, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 12725 + }, + { + "epoch": 0.12726, + "grad_norm": 0.5746616551559309, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 12726 + }, + { + "epoch": 0.12727, + "grad_norm": 0.6458643593764009, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 12727 + }, + { + "epoch": 0.12728, + "grad_norm": 0.6485488933465353, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 12728 + }, + { + "epoch": 0.12729, + "grad_norm": 0.7344200132381935, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 12729 + }, + { + "epoch": 0.1273, + "grad_norm": 0.8539964531450729, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 12730 + }, + { + "epoch": 0.12731, + "grad_norm": 0.9030966671673113, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 12731 + }, + { + "epoch": 0.12732, + "grad_norm": 0.838764038230571, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 12732 + }, + { + "epoch": 0.12733, + "grad_norm": 0.8876353879143315, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 12733 + }, + { + "epoch": 0.12734, + "grad_norm": 0.8867284740798884, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 12734 + }, + { + "epoch": 0.12735, + "grad_norm": 0.8226392593998617, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 12735 + }, + { + "epoch": 0.12736, + "grad_norm": 0.801772278063402, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 12736 + }, + { + "epoch": 0.12737, + "grad_norm": 0.830066137163045, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 12737 + }, + { + "epoch": 0.12738, + "grad_norm": 0.9367116476018525, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 12738 + }, + { + "epoch": 0.12739, + "grad_norm": 0.9649378760880264, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 12739 + }, + { + "epoch": 0.1274, + "grad_norm": 0.7766361214920136, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 12740 + }, + { + "epoch": 0.12741, + "grad_norm": 0.7393141848306055, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 12741 + }, + { + "epoch": 0.12742, + "grad_norm": 0.7449959122649397, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 12742 + }, + { + "epoch": 0.12743, + "grad_norm": 0.750916357322704, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 12743 + }, + { + "epoch": 0.12744, + "grad_norm": 0.8140116290990219, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 12744 + }, + { + "epoch": 0.12745, + "grad_norm": 1.0062345155438677, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 12745 + }, + { + "epoch": 0.12746, + "grad_norm": 1.026672366367907, + "learning_rate": 0.003, + "loss": 4.0677, + "step": 12746 + }, + { + "epoch": 0.12747, + "grad_norm": 0.9282474901807517, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 12747 + }, + { + "epoch": 0.12748, + "grad_norm": 0.886912636801055, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 12748 + }, + { + "epoch": 0.12749, + "grad_norm": 0.812533773345398, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 12749 + }, + { + "epoch": 0.1275, + "grad_norm": 0.709926806084665, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 12750 + }, + { + "epoch": 0.12751, + "grad_norm": 0.8018432475485396, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 12751 + }, + { + "epoch": 0.12752, + "grad_norm": 0.7432747162665548, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 12752 + }, + { + "epoch": 0.12753, + "grad_norm": 0.6634584232213494, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 12753 + }, + { + "epoch": 0.12754, + "grad_norm": 0.679858724324728, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 12754 + }, + { + "epoch": 0.12755, + "grad_norm": 0.6298249575343915, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 12755 + }, + { + "epoch": 0.12756, + "grad_norm": 0.5236454872388505, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 12756 + }, + { + "epoch": 0.12757, + "grad_norm": 0.5902992519693165, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 12757 + }, + { + "epoch": 0.12758, + "grad_norm": 0.6588957361164748, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 12758 + }, + { + "epoch": 0.12759, + "grad_norm": 0.7552125222490269, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 12759 + }, + { + "epoch": 0.1276, + "grad_norm": 0.8456259236454896, + "learning_rate": 0.003, + "loss": 4.039, + "step": 12760 + }, + { + "epoch": 0.12761, + "grad_norm": 0.7585831804193502, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 12761 + }, + { + "epoch": 0.12762, + "grad_norm": 0.6343119640050029, + "learning_rate": 0.003, + "loss": 3.992, + "step": 12762 + }, + { + "epoch": 0.12763, + "grad_norm": 0.6254420607399649, + "learning_rate": 0.003, + "loss": 3.985, + "step": 12763 + }, + { + "epoch": 0.12764, + "grad_norm": 0.5606964679661799, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 12764 + }, + { + "epoch": 0.12765, + "grad_norm": 0.6041953608776169, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 12765 + }, + { + "epoch": 0.12766, + "grad_norm": 0.7041023223499858, + "learning_rate": 0.003, + "loss": 4.03, + "step": 12766 + }, + { + "epoch": 0.12767, + "grad_norm": 0.8286336158906576, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 12767 + }, + { + "epoch": 0.12768, + "grad_norm": 0.9727560884574205, + "learning_rate": 0.003, + "loss": 4.047, + "step": 12768 + }, + { + "epoch": 0.12769, + "grad_norm": 1.03739254830677, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 12769 + }, + { + "epoch": 0.1277, + "grad_norm": 0.8863843020624801, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 12770 + }, + { + "epoch": 0.12771, + "grad_norm": 0.5544440442155633, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 12771 + }, + { + "epoch": 0.12772, + "grad_norm": 0.5132371144264126, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 12772 + }, + { + "epoch": 0.12773, + "grad_norm": 0.5484142352484586, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 12773 + }, + { + "epoch": 0.12774, + "grad_norm": 0.5817972623489164, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 12774 + }, + { + "epoch": 0.12775, + "grad_norm": 0.5763681600349856, + "learning_rate": 0.003, + "loss": 3.993, + "step": 12775 + }, + { + "epoch": 0.12776, + "grad_norm": 0.48588830975024416, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 12776 + }, + { + "epoch": 0.12777, + "grad_norm": 0.4668867465556421, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 12777 + }, + { + "epoch": 0.12778, + "grad_norm": 0.5008443295526123, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 12778 + }, + { + "epoch": 0.12779, + "grad_norm": 0.449214506898142, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 12779 + }, + { + "epoch": 0.1278, + "grad_norm": 0.4415573274989028, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 12780 + }, + { + "epoch": 0.12781, + "grad_norm": 0.4496163535869851, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 12781 + }, + { + "epoch": 0.12782, + "grad_norm": 0.5056258611454915, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 12782 + }, + { + "epoch": 0.12783, + "grad_norm": 0.5541103062913432, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 12783 + }, + { + "epoch": 0.12784, + "grad_norm": 0.6045442986937035, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 12784 + }, + { + "epoch": 0.12785, + "grad_norm": 0.5510028371799175, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 12785 + }, + { + "epoch": 0.12786, + "grad_norm": 0.5771857177843766, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 12786 + }, + { + "epoch": 0.12787, + "grad_norm": 0.6268299410969302, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 12787 + }, + { + "epoch": 0.12788, + "grad_norm": 0.6621076775357674, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 12788 + }, + { + "epoch": 0.12789, + "grad_norm": 0.7446507435005899, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 12789 + }, + { + "epoch": 0.1279, + "grad_norm": 0.7808997262573284, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 12790 + }, + { + "epoch": 0.12791, + "grad_norm": 0.7734341945645656, + "learning_rate": 0.003, + "loss": 4.019, + "step": 12791 + }, + { + "epoch": 0.12792, + "grad_norm": 0.7980004993198647, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 12792 + }, + { + "epoch": 0.12793, + "grad_norm": 0.7973649119592662, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 12793 + }, + { + "epoch": 0.12794, + "grad_norm": 0.7078338457446883, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 12794 + }, + { + "epoch": 0.12795, + "grad_norm": 0.7335016168540234, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 12795 + }, + { + "epoch": 0.12796, + "grad_norm": 0.8078914763078121, + "learning_rate": 0.003, + "loss": 3.995, + "step": 12796 + }, + { + "epoch": 0.12797, + "grad_norm": 0.9366536023811256, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 12797 + }, + { + "epoch": 0.12798, + "grad_norm": 1.040605397206868, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 12798 + }, + { + "epoch": 0.12799, + "grad_norm": 0.9310011089041549, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 12799 + }, + { + "epoch": 0.128, + "grad_norm": 0.9608225192218678, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 12800 + }, + { + "epoch": 0.12801, + "grad_norm": 0.9147218507209992, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 12801 + }, + { + "epoch": 0.12802, + "grad_norm": 0.7633068378413946, + "learning_rate": 0.003, + "loss": 4.0691, + "step": 12802 + }, + { + "epoch": 0.12803, + "grad_norm": 0.6940080940415748, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 12803 + }, + { + "epoch": 0.12804, + "grad_norm": 0.690025120844021, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 12804 + }, + { + "epoch": 0.12805, + "grad_norm": 0.8083175609012657, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 12805 + }, + { + "epoch": 0.12806, + "grad_norm": 0.7816407579127554, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 12806 + }, + { + "epoch": 0.12807, + "grad_norm": 0.732535572697184, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 12807 + }, + { + "epoch": 0.12808, + "grad_norm": 0.6169788129004569, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 12808 + }, + { + "epoch": 0.12809, + "grad_norm": 0.5862401793492781, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 12809 + }, + { + "epoch": 0.1281, + "grad_norm": 0.5018038154517552, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 12810 + }, + { + "epoch": 0.12811, + "grad_norm": 0.5837106340862616, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 12811 + }, + { + "epoch": 0.12812, + "grad_norm": 0.6033913724152268, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 12812 + }, + { + "epoch": 0.12813, + "grad_norm": 0.6727379959751584, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 12813 + }, + { + "epoch": 0.12814, + "grad_norm": 0.7526295780583007, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 12814 + }, + { + "epoch": 0.12815, + "grad_norm": 0.7411533980794333, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 12815 + }, + { + "epoch": 0.12816, + "grad_norm": 0.7359784318343954, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 12816 + }, + { + "epoch": 0.12817, + "grad_norm": 0.7649852816281582, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12817 + }, + { + "epoch": 0.12818, + "grad_norm": 0.7462998008644043, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 12818 + }, + { + "epoch": 0.12819, + "grad_norm": 0.6948293927589271, + "learning_rate": 0.003, + "loss": 4.029, + "step": 12819 + }, + { + "epoch": 0.1282, + "grad_norm": 0.7196694348611684, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 12820 + }, + { + "epoch": 0.12821, + "grad_norm": 0.7251667993623546, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 12821 + }, + { + "epoch": 0.12822, + "grad_norm": 0.7094570561743686, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 12822 + }, + { + "epoch": 0.12823, + "grad_norm": 0.6800985623775413, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 12823 + }, + { + "epoch": 0.12824, + "grad_norm": 0.6067574811778009, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 12824 + }, + { + "epoch": 0.12825, + "grad_norm": 0.6629286589187029, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 12825 + }, + { + "epoch": 0.12826, + "grad_norm": 0.7315297853737511, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 12826 + }, + { + "epoch": 0.12827, + "grad_norm": 0.691127523732906, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 12827 + }, + { + "epoch": 0.12828, + "grad_norm": 0.5719891067285455, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 12828 + }, + { + "epoch": 0.12829, + "grad_norm": 0.7001829126532528, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 12829 + }, + { + "epoch": 0.1283, + "grad_norm": 0.7322265856819983, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 12830 + }, + { + "epoch": 0.12831, + "grad_norm": 0.7101660547850247, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12831 + }, + { + "epoch": 0.12832, + "grad_norm": 0.7217242755328721, + "learning_rate": 0.003, + "loss": 4.016, + "step": 12832 + }, + { + "epoch": 0.12833, + "grad_norm": 0.6704298715919287, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 12833 + }, + { + "epoch": 0.12834, + "grad_norm": 0.7191903431362907, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 12834 + }, + { + "epoch": 0.12835, + "grad_norm": 0.8696372923248408, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 12835 + }, + { + "epoch": 0.12836, + "grad_norm": 1.0443626397943426, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 12836 + }, + { + "epoch": 0.12837, + "grad_norm": 0.932807183264994, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 12837 + }, + { + "epoch": 0.12838, + "grad_norm": 0.9084016711408195, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 12838 + }, + { + "epoch": 0.12839, + "grad_norm": 0.9552200364104015, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 12839 + }, + { + "epoch": 0.1284, + "grad_norm": 0.8650822669984716, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 12840 + }, + { + "epoch": 0.12841, + "grad_norm": 0.7954972964821377, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 12841 + }, + { + "epoch": 0.12842, + "grad_norm": 0.9640268483215286, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 12842 + }, + { + "epoch": 0.12843, + "grad_norm": 1.1693071544943237, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 12843 + }, + { + "epoch": 0.12844, + "grad_norm": 0.7041457068515213, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 12844 + }, + { + "epoch": 0.12845, + "grad_norm": 0.6443959415652293, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 12845 + }, + { + "epoch": 0.12846, + "grad_norm": 0.6333745837098246, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 12846 + }, + { + "epoch": 0.12847, + "grad_norm": 0.6679204415701047, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 12847 + }, + { + "epoch": 0.12848, + "grad_norm": 0.7998991513133767, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 12848 + }, + { + "epoch": 0.12849, + "grad_norm": 0.8557642979220237, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 12849 + }, + { + "epoch": 0.1285, + "grad_norm": 0.9063045562951076, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 12850 + }, + { + "epoch": 0.12851, + "grad_norm": 0.8103588037736258, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 12851 + }, + { + "epoch": 0.12852, + "grad_norm": 0.6967453772989425, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 12852 + }, + { + "epoch": 0.12853, + "grad_norm": 0.6504583538926301, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 12853 + }, + { + "epoch": 0.12854, + "grad_norm": 0.5851143454815028, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 12854 + }, + { + "epoch": 0.12855, + "grad_norm": 0.5293135387759583, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 12855 + }, + { + "epoch": 0.12856, + "grad_norm": 0.5218290788309927, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 12856 + }, + { + "epoch": 0.12857, + "grad_norm": 0.5099112209497075, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 12857 + }, + { + "epoch": 0.12858, + "grad_norm": 0.5045743127718421, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 12858 + }, + { + "epoch": 0.12859, + "grad_norm": 0.46583717494716537, + "learning_rate": 0.003, + "loss": 3.993, + "step": 12859 + }, + { + "epoch": 0.1286, + "grad_norm": 0.46527605569407277, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 12860 + }, + { + "epoch": 0.12861, + "grad_norm": 0.45878459539391453, + "learning_rate": 0.003, + "loss": 4.017, + "step": 12861 + }, + { + "epoch": 0.12862, + "grad_norm": 0.45451856339368796, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 12862 + }, + { + "epoch": 0.12863, + "grad_norm": 0.4787296313867782, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12863 + }, + { + "epoch": 0.12864, + "grad_norm": 0.5054658473021199, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 12864 + }, + { + "epoch": 0.12865, + "grad_norm": 0.49881891050676785, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 12865 + }, + { + "epoch": 0.12866, + "grad_norm": 0.5100686054895698, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 12866 + }, + { + "epoch": 0.12867, + "grad_norm": 0.6349821303139059, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 12867 + }, + { + "epoch": 0.12868, + "grad_norm": 0.8448600416196349, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 12868 + }, + { + "epoch": 0.12869, + "grad_norm": 1.0477845884556118, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 12869 + }, + { + "epoch": 0.1287, + "grad_norm": 0.8500576366326791, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 12870 + }, + { + "epoch": 0.12871, + "grad_norm": 0.6694868717011907, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 12871 + }, + { + "epoch": 0.12872, + "grad_norm": 0.7951056334079175, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 12872 + }, + { + "epoch": 0.12873, + "grad_norm": 0.8168633266695264, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 12873 + }, + { + "epoch": 0.12874, + "grad_norm": 0.7063518796710858, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 12874 + }, + { + "epoch": 0.12875, + "grad_norm": 0.7669538568379751, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 12875 + }, + { + "epoch": 0.12876, + "grad_norm": 0.8045411430061095, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 12876 + }, + { + "epoch": 0.12877, + "grad_norm": 0.6956879426522765, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 12877 + }, + { + "epoch": 0.12878, + "grad_norm": 0.769859696428875, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 12878 + }, + { + "epoch": 0.12879, + "grad_norm": 0.788028259283368, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 12879 + }, + { + "epoch": 0.1288, + "grad_norm": 0.8964095738125816, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 12880 + }, + { + "epoch": 0.12881, + "grad_norm": 1.0597377922399067, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 12881 + }, + { + "epoch": 0.12882, + "grad_norm": 0.9001596637121585, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 12882 + }, + { + "epoch": 0.12883, + "grad_norm": 0.9262507403679164, + "learning_rate": 0.003, + "loss": 4.045, + "step": 12883 + }, + { + "epoch": 0.12884, + "grad_norm": 0.7753934127551037, + "learning_rate": 0.003, + "loss": 4.04, + "step": 12884 + }, + { + "epoch": 0.12885, + "grad_norm": 0.7950755125041752, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 12885 + }, + { + "epoch": 0.12886, + "grad_norm": 0.7868687412827193, + "learning_rate": 0.003, + "loss": 4.031, + "step": 12886 + }, + { + "epoch": 0.12887, + "grad_norm": 0.6721665053724442, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 12887 + }, + { + "epoch": 0.12888, + "grad_norm": 0.751929022709998, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 12888 + }, + { + "epoch": 0.12889, + "grad_norm": 0.8250259541816732, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 12889 + }, + { + "epoch": 0.1289, + "grad_norm": 0.9785803264767253, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 12890 + }, + { + "epoch": 0.12891, + "grad_norm": 0.9466905426891853, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 12891 + }, + { + "epoch": 0.12892, + "grad_norm": 0.8754686781287974, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 12892 + }, + { + "epoch": 0.12893, + "grad_norm": 0.897530915644138, + "learning_rate": 0.003, + "loss": 4.0812, + "step": 12893 + }, + { + "epoch": 0.12894, + "grad_norm": 0.92005308323187, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 12894 + }, + { + "epoch": 0.12895, + "grad_norm": 0.8274582389676985, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 12895 + }, + { + "epoch": 0.12896, + "grad_norm": 0.7676090662900501, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12896 + }, + { + "epoch": 0.12897, + "grad_norm": 0.6578994618579546, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 12897 + }, + { + "epoch": 0.12898, + "grad_norm": 0.6199427583597847, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 12898 + }, + { + "epoch": 0.12899, + "grad_norm": 0.5768600245219924, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 12899 + }, + { + "epoch": 0.129, + "grad_norm": 0.5352569296902703, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 12900 + }, + { + "epoch": 0.12901, + "grad_norm": 0.5748119055592369, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 12901 + }, + { + "epoch": 0.12902, + "grad_norm": 0.5643088716982332, + "learning_rate": 0.003, + "loss": 4.022, + "step": 12902 + }, + { + "epoch": 0.12903, + "grad_norm": 0.6296497936746063, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 12903 + }, + { + "epoch": 0.12904, + "grad_norm": 0.6860482670781219, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 12904 + }, + { + "epoch": 0.12905, + "grad_norm": 0.77604738127635, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 12905 + }, + { + "epoch": 0.12906, + "grad_norm": 0.7109518764833995, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12906 + }, + { + "epoch": 0.12907, + "grad_norm": 0.7370479588780612, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 12907 + }, + { + "epoch": 0.12908, + "grad_norm": 0.8433112351788341, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 12908 + }, + { + "epoch": 0.12909, + "grad_norm": 0.8918979595823009, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 12909 + }, + { + "epoch": 0.1291, + "grad_norm": 0.8316680849682383, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 12910 + }, + { + "epoch": 0.12911, + "grad_norm": 0.7406901188702144, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 12911 + }, + { + "epoch": 0.12912, + "grad_norm": 0.6281680548589542, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 12912 + }, + { + "epoch": 0.12913, + "grad_norm": 0.6144792657152653, + "learning_rate": 0.003, + "loss": 4.024, + "step": 12913 + }, + { + "epoch": 0.12914, + "grad_norm": 0.6271154854609691, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 12914 + }, + { + "epoch": 0.12915, + "grad_norm": 0.749070425204854, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 12915 + }, + { + "epoch": 0.12916, + "grad_norm": 0.8096024127514485, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 12916 + }, + { + "epoch": 0.12917, + "grad_norm": 0.8108119758898933, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 12917 + }, + { + "epoch": 0.12918, + "grad_norm": 0.7913998307170463, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 12918 + }, + { + "epoch": 0.12919, + "grad_norm": 0.7281020872279017, + "learning_rate": 0.003, + "loss": 4.023, + "step": 12919 + }, + { + "epoch": 0.1292, + "grad_norm": 0.6904685049692444, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 12920 + }, + { + "epoch": 0.12921, + "grad_norm": 0.6258500837117931, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 12921 + }, + { + "epoch": 0.12922, + "grad_norm": 0.6369091133069588, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 12922 + }, + { + "epoch": 0.12923, + "grad_norm": 0.620983855088843, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 12923 + }, + { + "epoch": 0.12924, + "grad_norm": 0.6804343872329571, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 12924 + }, + { + "epoch": 0.12925, + "grad_norm": 0.7911673663836201, + "learning_rate": 0.003, + "loss": 4.021, + "step": 12925 + }, + { + "epoch": 0.12926, + "grad_norm": 0.7986118800599354, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 12926 + }, + { + "epoch": 0.12927, + "grad_norm": 0.8189989446727631, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 12927 + }, + { + "epoch": 0.12928, + "grad_norm": 0.8484408694803683, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 12928 + }, + { + "epoch": 0.12929, + "grad_norm": 0.7752746099420343, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 12929 + }, + { + "epoch": 0.1293, + "grad_norm": 0.7097282619607437, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 12930 + }, + { + "epoch": 0.12931, + "grad_norm": 0.6105403853678485, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 12931 + }, + { + "epoch": 0.12932, + "grad_norm": 0.4812300907818467, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 12932 + }, + { + "epoch": 0.12933, + "grad_norm": 0.45662882048579106, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 12933 + }, + { + "epoch": 0.12934, + "grad_norm": 0.4753744168040995, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 12934 + }, + { + "epoch": 0.12935, + "grad_norm": 0.5380947340058946, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 12935 + }, + { + "epoch": 0.12936, + "grad_norm": 0.5447943118408797, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 12936 + }, + { + "epoch": 0.12937, + "grad_norm": 0.5618616756094406, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 12937 + }, + { + "epoch": 0.12938, + "grad_norm": 0.5933735594695718, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 12938 + }, + { + "epoch": 0.12939, + "grad_norm": 0.6853900464567859, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 12939 + }, + { + "epoch": 0.1294, + "grad_norm": 0.7445650286991727, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 12940 + }, + { + "epoch": 0.12941, + "grad_norm": 0.8276182621148829, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 12941 + }, + { + "epoch": 0.12942, + "grad_norm": 1.0094312356499089, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 12942 + }, + { + "epoch": 0.12943, + "grad_norm": 0.9951504843932714, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 12943 + }, + { + "epoch": 0.12944, + "grad_norm": 0.8156155584131042, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 12944 + }, + { + "epoch": 0.12945, + "grad_norm": 0.8779581741629432, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 12945 + }, + { + "epoch": 0.12946, + "grad_norm": 0.8332588011162689, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 12946 + }, + { + "epoch": 0.12947, + "grad_norm": 0.6712892365626068, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 12947 + }, + { + "epoch": 0.12948, + "grad_norm": 0.6837094921303175, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 12948 + }, + { + "epoch": 0.12949, + "grad_norm": 0.6010588655556556, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 12949 + }, + { + "epoch": 0.1295, + "grad_norm": 0.6147569500556734, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 12950 + }, + { + "epoch": 0.12951, + "grad_norm": 0.6028397234904259, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 12951 + }, + { + "epoch": 0.12952, + "grad_norm": 0.5472564757175814, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 12952 + }, + { + "epoch": 0.12953, + "grad_norm": 0.46602957108127224, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 12953 + }, + { + "epoch": 0.12954, + "grad_norm": 0.5623451614720343, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 12954 + }, + { + "epoch": 0.12955, + "grad_norm": 0.6807037706662017, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 12955 + }, + { + "epoch": 0.12956, + "grad_norm": 0.8320372672413828, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 12956 + }, + { + "epoch": 0.12957, + "grad_norm": 1.0452134141171572, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 12957 + }, + { + "epoch": 0.12958, + "grad_norm": 0.9319508190829566, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 12958 + }, + { + "epoch": 0.12959, + "grad_norm": 0.7497749540838273, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 12959 + }, + { + "epoch": 0.1296, + "grad_norm": 0.6836170369099951, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 12960 + }, + { + "epoch": 0.12961, + "grad_norm": 0.7830904326363355, + "learning_rate": 0.003, + "loss": 4.051, + "step": 12961 + }, + { + "epoch": 0.12962, + "grad_norm": 0.7781444488572186, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 12962 + }, + { + "epoch": 0.12963, + "grad_norm": 0.6446418924943585, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 12963 + }, + { + "epoch": 0.12964, + "grad_norm": 0.6289628706455141, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 12964 + }, + { + "epoch": 0.12965, + "grad_norm": 0.7244716278361892, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 12965 + }, + { + "epoch": 0.12966, + "grad_norm": 0.7414833603201494, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 12966 + }, + { + "epoch": 0.12967, + "grad_norm": 0.7087840728123121, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 12967 + }, + { + "epoch": 0.12968, + "grad_norm": 0.6341303530418906, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12968 + }, + { + "epoch": 0.12969, + "grad_norm": 0.6094846626513098, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 12969 + }, + { + "epoch": 0.1297, + "grad_norm": 0.6089749156451588, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 12970 + }, + { + "epoch": 0.12971, + "grad_norm": 0.6544196653569223, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 12971 + }, + { + "epoch": 0.12972, + "grad_norm": 0.7204901657240181, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 12972 + }, + { + "epoch": 0.12973, + "grad_norm": 0.7594375019086355, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 12973 + }, + { + "epoch": 0.12974, + "grad_norm": 0.8099546933515243, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 12974 + }, + { + "epoch": 0.12975, + "grad_norm": 0.7457068257157814, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 12975 + }, + { + "epoch": 0.12976, + "grad_norm": 0.6189319316508349, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 12976 + }, + { + "epoch": 0.12977, + "grad_norm": 0.6292336046391308, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 12977 + }, + { + "epoch": 0.12978, + "grad_norm": 0.7248310219716132, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 12978 + }, + { + "epoch": 0.12979, + "grad_norm": 0.8632308818012894, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 12979 + }, + { + "epoch": 0.1298, + "grad_norm": 0.9691297764145118, + "learning_rate": 0.003, + "loss": 4.006, + "step": 12980 + }, + { + "epoch": 0.12981, + "grad_norm": 0.9256216114442899, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 12981 + }, + { + "epoch": 0.12982, + "grad_norm": 0.8077111408738972, + "learning_rate": 0.003, + "loss": 4.0573, + "step": 12982 + }, + { + "epoch": 0.12983, + "grad_norm": 0.7588444972149183, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 12983 + }, + { + "epoch": 0.12984, + "grad_norm": 0.715971792073416, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 12984 + }, + { + "epoch": 0.12985, + "grad_norm": 0.6979248581354024, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 12985 + }, + { + "epoch": 0.12986, + "grad_norm": 0.7219755527767605, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 12986 + }, + { + "epoch": 0.12987, + "grad_norm": 0.7288867907686327, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 12987 + }, + { + "epoch": 0.12988, + "grad_norm": 0.6319448696261232, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 12988 + }, + { + "epoch": 0.12989, + "grad_norm": 0.6356679049391042, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 12989 + }, + { + "epoch": 0.1299, + "grad_norm": 0.6663574077595541, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 12990 + }, + { + "epoch": 0.12991, + "grad_norm": 0.590958629217494, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 12991 + }, + { + "epoch": 0.12992, + "grad_norm": 0.5522225810400011, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 12992 + }, + { + "epoch": 0.12993, + "grad_norm": 0.5053439384810064, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 12993 + }, + { + "epoch": 0.12994, + "grad_norm": 0.5561026954827961, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 12994 + }, + { + "epoch": 0.12995, + "grad_norm": 0.6889948595233923, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 12995 + }, + { + "epoch": 0.12996, + "grad_norm": 0.893143802297226, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 12996 + }, + { + "epoch": 0.12997, + "grad_norm": 1.0231962861020802, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 12997 + }, + { + "epoch": 0.12998, + "grad_norm": 0.7678060044272385, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 12998 + }, + { + "epoch": 0.12999, + "grad_norm": 0.5786270014604347, + "learning_rate": 0.003, + "loss": 3.991, + "step": 12999 + }, + { + "epoch": 0.13, + "grad_norm": 0.7109288076825049, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13000 + }, + { + "epoch": 0.13001, + "grad_norm": 0.6981607360784682, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 13001 + }, + { + "epoch": 0.13002, + "grad_norm": 0.7299159917732377, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 13002 + }, + { + "epoch": 0.13003, + "grad_norm": 0.7263194516836214, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 13003 + }, + { + "epoch": 0.13004, + "grad_norm": 0.7929285938158723, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 13004 + }, + { + "epoch": 0.13005, + "grad_norm": 0.8922524747107461, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 13005 + }, + { + "epoch": 0.13006, + "grad_norm": 1.084291677040571, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 13006 + }, + { + "epoch": 0.13007, + "grad_norm": 0.8910252868405993, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 13007 + }, + { + "epoch": 0.13008, + "grad_norm": 0.7728435026649844, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 13008 + }, + { + "epoch": 0.13009, + "grad_norm": 0.7352109097475028, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 13009 + }, + { + "epoch": 0.1301, + "grad_norm": 0.7793337645993438, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 13010 + }, + { + "epoch": 0.13011, + "grad_norm": 0.7371021048354124, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 13011 + }, + { + "epoch": 0.13012, + "grad_norm": 0.7648052794965128, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 13012 + }, + { + "epoch": 0.13013, + "grad_norm": 0.7348212081349208, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 13013 + }, + { + "epoch": 0.13014, + "grad_norm": 0.7143209874122624, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 13014 + }, + { + "epoch": 0.13015, + "grad_norm": 0.7744834049983579, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 13015 + }, + { + "epoch": 0.13016, + "grad_norm": 0.6376247554123774, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 13016 + }, + { + "epoch": 0.13017, + "grad_norm": 0.5174179176907554, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 13017 + }, + { + "epoch": 0.13018, + "grad_norm": 0.46350508185315903, + "learning_rate": 0.003, + "loss": 4.004, + "step": 13018 + }, + { + "epoch": 0.13019, + "grad_norm": 0.4640691105491586, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 13019 + }, + { + "epoch": 0.1302, + "grad_norm": 0.47773050487746255, + "learning_rate": 0.003, + "loss": 4.009, + "step": 13020 + }, + { + "epoch": 0.13021, + "grad_norm": 0.5148979356286048, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13021 + }, + { + "epoch": 0.13022, + "grad_norm": 0.4596946829929602, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 13022 + }, + { + "epoch": 0.13023, + "grad_norm": 0.47910884540200954, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 13023 + }, + { + "epoch": 0.13024, + "grad_norm": 0.5705935774693228, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 13024 + }, + { + "epoch": 0.13025, + "grad_norm": 0.6405142429842138, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 13025 + }, + { + "epoch": 0.13026, + "grad_norm": 0.6782919873557559, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 13026 + }, + { + "epoch": 0.13027, + "grad_norm": 0.7507059604204482, + "learning_rate": 0.003, + "loss": 3.996, + "step": 13027 + }, + { + "epoch": 0.13028, + "grad_norm": 0.8091092808357904, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 13028 + }, + { + "epoch": 0.13029, + "grad_norm": 0.8803446793996493, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 13029 + }, + { + "epoch": 0.1303, + "grad_norm": 0.8267049526699398, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 13030 + }, + { + "epoch": 0.13031, + "grad_norm": 0.7447051373492657, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 13031 + }, + { + "epoch": 0.13032, + "grad_norm": 0.8129181787567167, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 13032 + }, + { + "epoch": 0.13033, + "grad_norm": 0.7020540661236004, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 13033 + }, + { + "epoch": 0.13034, + "grad_norm": 0.7884444190331344, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 13034 + }, + { + "epoch": 0.13035, + "grad_norm": 0.7902193605785285, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 13035 + }, + { + "epoch": 0.13036, + "grad_norm": 0.9531072890771126, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 13036 + }, + { + "epoch": 0.13037, + "grad_norm": 0.897961626336336, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 13037 + }, + { + "epoch": 0.13038, + "grad_norm": 0.9404628909938803, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 13038 + }, + { + "epoch": 0.13039, + "grad_norm": 0.8469480770018044, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 13039 + }, + { + "epoch": 0.1304, + "grad_norm": 0.9311320611463285, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13040 + }, + { + "epoch": 0.13041, + "grad_norm": 1.050907117715992, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 13041 + }, + { + "epoch": 0.13042, + "grad_norm": 1.0210476512986348, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 13042 + }, + { + "epoch": 0.13043, + "grad_norm": 0.9245178202874119, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 13043 + }, + { + "epoch": 0.13044, + "grad_norm": 0.9650391443399822, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 13044 + }, + { + "epoch": 0.13045, + "grad_norm": 0.881787588909884, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13045 + }, + { + "epoch": 0.13046, + "grad_norm": 0.9420113240302745, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 13046 + }, + { + "epoch": 0.13047, + "grad_norm": 1.1036402694748901, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 13047 + }, + { + "epoch": 0.13048, + "grad_norm": 1.0083079989563326, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 13048 + }, + { + "epoch": 0.13049, + "grad_norm": 0.9473636011883718, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 13049 + }, + { + "epoch": 0.1305, + "grad_norm": 0.9246437454053951, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 13050 + }, + { + "epoch": 0.13051, + "grad_norm": 0.8912866059426476, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 13051 + }, + { + "epoch": 0.13052, + "grad_norm": 0.8052280461033229, + "learning_rate": 0.003, + "loss": 4.028, + "step": 13052 + }, + { + "epoch": 0.13053, + "grad_norm": 0.6220823003884444, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13053 + }, + { + "epoch": 0.13054, + "grad_norm": 0.5830405690523822, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 13054 + }, + { + "epoch": 0.13055, + "grad_norm": 0.5682682258781175, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 13055 + }, + { + "epoch": 0.13056, + "grad_norm": 0.5678453114946214, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 13056 + }, + { + "epoch": 0.13057, + "grad_norm": 0.5391041765571322, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 13057 + }, + { + "epoch": 0.13058, + "grad_norm": 0.500493706084591, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 13058 + }, + { + "epoch": 0.13059, + "grad_norm": 0.5744826884808235, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 13059 + }, + { + "epoch": 0.1306, + "grad_norm": 0.6195945339173821, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 13060 + }, + { + "epoch": 0.13061, + "grad_norm": 0.6418430537317352, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 13061 + }, + { + "epoch": 0.13062, + "grad_norm": 0.610866384083531, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13062 + }, + { + "epoch": 0.13063, + "grad_norm": 0.5195922433496627, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 13063 + }, + { + "epoch": 0.13064, + "grad_norm": 0.4408966556288092, + "learning_rate": 0.003, + "loss": 4.049, + "step": 13064 + }, + { + "epoch": 0.13065, + "grad_norm": 0.46002485098401047, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 13065 + }, + { + "epoch": 0.13066, + "grad_norm": 0.49239513778110705, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 13066 + }, + { + "epoch": 0.13067, + "grad_norm": 0.5719396994317322, + "learning_rate": 0.003, + "loss": 4.062, + "step": 13067 + }, + { + "epoch": 0.13068, + "grad_norm": 0.6312822934977946, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 13068 + }, + { + "epoch": 0.13069, + "grad_norm": 0.6341401260342081, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 13069 + }, + { + "epoch": 0.1307, + "grad_norm": 0.6262110755991823, + "learning_rate": 0.003, + "loss": 4.022, + "step": 13070 + }, + { + "epoch": 0.13071, + "grad_norm": 0.6224584909025384, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 13071 + }, + { + "epoch": 0.13072, + "grad_norm": 0.5687706861664837, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 13072 + }, + { + "epoch": 0.13073, + "grad_norm": 0.5155985121267158, + "learning_rate": 0.003, + "loss": 4.0, + "step": 13073 + }, + { + "epoch": 0.13074, + "grad_norm": 0.5891973540208136, + "learning_rate": 0.003, + "loss": 3.995, + "step": 13074 + }, + { + "epoch": 0.13075, + "grad_norm": 0.5864304535286806, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13075 + }, + { + "epoch": 0.13076, + "grad_norm": 0.6104923637808817, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 13076 + }, + { + "epoch": 0.13077, + "grad_norm": 0.6409796150961073, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 13077 + }, + { + "epoch": 0.13078, + "grad_norm": 0.6567824293157022, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 13078 + }, + { + "epoch": 0.13079, + "grad_norm": 0.6122456375270453, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 13079 + }, + { + "epoch": 0.1308, + "grad_norm": 0.624629791967273, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 13080 + }, + { + "epoch": 0.13081, + "grad_norm": 0.6573815040481957, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 13081 + }, + { + "epoch": 0.13082, + "grad_norm": 0.7914492823483801, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 13082 + }, + { + "epoch": 0.13083, + "grad_norm": 0.999688600027972, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13083 + }, + { + "epoch": 0.13084, + "grad_norm": 1.2978652026852164, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 13084 + }, + { + "epoch": 0.13085, + "grad_norm": 0.7954688971541304, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 13085 + }, + { + "epoch": 0.13086, + "grad_norm": 0.6171605723407216, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 13086 + }, + { + "epoch": 0.13087, + "grad_norm": 0.6531446045725782, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 13087 + }, + { + "epoch": 0.13088, + "grad_norm": 0.6463084315335952, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 13088 + }, + { + "epoch": 0.13089, + "grad_norm": 0.6422271113075354, + "learning_rate": 0.003, + "loss": 4.019, + "step": 13089 + }, + { + "epoch": 0.1309, + "grad_norm": 0.5503024836641156, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 13090 + }, + { + "epoch": 0.13091, + "grad_norm": 0.6042461273160867, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13091 + }, + { + "epoch": 0.13092, + "grad_norm": 0.564378693595355, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 13092 + }, + { + "epoch": 0.13093, + "grad_norm": 0.5688362630929851, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 13093 + }, + { + "epoch": 0.13094, + "grad_norm": 0.6351600903854989, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 13094 + }, + { + "epoch": 0.13095, + "grad_norm": 0.7615543626980247, + "learning_rate": 0.003, + "loss": 4.019, + "step": 13095 + }, + { + "epoch": 0.13096, + "grad_norm": 0.8080690872305026, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 13096 + }, + { + "epoch": 0.13097, + "grad_norm": 0.7690702529585823, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 13097 + }, + { + "epoch": 0.13098, + "grad_norm": 0.7317545662252383, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 13098 + }, + { + "epoch": 0.13099, + "grad_norm": 0.7412372962400838, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 13099 + }, + { + "epoch": 0.131, + "grad_norm": 0.8017974800322087, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 13100 + }, + { + "epoch": 0.13101, + "grad_norm": 0.7333187387906696, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 13101 + }, + { + "epoch": 0.13102, + "grad_norm": 0.7906645873869919, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 13102 + }, + { + "epoch": 0.13103, + "grad_norm": 0.8129707288050533, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 13103 + }, + { + "epoch": 0.13104, + "grad_norm": 0.7370998285868408, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 13104 + }, + { + "epoch": 0.13105, + "grad_norm": 0.7178769574459251, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 13105 + }, + { + "epoch": 0.13106, + "grad_norm": 0.8091688666006163, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 13106 + }, + { + "epoch": 0.13107, + "grad_norm": 0.9562098895817174, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13107 + }, + { + "epoch": 0.13108, + "grad_norm": 0.9017954427339477, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 13108 + }, + { + "epoch": 0.13109, + "grad_norm": 0.8169540256094443, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13109 + }, + { + "epoch": 0.1311, + "grad_norm": 0.8519155585272256, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 13110 + }, + { + "epoch": 0.13111, + "grad_norm": 0.7971099311137243, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 13111 + }, + { + "epoch": 0.13112, + "grad_norm": 0.7168145009591058, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 13112 + }, + { + "epoch": 0.13113, + "grad_norm": 0.7630449362098752, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 13113 + }, + { + "epoch": 0.13114, + "grad_norm": 0.8115685652496063, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 13114 + }, + { + "epoch": 0.13115, + "grad_norm": 0.8109768055856209, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 13115 + }, + { + "epoch": 0.13116, + "grad_norm": 0.7515666906400302, + "learning_rate": 0.003, + "loss": 4.026, + "step": 13116 + }, + { + "epoch": 0.13117, + "grad_norm": 0.7574139787811146, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 13117 + }, + { + "epoch": 0.13118, + "grad_norm": 0.7437710506862553, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 13118 + }, + { + "epoch": 0.13119, + "grad_norm": 0.769437318635099, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 13119 + }, + { + "epoch": 0.1312, + "grad_norm": 0.7690496791055725, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 13120 + }, + { + "epoch": 0.13121, + "grad_norm": 0.6916807674531394, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 13121 + }, + { + "epoch": 0.13122, + "grad_norm": 0.7754885464487861, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13122 + }, + { + "epoch": 0.13123, + "grad_norm": 0.6821521586496545, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 13123 + }, + { + "epoch": 0.13124, + "grad_norm": 0.537844351830531, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 13124 + }, + { + "epoch": 0.13125, + "grad_norm": 0.5157142027171415, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 13125 + }, + { + "epoch": 0.13126, + "grad_norm": 0.5747811378847046, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 13126 + }, + { + "epoch": 0.13127, + "grad_norm": 0.6143904821387517, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 13127 + }, + { + "epoch": 0.13128, + "grad_norm": 0.5499561899396731, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 13128 + }, + { + "epoch": 0.13129, + "grad_norm": 0.5261844499358456, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 13129 + }, + { + "epoch": 0.1313, + "grad_norm": 0.495898558570134, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 13130 + }, + { + "epoch": 0.13131, + "grad_norm": 0.47433482205822103, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 13131 + }, + { + "epoch": 0.13132, + "grad_norm": 0.5682622261576157, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 13132 + }, + { + "epoch": 0.13133, + "grad_norm": 0.6597037748684665, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 13133 + }, + { + "epoch": 0.13134, + "grad_norm": 0.7895367725805804, + "learning_rate": 0.003, + "loss": 4.004, + "step": 13134 + }, + { + "epoch": 0.13135, + "grad_norm": 0.8863968123250583, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 13135 + }, + { + "epoch": 0.13136, + "grad_norm": 0.9239751268063233, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 13136 + }, + { + "epoch": 0.13137, + "grad_norm": 1.0181584702480029, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 13137 + }, + { + "epoch": 0.13138, + "grad_norm": 0.8539942665969072, + "learning_rate": 0.003, + "loss": 4.008, + "step": 13138 + }, + { + "epoch": 0.13139, + "grad_norm": 0.7668750677508244, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 13139 + }, + { + "epoch": 0.1314, + "grad_norm": 0.8018144457087006, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 13140 + }, + { + "epoch": 0.13141, + "grad_norm": 0.878838473282242, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 13141 + }, + { + "epoch": 0.13142, + "grad_norm": 0.9076214251103577, + "learning_rate": 0.003, + "loss": 4.0611, + "step": 13142 + }, + { + "epoch": 0.13143, + "grad_norm": 0.8428378131867053, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 13143 + }, + { + "epoch": 0.13144, + "grad_norm": 0.8375042295980435, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 13144 + }, + { + "epoch": 0.13145, + "grad_norm": 0.9410719178263584, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 13145 + }, + { + "epoch": 0.13146, + "grad_norm": 0.8714867822178017, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 13146 + }, + { + "epoch": 0.13147, + "grad_norm": 0.912281486654637, + "learning_rate": 0.003, + "loss": 4.045, + "step": 13147 + }, + { + "epoch": 0.13148, + "grad_norm": 0.899690635061988, + "learning_rate": 0.003, + "loss": 4.055, + "step": 13148 + }, + { + "epoch": 0.13149, + "grad_norm": 0.8625502274575828, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 13149 + }, + { + "epoch": 0.1315, + "grad_norm": 0.7589010276317117, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 13150 + }, + { + "epoch": 0.13151, + "grad_norm": 0.6518248000265626, + "learning_rate": 0.003, + "loss": 4.024, + "step": 13151 + }, + { + "epoch": 0.13152, + "grad_norm": 0.6868569781202374, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 13152 + }, + { + "epoch": 0.13153, + "grad_norm": 0.7778054946109996, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 13153 + }, + { + "epoch": 0.13154, + "grad_norm": 0.8471677171300068, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 13154 + }, + { + "epoch": 0.13155, + "grad_norm": 0.8792598916011967, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 13155 + }, + { + "epoch": 0.13156, + "grad_norm": 0.8019416326117923, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13156 + }, + { + "epoch": 0.13157, + "grad_norm": 0.5933806952173072, + "learning_rate": 0.003, + "loss": 4.062, + "step": 13157 + }, + { + "epoch": 0.13158, + "grad_norm": 0.5325928261496442, + "learning_rate": 0.003, + "loss": 4.012, + "step": 13158 + }, + { + "epoch": 0.13159, + "grad_norm": 0.6588591025982233, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 13159 + }, + { + "epoch": 0.1316, + "grad_norm": 0.7359945194021863, + "learning_rate": 0.003, + "loss": 3.984, + "step": 13160 + }, + { + "epoch": 0.13161, + "grad_norm": 0.6625400525891768, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13161 + }, + { + "epoch": 0.13162, + "grad_norm": 0.5438771098130668, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 13162 + }, + { + "epoch": 0.13163, + "grad_norm": 0.5514765288411507, + "learning_rate": 0.003, + "loss": 4.0561, + "step": 13163 + }, + { + "epoch": 0.13164, + "grad_norm": 0.5707267675838418, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 13164 + }, + { + "epoch": 0.13165, + "grad_norm": 0.589743830874756, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 13165 + }, + { + "epoch": 0.13166, + "grad_norm": 0.6489327842113343, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 13166 + }, + { + "epoch": 0.13167, + "grad_norm": 0.6304342993699401, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 13167 + }, + { + "epoch": 0.13168, + "grad_norm": 0.6032529034819337, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 13168 + }, + { + "epoch": 0.13169, + "grad_norm": 0.6263347707786251, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 13169 + }, + { + "epoch": 0.1317, + "grad_norm": 0.6415209488017183, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 13170 + }, + { + "epoch": 0.13171, + "grad_norm": 0.6800591771396955, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 13171 + }, + { + "epoch": 0.13172, + "grad_norm": 0.6377660575264066, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 13172 + }, + { + "epoch": 0.13173, + "grad_norm": 0.6411092456815173, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 13173 + }, + { + "epoch": 0.13174, + "grad_norm": 0.6049844159799904, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 13174 + }, + { + "epoch": 0.13175, + "grad_norm": 0.6620680420608737, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 13175 + }, + { + "epoch": 0.13176, + "grad_norm": 0.6415904414024663, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 13176 + }, + { + "epoch": 0.13177, + "grad_norm": 0.5448000333112191, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 13177 + }, + { + "epoch": 0.13178, + "grad_norm": 0.5496579035623841, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 13178 + }, + { + "epoch": 0.13179, + "grad_norm": 0.6282770338466882, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 13179 + }, + { + "epoch": 0.1318, + "grad_norm": 0.7974380560934834, + "learning_rate": 0.003, + "loss": 4.003, + "step": 13180 + }, + { + "epoch": 0.13181, + "grad_norm": 0.7930220023633505, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 13181 + }, + { + "epoch": 0.13182, + "grad_norm": 0.815959700868477, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 13182 + }, + { + "epoch": 0.13183, + "grad_norm": 0.9290121774889113, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 13183 + }, + { + "epoch": 0.13184, + "grad_norm": 0.8852688150435432, + "learning_rate": 0.003, + "loss": 4.028, + "step": 13184 + }, + { + "epoch": 0.13185, + "grad_norm": 0.8874922288115114, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 13185 + }, + { + "epoch": 0.13186, + "grad_norm": 0.9006073104993327, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 13186 + }, + { + "epoch": 0.13187, + "grad_norm": 0.9348725342718891, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 13187 + }, + { + "epoch": 0.13188, + "grad_norm": 1.0422044547016631, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 13188 + }, + { + "epoch": 0.13189, + "grad_norm": 1.0918736405416838, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 13189 + }, + { + "epoch": 0.1319, + "grad_norm": 0.8263240714376336, + "learning_rate": 0.003, + "loss": 4.068, + "step": 13190 + }, + { + "epoch": 0.13191, + "grad_norm": 0.7957076355936558, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 13191 + }, + { + "epoch": 0.13192, + "grad_norm": 0.9711512861061377, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 13192 + }, + { + "epoch": 0.13193, + "grad_norm": 0.936365848827128, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 13193 + }, + { + "epoch": 0.13194, + "grad_norm": 0.7812974065897336, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 13194 + }, + { + "epoch": 0.13195, + "grad_norm": 0.7283372704458176, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 13195 + }, + { + "epoch": 0.13196, + "grad_norm": 0.6738165725111824, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 13196 + }, + { + "epoch": 0.13197, + "grad_norm": 0.5321549363243743, + "learning_rate": 0.003, + "loss": 4.039, + "step": 13197 + }, + { + "epoch": 0.13198, + "grad_norm": 0.48934484383773347, + "learning_rate": 0.003, + "loss": 4.021, + "step": 13198 + }, + { + "epoch": 0.13199, + "grad_norm": 0.4924764578861122, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 13199 + }, + { + "epoch": 0.132, + "grad_norm": 0.48336035682382755, + "learning_rate": 0.003, + "loss": 4.033, + "step": 13200 + }, + { + "epoch": 0.13201, + "grad_norm": 0.4887888349440515, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 13201 + }, + { + "epoch": 0.13202, + "grad_norm": 0.5327191702776267, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 13202 + }, + { + "epoch": 0.13203, + "grad_norm": 0.5773222464671407, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 13203 + }, + { + "epoch": 0.13204, + "grad_norm": 0.5825887604310982, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 13204 + }, + { + "epoch": 0.13205, + "grad_norm": 0.5328766468627328, + "learning_rate": 0.003, + "loss": 4.024, + "step": 13205 + }, + { + "epoch": 0.13206, + "grad_norm": 0.6829973949985749, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 13206 + }, + { + "epoch": 0.13207, + "grad_norm": 0.8018537235434904, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 13207 + }, + { + "epoch": 0.13208, + "grad_norm": 0.8174783682757265, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 13208 + }, + { + "epoch": 0.13209, + "grad_norm": 0.7210477974748544, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 13209 + }, + { + "epoch": 0.1321, + "grad_norm": 0.6993819744186816, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 13210 + }, + { + "epoch": 0.13211, + "grad_norm": 0.693217359820719, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 13211 + }, + { + "epoch": 0.13212, + "grad_norm": 0.7914056809417744, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 13212 + }, + { + "epoch": 0.13213, + "grad_norm": 0.8193869442270668, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 13213 + }, + { + "epoch": 0.13214, + "grad_norm": 0.8254721604062054, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 13214 + }, + { + "epoch": 0.13215, + "grad_norm": 0.8123585132447667, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 13215 + }, + { + "epoch": 0.13216, + "grad_norm": 0.7777789606256749, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 13216 + }, + { + "epoch": 0.13217, + "grad_norm": 0.6998713902105538, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 13217 + }, + { + "epoch": 0.13218, + "grad_norm": 0.6731111490760048, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 13218 + }, + { + "epoch": 0.13219, + "grad_norm": 0.6277063458139682, + "learning_rate": 0.003, + "loss": 4.0701, + "step": 13219 + }, + { + "epoch": 0.1322, + "grad_norm": 0.6155974754523498, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 13220 + }, + { + "epoch": 0.13221, + "grad_norm": 0.6611179003052993, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 13221 + }, + { + "epoch": 0.13222, + "grad_norm": 0.676076330427607, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 13222 + }, + { + "epoch": 0.13223, + "grad_norm": 0.6536045563591315, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 13223 + }, + { + "epoch": 0.13224, + "grad_norm": 0.6603525664910237, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 13224 + }, + { + "epoch": 0.13225, + "grad_norm": 0.6902724350817997, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 13225 + }, + { + "epoch": 0.13226, + "grad_norm": 0.6345130142842084, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 13226 + }, + { + "epoch": 0.13227, + "grad_norm": 0.6579617700964082, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 13227 + }, + { + "epoch": 0.13228, + "grad_norm": 0.7208190141118029, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13228 + }, + { + "epoch": 0.13229, + "grad_norm": 0.6922326940414962, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 13229 + }, + { + "epoch": 0.1323, + "grad_norm": 0.6786792605165282, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 13230 + }, + { + "epoch": 0.13231, + "grad_norm": 0.7178288674108746, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 13231 + }, + { + "epoch": 0.13232, + "grad_norm": 0.8666306180913271, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 13232 + }, + { + "epoch": 0.13233, + "grad_norm": 1.0240246021930932, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 13233 + }, + { + "epoch": 0.13234, + "grad_norm": 1.1578502720942763, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 13234 + }, + { + "epoch": 0.13235, + "grad_norm": 0.8608963660303678, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 13235 + }, + { + "epoch": 0.13236, + "grad_norm": 0.7609641645258582, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 13236 + }, + { + "epoch": 0.13237, + "grad_norm": 0.7541200894914037, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 13237 + }, + { + "epoch": 0.13238, + "grad_norm": 0.6638839094393235, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 13238 + }, + { + "epoch": 0.13239, + "grad_norm": 0.7444404217670486, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 13239 + }, + { + "epoch": 0.1324, + "grad_norm": 0.6667019610975216, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 13240 + }, + { + "epoch": 0.13241, + "grad_norm": 0.6271317507586476, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 13241 + }, + { + "epoch": 0.13242, + "grad_norm": 0.6621245030953702, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 13242 + }, + { + "epoch": 0.13243, + "grad_norm": 0.7348265157042877, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 13243 + }, + { + "epoch": 0.13244, + "grad_norm": 0.7811926371848927, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 13244 + }, + { + "epoch": 0.13245, + "grad_norm": 0.82357873952852, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 13245 + }, + { + "epoch": 0.13246, + "grad_norm": 0.8407187927389124, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 13246 + }, + { + "epoch": 0.13247, + "grad_norm": 0.8321443705291162, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 13247 + }, + { + "epoch": 0.13248, + "grad_norm": 0.8523205853173546, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 13248 + }, + { + "epoch": 0.13249, + "grad_norm": 0.8357349917382598, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13249 + }, + { + "epoch": 0.1325, + "grad_norm": 0.7850156984743906, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 13250 + }, + { + "epoch": 0.13251, + "grad_norm": 0.6905340657851623, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 13251 + }, + { + "epoch": 0.13252, + "grad_norm": 0.6531893350841619, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 13252 + }, + { + "epoch": 0.13253, + "grad_norm": 0.6629827582961826, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 13253 + }, + { + "epoch": 0.13254, + "grad_norm": 0.8290655290774405, + "learning_rate": 0.003, + "loss": 4.046, + "step": 13254 + }, + { + "epoch": 0.13255, + "grad_norm": 1.0307699641988233, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 13255 + }, + { + "epoch": 0.13256, + "grad_norm": 0.8237561281252, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 13256 + }, + { + "epoch": 0.13257, + "grad_norm": 0.6250873918265447, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 13257 + }, + { + "epoch": 0.13258, + "grad_norm": 0.5890807628522355, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 13258 + }, + { + "epoch": 0.13259, + "grad_norm": 0.7327973081603304, + "learning_rate": 0.003, + "loss": 4.043, + "step": 13259 + }, + { + "epoch": 0.1326, + "grad_norm": 0.7138912115728967, + "learning_rate": 0.003, + "loss": 4.028, + "step": 13260 + }, + { + "epoch": 0.13261, + "grad_norm": 0.6151709849552213, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 13261 + }, + { + "epoch": 0.13262, + "grad_norm": 0.4857210730618576, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 13262 + }, + { + "epoch": 0.13263, + "grad_norm": 0.4185697286861819, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 13263 + }, + { + "epoch": 0.13264, + "grad_norm": 0.45545708328358636, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 13264 + }, + { + "epoch": 0.13265, + "grad_norm": 0.41918508125223963, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 13265 + }, + { + "epoch": 0.13266, + "grad_norm": 0.4136604724535879, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 13266 + }, + { + "epoch": 0.13267, + "grad_norm": 0.4423089979458951, + "learning_rate": 0.003, + "loss": 4.011, + "step": 13267 + }, + { + "epoch": 0.13268, + "grad_norm": 0.49262933330146225, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13268 + }, + { + "epoch": 0.13269, + "grad_norm": 0.5209959988921335, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 13269 + }, + { + "epoch": 0.1327, + "grad_norm": 0.5943990878740613, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 13270 + }, + { + "epoch": 0.13271, + "grad_norm": 0.6720960607971912, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 13271 + }, + { + "epoch": 0.13272, + "grad_norm": 0.68088322781504, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 13272 + }, + { + "epoch": 0.13273, + "grad_norm": 0.6663657474690332, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 13273 + }, + { + "epoch": 0.13274, + "grad_norm": 0.8214112728153096, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 13274 + }, + { + "epoch": 0.13275, + "grad_norm": 1.0642486288211563, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 13275 + }, + { + "epoch": 0.13276, + "grad_norm": 1.030946028918972, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 13276 + }, + { + "epoch": 0.13277, + "grad_norm": 1.0186251823217087, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 13277 + }, + { + "epoch": 0.13278, + "grad_norm": 0.8773491253578496, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 13278 + }, + { + "epoch": 0.13279, + "grad_norm": 0.7969463636142594, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 13279 + }, + { + "epoch": 0.1328, + "grad_norm": 0.8456128229276985, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 13280 + }, + { + "epoch": 0.13281, + "grad_norm": 0.8477648767820268, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13281 + }, + { + "epoch": 0.13282, + "grad_norm": 0.7438211433862554, + "learning_rate": 0.003, + "loss": 4.05, + "step": 13282 + }, + { + "epoch": 0.13283, + "grad_norm": 0.8067074136362178, + "learning_rate": 0.003, + "loss": 4.038, + "step": 13283 + }, + { + "epoch": 0.13284, + "grad_norm": 0.9172678733504603, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 13284 + }, + { + "epoch": 0.13285, + "grad_norm": 1.0594545063224847, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 13285 + }, + { + "epoch": 0.13286, + "grad_norm": 1.0449332881902562, + "learning_rate": 0.003, + "loss": 4.0673, + "step": 13286 + }, + { + "epoch": 0.13287, + "grad_norm": 1.0118159529146085, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 13287 + }, + { + "epoch": 0.13288, + "grad_norm": 0.86345516105329, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 13288 + }, + { + "epoch": 0.13289, + "grad_norm": 0.7564045911825217, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 13289 + }, + { + "epoch": 0.1329, + "grad_norm": 0.7157347831768205, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 13290 + }, + { + "epoch": 0.13291, + "grad_norm": 0.7015403193016375, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 13291 + }, + { + "epoch": 0.13292, + "grad_norm": 0.7508439024577764, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 13292 + }, + { + "epoch": 0.13293, + "grad_norm": 0.9090466622578697, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 13293 + }, + { + "epoch": 0.13294, + "grad_norm": 1.0039518224849016, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 13294 + }, + { + "epoch": 0.13295, + "grad_norm": 0.8481214703417205, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 13295 + }, + { + "epoch": 0.13296, + "grad_norm": 0.8859792203207714, + "learning_rate": 0.003, + "loss": 4.0769, + "step": 13296 + }, + { + "epoch": 0.13297, + "grad_norm": 0.8695967208426048, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 13297 + }, + { + "epoch": 0.13298, + "grad_norm": 0.7376590401671862, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 13298 + }, + { + "epoch": 0.13299, + "grad_norm": 0.6100224287440704, + "learning_rate": 0.003, + "loss": 4.0784, + "step": 13299 + }, + { + "epoch": 0.133, + "grad_norm": 0.5763998250896152, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 13300 + }, + { + "epoch": 0.13301, + "grad_norm": 0.5409155109763648, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 13301 + }, + { + "epoch": 0.13302, + "grad_norm": 0.5040129351453145, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 13302 + }, + { + "epoch": 0.13303, + "grad_norm": 0.5228250499571702, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13303 + }, + { + "epoch": 0.13304, + "grad_norm": 0.5831687912910107, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 13304 + }, + { + "epoch": 0.13305, + "grad_norm": 0.5822302831686085, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 13305 + }, + { + "epoch": 0.13306, + "grad_norm": 0.5792340076338153, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 13306 + }, + { + "epoch": 0.13307, + "grad_norm": 0.551526752183952, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 13307 + }, + { + "epoch": 0.13308, + "grad_norm": 0.5313280348644863, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 13308 + }, + { + "epoch": 0.13309, + "grad_norm": 0.596523948376668, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 13309 + }, + { + "epoch": 0.1331, + "grad_norm": 0.6384646778430016, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 13310 + }, + { + "epoch": 0.13311, + "grad_norm": 0.7361746037344366, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 13311 + }, + { + "epoch": 0.13312, + "grad_norm": 0.745805364207279, + "learning_rate": 0.003, + "loss": 4.027, + "step": 13312 + }, + { + "epoch": 0.13313, + "grad_norm": 0.7616602992806318, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 13313 + }, + { + "epoch": 0.13314, + "grad_norm": 0.7345384944914312, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 13314 + }, + { + "epoch": 0.13315, + "grad_norm": 0.6796519471686316, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 13315 + }, + { + "epoch": 0.13316, + "grad_norm": 0.6455884114421527, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 13316 + }, + { + "epoch": 0.13317, + "grad_norm": 0.6351329368199424, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 13317 + }, + { + "epoch": 0.13318, + "grad_norm": 0.6784773221801946, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 13318 + }, + { + "epoch": 0.13319, + "grad_norm": 0.5953181163262075, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 13319 + }, + { + "epoch": 0.1332, + "grad_norm": 0.6306705947831605, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 13320 + }, + { + "epoch": 0.13321, + "grad_norm": 0.6683035626289083, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 13321 + }, + { + "epoch": 0.13322, + "grad_norm": 0.6011084191147845, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 13322 + }, + { + "epoch": 0.13323, + "grad_norm": 0.6195430895587234, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 13323 + }, + { + "epoch": 0.13324, + "grad_norm": 0.6446533425822164, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 13324 + }, + { + "epoch": 0.13325, + "grad_norm": 0.6715400130562317, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 13325 + }, + { + "epoch": 0.13326, + "grad_norm": 0.7176194723069367, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 13326 + }, + { + "epoch": 0.13327, + "grad_norm": 0.8604664576072841, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 13327 + }, + { + "epoch": 0.13328, + "grad_norm": 0.9208763128019319, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 13328 + }, + { + "epoch": 0.13329, + "grad_norm": 0.8003456516010723, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 13329 + }, + { + "epoch": 0.1333, + "grad_norm": 0.7073985682427617, + "learning_rate": 0.003, + "loss": 4.04, + "step": 13330 + }, + { + "epoch": 0.13331, + "grad_norm": 0.7261609000203826, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 13331 + }, + { + "epoch": 0.13332, + "grad_norm": 0.7400216788537648, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 13332 + }, + { + "epoch": 0.13333, + "grad_norm": 0.7603102005751108, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 13333 + }, + { + "epoch": 0.13334, + "grad_norm": 0.805022550522761, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 13334 + }, + { + "epoch": 0.13335, + "grad_norm": 0.7770576573192901, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 13335 + }, + { + "epoch": 0.13336, + "grad_norm": 0.7086796125667405, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 13336 + }, + { + "epoch": 0.13337, + "grad_norm": 0.6603866887808134, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13337 + }, + { + "epoch": 0.13338, + "grad_norm": 0.6303932504142032, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 13338 + }, + { + "epoch": 0.13339, + "grad_norm": 0.6087894069742235, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 13339 + }, + { + "epoch": 0.1334, + "grad_norm": 0.6523292346958911, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 13340 + }, + { + "epoch": 0.13341, + "grad_norm": 0.7014432516531423, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 13341 + }, + { + "epoch": 0.13342, + "grad_norm": 0.6726405475294195, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 13342 + }, + { + "epoch": 0.13343, + "grad_norm": 0.6275497438531747, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 13343 + }, + { + "epoch": 0.13344, + "grad_norm": 0.6497414917330657, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 13344 + }, + { + "epoch": 0.13345, + "grad_norm": 0.6946110337475737, + "learning_rate": 0.003, + "loss": 4.0546, + "step": 13345 + }, + { + "epoch": 0.13346, + "grad_norm": 0.8175280503005853, + "learning_rate": 0.003, + "loss": 4.025, + "step": 13346 + }, + { + "epoch": 0.13347, + "grad_norm": 0.8495841209979151, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 13347 + }, + { + "epoch": 0.13348, + "grad_norm": 0.8425296465080921, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 13348 + }, + { + "epoch": 0.13349, + "grad_norm": 0.8252295210679658, + "learning_rate": 0.003, + "loss": 4.03, + "step": 13349 + }, + { + "epoch": 0.1335, + "grad_norm": 0.7886344900049999, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 13350 + }, + { + "epoch": 0.13351, + "grad_norm": 0.6803191572213625, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 13351 + }, + { + "epoch": 0.13352, + "grad_norm": 0.6210560992596801, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 13352 + }, + { + "epoch": 0.13353, + "grad_norm": 0.6155851953326842, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 13353 + }, + { + "epoch": 0.13354, + "grad_norm": 0.7267687331885052, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 13354 + }, + { + "epoch": 0.13355, + "grad_norm": 0.8421320015371593, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 13355 + }, + { + "epoch": 0.13356, + "grad_norm": 0.9841588601985528, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 13356 + }, + { + "epoch": 0.13357, + "grad_norm": 1.0089856025218544, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 13357 + }, + { + "epoch": 0.13358, + "grad_norm": 0.7561577273857809, + "learning_rate": 0.003, + "loss": 4.013, + "step": 13358 + }, + { + "epoch": 0.13359, + "grad_norm": 0.7250226546460137, + "learning_rate": 0.003, + "loss": 4.039, + "step": 13359 + }, + { + "epoch": 0.1336, + "grad_norm": 0.7821581115257211, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 13360 + }, + { + "epoch": 0.13361, + "grad_norm": 0.7403566385731374, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 13361 + }, + { + "epoch": 0.13362, + "grad_norm": 0.7060128702827391, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 13362 + }, + { + "epoch": 0.13363, + "grad_norm": 0.7081330170886662, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 13363 + }, + { + "epoch": 0.13364, + "grad_norm": 0.776194642348398, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 13364 + }, + { + "epoch": 0.13365, + "grad_norm": 0.7369545091646295, + "learning_rate": 0.003, + "loss": 4.017, + "step": 13365 + }, + { + "epoch": 0.13366, + "grad_norm": 0.754257424184205, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 13366 + }, + { + "epoch": 0.13367, + "grad_norm": 0.7902577426107843, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 13367 + }, + { + "epoch": 0.13368, + "grad_norm": 0.7348105516063611, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 13368 + }, + { + "epoch": 0.13369, + "grad_norm": 0.83475536372394, + "learning_rate": 0.003, + "loss": 4.031, + "step": 13369 + }, + { + "epoch": 0.1337, + "grad_norm": 0.8609837269648364, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 13370 + }, + { + "epoch": 0.13371, + "grad_norm": 0.9004197836070135, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 13371 + }, + { + "epoch": 0.13372, + "grad_norm": 0.9008335615208588, + "learning_rate": 0.003, + "loss": 4.03, + "step": 13372 + }, + { + "epoch": 0.13373, + "grad_norm": 0.8489284449473385, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 13373 + }, + { + "epoch": 0.13374, + "grad_norm": 0.6832878918892245, + "learning_rate": 0.003, + "loss": 4.052, + "step": 13374 + }, + { + "epoch": 0.13375, + "grad_norm": 0.6557433555351565, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 13375 + }, + { + "epoch": 0.13376, + "grad_norm": 0.5785557515704941, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 13376 + }, + { + "epoch": 0.13377, + "grad_norm": 0.7001516423057923, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 13377 + }, + { + "epoch": 0.13378, + "grad_norm": 0.7009587029611405, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 13378 + }, + { + "epoch": 0.13379, + "grad_norm": 0.5878700452480461, + "learning_rate": 0.003, + "loss": 4.011, + "step": 13379 + }, + { + "epoch": 0.1338, + "grad_norm": 0.5968695840649729, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13380 + }, + { + "epoch": 0.13381, + "grad_norm": 0.6102779714942189, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 13381 + }, + { + "epoch": 0.13382, + "grad_norm": 0.6874696309378021, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 13382 + }, + { + "epoch": 0.13383, + "grad_norm": 0.7409014102944833, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 13383 + }, + { + "epoch": 0.13384, + "grad_norm": 0.7348420933091394, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 13384 + }, + { + "epoch": 0.13385, + "grad_norm": 0.6793991602450019, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 13385 + }, + { + "epoch": 0.13386, + "grad_norm": 0.7449724002446617, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 13386 + }, + { + "epoch": 0.13387, + "grad_norm": 0.8148673525554917, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 13387 + }, + { + "epoch": 0.13388, + "grad_norm": 0.8958830874875214, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 13388 + }, + { + "epoch": 0.13389, + "grad_norm": 0.8300570325053623, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 13389 + }, + { + "epoch": 0.1339, + "grad_norm": 0.7522843358332626, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 13390 + }, + { + "epoch": 0.13391, + "grad_norm": 0.7769600545299576, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 13391 + }, + { + "epoch": 0.13392, + "grad_norm": 0.8309911597173285, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 13392 + }, + { + "epoch": 0.13393, + "grad_norm": 0.6847221965670388, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 13393 + }, + { + "epoch": 0.13394, + "grad_norm": 0.5925295944942925, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 13394 + }, + { + "epoch": 0.13395, + "grad_norm": 0.5912887965228717, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 13395 + }, + { + "epoch": 0.13396, + "grad_norm": 0.6055279485995922, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 13396 + }, + { + "epoch": 0.13397, + "grad_norm": 0.6856029728593821, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 13397 + }, + { + "epoch": 0.13398, + "grad_norm": 0.676024219291663, + "learning_rate": 0.003, + "loss": 4.006, + "step": 13398 + }, + { + "epoch": 0.13399, + "grad_norm": 0.6891919648718567, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 13399 + }, + { + "epoch": 0.134, + "grad_norm": 0.6929599318592037, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 13400 + }, + { + "epoch": 0.13401, + "grad_norm": 0.837576438620971, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13401 + }, + { + "epoch": 0.13402, + "grad_norm": 0.8097344545444961, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 13402 + }, + { + "epoch": 0.13403, + "grad_norm": 0.7752320172549755, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 13403 + }, + { + "epoch": 0.13404, + "grad_norm": 0.8048333137554738, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 13404 + }, + { + "epoch": 0.13405, + "grad_norm": 0.7961520047772, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 13405 + }, + { + "epoch": 0.13406, + "grad_norm": 0.7602080566304474, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 13406 + }, + { + "epoch": 0.13407, + "grad_norm": 0.7697049137286447, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 13407 + }, + { + "epoch": 0.13408, + "grad_norm": 0.6945769800147242, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 13408 + }, + { + "epoch": 0.13409, + "grad_norm": 0.6281887061057714, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 13409 + }, + { + "epoch": 0.1341, + "grad_norm": 0.5990773540740792, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 13410 + }, + { + "epoch": 0.13411, + "grad_norm": 0.6224540673906073, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 13411 + }, + { + "epoch": 0.13412, + "grad_norm": 0.5942664014695002, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 13412 + }, + { + "epoch": 0.13413, + "grad_norm": 0.6056589650476428, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 13413 + }, + { + "epoch": 0.13414, + "grad_norm": 0.6676930680964502, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 13414 + }, + { + "epoch": 0.13415, + "grad_norm": 0.7769239092289175, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 13415 + }, + { + "epoch": 0.13416, + "grad_norm": 0.8316262024330262, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 13416 + }, + { + "epoch": 0.13417, + "grad_norm": 0.9432694923201876, + "learning_rate": 0.003, + "loss": 4.003, + "step": 13417 + }, + { + "epoch": 0.13418, + "grad_norm": 0.8889913225861839, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 13418 + }, + { + "epoch": 0.13419, + "grad_norm": 0.7862031629269289, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 13419 + }, + { + "epoch": 0.1342, + "grad_norm": 0.8096124571000916, + "learning_rate": 0.003, + "loss": 4.016, + "step": 13420 + }, + { + "epoch": 0.13421, + "grad_norm": 0.7217621113934771, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 13421 + }, + { + "epoch": 0.13422, + "grad_norm": 0.6640838383150456, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 13422 + }, + { + "epoch": 0.13423, + "grad_norm": 0.6831748457018458, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 13423 + }, + { + "epoch": 0.13424, + "grad_norm": 0.6435633444838272, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 13424 + }, + { + "epoch": 0.13425, + "grad_norm": 0.7278582279133111, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 13425 + }, + { + "epoch": 0.13426, + "grad_norm": 0.7640851335394583, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 13426 + }, + { + "epoch": 0.13427, + "grad_norm": 0.7419882661724222, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 13427 + }, + { + "epoch": 0.13428, + "grad_norm": 0.8392794187179503, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 13428 + }, + { + "epoch": 0.13429, + "grad_norm": 0.8295626704392255, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 13429 + }, + { + "epoch": 0.1343, + "grad_norm": 0.8340320129725214, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 13430 + }, + { + "epoch": 0.13431, + "grad_norm": 0.7637343852701431, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 13431 + }, + { + "epoch": 0.13432, + "grad_norm": 0.6518080743268002, + "learning_rate": 0.003, + "loss": 4.017, + "step": 13432 + }, + { + "epoch": 0.13433, + "grad_norm": 0.604948798074213, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 13433 + }, + { + "epoch": 0.13434, + "grad_norm": 0.6299465950661489, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 13434 + }, + { + "epoch": 0.13435, + "grad_norm": 0.6634854950311903, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 13435 + }, + { + "epoch": 0.13436, + "grad_norm": 0.6550941190915326, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 13436 + }, + { + "epoch": 0.13437, + "grad_norm": 0.641809332305717, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 13437 + }, + { + "epoch": 0.13438, + "grad_norm": 0.5741947613674069, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 13438 + }, + { + "epoch": 0.13439, + "grad_norm": 0.5138567084848007, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 13439 + }, + { + "epoch": 0.1344, + "grad_norm": 0.5131691122007062, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 13440 + }, + { + "epoch": 0.13441, + "grad_norm": 0.614115476647699, + "learning_rate": 0.003, + "loss": 3.996, + "step": 13441 + }, + { + "epoch": 0.13442, + "grad_norm": 0.7499888238190047, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 13442 + }, + { + "epoch": 0.13443, + "grad_norm": 0.8464783444404207, + "learning_rate": 0.003, + "loss": 4.03, + "step": 13443 + }, + { + "epoch": 0.13444, + "grad_norm": 0.9537198137072744, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13444 + }, + { + "epoch": 0.13445, + "grad_norm": 0.8942029221206805, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 13445 + }, + { + "epoch": 0.13446, + "grad_norm": 0.8335364212269964, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 13446 + }, + { + "epoch": 0.13447, + "grad_norm": 0.836229927138819, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 13447 + }, + { + "epoch": 0.13448, + "grad_norm": 0.7448587005061763, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 13448 + }, + { + "epoch": 0.13449, + "grad_norm": 0.7260862894357353, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 13449 + }, + { + "epoch": 0.1345, + "grad_norm": 0.7446120631688947, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 13450 + }, + { + "epoch": 0.13451, + "grad_norm": 0.6868547884144315, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 13451 + }, + { + "epoch": 0.13452, + "grad_norm": 0.7030653360867052, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 13452 + }, + { + "epoch": 0.13453, + "grad_norm": 0.8256627801941999, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 13453 + }, + { + "epoch": 0.13454, + "grad_norm": 0.8924510701172372, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 13454 + }, + { + "epoch": 0.13455, + "grad_norm": 0.8778422289602265, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 13455 + }, + { + "epoch": 0.13456, + "grad_norm": 0.793217454934818, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 13456 + }, + { + "epoch": 0.13457, + "grad_norm": 0.6895044118327128, + "learning_rate": 0.003, + "loss": 3.997, + "step": 13457 + }, + { + "epoch": 0.13458, + "grad_norm": 0.6736847891702478, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 13458 + }, + { + "epoch": 0.13459, + "grad_norm": 0.6310289419850681, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 13459 + }, + { + "epoch": 0.1346, + "grad_norm": 0.656795600345577, + "learning_rate": 0.003, + "loss": 4.029, + "step": 13460 + }, + { + "epoch": 0.13461, + "grad_norm": 0.769752164940889, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 13461 + }, + { + "epoch": 0.13462, + "grad_norm": 0.7873998331895217, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 13462 + }, + { + "epoch": 0.13463, + "grad_norm": 0.7879625609415087, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 13463 + }, + { + "epoch": 0.13464, + "grad_norm": 0.8467702680515995, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 13464 + }, + { + "epoch": 0.13465, + "grad_norm": 0.776550321275409, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 13465 + }, + { + "epoch": 0.13466, + "grad_norm": 0.7290941861204437, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13466 + }, + { + "epoch": 0.13467, + "grad_norm": 0.744372968003231, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 13467 + }, + { + "epoch": 0.13468, + "grad_norm": 0.7513853517889975, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 13468 + }, + { + "epoch": 0.13469, + "grad_norm": 0.6623743503029613, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 13469 + }, + { + "epoch": 0.1347, + "grad_norm": 0.6243711647641526, + "learning_rate": 0.003, + "loss": 4.013, + "step": 13470 + }, + { + "epoch": 0.13471, + "grad_norm": 0.7081726331170599, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 13471 + }, + { + "epoch": 0.13472, + "grad_norm": 0.7039542917849023, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 13472 + }, + { + "epoch": 0.13473, + "grad_norm": 0.6949501306927071, + "learning_rate": 0.003, + "loss": 3.987, + "step": 13473 + }, + { + "epoch": 0.13474, + "grad_norm": 0.6601401822117123, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 13474 + }, + { + "epoch": 0.13475, + "grad_norm": 0.5981993437479421, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 13475 + }, + { + "epoch": 0.13476, + "grad_norm": 0.6102597931064876, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13476 + }, + { + "epoch": 0.13477, + "grad_norm": 0.5820011471431155, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 13477 + }, + { + "epoch": 0.13478, + "grad_norm": 0.5566297606942044, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 13478 + }, + { + "epoch": 0.13479, + "grad_norm": 0.577352665206282, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 13479 + }, + { + "epoch": 0.1348, + "grad_norm": 0.702640613658655, + "learning_rate": 0.003, + "loss": 4.027, + "step": 13480 + }, + { + "epoch": 0.13481, + "grad_norm": 0.8864518607720958, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 13481 + }, + { + "epoch": 0.13482, + "grad_norm": 1.0121781661844924, + "learning_rate": 0.003, + "loss": 4.007, + "step": 13482 + }, + { + "epoch": 0.13483, + "grad_norm": 0.9477087204554271, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 13483 + }, + { + "epoch": 0.13484, + "grad_norm": 0.8070796115186736, + "learning_rate": 0.003, + "loss": 4.021, + "step": 13484 + }, + { + "epoch": 0.13485, + "grad_norm": 0.8627477072473649, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 13485 + }, + { + "epoch": 0.13486, + "grad_norm": 0.9430295209195931, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 13486 + }, + { + "epoch": 0.13487, + "grad_norm": 0.9464600898192109, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 13487 + }, + { + "epoch": 0.13488, + "grad_norm": 1.0017505991638005, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 13488 + }, + { + "epoch": 0.13489, + "grad_norm": 0.8931123675848208, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 13489 + }, + { + "epoch": 0.1349, + "grad_norm": 0.8155822970486017, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 13490 + }, + { + "epoch": 0.13491, + "grad_norm": 0.7399631702927486, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13491 + }, + { + "epoch": 0.13492, + "grad_norm": 0.739213586779716, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 13492 + }, + { + "epoch": 0.13493, + "grad_norm": 0.7189731404311065, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 13493 + }, + { + "epoch": 0.13494, + "grad_norm": 0.6372689743022081, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13494 + }, + { + "epoch": 0.13495, + "grad_norm": 0.5721702962705931, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 13495 + }, + { + "epoch": 0.13496, + "grad_norm": 0.5534805570193871, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 13496 + }, + { + "epoch": 0.13497, + "grad_norm": 0.5302831816200301, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 13497 + }, + { + "epoch": 0.13498, + "grad_norm": 0.48447725883569237, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 13498 + }, + { + "epoch": 0.13499, + "grad_norm": 0.5322191850735591, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 13499 + }, + { + "epoch": 0.135, + "grad_norm": 0.6494337901197486, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13500 + }, + { + "epoch": 0.13501, + "grad_norm": 0.6520629432559513, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13501 + }, + { + "epoch": 0.13502, + "grad_norm": 0.5846229516670286, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 13502 + }, + { + "epoch": 0.13503, + "grad_norm": 0.4995616914237184, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 13503 + }, + { + "epoch": 0.13504, + "grad_norm": 0.4900150457870593, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 13504 + }, + { + "epoch": 0.13505, + "grad_norm": 0.5358083876566272, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 13505 + }, + { + "epoch": 0.13506, + "grad_norm": 0.6320758203399659, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 13506 + }, + { + "epoch": 0.13507, + "grad_norm": 0.7853664164015811, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 13507 + }, + { + "epoch": 0.13508, + "grad_norm": 0.8833505783652318, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 13508 + }, + { + "epoch": 0.13509, + "grad_norm": 0.7656484275878513, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 13509 + }, + { + "epoch": 0.1351, + "grad_norm": 0.6428533598852069, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 13510 + }, + { + "epoch": 0.13511, + "grad_norm": 0.6602775103318901, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 13511 + }, + { + "epoch": 0.13512, + "grad_norm": 0.7278644896010351, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 13512 + }, + { + "epoch": 0.13513, + "grad_norm": 0.7487013116649441, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 13513 + }, + { + "epoch": 0.13514, + "grad_norm": 0.9257454689585175, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 13514 + }, + { + "epoch": 0.13515, + "grad_norm": 0.9617624634183594, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 13515 + }, + { + "epoch": 0.13516, + "grad_norm": 0.8312630714375736, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 13516 + }, + { + "epoch": 0.13517, + "grad_norm": 0.7819111482805173, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 13517 + }, + { + "epoch": 0.13518, + "grad_norm": 0.7933371916300185, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 13518 + }, + { + "epoch": 0.13519, + "grad_norm": 0.7420553935297742, + "learning_rate": 0.003, + "loss": 4.035, + "step": 13519 + }, + { + "epoch": 0.1352, + "grad_norm": 0.6354156164194275, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 13520 + }, + { + "epoch": 0.13521, + "grad_norm": 0.6575110883014428, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 13521 + }, + { + "epoch": 0.13522, + "grad_norm": 0.7759501726438426, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 13522 + }, + { + "epoch": 0.13523, + "grad_norm": 0.7563821119798941, + "learning_rate": 0.003, + "loss": 4.023, + "step": 13523 + }, + { + "epoch": 0.13524, + "grad_norm": 0.6817713484754429, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 13524 + }, + { + "epoch": 0.13525, + "grad_norm": 0.6661694264531044, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 13525 + }, + { + "epoch": 0.13526, + "grad_norm": 0.7039595688538745, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 13526 + }, + { + "epoch": 0.13527, + "grad_norm": 0.8041639637220914, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 13527 + }, + { + "epoch": 0.13528, + "grad_norm": 0.8435626006739567, + "learning_rate": 0.003, + "loss": 4.02, + "step": 13528 + }, + { + "epoch": 0.13529, + "grad_norm": 0.8358911918467196, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 13529 + }, + { + "epoch": 0.1353, + "grad_norm": 0.7273551153961263, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 13530 + }, + { + "epoch": 0.13531, + "grad_norm": 0.6910686472721088, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 13531 + }, + { + "epoch": 0.13532, + "grad_norm": 0.683178135328063, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 13532 + }, + { + "epoch": 0.13533, + "grad_norm": 0.7292357242700306, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 13533 + }, + { + "epoch": 0.13534, + "grad_norm": 0.7143540248443992, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 13534 + }, + { + "epoch": 0.13535, + "grad_norm": 0.7279706756232983, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 13535 + }, + { + "epoch": 0.13536, + "grad_norm": 0.6830837045552787, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 13536 + }, + { + "epoch": 0.13537, + "grad_norm": 0.705967767744165, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 13537 + }, + { + "epoch": 0.13538, + "grad_norm": 0.7498991585136916, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 13538 + }, + { + "epoch": 0.13539, + "grad_norm": 0.9325513598513268, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 13539 + }, + { + "epoch": 0.1354, + "grad_norm": 1.057142831021509, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 13540 + }, + { + "epoch": 0.13541, + "grad_norm": 1.0335535034266436, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 13541 + }, + { + "epoch": 0.13542, + "grad_norm": 1.035806544741675, + "learning_rate": 0.003, + "loss": 4.0531, + "step": 13542 + }, + { + "epoch": 0.13543, + "grad_norm": 1.0217318238099917, + "learning_rate": 0.003, + "loss": 4.029, + "step": 13543 + }, + { + "epoch": 0.13544, + "grad_norm": 0.9586716042659136, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 13544 + }, + { + "epoch": 0.13545, + "grad_norm": 1.1375979316719915, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 13545 + }, + { + "epoch": 0.13546, + "grad_norm": 0.9553766695752506, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 13546 + }, + { + "epoch": 0.13547, + "grad_norm": 0.8905858708481482, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 13547 + }, + { + "epoch": 0.13548, + "grad_norm": 0.9924130859451529, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 13548 + }, + { + "epoch": 0.13549, + "grad_norm": 1.1586026659588229, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 13549 + }, + { + "epoch": 0.1355, + "grad_norm": 0.723357706979056, + "learning_rate": 0.003, + "loss": 4.021, + "step": 13550 + }, + { + "epoch": 0.13551, + "grad_norm": 0.6911780206461301, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 13551 + }, + { + "epoch": 0.13552, + "grad_norm": 0.6937834987433363, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 13552 + }, + { + "epoch": 0.13553, + "grad_norm": 0.6695654331369567, + "learning_rate": 0.003, + "loss": 4.0688, + "step": 13553 + }, + { + "epoch": 0.13554, + "grad_norm": 0.7008910758246419, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 13554 + }, + { + "epoch": 0.13555, + "grad_norm": 0.7054607438592663, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 13555 + }, + { + "epoch": 0.13556, + "grad_norm": 0.7022498775359609, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 13556 + }, + { + "epoch": 0.13557, + "grad_norm": 0.6498048616623066, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 13557 + }, + { + "epoch": 0.13558, + "grad_norm": 0.6714663382644381, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 13558 + }, + { + "epoch": 0.13559, + "grad_norm": 0.6352873961325957, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 13559 + }, + { + "epoch": 0.1356, + "grad_norm": 0.5928362743283521, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 13560 + }, + { + "epoch": 0.13561, + "grad_norm": 0.5747226298112652, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 13561 + }, + { + "epoch": 0.13562, + "grad_norm": 0.4961150790871747, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 13562 + }, + { + "epoch": 0.13563, + "grad_norm": 0.49699228584647365, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 13563 + }, + { + "epoch": 0.13564, + "grad_norm": 0.4707779767254588, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 13564 + }, + { + "epoch": 0.13565, + "grad_norm": 0.47919612106878723, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 13565 + }, + { + "epoch": 0.13566, + "grad_norm": 0.4750782750020157, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 13566 + }, + { + "epoch": 0.13567, + "grad_norm": 0.4197294372982066, + "learning_rate": 0.003, + "loss": 4.011, + "step": 13567 + }, + { + "epoch": 0.13568, + "grad_norm": 0.4131336057724813, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 13568 + }, + { + "epoch": 0.13569, + "grad_norm": 0.40016593773855896, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 13569 + }, + { + "epoch": 0.1357, + "grad_norm": 0.35558229814398135, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 13570 + }, + { + "epoch": 0.13571, + "grad_norm": 0.3473165995348196, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 13571 + }, + { + "epoch": 0.13572, + "grad_norm": 0.39190440026762535, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 13572 + }, + { + "epoch": 0.13573, + "grad_norm": 0.46018677925228246, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 13573 + }, + { + "epoch": 0.13574, + "grad_norm": 0.600459448472703, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 13574 + }, + { + "epoch": 0.13575, + "grad_norm": 0.8267155015292048, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 13575 + }, + { + "epoch": 0.13576, + "grad_norm": 0.9468441101792817, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 13576 + }, + { + "epoch": 0.13577, + "grad_norm": 0.9759688550630129, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 13577 + }, + { + "epoch": 0.13578, + "grad_norm": 0.9268819098299951, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 13578 + }, + { + "epoch": 0.13579, + "grad_norm": 0.7790895340109626, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13579 + }, + { + "epoch": 0.1358, + "grad_norm": 0.7411186778514994, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 13580 + }, + { + "epoch": 0.13581, + "grad_norm": 0.7325819078362146, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 13581 + }, + { + "epoch": 0.13582, + "grad_norm": 0.7656337496825932, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 13582 + }, + { + "epoch": 0.13583, + "grad_norm": 0.6814832777837132, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 13583 + }, + { + "epoch": 0.13584, + "grad_norm": 0.665422067577505, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 13584 + }, + { + "epoch": 0.13585, + "grad_norm": 0.769215816751987, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 13585 + }, + { + "epoch": 0.13586, + "grad_norm": 0.845055805095514, + "learning_rate": 0.003, + "loss": 4.0649, + "step": 13586 + }, + { + "epoch": 0.13587, + "grad_norm": 0.9054701282785943, + "learning_rate": 0.003, + "loss": 4.012, + "step": 13587 + }, + { + "epoch": 0.13588, + "grad_norm": 0.813941190874482, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 13588 + }, + { + "epoch": 0.13589, + "grad_norm": 0.8218189237080459, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 13589 + }, + { + "epoch": 0.1359, + "grad_norm": 0.9311479802883268, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 13590 + }, + { + "epoch": 0.13591, + "grad_norm": 0.9355047806562907, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13591 + }, + { + "epoch": 0.13592, + "grad_norm": 1.075667295044156, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 13592 + }, + { + "epoch": 0.13593, + "grad_norm": 0.9870468875705783, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 13593 + }, + { + "epoch": 0.13594, + "grad_norm": 0.8297401624503484, + "learning_rate": 0.003, + "loss": 4.05, + "step": 13594 + }, + { + "epoch": 0.13595, + "grad_norm": 0.7013940354439938, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 13595 + }, + { + "epoch": 0.13596, + "grad_norm": 0.6920665611834944, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 13596 + }, + { + "epoch": 0.13597, + "grad_norm": 0.658143817420989, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 13597 + }, + { + "epoch": 0.13598, + "grad_norm": 0.6465700143421919, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 13598 + }, + { + "epoch": 0.13599, + "grad_norm": 0.6243052792023788, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 13599 + }, + { + "epoch": 0.136, + "grad_norm": 0.5781124603637869, + "learning_rate": 0.003, + "loss": 3.978, + "step": 13600 + }, + { + "epoch": 0.13601, + "grad_norm": 0.5818583801625775, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 13601 + }, + { + "epoch": 0.13602, + "grad_norm": 0.6624294211741928, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 13602 + }, + { + "epoch": 0.13603, + "grad_norm": 0.8317590648579457, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 13603 + }, + { + "epoch": 0.13604, + "grad_norm": 0.8312124979025683, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 13604 + }, + { + "epoch": 0.13605, + "grad_norm": 0.7108264042312147, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 13605 + }, + { + "epoch": 0.13606, + "grad_norm": 0.6216397194110566, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 13606 + }, + { + "epoch": 0.13607, + "grad_norm": 0.6297678901143737, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 13607 + }, + { + "epoch": 0.13608, + "grad_norm": 0.6226773229789108, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 13608 + }, + { + "epoch": 0.13609, + "grad_norm": 0.6230817598000129, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 13609 + }, + { + "epoch": 0.1361, + "grad_norm": 0.6468217082312046, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 13610 + }, + { + "epoch": 0.13611, + "grad_norm": 0.6661899145338319, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 13611 + }, + { + "epoch": 0.13612, + "grad_norm": 0.6384022730170567, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 13612 + }, + { + "epoch": 0.13613, + "grad_norm": 0.6996568830180201, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 13613 + }, + { + "epoch": 0.13614, + "grad_norm": 0.8853728876271569, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 13614 + }, + { + "epoch": 0.13615, + "grad_norm": 0.9784306929784484, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 13615 + }, + { + "epoch": 0.13616, + "grad_norm": 0.7922475392827011, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 13616 + }, + { + "epoch": 0.13617, + "grad_norm": 0.7365069575333139, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 13617 + }, + { + "epoch": 0.13618, + "grad_norm": 0.7106666264486423, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 13618 + }, + { + "epoch": 0.13619, + "grad_norm": 0.7518808864625118, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 13619 + }, + { + "epoch": 0.1362, + "grad_norm": 0.7584599757240232, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 13620 + }, + { + "epoch": 0.13621, + "grad_norm": 0.6610391598682382, + "learning_rate": 0.003, + "loss": 4.007, + "step": 13621 + }, + { + "epoch": 0.13622, + "grad_norm": 0.5407256086317146, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 13622 + }, + { + "epoch": 0.13623, + "grad_norm": 0.5919252722507112, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 13623 + }, + { + "epoch": 0.13624, + "grad_norm": 0.6102125663938333, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 13624 + }, + { + "epoch": 0.13625, + "grad_norm": 0.6164825487579109, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 13625 + }, + { + "epoch": 0.13626, + "grad_norm": 0.598667855596343, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 13626 + }, + { + "epoch": 0.13627, + "grad_norm": 0.6022913331974324, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 13627 + }, + { + "epoch": 0.13628, + "grad_norm": 0.6076674253888676, + "learning_rate": 0.003, + "loss": 4.021, + "step": 13628 + }, + { + "epoch": 0.13629, + "grad_norm": 0.6627390062830559, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 13629 + }, + { + "epoch": 0.1363, + "grad_norm": 0.662445434314985, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 13630 + }, + { + "epoch": 0.13631, + "grad_norm": 0.6447495478508664, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 13631 + }, + { + "epoch": 0.13632, + "grad_norm": 0.693485648234495, + "learning_rate": 0.003, + "loss": 4.042, + "step": 13632 + }, + { + "epoch": 0.13633, + "grad_norm": 0.7849047544921176, + "learning_rate": 0.003, + "loss": 4.01, + "step": 13633 + }, + { + "epoch": 0.13634, + "grad_norm": 0.9094419803089064, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 13634 + }, + { + "epoch": 0.13635, + "grad_norm": 0.9268330287550955, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 13635 + }, + { + "epoch": 0.13636, + "grad_norm": 0.9968731824320711, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 13636 + }, + { + "epoch": 0.13637, + "grad_norm": 1.136306179230646, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 13637 + }, + { + "epoch": 0.13638, + "grad_norm": 0.813104309313599, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 13638 + }, + { + "epoch": 0.13639, + "grad_norm": 0.6560460465377115, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 13639 + }, + { + "epoch": 0.1364, + "grad_norm": 0.6256705246202008, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 13640 + }, + { + "epoch": 0.13641, + "grad_norm": 0.66500192280199, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 13641 + }, + { + "epoch": 0.13642, + "grad_norm": 0.7841297792717713, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 13642 + }, + { + "epoch": 0.13643, + "grad_norm": 0.8470138356529374, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 13643 + }, + { + "epoch": 0.13644, + "grad_norm": 0.8259123924217598, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 13644 + }, + { + "epoch": 0.13645, + "grad_norm": 0.6544581315622752, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 13645 + }, + { + "epoch": 0.13646, + "grad_norm": 0.5809390585421675, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 13646 + }, + { + "epoch": 0.13647, + "grad_norm": 0.6213784958450691, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 13647 + }, + { + "epoch": 0.13648, + "grad_norm": 0.5644188403857194, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 13648 + }, + { + "epoch": 0.13649, + "grad_norm": 0.5166327197251017, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13649 + }, + { + "epoch": 0.1365, + "grad_norm": 0.592825443572216, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 13650 + }, + { + "epoch": 0.13651, + "grad_norm": 0.7314565738431891, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 13651 + }, + { + "epoch": 0.13652, + "grad_norm": 0.8376529112272527, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 13652 + }, + { + "epoch": 0.13653, + "grad_norm": 0.8751704544218128, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 13653 + }, + { + "epoch": 0.13654, + "grad_norm": 0.9222294509627254, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 13654 + }, + { + "epoch": 0.13655, + "grad_norm": 1.0124381976655445, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 13655 + }, + { + "epoch": 0.13656, + "grad_norm": 0.8928368651835986, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 13656 + }, + { + "epoch": 0.13657, + "grad_norm": 0.7473318353120795, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 13657 + }, + { + "epoch": 0.13658, + "grad_norm": 0.6662727623806982, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 13658 + }, + { + "epoch": 0.13659, + "grad_norm": 0.5974735934036972, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 13659 + }, + { + "epoch": 0.1366, + "grad_norm": 0.6208916853993167, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 13660 + }, + { + "epoch": 0.13661, + "grad_norm": 0.5949643152869142, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 13661 + }, + { + "epoch": 0.13662, + "grad_norm": 0.640664525359715, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 13662 + }, + { + "epoch": 0.13663, + "grad_norm": 0.6793404264266538, + "learning_rate": 0.003, + "loss": 4.03, + "step": 13663 + }, + { + "epoch": 0.13664, + "grad_norm": 0.6981956075556591, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 13664 + }, + { + "epoch": 0.13665, + "grad_norm": 0.7835365812090614, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 13665 + }, + { + "epoch": 0.13666, + "grad_norm": 0.7618174229887174, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 13666 + }, + { + "epoch": 0.13667, + "grad_norm": 0.679937087845774, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 13667 + }, + { + "epoch": 0.13668, + "grad_norm": 0.5745918172024777, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 13668 + }, + { + "epoch": 0.13669, + "grad_norm": 0.6065213035005855, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 13669 + }, + { + "epoch": 0.1367, + "grad_norm": 0.6549879717143664, + "learning_rate": 0.003, + "loss": 4.025, + "step": 13670 + }, + { + "epoch": 0.13671, + "grad_norm": 0.6234225890487202, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 13671 + }, + { + "epoch": 0.13672, + "grad_norm": 0.7167107275888067, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 13672 + }, + { + "epoch": 0.13673, + "grad_norm": 0.7817779832270885, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 13673 + }, + { + "epoch": 0.13674, + "grad_norm": 0.8592302240983272, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 13674 + }, + { + "epoch": 0.13675, + "grad_norm": 0.8474463515652807, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 13675 + }, + { + "epoch": 0.13676, + "grad_norm": 0.7508599270008647, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 13676 + }, + { + "epoch": 0.13677, + "grad_norm": 0.7681990443611532, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 13677 + }, + { + "epoch": 0.13678, + "grad_norm": 0.7993575590606922, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13678 + }, + { + "epoch": 0.13679, + "grad_norm": 0.80411893558822, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 13679 + }, + { + "epoch": 0.1368, + "grad_norm": 0.6971302956626652, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 13680 + }, + { + "epoch": 0.13681, + "grad_norm": 0.6200184176747422, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 13681 + }, + { + "epoch": 0.13682, + "grad_norm": 0.5467294256647931, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 13682 + }, + { + "epoch": 0.13683, + "grad_norm": 0.592722328613346, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 13683 + }, + { + "epoch": 0.13684, + "grad_norm": 0.70057808526398, + "learning_rate": 0.003, + "loss": 4.019, + "step": 13684 + }, + { + "epoch": 0.13685, + "grad_norm": 0.7924542682615539, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 13685 + }, + { + "epoch": 0.13686, + "grad_norm": 0.8856735300566443, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 13686 + }, + { + "epoch": 0.13687, + "grad_norm": 0.906658678579654, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13687 + }, + { + "epoch": 0.13688, + "grad_norm": 0.912359871289803, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 13688 + }, + { + "epoch": 0.13689, + "grad_norm": 0.8362278163963015, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 13689 + }, + { + "epoch": 0.1369, + "grad_norm": 0.7977891292072903, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 13690 + }, + { + "epoch": 0.13691, + "grad_norm": 0.7431998694559533, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 13691 + }, + { + "epoch": 0.13692, + "grad_norm": 0.76416432846593, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 13692 + }, + { + "epoch": 0.13693, + "grad_norm": 0.6968884810853461, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 13693 + }, + { + "epoch": 0.13694, + "grad_norm": 0.6478088312773628, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 13694 + }, + { + "epoch": 0.13695, + "grad_norm": 0.6318247493311704, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 13695 + }, + { + "epoch": 0.13696, + "grad_norm": 0.6263600972013175, + "learning_rate": 0.003, + "loss": 4.022, + "step": 13696 + }, + { + "epoch": 0.13697, + "grad_norm": 0.6791284173429935, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 13697 + }, + { + "epoch": 0.13698, + "grad_norm": 0.6957041387391977, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 13698 + }, + { + "epoch": 0.13699, + "grad_norm": 0.7047274891351449, + "learning_rate": 0.003, + "loss": 4.026, + "step": 13699 + }, + { + "epoch": 0.137, + "grad_norm": 0.7180654881779267, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 13700 + }, + { + "epoch": 0.13701, + "grad_norm": 0.7789974275788147, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 13701 + }, + { + "epoch": 0.13702, + "grad_norm": 0.7324224357087941, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 13702 + }, + { + "epoch": 0.13703, + "grad_norm": 0.7934964566790187, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 13703 + }, + { + "epoch": 0.13704, + "grad_norm": 0.9007785655608631, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 13704 + }, + { + "epoch": 0.13705, + "grad_norm": 1.0853051015256747, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 13705 + }, + { + "epoch": 0.13706, + "grad_norm": 1.0521224141250152, + "learning_rate": 0.003, + "loss": 4.026, + "step": 13706 + }, + { + "epoch": 0.13707, + "grad_norm": 0.9654048557664204, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 13707 + }, + { + "epoch": 0.13708, + "grad_norm": 0.8097248608999777, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 13708 + }, + { + "epoch": 0.13709, + "grad_norm": 0.7639670715202757, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13709 + }, + { + "epoch": 0.1371, + "grad_norm": 0.8879301911129639, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 13710 + }, + { + "epoch": 0.13711, + "grad_norm": 0.9210861560555238, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 13711 + }, + { + "epoch": 0.13712, + "grad_norm": 0.7409477098740044, + "learning_rate": 0.003, + "loss": 4.005, + "step": 13712 + }, + { + "epoch": 0.13713, + "grad_norm": 0.64405543160624, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 13713 + }, + { + "epoch": 0.13714, + "grad_norm": 0.6821642018353136, + "learning_rate": 0.003, + "loss": 4.013, + "step": 13714 + }, + { + "epoch": 0.13715, + "grad_norm": 0.7813606443781443, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 13715 + }, + { + "epoch": 0.13716, + "grad_norm": 0.7493916034635064, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 13716 + }, + { + "epoch": 0.13717, + "grad_norm": 0.7486198659803514, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 13717 + }, + { + "epoch": 0.13718, + "grad_norm": 0.773480074813737, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 13718 + }, + { + "epoch": 0.13719, + "grad_norm": 0.720930376427238, + "learning_rate": 0.003, + "loss": 4.015, + "step": 13719 + }, + { + "epoch": 0.1372, + "grad_norm": 0.6184031573369321, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 13720 + }, + { + "epoch": 0.13721, + "grad_norm": 0.6246672794333823, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 13721 + }, + { + "epoch": 0.13722, + "grad_norm": 0.5456587610753109, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 13722 + }, + { + "epoch": 0.13723, + "grad_norm": 0.5086301233464817, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 13723 + }, + { + "epoch": 0.13724, + "grad_norm": 0.5407213366922875, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 13724 + }, + { + "epoch": 0.13725, + "grad_norm": 0.5419147266873805, + "learning_rate": 0.003, + "loss": 4.035, + "step": 13725 + }, + { + "epoch": 0.13726, + "grad_norm": 0.5719940829269515, + "learning_rate": 0.003, + "loss": 4.018, + "step": 13726 + }, + { + "epoch": 0.13727, + "grad_norm": 0.6318436821645593, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 13727 + }, + { + "epoch": 0.13728, + "grad_norm": 0.6791168479206282, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 13728 + }, + { + "epoch": 0.13729, + "grad_norm": 0.7126809533593456, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 13729 + }, + { + "epoch": 0.1373, + "grad_norm": 0.7555875261771022, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 13730 + }, + { + "epoch": 0.13731, + "grad_norm": 0.6796972676402867, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 13731 + }, + { + "epoch": 0.13732, + "grad_norm": 0.6064545581044645, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 13732 + }, + { + "epoch": 0.13733, + "grad_norm": 0.5373380797268152, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 13733 + }, + { + "epoch": 0.13734, + "grad_norm": 0.5672963908442409, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 13734 + }, + { + "epoch": 0.13735, + "grad_norm": 0.5991802681703279, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 13735 + }, + { + "epoch": 0.13736, + "grad_norm": 0.650246316519405, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 13736 + }, + { + "epoch": 0.13737, + "grad_norm": 0.7213250295765571, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 13737 + }, + { + "epoch": 0.13738, + "grad_norm": 0.8711766918686296, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 13738 + }, + { + "epoch": 0.13739, + "grad_norm": 1.0251026371916467, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 13739 + }, + { + "epoch": 0.1374, + "grad_norm": 1.0274333761000687, + "learning_rate": 0.003, + "loss": 4.009, + "step": 13740 + }, + { + "epoch": 0.13741, + "grad_norm": 0.8540783524339859, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 13741 + }, + { + "epoch": 0.13742, + "grad_norm": 0.6503518938436973, + "learning_rate": 0.003, + "loss": 4.002, + "step": 13742 + }, + { + "epoch": 0.13743, + "grad_norm": 0.5698580378891415, + "learning_rate": 0.003, + "loss": 3.984, + "step": 13743 + }, + { + "epoch": 0.13744, + "grad_norm": 0.5460997642039072, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 13744 + }, + { + "epoch": 0.13745, + "grad_norm": 0.6177101605219046, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 13745 + }, + { + "epoch": 0.13746, + "grad_norm": 0.6164293807942887, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 13746 + }, + { + "epoch": 0.13747, + "grad_norm": 0.7046244957338004, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 13747 + }, + { + "epoch": 0.13748, + "grad_norm": 0.7955755154571209, + "learning_rate": 0.003, + "loss": 4.0547, + "step": 13748 + }, + { + "epoch": 0.13749, + "grad_norm": 0.8564609126096621, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 13749 + }, + { + "epoch": 0.1375, + "grad_norm": 0.8197712893209731, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 13750 + }, + { + "epoch": 0.13751, + "grad_norm": 0.8753520764219259, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 13751 + }, + { + "epoch": 0.13752, + "grad_norm": 0.8440601682100642, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 13752 + }, + { + "epoch": 0.13753, + "grad_norm": 0.8082724559338514, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 13753 + }, + { + "epoch": 0.13754, + "grad_norm": 0.8010457209655245, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 13754 + }, + { + "epoch": 0.13755, + "grad_norm": 0.7431692408458762, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 13755 + }, + { + "epoch": 0.13756, + "grad_norm": 0.6471189016728103, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 13756 + }, + { + "epoch": 0.13757, + "grad_norm": 0.6427678207163354, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 13757 + }, + { + "epoch": 0.13758, + "grad_norm": 0.5788470424691938, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 13758 + }, + { + "epoch": 0.13759, + "grad_norm": 0.6246657370515148, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 13759 + }, + { + "epoch": 0.1376, + "grad_norm": 0.6937896394472727, + "learning_rate": 0.003, + "loss": 4.039, + "step": 13760 + }, + { + "epoch": 0.13761, + "grad_norm": 0.7459825528070371, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 13761 + }, + { + "epoch": 0.13762, + "grad_norm": 0.7850052246779796, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 13762 + }, + { + "epoch": 0.13763, + "grad_norm": 0.8699389671759125, + "learning_rate": 0.003, + "loss": 4.0566, + "step": 13763 + }, + { + "epoch": 0.13764, + "grad_norm": 0.9159651005475722, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 13764 + }, + { + "epoch": 0.13765, + "grad_norm": 0.8585752455107407, + "learning_rate": 0.003, + "loss": 4.026, + "step": 13765 + }, + { + "epoch": 0.13766, + "grad_norm": 0.7458975898531838, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 13766 + }, + { + "epoch": 0.13767, + "grad_norm": 0.7265457202198387, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 13767 + }, + { + "epoch": 0.13768, + "grad_norm": 0.6547244851041687, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 13768 + }, + { + "epoch": 0.13769, + "grad_norm": 0.6351582059135239, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 13769 + }, + { + "epoch": 0.1377, + "grad_norm": 0.5757891786813515, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13770 + }, + { + "epoch": 0.13771, + "grad_norm": 0.5362808383385976, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 13771 + }, + { + "epoch": 0.13772, + "grad_norm": 0.5286146961338919, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 13772 + }, + { + "epoch": 0.13773, + "grad_norm": 0.5942756231808635, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 13773 + }, + { + "epoch": 0.13774, + "grad_norm": 0.6447099495782918, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 13774 + }, + { + "epoch": 0.13775, + "grad_norm": 0.7561193849236731, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 13775 + }, + { + "epoch": 0.13776, + "grad_norm": 0.8202906624453573, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 13776 + }, + { + "epoch": 0.13777, + "grad_norm": 0.7994042899742461, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 13777 + }, + { + "epoch": 0.13778, + "grad_norm": 0.7485461498378547, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 13778 + }, + { + "epoch": 0.13779, + "grad_norm": 0.7574757827130447, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 13779 + }, + { + "epoch": 0.1378, + "grad_norm": 0.777440005590254, + "learning_rate": 0.003, + "loss": 4.023, + "step": 13780 + }, + { + "epoch": 0.13781, + "grad_norm": 0.8066005281188913, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 13781 + }, + { + "epoch": 0.13782, + "grad_norm": 0.7963540446439237, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 13782 + }, + { + "epoch": 0.13783, + "grad_norm": 0.8862950427697684, + "learning_rate": 0.003, + "loss": 4.0679, + "step": 13783 + }, + { + "epoch": 0.13784, + "grad_norm": 0.8732138929164774, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 13784 + }, + { + "epoch": 0.13785, + "grad_norm": 0.842097983121478, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 13785 + }, + { + "epoch": 0.13786, + "grad_norm": 0.8545822720429108, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 13786 + }, + { + "epoch": 0.13787, + "grad_norm": 0.8170618628197434, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 13787 + }, + { + "epoch": 0.13788, + "grad_norm": 0.8388319577765193, + "learning_rate": 0.003, + "loss": 3.993, + "step": 13788 + }, + { + "epoch": 0.13789, + "grad_norm": 0.9489768919417784, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 13789 + }, + { + "epoch": 0.1379, + "grad_norm": 0.8322860899032397, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 13790 + }, + { + "epoch": 0.13791, + "grad_norm": 0.8132763186890463, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 13791 + }, + { + "epoch": 0.13792, + "grad_norm": 0.723346891019563, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 13792 + }, + { + "epoch": 0.13793, + "grad_norm": 0.7758246019151088, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 13793 + }, + { + "epoch": 0.13794, + "grad_norm": 0.8186145820676413, + "learning_rate": 0.003, + "loss": 4.0513, + "step": 13794 + }, + { + "epoch": 0.13795, + "grad_norm": 0.7775806820765366, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 13795 + }, + { + "epoch": 0.13796, + "grad_norm": 0.7316697443602899, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 13796 + }, + { + "epoch": 0.13797, + "grad_norm": 0.6380722611345784, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 13797 + }, + { + "epoch": 0.13798, + "grad_norm": 0.5707142621970207, + "learning_rate": 0.003, + "loss": 4.013, + "step": 13798 + }, + { + "epoch": 0.13799, + "grad_norm": 0.5084406928440293, + "learning_rate": 0.003, + "loss": 4.008, + "step": 13799 + }, + { + "epoch": 0.138, + "grad_norm": 0.4695012049150661, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 13800 + }, + { + "epoch": 0.13801, + "grad_norm": 0.4391919974114524, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 13801 + }, + { + "epoch": 0.13802, + "grad_norm": 0.44390544543846594, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 13802 + }, + { + "epoch": 0.13803, + "grad_norm": 0.4493078340426989, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 13803 + }, + { + "epoch": 0.13804, + "grad_norm": 0.4733760822903776, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 13804 + }, + { + "epoch": 0.13805, + "grad_norm": 0.5080228969284961, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 13805 + }, + { + "epoch": 0.13806, + "grad_norm": 0.5299254506276755, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 13806 + }, + { + "epoch": 0.13807, + "grad_norm": 0.5625092971658748, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 13807 + }, + { + "epoch": 0.13808, + "grad_norm": 0.6506318476284719, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 13808 + }, + { + "epoch": 0.13809, + "grad_norm": 0.8477543643168416, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 13809 + }, + { + "epoch": 0.1381, + "grad_norm": 0.9481724777814104, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 13810 + }, + { + "epoch": 0.13811, + "grad_norm": 0.9608341734159059, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 13811 + }, + { + "epoch": 0.13812, + "grad_norm": 0.8875837772623899, + "learning_rate": 0.003, + "loss": 4.033, + "step": 13812 + }, + { + "epoch": 0.13813, + "grad_norm": 0.8150257112990713, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 13813 + }, + { + "epoch": 0.13814, + "grad_norm": 0.8498007326436635, + "learning_rate": 0.003, + "loss": 4.04, + "step": 13814 + }, + { + "epoch": 0.13815, + "grad_norm": 0.8577455036967565, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 13815 + }, + { + "epoch": 0.13816, + "grad_norm": 0.9333110818234218, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 13816 + }, + { + "epoch": 0.13817, + "grad_norm": 0.9692016433934236, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 13817 + }, + { + "epoch": 0.13818, + "grad_norm": 1.1442781920548215, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 13818 + }, + { + "epoch": 0.13819, + "grad_norm": 0.8733598022836817, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 13819 + }, + { + "epoch": 0.1382, + "grad_norm": 0.685056559270909, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 13820 + }, + { + "epoch": 0.13821, + "grad_norm": 0.6140523410449736, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 13821 + }, + { + "epoch": 0.13822, + "grad_norm": 0.5848304592564204, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 13822 + }, + { + "epoch": 0.13823, + "grad_norm": 0.6127160037547363, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 13823 + }, + { + "epoch": 0.13824, + "grad_norm": 0.6333976376594586, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 13824 + }, + { + "epoch": 0.13825, + "grad_norm": 0.543586500441175, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 13825 + }, + { + "epoch": 0.13826, + "grad_norm": 0.5182856273331654, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 13826 + }, + { + "epoch": 0.13827, + "grad_norm": 0.5061680203589085, + "learning_rate": 0.003, + "loss": 4.025, + "step": 13827 + }, + { + "epoch": 0.13828, + "grad_norm": 0.5568282471149891, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 13828 + }, + { + "epoch": 0.13829, + "grad_norm": 0.41169997444987005, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 13829 + }, + { + "epoch": 0.1383, + "grad_norm": 0.4716590692160127, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 13830 + }, + { + "epoch": 0.13831, + "grad_norm": 0.5812486401050767, + "learning_rate": 0.003, + "loss": 3.979, + "step": 13831 + }, + { + "epoch": 0.13832, + "grad_norm": 0.6944882798305106, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 13832 + }, + { + "epoch": 0.13833, + "grad_norm": 0.7592330595640817, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 13833 + }, + { + "epoch": 0.13834, + "grad_norm": 0.7665140744905488, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 13834 + }, + { + "epoch": 0.13835, + "grad_norm": 0.7263292152741246, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 13835 + }, + { + "epoch": 0.13836, + "grad_norm": 0.6454184229116936, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 13836 + }, + { + "epoch": 0.13837, + "grad_norm": 0.613749575005249, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 13837 + }, + { + "epoch": 0.13838, + "grad_norm": 0.6567629006487988, + "learning_rate": 0.003, + "loss": 3.97, + "step": 13838 + }, + { + "epoch": 0.13839, + "grad_norm": 0.7392725200422674, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 13839 + }, + { + "epoch": 0.1384, + "grad_norm": 0.7991623545756277, + "learning_rate": 0.003, + "loss": 4.023, + "step": 13840 + }, + { + "epoch": 0.13841, + "grad_norm": 0.7910932504081978, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 13841 + }, + { + "epoch": 0.13842, + "grad_norm": 0.8303591037665269, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 13842 + }, + { + "epoch": 0.13843, + "grad_norm": 0.7627688707811032, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13843 + }, + { + "epoch": 0.13844, + "grad_norm": 0.758503972766407, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 13844 + }, + { + "epoch": 0.13845, + "grad_norm": 0.6955206283207671, + "learning_rate": 0.003, + "loss": 4.023, + "step": 13845 + }, + { + "epoch": 0.13846, + "grad_norm": 0.5551300926403062, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 13846 + }, + { + "epoch": 0.13847, + "grad_norm": 0.5492052970197238, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 13847 + }, + { + "epoch": 0.13848, + "grad_norm": 0.5005875004377431, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 13848 + }, + { + "epoch": 0.13849, + "grad_norm": 0.5737264294903552, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 13849 + }, + { + "epoch": 0.1385, + "grad_norm": 0.7793147562714341, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 13850 + }, + { + "epoch": 0.13851, + "grad_norm": 0.9499633860132708, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 13851 + }, + { + "epoch": 0.13852, + "grad_norm": 0.9297088743980244, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 13852 + }, + { + "epoch": 0.13853, + "grad_norm": 0.952421213359645, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 13853 + }, + { + "epoch": 0.13854, + "grad_norm": 0.8270435645342289, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 13854 + }, + { + "epoch": 0.13855, + "grad_norm": 0.7983426803422663, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13855 + }, + { + "epoch": 0.13856, + "grad_norm": 0.915572254660217, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 13856 + }, + { + "epoch": 0.13857, + "grad_norm": 0.9896093736588405, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 13857 + }, + { + "epoch": 0.13858, + "grad_norm": 0.9418687578222484, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 13858 + }, + { + "epoch": 0.13859, + "grad_norm": 0.9028053954795777, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 13859 + }, + { + "epoch": 0.1386, + "grad_norm": 0.8586334538530608, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 13860 + }, + { + "epoch": 0.13861, + "grad_norm": 0.8583849559894254, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 13861 + }, + { + "epoch": 0.13862, + "grad_norm": 0.8420448933776572, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 13862 + }, + { + "epoch": 0.13863, + "grad_norm": 0.9501858985770738, + "learning_rate": 0.003, + "loss": 4.0738, + "step": 13863 + }, + { + "epoch": 0.13864, + "grad_norm": 1.0778290410002105, + "learning_rate": 0.003, + "loss": 4.0665, + "step": 13864 + }, + { + "epoch": 0.13865, + "grad_norm": 0.928273025752994, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 13865 + }, + { + "epoch": 0.13866, + "grad_norm": 0.8501839988693616, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 13866 + }, + { + "epoch": 0.13867, + "grad_norm": 0.8951481797102301, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 13867 + }, + { + "epoch": 0.13868, + "grad_norm": 0.8299535378881177, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 13868 + }, + { + "epoch": 0.13869, + "grad_norm": 0.7337420772854183, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 13869 + }, + { + "epoch": 0.1387, + "grad_norm": 0.7133730430104673, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 13870 + }, + { + "epoch": 0.13871, + "grad_norm": 0.7350005803300973, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 13871 + }, + { + "epoch": 0.13872, + "grad_norm": 0.8089717509738659, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 13872 + }, + { + "epoch": 0.13873, + "grad_norm": 1.0062445432603773, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 13873 + }, + { + "epoch": 0.13874, + "grad_norm": 1.0004706912218573, + "learning_rate": 0.003, + "loss": 4.035, + "step": 13874 + }, + { + "epoch": 0.13875, + "grad_norm": 0.9321175366457213, + "learning_rate": 0.003, + "loss": 4.031, + "step": 13875 + }, + { + "epoch": 0.13876, + "grad_norm": 0.792041266193824, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 13876 + }, + { + "epoch": 0.13877, + "grad_norm": 0.7631868357238683, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 13877 + }, + { + "epoch": 0.13878, + "grad_norm": 0.6928987276198098, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 13878 + }, + { + "epoch": 0.13879, + "grad_norm": 0.6155992045967108, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 13879 + }, + { + "epoch": 0.1388, + "grad_norm": 0.532596623784297, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 13880 + }, + { + "epoch": 0.13881, + "grad_norm": 0.5135103334047668, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 13881 + }, + { + "epoch": 0.13882, + "grad_norm": 0.45308442494844925, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 13882 + }, + { + "epoch": 0.13883, + "grad_norm": 0.4085898811644597, + "learning_rate": 0.003, + "loss": 3.997, + "step": 13883 + }, + { + "epoch": 0.13884, + "grad_norm": 0.38729237784457066, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 13884 + }, + { + "epoch": 0.13885, + "grad_norm": 0.4000561041514549, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 13885 + }, + { + "epoch": 0.13886, + "grad_norm": 0.4398873984499189, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 13886 + }, + { + "epoch": 0.13887, + "grad_norm": 0.5375451666512951, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 13887 + }, + { + "epoch": 0.13888, + "grad_norm": 0.6229638055357107, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 13888 + }, + { + "epoch": 0.13889, + "grad_norm": 0.718864459504458, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 13889 + }, + { + "epoch": 0.1389, + "grad_norm": 0.7512869554936539, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 13890 + }, + { + "epoch": 0.13891, + "grad_norm": 0.8375363186941925, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 13891 + }, + { + "epoch": 0.13892, + "grad_norm": 0.9240838455375824, + "learning_rate": 0.003, + "loss": 4.031, + "step": 13892 + }, + { + "epoch": 0.13893, + "grad_norm": 0.8622220955536299, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 13893 + }, + { + "epoch": 0.13894, + "grad_norm": 0.7372979204964413, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 13894 + }, + { + "epoch": 0.13895, + "grad_norm": 0.7062160744261561, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 13895 + }, + { + "epoch": 0.13896, + "grad_norm": 0.6516650324483099, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 13896 + }, + { + "epoch": 0.13897, + "grad_norm": 0.655164673540387, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 13897 + }, + { + "epoch": 0.13898, + "grad_norm": 0.6263595290517927, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 13898 + }, + { + "epoch": 0.13899, + "grad_norm": 0.6756862794165895, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 13899 + }, + { + "epoch": 0.139, + "grad_norm": 0.8022896972477347, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 13900 + }, + { + "epoch": 0.13901, + "grad_norm": 0.8499487413858064, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 13901 + }, + { + "epoch": 0.13902, + "grad_norm": 0.8031261194733184, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 13902 + }, + { + "epoch": 0.13903, + "grad_norm": 0.673057759461689, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 13903 + }, + { + "epoch": 0.13904, + "grad_norm": 0.7144434747189939, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 13904 + }, + { + "epoch": 0.13905, + "grad_norm": 0.7003020584966921, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 13905 + }, + { + "epoch": 0.13906, + "grad_norm": 0.7911494271931697, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 13906 + }, + { + "epoch": 0.13907, + "grad_norm": 0.8401664899785832, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 13907 + }, + { + "epoch": 0.13908, + "grad_norm": 0.7420474522601677, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 13908 + }, + { + "epoch": 0.13909, + "grad_norm": 0.644471110413103, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 13909 + }, + { + "epoch": 0.1391, + "grad_norm": 0.5693095066618481, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 13910 + }, + { + "epoch": 0.13911, + "grad_norm": 0.5579622594902994, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 13911 + }, + { + "epoch": 0.13912, + "grad_norm": 0.528341378614972, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 13912 + }, + { + "epoch": 0.13913, + "grad_norm": 0.5497026473003875, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 13913 + }, + { + "epoch": 0.13914, + "grad_norm": 0.5115870726674325, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 13914 + }, + { + "epoch": 0.13915, + "grad_norm": 0.5173222742597209, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 13915 + }, + { + "epoch": 0.13916, + "grad_norm": 0.5355556699921926, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 13916 + }, + { + "epoch": 0.13917, + "grad_norm": 0.5337236169567687, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 13917 + }, + { + "epoch": 0.13918, + "grad_norm": 0.5657297891219311, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 13918 + }, + { + "epoch": 0.13919, + "grad_norm": 0.6457500390321996, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 13919 + }, + { + "epoch": 0.1392, + "grad_norm": 0.7364335707061344, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 13920 + }, + { + "epoch": 0.13921, + "grad_norm": 0.8318778104086771, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 13921 + }, + { + "epoch": 0.13922, + "grad_norm": 0.9191923688095273, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 13922 + }, + { + "epoch": 0.13923, + "grad_norm": 0.9213347011675236, + "learning_rate": 0.003, + "loss": 4.031, + "step": 13923 + }, + { + "epoch": 0.13924, + "grad_norm": 1.0052608887363743, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 13924 + }, + { + "epoch": 0.13925, + "grad_norm": 1.0217810639530271, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 13925 + }, + { + "epoch": 0.13926, + "grad_norm": 0.9983535186320083, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 13926 + }, + { + "epoch": 0.13927, + "grad_norm": 0.9244203443809373, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 13927 + }, + { + "epoch": 0.13928, + "grad_norm": 0.9087500476884947, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 13928 + }, + { + "epoch": 0.13929, + "grad_norm": 0.8559120369144148, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 13929 + }, + { + "epoch": 0.1393, + "grad_norm": 0.7963631265556835, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 13930 + }, + { + "epoch": 0.13931, + "grad_norm": 0.7760986770425912, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 13931 + }, + { + "epoch": 0.13932, + "grad_norm": 0.7241605437528538, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 13932 + }, + { + "epoch": 0.13933, + "grad_norm": 0.6590758119698631, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 13933 + }, + { + "epoch": 0.13934, + "grad_norm": 0.5789588916972791, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 13934 + }, + { + "epoch": 0.13935, + "grad_norm": 0.514777367176603, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 13935 + }, + { + "epoch": 0.13936, + "grad_norm": 0.5563143949624489, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 13936 + }, + { + "epoch": 0.13937, + "grad_norm": 0.541611677654234, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 13937 + }, + { + "epoch": 0.13938, + "grad_norm": 0.5462728316528889, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 13938 + }, + { + "epoch": 0.13939, + "grad_norm": 0.5671917167401295, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 13939 + }, + { + "epoch": 0.1394, + "grad_norm": 0.604297106424633, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 13940 + }, + { + "epoch": 0.13941, + "grad_norm": 0.705725801357931, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 13941 + }, + { + "epoch": 0.13942, + "grad_norm": 0.7091960193693262, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 13942 + }, + { + "epoch": 0.13943, + "grad_norm": 0.6804222607768725, + "learning_rate": 0.003, + "loss": 4.061, + "step": 13943 + }, + { + "epoch": 0.13944, + "grad_norm": 0.7292816193966758, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 13944 + }, + { + "epoch": 0.13945, + "grad_norm": 0.8625961079893217, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 13945 + }, + { + "epoch": 0.13946, + "grad_norm": 0.7848240747458877, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 13946 + }, + { + "epoch": 0.13947, + "grad_norm": 0.6375300298726121, + "learning_rate": 0.003, + "loss": 4.031, + "step": 13947 + }, + { + "epoch": 0.13948, + "grad_norm": 0.708315587351151, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 13948 + }, + { + "epoch": 0.13949, + "grad_norm": 0.6703724015598966, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 13949 + }, + { + "epoch": 0.1395, + "grad_norm": 0.546331777673317, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 13950 + }, + { + "epoch": 0.13951, + "grad_norm": 0.6511708882111715, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 13951 + }, + { + "epoch": 0.13952, + "grad_norm": 0.7472742274759865, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 13952 + }, + { + "epoch": 0.13953, + "grad_norm": 0.8371799978269091, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 13953 + }, + { + "epoch": 0.13954, + "grad_norm": 1.0852170097220617, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 13954 + }, + { + "epoch": 0.13955, + "grad_norm": 1.030544242932612, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 13955 + }, + { + "epoch": 0.13956, + "grad_norm": 0.8955812120958396, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 13956 + }, + { + "epoch": 0.13957, + "grad_norm": 0.8436899296944854, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 13957 + }, + { + "epoch": 0.13958, + "grad_norm": 0.7848697534499196, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 13958 + }, + { + "epoch": 0.13959, + "grad_norm": 0.8120349865864046, + "learning_rate": 0.003, + "loss": 4.027, + "step": 13959 + }, + { + "epoch": 0.1396, + "grad_norm": 0.8830648143241308, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 13960 + }, + { + "epoch": 0.13961, + "grad_norm": 1.0010682106626, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 13961 + }, + { + "epoch": 0.13962, + "grad_norm": 0.8922442996404836, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 13962 + }, + { + "epoch": 0.13963, + "grad_norm": 0.8176984221510363, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 13963 + }, + { + "epoch": 0.13964, + "grad_norm": 0.7422946542925123, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 13964 + }, + { + "epoch": 0.13965, + "grad_norm": 0.7835137951555141, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 13965 + }, + { + "epoch": 0.13966, + "grad_norm": 0.7491252342121353, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 13966 + }, + { + "epoch": 0.13967, + "grad_norm": 0.774356029275904, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 13967 + }, + { + "epoch": 0.13968, + "grad_norm": 0.8320601206532878, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 13968 + }, + { + "epoch": 0.13969, + "grad_norm": 0.850979843082506, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 13969 + }, + { + "epoch": 0.1397, + "grad_norm": 0.8282546334597546, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 13970 + }, + { + "epoch": 0.13971, + "grad_norm": 0.9072078198096526, + "learning_rate": 0.003, + "loss": 4.046, + "step": 13971 + }, + { + "epoch": 0.13972, + "grad_norm": 0.8340468156950211, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 13972 + }, + { + "epoch": 0.13973, + "grad_norm": 0.7398403188808734, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 13973 + }, + { + "epoch": 0.13974, + "grad_norm": 0.6794361793017033, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 13974 + }, + { + "epoch": 0.13975, + "grad_norm": 0.5658598269918268, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 13975 + }, + { + "epoch": 0.13976, + "grad_norm": 0.5015124013663629, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 13976 + }, + { + "epoch": 0.13977, + "grad_norm": 0.4837694493037047, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 13977 + }, + { + "epoch": 0.13978, + "grad_norm": 0.48463575464031744, + "learning_rate": 0.003, + "loss": 4.0699, + "step": 13978 + }, + { + "epoch": 0.13979, + "grad_norm": 0.5052168809122427, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 13979 + }, + { + "epoch": 0.1398, + "grad_norm": 0.5529967129524723, + "learning_rate": 0.003, + "loss": 4.014, + "step": 13980 + }, + { + "epoch": 0.13981, + "grad_norm": 0.5648181707103027, + "learning_rate": 0.003, + "loss": 4.01, + "step": 13981 + }, + { + "epoch": 0.13982, + "grad_norm": 0.6123195266647571, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 13982 + }, + { + "epoch": 0.13983, + "grad_norm": 0.5843983790628021, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 13983 + }, + { + "epoch": 0.13984, + "grad_norm": 0.5020040777073396, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 13984 + }, + { + "epoch": 0.13985, + "grad_norm": 0.48394399121040044, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 13985 + }, + { + "epoch": 0.13986, + "grad_norm": 0.5592083644848469, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 13986 + }, + { + "epoch": 0.13987, + "grad_norm": 0.645010110793866, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 13987 + }, + { + "epoch": 0.13988, + "grad_norm": 0.7544816545627352, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 13988 + }, + { + "epoch": 0.13989, + "grad_norm": 0.7130527292219444, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 13989 + }, + { + "epoch": 0.1399, + "grad_norm": 0.6439136532951192, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 13990 + }, + { + "epoch": 0.13991, + "grad_norm": 0.6797128045562497, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 13991 + }, + { + "epoch": 0.13992, + "grad_norm": 0.6916178598499666, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 13992 + }, + { + "epoch": 0.13993, + "grad_norm": 0.7338091391015132, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 13993 + }, + { + "epoch": 0.13994, + "grad_norm": 0.7942022076642611, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 13994 + }, + { + "epoch": 0.13995, + "grad_norm": 0.8939224347112373, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 13995 + }, + { + "epoch": 0.13996, + "grad_norm": 0.8505035272582242, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 13996 + }, + { + "epoch": 0.13997, + "grad_norm": 0.7695966244444477, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 13997 + }, + { + "epoch": 0.13998, + "grad_norm": 0.6777933903951951, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 13998 + }, + { + "epoch": 0.13999, + "grad_norm": 0.6683821851400994, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 13999 + }, + { + "epoch": 0.14, + "grad_norm": 0.7252972628093862, + "learning_rate": 0.003, + "loss": 3.992, + "step": 14000 + }, + { + "epoch": 0.14001, + "grad_norm": 0.7607482507173805, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 14001 + }, + { + "epoch": 0.14002, + "grad_norm": 0.7085410233753271, + "learning_rate": 0.003, + "loss": 4.026, + "step": 14002 + }, + { + "epoch": 0.14003, + "grad_norm": 0.8067904804101194, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 14003 + }, + { + "epoch": 0.14004, + "grad_norm": 0.8578840287719772, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14004 + }, + { + "epoch": 0.14005, + "grad_norm": 0.7463033767970478, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 14005 + }, + { + "epoch": 0.14006, + "grad_norm": 0.7242758968838895, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 14006 + }, + { + "epoch": 0.14007, + "grad_norm": 0.6967316371494908, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 14007 + }, + { + "epoch": 0.14008, + "grad_norm": 0.6613412712362361, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 14008 + }, + { + "epoch": 0.14009, + "grad_norm": 0.7193326506155809, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 14009 + }, + { + "epoch": 0.1401, + "grad_norm": 0.7354287887689765, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 14010 + }, + { + "epoch": 0.14011, + "grad_norm": 0.712521249155089, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 14011 + }, + { + "epoch": 0.14012, + "grad_norm": 0.7283959108627545, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 14012 + }, + { + "epoch": 0.14013, + "grad_norm": 0.7902141098195815, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 14013 + }, + { + "epoch": 0.14014, + "grad_norm": 0.8614493580617998, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 14014 + }, + { + "epoch": 0.14015, + "grad_norm": 0.9501345268268907, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 14015 + }, + { + "epoch": 0.14016, + "grad_norm": 0.9299669400077956, + "learning_rate": 0.003, + "loss": 4.039, + "step": 14016 + }, + { + "epoch": 0.14017, + "grad_norm": 0.9195512662773537, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 14017 + }, + { + "epoch": 0.14018, + "grad_norm": 0.9681109955473526, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 14018 + }, + { + "epoch": 0.14019, + "grad_norm": 0.9849712042348168, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 14019 + }, + { + "epoch": 0.1402, + "grad_norm": 0.9040973479793193, + "learning_rate": 0.003, + "loss": 3.997, + "step": 14020 + }, + { + "epoch": 0.14021, + "grad_norm": 0.9460625749295345, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 14021 + }, + { + "epoch": 0.14022, + "grad_norm": 0.9229715555892799, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 14022 + }, + { + "epoch": 0.14023, + "grad_norm": 0.9088769161924346, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 14023 + }, + { + "epoch": 0.14024, + "grad_norm": 0.893182095912496, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 14024 + }, + { + "epoch": 0.14025, + "grad_norm": 0.8119933804062973, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 14025 + }, + { + "epoch": 0.14026, + "grad_norm": 0.7180023185563564, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 14026 + }, + { + "epoch": 0.14027, + "grad_norm": 0.6633308211838227, + "learning_rate": 0.003, + "loss": 3.995, + "step": 14027 + }, + { + "epoch": 0.14028, + "grad_norm": 0.6275720999584222, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 14028 + }, + { + "epoch": 0.14029, + "grad_norm": 0.6103053667318443, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 14029 + }, + { + "epoch": 0.1403, + "grad_norm": 0.6941026222357968, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 14030 + }, + { + "epoch": 0.14031, + "grad_norm": 0.7666685822687981, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 14031 + }, + { + "epoch": 0.14032, + "grad_norm": 0.7836055598213464, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 14032 + }, + { + "epoch": 0.14033, + "grad_norm": 0.8140661329152998, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 14033 + }, + { + "epoch": 0.14034, + "grad_norm": 0.761726322662932, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 14034 + }, + { + "epoch": 0.14035, + "grad_norm": 0.7401101478472556, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 14035 + }, + { + "epoch": 0.14036, + "grad_norm": 0.8015474137519827, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 14036 + }, + { + "epoch": 0.14037, + "grad_norm": 0.7858720166559184, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 14037 + }, + { + "epoch": 0.14038, + "grad_norm": 0.7160577666484318, + "learning_rate": 0.003, + "loss": 4.0562, + "step": 14038 + }, + { + "epoch": 0.14039, + "grad_norm": 0.7027171626651753, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 14039 + }, + { + "epoch": 0.1404, + "grad_norm": 0.6182099081375584, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 14040 + }, + { + "epoch": 0.14041, + "grad_norm": 0.6539741175044093, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 14041 + }, + { + "epoch": 0.14042, + "grad_norm": 0.6502734001613191, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 14042 + }, + { + "epoch": 0.14043, + "grad_norm": 0.6398023042562474, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 14043 + }, + { + "epoch": 0.14044, + "grad_norm": 0.5907921435180011, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 14044 + }, + { + "epoch": 0.14045, + "grad_norm": 0.5512691976390254, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 14045 + }, + { + "epoch": 0.14046, + "grad_norm": 0.4634233832216649, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 14046 + }, + { + "epoch": 0.14047, + "grad_norm": 0.5138852860830836, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 14047 + }, + { + "epoch": 0.14048, + "grad_norm": 0.5717923824171107, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 14048 + }, + { + "epoch": 0.14049, + "grad_norm": 0.6658684060606377, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 14049 + }, + { + "epoch": 0.1405, + "grad_norm": 0.7516304757766884, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 14050 + }, + { + "epoch": 0.14051, + "grad_norm": 0.7109081789653716, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 14051 + }, + { + "epoch": 0.14052, + "grad_norm": 0.6646154704922388, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14052 + }, + { + "epoch": 0.14053, + "grad_norm": 0.6182306437182861, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 14053 + }, + { + "epoch": 0.14054, + "grad_norm": 0.5470199970499583, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 14054 + }, + { + "epoch": 0.14055, + "grad_norm": 0.5426450945175538, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 14055 + }, + { + "epoch": 0.14056, + "grad_norm": 0.5571005154958352, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 14056 + }, + { + "epoch": 0.14057, + "grad_norm": 0.6876160658434394, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 14057 + }, + { + "epoch": 0.14058, + "grad_norm": 0.9069604285069431, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 14058 + }, + { + "epoch": 0.14059, + "grad_norm": 1.123715019111587, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 14059 + }, + { + "epoch": 0.1406, + "grad_norm": 0.7001411305574646, + "learning_rate": 0.003, + "loss": 3.996, + "step": 14060 + }, + { + "epoch": 0.14061, + "grad_norm": 0.6473887195931202, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 14061 + }, + { + "epoch": 0.14062, + "grad_norm": 0.8055069993105584, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 14062 + }, + { + "epoch": 0.14063, + "grad_norm": 0.7528594911072536, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 14063 + }, + { + "epoch": 0.14064, + "grad_norm": 0.6700016662291299, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 14064 + }, + { + "epoch": 0.14065, + "grad_norm": 0.6327213510306543, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 14065 + }, + { + "epoch": 0.14066, + "grad_norm": 0.5869190295713974, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 14066 + }, + { + "epoch": 0.14067, + "grad_norm": 0.64276404649615, + "learning_rate": 0.003, + "loss": 4.004, + "step": 14067 + }, + { + "epoch": 0.14068, + "grad_norm": 0.7707726525304692, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 14068 + }, + { + "epoch": 0.14069, + "grad_norm": 0.7263929554952604, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 14069 + }, + { + "epoch": 0.1407, + "grad_norm": 0.5792193599116382, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 14070 + }, + { + "epoch": 0.14071, + "grad_norm": 0.572638558915296, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 14071 + }, + { + "epoch": 0.14072, + "grad_norm": 0.61031354299336, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 14072 + }, + { + "epoch": 0.14073, + "grad_norm": 0.6486875858202419, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 14073 + }, + { + "epoch": 0.14074, + "grad_norm": 0.7029316262058838, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 14074 + }, + { + "epoch": 0.14075, + "grad_norm": 0.8081678931730886, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 14075 + }, + { + "epoch": 0.14076, + "grad_norm": 0.8463286894283335, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 14076 + }, + { + "epoch": 0.14077, + "grad_norm": 0.9325790298118452, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 14077 + }, + { + "epoch": 0.14078, + "grad_norm": 0.922365797657744, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 14078 + }, + { + "epoch": 0.14079, + "grad_norm": 0.9420321395405301, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 14079 + }, + { + "epoch": 0.1408, + "grad_norm": 0.857728087631492, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 14080 + }, + { + "epoch": 0.14081, + "grad_norm": 0.8843713837360134, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 14081 + }, + { + "epoch": 0.14082, + "grad_norm": 0.8703391559465355, + "learning_rate": 0.003, + "loss": 4.042, + "step": 14082 + }, + { + "epoch": 0.14083, + "grad_norm": 0.9266068963748381, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 14083 + }, + { + "epoch": 0.14084, + "grad_norm": 0.9176892295929161, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 14084 + }, + { + "epoch": 0.14085, + "grad_norm": 0.8631726833196216, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14085 + }, + { + "epoch": 0.14086, + "grad_norm": 0.7930886889989198, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 14086 + }, + { + "epoch": 0.14087, + "grad_norm": 0.7342035274285613, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 14087 + }, + { + "epoch": 0.14088, + "grad_norm": 0.7721692164571742, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 14088 + }, + { + "epoch": 0.14089, + "grad_norm": 0.8103488158539102, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 14089 + }, + { + "epoch": 0.1409, + "grad_norm": 0.8003955314305883, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 14090 + }, + { + "epoch": 0.14091, + "grad_norm": 0.9073879274102784, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 14091 + }, + { + "epoch": 0.14092, + "grad_norm": 1.0295425570612085, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 14092 + }, + { + "epoch": 0.14093, + "grad_norm": 1.2239576292291843, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 14093 + }, + { + "epoch": 0.14094, + "grad_norm": 0.8255582577911282, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14094 + }, + { + "epoch": 0.14095, + "grad_norm": 0.7674181163380348, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 14095 + }, + { + "epoch": 0.14096, + "grad_norm": 0.7814802055588314, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 14096 + }, + { + "epoch": 0.14097, + "grad_norm": 0.6992508928655838, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 14097 + }, + { + "epoch": 0.14098, + "grad_norm": 0.5494557232487433, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 14098 + }, + { + "epoch": 0.14099, + "grad_norm": 0.5179618540844614, + "learning_rate": 0.003, + "loss": 4.035, + "step": 14099 + }, + { + "epoch": 0.141, + "grad_norm": 0.5402575428756918, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 14100 + }, + { + "epoch": 0.14101, + "grad_norm": 0.6150491167045326, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 14101 + }, + { + "epoch": 0.14102, + "grad_norm": 0.6164004199624318, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 14102 + }, + { + "epoch": 0.14103, + "grad_norm": 0.5751571608089439, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 14103 + }, + { + "epoch": 0.14104, + "grad_norm": 0.49965413824361765, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 14104 + }, + { + "epoch": 0.14105, + "grad_norm": 0.4946341236564402, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 14105 + }, + { + "epoch": 0.14106, + "grad_norm": 0.506842225473271, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 14106 + }, + { + "epoch": 0.14107, + "grad_norm": 0.5143418457351441, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 14107 + }, + { + "epoch": 0.14108, + "grad_norm": 0.5138621276059216, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 14108 + }, + { + "epoch": 0.14109, + "grad_norm": 0.49937234880337467, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14109 + }, + { + "epoch": 0.1411, + "grad_norm": 0.5332834435539163, + "learning_rate": 0.003, + "loss": 3.994, + "step": 14110 + }, + { + "epoch": 0.14111, + "grad_norm": 0.5829395952724279, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 14111 + }, + { + "epoch": 0.14112, + "grad_norm": 0.5634358223986027, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14112 + }, + { + "epoch": 0.14113, + "grad_norm": 0.59248152359256, + "learning_rate": 0.003, + "loss": 3.989, + "step": 14113 + }, + { + "epoch": 0.14114, + "grad_norm": 0.6099824186543078, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 14114 + }, + { + "epoch": 0.14115, + "grad_norm": 0.5700095859862626, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 14115 + }, + { + "epoch": 0.14116, + "grad_norm": 0.5710857001422991, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 14116 + }, + { + "epoch": 0.14117, + "grad_norm": 0.6482407595654168, + "learning_rate": 0.003, + "loss": 4.003, + "step": 14117 + }, + { + "epoch": 0.14118, + "grad_norm": 0.5971154775010189, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 14118 + }, + { + "epoch": 0.14119, + "grad_norm": 0.5485352306193342, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14119 + }, + { + "epoch": 0.1412, + "grad_norm": 0.6086022122416116, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 14120 + }, + { + "epoch": 0.14121, + "grad_norm": 0.7483377342911751, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14121 + }, + { + "epoch": 0.14122, + "grad_norm": 0.8937062820448419, + "learning_rate": 0.003, + "loss": 4.036, + "step": 14122 + }, + { + "epoch": 0.14123, + "grad_norm": 1.1429598488511874, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 14123 + }, + { + "epoch": 0.14124, + "grad_norm": 0.8374526372559008, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14124 + }, + { + "epoch": 0.14125, + "grad_norm": 0.7148167684016176, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 14125 + }, + { + "epoch": 0.14126, + "grad_norm": 0.7499453510044909, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 14126 + }, + { + "epoch": 0.14127, + "grad_norm": 0.7174348829290581, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 14127 + }, + { + "epoch": 0.14128, + "grad_norm": 0.6916563315737138, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 14128 + }, + { + "epoch": 0.14129, + "grad_norm": 0.6291421820034309, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 14129 + }, + { + "epoch": 0.1413, + "grad_norm": 0.6314758872532292, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 14130 + }, + { + "epoch": 0.14131, + "grad_norm": 0.6846908067770962, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 14131 + }, + { + "epoch": 0.14132, + "grad_norm": 0.7898201413233144, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 14132 + }, + { + "epoch": 0.14133, + "grad_norm": 0.7724325978180354, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 14133 + }, + { + "epoch": 0.14134, + "grad_norm": 0.7911870554106606, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 14134 + }, + { + "epoch": 0.14135, + "grad_norm": 0.774994929614526, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 14135 + }, + { + "epoch": 0.14136, + "grad_norm": 0.8779175345198498, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 14136 + }, + { + "epoch": 0.14137, + "grad_norm": 0.9517712128026882, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 14137 + }, + { + "epoch": 0.14138, + "grad_norm": 1.010967025186217, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 14138 + }, + { + "epoch": 0.14139, + "grad_norm": 1.1278987886801999, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 14139 + }, + { + "epoch": 0.1414, + "grad_norm": 0.8541035012331886, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 14140 + }, + { + "epoch": 0.14141, + "grad_norm": 0.8833792177455995, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 14141 + }, + { + "epoch": 0.14142, + "grad_norm": 0.7738278389390866, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 14142 + }, + { + "epoch": 0.14143, + "grad_norm": 0.7202938085683112, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 14143 + }, + { + "epoch": 0.14144, + "grad_norm": 0.7261226980837885, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 14144 + }, + { + "epoch": 0.14145, + "grad_norm": 0.6676760904151481, + "learning_rate": 0.003, + "loss": 4.036, + "step": 14145 + }, + { + "epoch": 0.14146, + "grad_norm": 0.6239302592819462, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 14146 + }, + { + "epoch": 0.14147, + "grad_norm": 0.5743088915952748, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 14147 + }, + { + "epoch": 0.14148, + "grad_norm": 0.4713025840500045, + "learning_rate": 0.003, + "loss": 4.022, + "step": 14148 + }, + { + "epoch": 0.14149, + "grad_norm": 0.50778283273304, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 14149 + }, + { + "epoch": 0.1415, + "grad_norm": 0.534208558143412, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 14150 + }, + { + "epoch": 0.14151, + "grad_norm": 0.5836024932513526, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 14151 + }, + { + "epoch": 0.14152, + "grad_norm": 0.707029925363811, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 14152 + }, + { + "epoch": 0.14153, + "grad_norm": 0.751215104459904, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 14153 + }, + { + "epoch": 0.14154, + "grad_norm": 0.8159419634011694, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 14154 + }, + { + "epoch": 0.14155, + "grad_norm": 0.8944179744821371, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 14155 + }, + { + "epoch": 0.14156, + "grad_norm": 1.115413094187359, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 14156 + }, + { + "epoch": 0.14157, + "grad_norm": 1.0644981345210291, + "learning_rate": 0.003, + "loss": 4.0485, + "step": 14157 + }, + { + "epoch": 0.14158, + "grad_norm": 0.8114976598098594, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 14158 + }, + { + "epoch": 0.14159, + "grad_norm": 0.785530252307587, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 14159 + }, + { + "epoch": 0.1416, + "grad_norm": 0.6989569983166278, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 14160 + }, + { + "epoch": 0.14161, + "grad_norm": 0.677755676693073, + "learning_rate": 0.003, + "loss": 4.0635, + "step": 14161 + }, + { + "epoch": 0.14162, + "grad_norm": 0.6321650058250659, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 14162 + }, + { + "epoch": 0.14163, + "grad_norm": 0.7627306579117497, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 14163 + }, + { + "epoch": 0.14164, + "grad_norm": 0.8306454450667825, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 14164 + }, + { + "epoch": 0.14165, + "grad_norm": 0.8280696260633121, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 14165 + }, + { + "epoch": 0.14166, + "grad_norm": 0.8353263298155297, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 14166 + }, + { + "epoch": 0.14167, + "grad_norm": 0.7618482567492684, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 14167 + }, + { + "epoch": 0.14168, + "grad_norm": 0.6136717919229085, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14168 + }, + { + "epoch": 0.14169, + "grad_norm": 0.6033853485903198, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 14169 + }, + { + "epoch": 0.1417, + "grad_norm": 0.5611070243360949, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 14170 + }, + { + "epoch": 0.14171, + "grad_norm": 0.4979086059966332, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14171 + }, + { + "epoch": 0.14172, + "grad_norm": 0.5461167173609474, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 14172 + }, + { + "epoch": 0.14173, + "grad_norm": 0.5728675889532433, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 14173 + }, + { + "epoch": 0.14174, + "grad_norm": 0.6673627767288272, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14174 + }, + { + "epoch": 0.14175, + "grad_norm": 0.9306340273961583, + "learning_rate": 0.003, + "loss": 4.039, + "step": 14175 + }, + { + "epoch": 0.14176, + "grad_norm": 1.126730626310538, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 14176 + }, + { + "epoch": 0.14177, + "grad_norm": 0.9259435299585543, + "learning_rate": 0.003, + "loss": 4.0585, + "step": 14177 + }, + { + "epoch": 0.14178, + "grad_norm": 0.7416609584414927, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 14178 + }, + { + "epoch": 0.14179, + "grad_norm": 0.6461180245304, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 14179 + }, + { + "epoch": 0.1418, + "grad_norm": 0.5834139813693796, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 14180 + }, + { + "epoch": 0.14181, + "grad_norm": 0.6556918083655104, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 14181 + }, + { + "epoch": 0.14182, + "grad_norm": 0.6790823222017613, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 14182 + }, + { + "epoch": 0.14183, + "grad_norm": 0.7283849850808959, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 14183 + }, + { + "epoch": 0.14184, + "grad_norm": 0.682170703152437, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 14184 + }, + { + "epoch": 0.14185, + "grad_norm": 0.6978363262036277, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 14185 + }, + { + "epoch": 0.14186, + "grad_norm": 0.8095372955438395, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14186 + }, + { + "epoch": 0.14187, + "grad_norm": 0.871560728882434, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 14187 + }, + { + "epoch": 0.14188, + "grad_norm": 0.818786205651462, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 14188 + }, + { + "epoch": 0.14189, + "grad_norm": 0.6588655272294687, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 14189 + }, + { + "epoch": 0.1419, + "grad_norm": 0.6412196879904046, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 14190 + }, + { + "epoch": 0.14191, + "grad_norm": 0.5793621757507376, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 14191 + }, + { + "epoch": 0.14192, + "grad_norm": 0.5316567197485605, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 14192 + }, + { + "epoch": 0.14193, + "grad_norm": 0.5485118159571154, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 14193 + }, + { + "epoch": 0.14194, + "grad_norm": 0.5827844911893088, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 14194 + }, + { + "epoch": 0.14195, + "grad_norm": 0.5633502325058611, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 14195 + }, + { + "epoch": 0.14196, + "grad_norm": 0.6184508416721164, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 14196 + }, + { + "epoch": 0.14197, + "grad_norm": 0.596989395909945, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 14197 + }, + { + "epoch": 0.14198, + "grad_norm": 0.5971917217070338, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 14198 + }, + { + "epoch": 0.14199, + "grad_norm": 0.7070962766371606, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 14199 + }, + { + "epoch": 0.142, + "grad_norm": 0.9609410219703043, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 14200 + }, + { + "epoch": 0.14201, + "grad_norm": 1.1159956486553615, + "learning_rate": 0.003, + "loss": 3.996, + "step": 14201 + }, + { + "epoch": 0.14202, + "grad_norm": 0.7429306329514537, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 14202 + }, + { + "epoch": 0.14203, + "grad_norm": 0.6223193700547376, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 14203 + }, + { + "epoch": 0.14204, + "grad_norm": 0.6287717346334032, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 14204 + }, + { + "epoch": 0.14205, + "grad_norm": 0.6304774788499377, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 14205 + }, + { + "epoch": 0.14206, + "grad_norm": 0.6372385534342372, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 14206 + }, + { + "epoch": 0.14207, + "grad_norm": 0.6579911754320352, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 14207 + }, + { + "epoch": 0.14208, + "grad_norm": 0.807598249591931, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 14208 + }, + { + "epoch": 0.14209, + "grad_norm": 0.8915201444992573, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 14209 + }, + { + "epoch": 0.1421, + "grad_norm": 0.7137055466041897, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 14210 + }, + { + "epoch": 0.14211, + "grad_norm": 0.6060521849266625, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 14211 + }, + { + "epoch": 0.14212, + "grad_norm": 0.734612425200557, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 14212 + }, + { + "epoch": 0.14213, + "grad_norm": 0.8930853574754026, + "learning_rate": 0.003, + "loss": 4.03, + "step": 14213 + }, + { + "epoch": 0.14214, + "grad_norm": 0.9705717973381859, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 14214 + }, + { + "epoch": 0.14215, + "grad_norm": 1.0543903196210285, + "learning_rate": 0.003, + "loss": 4.013, + "step": 14215 + }, + { + "epoch": 0.14216, + "grad_norm": 1.1830770283557992, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 14216 + }, + { + "epoch": 0.14217, + "grad_norm": 0.7289596190596122, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 14217 + }, + { + "epoch": 0.14218, + "grad_norm": 0.8330484748119504, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 14218 + }, + { + "epoch": 0.14219, + "grad_norm": 0.8531352936205642, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 14219 + }, + { + "epoch": 0.1422, + "grad_norm": 0.8148766789774047, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 14220 + }, + { + "epoch": 0.14221, + "grad_norm": 0.8899057272301019, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14221 + }, + { + "epoch": 0.14222, + "grad_norm": 0.8767078173835834, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 14222 + }, + { + "epoch": 0.14223, + "grad_norm": 0.8507353903702835, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 14223 + }, + { + "epoch": 0.14224, + "grad_norm": 0.9115031711002365, + "learning_rate": 0.003, + "loss": 3.99, + "step": 14224 + }, + { + "epoch": 0.14225, + "grad_norm": 0.8643154949680751, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 14225 + }, + { + "epoch": 0.14226, + "grad_norm": 0.7334070643373405, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 14226 + }, + { + "epoch": 0.14227, + "grad_norm": 0.6500163392042758, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 14227 + }, + { + "epoch": 0.14228, + "grad_norm": 0.6903112074474267, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 14228 + }, + { + "epoch": 0.14229, + "grad_norm": 0.745673233096331, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14229 + }, + { + "epoch": 0.1423, + "grad_norm": 0.7265477873850201, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 14230 + }, + { + "epoch": 0.14231, + "grad_norm": 0.8297094492437758, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 14231 + }, + { + "epoch": 0.14232, + "grad_norm": 0.9270578353747414, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 14232 + }, + { + "epoch": 0.14233, + "grad_norm": 0.8604702036124002, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 14233 + }, + { + "epoch": 0.14234, + "grad_norm": 0.7156950462491759, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 14234 + }, + { + "epoch": 0.14235, + "grad_norm": 0.6395473914613451, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 14235 + }, + { + "epoch": 0.14236, + "grad_norm": 0.5058998979709237, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 14236 + }, + { + "epoch": 0.14237, + "grad_norm": 0.5412229549304455, + "learning_rate": 0.003, + "loss": 4.008, + "step": 14237 + }, + { + "epoch": 0.14238, + "grad_norm": 0.5267387657139021, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 14238 + }, + { + "epoch": 0.14239, + "grad_norm": 0.5743381911054157, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 14239 + }, + { + "epoch": 0.1424, + "grad_norm": 0.5703061353354971, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 14240 + }, + { + "epoch": 0.14241, + "grad_norm": 0.6632525031274826, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 14241 + }, + { + "epoch": 0.14242, + "grad_norm": 0.7647975567479239, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 14242 + }, + { + "epoch": 0.14243, + "grad_norm": 0.806423292261851, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 14243 + }, + { + "epoch": 0.14244, + "grad_norm": 0.7754244779299369, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 14244 + }, + { + "epoch": 0.14245, + "grad_norm": 0.6683139542039896, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 14245 + }, + { + "epoch": 0.14246, + "grad_norm": 0.6459091038191279, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 14246 + }, + { + "epoch": 0.14247, + "grad_norm": 0.5666016699180112, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 14247 + }, + { + "epoch": 0.14248, + "grad_norm": 0.5428952509109771, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 14248 + }, + { + "epoch": 0.14249, + "grad_norm": 0.5080120638547413, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 14249 + }, + { + "epoch": 0.1425, + "grad_norm": 0.6081778861714088, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 14250 + }, + { + "epoch": 0.14251, + "grad_norm": 0.6712746863401664, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 14251 + }, + { + "epoch": 0.14252, + "grad_norm": 0.836578312594993, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 14252 + }, + { + "epoch": 0.14253, + "grad_norm": 0.8739805480836639, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 14253 + }, + { + "epoch": 0.14254, + "grad_norm": 0.7421909571151157, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 14254 + }, + { + "epoch": 0.14255, + "grad_norm": 0.7075198641873567, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14255 + }, + { + "epoch": 0.14256, + "grad_norm": 0.8027004820221948, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 14256 + }, + { + "epoch": 0.14257, + "grad_norm": 0.8352779861563754, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 14257 + }, + { + "epoch": 0.14258, + "grad_norm": 0.8052964604697136, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14258 + }, + { + "epoch": 0.14259, + "grad_norm": 0.9222648371552653, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 14259 + }, + { + "epoch": 0.1426, + "grad_norm": 0.9052598643232188, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 14260 + }, + { + "epoch": 0.14261, + "grad_norm": 0.8331713731679048, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 14261 + }, + { + "epoch": 0.14262, + "grad_norm": 0.7797006361244381, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 14262 + }, + { + "epoch": 0.14263, + "grad_norm": 0.7637788748094652, + "learning_rate": 0.003, + "loss": 4.03, + "step": 14263 + }, + { + "epoch": 0.14264, + "grad_norm": 0.7444075291282837, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 14264 + }, + { + "epoch": 0.14265, + "grad_norm": 0.7413290170686683, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 14265 + }, + { + "epoch": 0.14266, + "grad_norm": 0.6168915053499598, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 14266 + }, + { + "epoch": 0.14267, + "grad_norm": 0.6308961545491731, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 14267 + }, + { + "epoch": 0.14268, + "grad_norm": 0.6459999802737858, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 14268 + }, + { + "epoch": 0.14269, + "grad_norm": 0.6807379102330972, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 14269 + }, + { + "epoch": 0.1427, + "grad_norm": 0.6723556304091756, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 14270 + }, + { + "epoch": 0.14271, + "grad_norm": 0.7396334483865694, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 14271 + }, + { + "epoch": 0.14272, + "grad_norm": 0.8588555017641745, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 14272 + }, + { + "epoch": 0.14273, + "grad_norm": 0.9361961946325524, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 14273 + }, + { + "epoch": 0.14274, + "grad_norm": 0.9107781135344597, + "learning_rate": 0.003, + "loss": 4.034, + "step": 14274 + }, + { + "epoch": 0.14275, + "grad_norm": 0.8566514892722046, + "learning_rate": 0.003, + "loss": 4.025, + "step": 14275 + }, + { + "epoch": 0.14276, + "grad_norm": 1.003451794019602, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 14276 + }, + { + "epoch": 0.14277, + "grad_norm": 1.0752809232740164, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 14277 + }, + { + "epoch": 0.14278, + "grad_norm": 0.9197547860559877, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 14278 + }, + { + "epoch": 0.14279, + "grad_norm": 0.7832756956147118, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 14279 + }, + { + "epoch": 0.1428, + "grad_norm": 0.7344768715645955, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 14280 + }, + { + "epoch": 0.14281, + "grad_norm": 0.6981647495728134, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 14281 + }, + { + "epoch": 0.14282, + "grad_norm": 0.6871150068469846, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14282 + }, + { + "epoch": 0.14283, + "grad_norm": 0.5855814049669025, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 14283 + }, + { + "epoch": 0.14284, + "grad_norm": 0.62065093509835, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 14284 + }, + { + "epoch": 0.14285, + "grad_norm": 0.5722578960586522, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 14285 + }, + { + "epoch": 0.14286, + "grad_norm": 0.5633843635342873, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 14286 + }, + { + "epoch": 0.14287, + "grad_norm": 0.5514732737224249, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 14287 + }, + { + "epoch": 0.14288, + "grad_norm": 0.5601812810282304, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 14288 + }, + { + "epoch": 0.14289, + "grad_norm": 0.556640787466573, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 14289 + }, + { + "epoch": 0.1429, + "grad_norm": 0.5818933021799905, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 14290 + }, + { + "epoch": 0.14291, + "grad_norm": 0.5446886103636069, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 14291 + }, + { + "epoch": 0.14292, + "grad_norm": 0.5866551031280806, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 14292 + }, + { + "epoch": 0.14293, + "grad_norm": 0.7190853868345451, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 14293 + }, + { + "epoch": 0.14294, + "grad_norm": 0.7548457709794104, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 14294 + }, + { + "epoch": 0.14295, + "grad_norm": 0.6511803624824845, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 14295 + }, + { + "epoch": 0.14296, + "grad_norm": 0.6071490526297096, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 14296 + }, + { + "epoch": 0.14297, + "grad_norm": 0.696050055154502, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 14297 + }, + { + "epoch": 0.14298, + "grad_norm": 0.7256522437414765, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 14298 + }, + { + "epoch": 0.14299, + "grad_norm": 0.6534420973309141, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 14299 + }, + { + "epoch": 0.143, + "grad_norm": 0.7143590489407827, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 14300 + }, + { + "epoch": 0.14301, + "grad_norm": 0.8289055440220596, + "learning_rate": 0.003, + "loss": 4.03, + "step": 14301 + }, + { + "epoch": 0.14302, + "grad_norm": 0.8710491348483447, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 14302 + }, + { + "epoch": 0.14303, + "grad_norm": 1.0638832821847246, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 14303 + }, + { + "epoch": 0.14304, + "grad_norm": 0.9908775283926752, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 14304 + }, + { + "epoch": 0.14305, + "grad_norm": 0.8389169425229182, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 14305 + }, + { + "epoch": 0.14306, + "grad_norm": 0.7924747032497808, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 14306 + }, + { + "epoch": 0.14307, + "grad_norm": 0.8372693565904421, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 14307 + }, + { + "epoch": 0.14308, + "grad_norm": 0.8418582864922485, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 14308 + }, + { + "epoch": 0.14309, + "grad_norm": 0.7315386510997703, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 14309 + }, + { + "epoch": 0.1431, + "grad_norm": 0.6465516572663705, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 14310 + }, + { + "epoch": 0.14311, + "grad_norm": 0.6587675428647326, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 14311 + }, + { + "epoch": 0.14312, + "grad_norm": 0.7454374032461575, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 14312 + }, + { + "epoch": 0.14313, + "grad_norm": 0.834764813839016, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 14313 + }, + { + "epoch": 0.14314, + "grad_norm": 0.8197532420618233, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 14314 + }, + { + "epoch": 0.14315, + "grad_norm": 0.6906437275354359, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 14315 + }, + { + "epoch": 0.14316, + "grad_norm": 0.5747761896397774, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 14316 + }, + { + "epoch": 0.14317, + "grad_norm": 0.5659353092553077, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 14317 + }, + { + "epoch": 0.14318, + "grad_norm": 0.5833249810058427, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 14318 + }, + { + "epoch": 0.14319, + "grad_norm": 0.5022000949565524, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 14319 + }, + { + "epoch": 0.1432, + "grad_norm": 0.5257663836574404, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 14320 + }, + { + "epoch": 0.14321, + "grad_norm": 0.5286946069301452, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 14321 + }, + { + "epoch": 0.14322, + "grad_norm": 0.5795059999810741, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 14322 + }, + { + "epoch": 0.14323, + "grad_norm": 0.6564119028228382, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 14323 + }, + { + "epoch": 0.14324, + "grad_norm": 0.571825612130656, + "learning_rate": 0.003, + "loss": 4.004, + "step": 14324 + }, + { + "epoch": 0.14325, + "grad_norm": 0.5237642840817812, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 14325 + }, + { + "epoch": 0.14326, + "grad_norm": 0.5367228713022815, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 14326 + }, + { + "epoch": 0.14327, + "grad_norm": 0.6638009556818326, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 14327 + }, + { + "epoch": 0.14328, + "grad_norm": 0.8396578442998005, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 14328 + }, + { + "epoch": 0.14329, + "grad_norm": 1.0842837692593907, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14329 + }, + { + "epoch": 0.1433, + "grad_norm": 1.033838273642069, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 14330 + }, + { + "epoch": 0.14331, + "grad_norm": 0.7331803781896546, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 14331 + }, + { + "epoch": 0.14332, + "grad_norm": 0.5770473008848732, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 14332 + }, + { + "epoch": 0.14333, + "grad_norm": 0.6367155616444657, + "learning_rate": 0.003, + "loss": 4.002, + "step": 14333 + }, + { + "epoch": 0.14334, + "grad_norm": 0.7604432694501541, + "learning_rate": 0.003, + "loss": 4.014, + "step": 14334 + }, + { + "epoch": 0.14335, + "grad_norm": 0.8525818342171101, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 14335 + }, + { + "epoch": 0.14336, + "grad_norm": 0.9523588054706074, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 14336 + }, + { + "epoch": 0.14337, + "grad_norm": 0.9919432696854219, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 14337 + }, + { + "epoch": 0.14338, + "grad_norm": 0.9926288134207266, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 14338 + }, + { + "epoch": 0.14339, + "grad_norm": 0.9290323613026896, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 14339 + }, + { + "epoch": 0.1434, + "grad_norm": 0.8255884474618908, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 14340 + }, + { + "epoch": 0.14341, + "grad_norm": 0.8081628085550768, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 14341 + }, + { + "epoch": 0.14342, + "grad_norm": 0.7595564409424598, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14342 + }, + { + "epoch": 0.14343, + "grad_norm": 0.70640443808789, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 14343 + }, + { + "epoch": 0.14344, + "grad_norm": 0.674449490028374, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 14344 + }, + { + "epoch": 0.14345, + "grad_norm": 0.6423485544730745, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 14345 + }, + { + "epoch": 0.14346, + "grad_norm": 0.6252481575311553, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 14346 + }, + { + "epoch": 0.14347, + "grad_norm": 0.6504470358208538, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 14347 + }, + { + "epoch": 0.14348, + "grad_norm": 0.68522131456657, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 14348 + }, + { + "epoch": 0.14349, + "grad_norm": 0.6613443942620433, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 14349 + }, + { + "epoch": 0.1435, + "grad_norm": 0.6309605640768441, + "learning_rate": 0.003, + "loss": 4.013, + "step": 14350 + }, + { + "epoch": 0.14351, + "grad_norm": 0.5576303993971314, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 14351 + }, + { + "epoch": 0.14352, + "grad_norm": 0.5051782223133772, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 14352 + }, + { + "epoch": 0.14353, + "grad_norm": 0.5385350726355483, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14353 + }, + { + "epoch": 0.14354, + "grad_norm": 0.519550755738529, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 14354 + }, + { + "epoch": 0.14355, + "grad_norm": 0.6046362633543464, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 14355 + }, + { + "epoch": 0.14356, + "grad_norm": 0.6763414605242551, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 14356 + }, + { + "epoch": 0.14357, + "grad_norm": 0.7569970380958595, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 14357 + }, + { + "epoch": 0.14358, + "grad_norm": 0.8308832019314129, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 14358 + }, + { + "epoch": 0.14359, + "grad_norm": 0.8960490114155335, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 14359 + }, + { + "epoch": 0.1436, + "grad_norm": 0.9977730225855522, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 14360 + }, + { + "epoch": 0.14361, + "grad_norm": 0.9815045551772417, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 14361 + }, + { + "epoch": 0.14362, + "grad_norm": 0.8832772140648506, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 14362 + }, + { + "epoch": 0.14363, + "grad_norm": 0.7579020945686126, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 14363 + }, + { + "epoch": 0.14364, + "grad_norm": 0.697255137921127, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 14364 + }, + { + "epoch": 0.14365, + "grad_norm": 0.6804948539251883, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 14365 + }, + { + "epoch": 0.14366, + "grad_norm": 0.7295459974036163, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 14366 + }, + { + "epoch": 0.14367, + "grad_norm": 0.7701135049853245, + "learning_rate": 0.003, + "loss": 3.997, + "step": 14367 + }, + { + "epoch": 0.14368, + "grad_norm": 0.6988859571431073, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 14368 + }, + { + "epoch": 0.14369, + "grad_norm": 0.657731513728954, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 14369 + }, + { + "epoch": 0.1437, + "grad_norm": 0.6187355375847645, + "learning_rate": 0.003, + "loss": 3.993, + "step": 14370 + }, + { + "epoch": 0.14371, + "grad_norm": 0.6264205878629374, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 14371 + }, + { + "epoch": 0.14372, + "grad_norm": 0.6526898107700825, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 14372 + }, + { + "epoch": 0.14373, + "grad_norm": 0.6449212086450993, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 14373 + }, + { + "epoch": 0.14374, + "grad_norm": 0.6853005501748762, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 14374 + }, + { + "epoch": 0.14375, + "grad_norm": 0.7546292719969678, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 14375 + }, + { + "epoch": 0.14376, + "grad_norm": 0.8659333073440202, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 14376 + }, + { + "epoch": 0.14377, + "grad_norm": 0.8879285827352197, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 14377 + }, + { + "epoch": 0.14378, + "grad_norm": 0.8797032563212951, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 14378 + }, + { + "epoch": 0.14379, + "grad_norm": 1.016343540534367, + "learning_rate": 0.003, + "loss": 4.028, + "step": 14379 + }, + { + "epoch": 0.1438, + "grad_norm": 1.0134708043261262, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14380 + }, + { + "epoch": 0.14381, + "grad_norm": 1.0423014153983454, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 14381 + }, + { + "epoch": 0.14382, + "grad_norm": 0.9118373010875825, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 14382 + }, + { + "epoch": 0.14383, + "grad_norm": 0.7647508243937382, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 14383 + }, + { + "epoch": 0.14384, + "grad_norm": 0.8185870416554186, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 14384 + }, + { + "epoch": 0.14385, + "grad_norm": 0.8774185551231878, + "learning_rate": 0.003, + "loss": 4.0405, + "step": 14385 + }, + { + "epoch": 0.14386, + "grad_norm": 0.8194698614813727, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 14386 + }, + { + "epoch": 0.14387, + "grad_norm": 0.6983194159128269, + "learning_rate": 0.003, + "loss": 4.004, + "step": 14387 + }, + { + "epoch": 0.14388, + "grad_norm": 0.7251611942517602, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 14388 + }, + { + "epoch": 0.14389, + "grad_norm": 0.576280642568433, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 14389 + }, + { + "epoch": 0.1439, + "grad_norm": 0.6823738263874107, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 14390 + }, + { + "epoch": 0.14391, + "grad_norm": 0.6399041834585388, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 14391 + }, + { + "epoch": 0.14392, + "grad_norm": 0.6599999878065873, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 14392 + }, + { + "epoch": 0.14393, + "grad_norm": 0.5397647246136799, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 14393 + }, + { + "epoch": 0.14394, + "grad_norm": 0.491310936507617, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 14394 + }, + { + "epoch": 0.14395, + "grad_norm": 0.555640077437065, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 14395 + }, + { + "epoch": 0.14396, + "grad_norm": 0.6722324860595342, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 14396 + }, + { + "epoch": 0.14397, + "grad_norm": 0.8514371108504781, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 14397 + }, + { + "epoch": 0.14398, + "grad_norm": 0.9523040996142792, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 14398 + }, + { + "epoch": 0.14399, + "grad_norm": 0.8288809289653792, + "learning_rate": 0.003, + "loss": 3.999, + "step": 14399 + }, + { + "epoch": 0.144, + "grad_norm": 0.6138134165503308, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 14400 + }, + { + "epoch": 0.14401, + "grad_norm": 0.603467522203823, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 14401 + }, + { + "epoch": 0.14402, + "grad_norm": 0.6700138793170857, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 14402 + }, + { + "epoch": 0.14403, + "grad_norm": 0.5864185594614535, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 14403 + }, + { + "epoch": 0.14404, + "grad_norm": 0.5317919767413087, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 14404 + }, + { + "epoch": 0.14405, + "grad_norm": 0.48373839064599455, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 14405 + }, + { + "epoch": 0.14406, + "grad_norm": 0.4271399390417837, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 14406 + }, + { + "epoch": 0.14407, + "grad_norm": 0.4781877423779355, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 14407 + }, + { + "epoch": 0.14408, + "grad_norm": 0.49956254584512005, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 14408 + }, + { + "epoch": 0.14409, + "grad_norm": 0.49945888716783615, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 14409 + }, + { + "epoch": 0.1441, + "grad_norm": 0.518336535107641, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 14410 + }, + { + "epoch": 0.14411, + "grad_norm": 0.6035220192637003, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 14411 + }, + { + "epoch": 0.14412, + "grad_norm": 0.6942120303082241, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 14412 + }, + { + "epoch": 0.14413, + "grad_norm": 0.6602261941849155, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 14413 + }, + { + "epoch": 0.14414, + "grad_norm": 0.6840727606624321, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 14414 + }, + { + "epoch": 0.14415, + "grad_norm": 0.7342216632017093, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 14415 + }, + { + "epoch": 0.14416, + "grad_norm": 0.7213065751840104, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 14416 + }, + { + "epoch": 0.14417, + "grad_norm": 0.7440599940951941, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 14417 + }, + { + "epoch": 0.14418, + "grad_norm": 0.776481087821698, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14418 + }, + { + "epoch": 0.14419, + "grad_norm": 0.806621678703376, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 14419 + }, + { + "epoch": 0.1442, + "grad_norm": 0.7204002888452455, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 14420 + }, + { + "epoch": 0.14421, + "grad_norm": 0.7945594160605913, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 14421 + }, + { + "epoch": 0.14422, + "grad_norm": 0.830559516157465, + "learning_rate": 0.003, + "loss": 4.015, + "step": 14422 + }, + { + "epoch": 0.14423, + "grad_norm": 0.9226045503715983, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 14423 + }, + { + "epoch": 0.14424, + "grad_norm": 0.9803009793741705, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 14424 + }, + { + "epoch": 0.14425, + "grad_norm": 1.0333005197895075, + "learning_rate": 0.003, + "loss": 4.006, + "step": 14425 + }, + { + "epoch": 0.14426, + "grad_norm": 0.890645417576854, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 14426 + }, + { + "epoch": 0.14427, + "grad_norm": 0.8522632888644105, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 14427 + }, + { + "epoch": 0.14428, + "grad_norm": 0.8818788175275317, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 14428 + }, + { + "epoch": 0.14429, + "grad_norm": 0.8566228725250281, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 14429 + }, + { + "epoch": 0.1443, + "grad_norm": 0.769432128007338, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 14430 + }, + { + "epoch": 0.14431, + "grad_norm": 0.7575090714894748, + "learning_rate": 0.003, + "loss": 4.0521, + "step": 14431 + }, + { + "epoch": 0.14432, + "grad_norm": 0.8111890799986995, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 14432 + }, + { + "epoch": 0.14433, + "grad_norm": 0.8058821940025932, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 14433 + }, + { + "epoch": 0.14434, + "grad_norm": 0.8375896367984508, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 14434 + }, + { + "epoch": 0.14435, + "grad_norm": 0.9952785701235625, + "learning_rate": 0.003, + "loss": 4.049, + "step": 14435 + }, + { + "epoch": 0.14436, + "grad_norm": 0.9319380250674258, + "learning_rate": 0.003, + "loss": 4.0603, + "step": 14436 + }, + { + "epoch": 0.14437, + "grad_norm": 0.8065484158500317, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 14437 + }, + { + "epoch": 0.14438, + "grad_norm": 0.8015493264675314, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 14438 + }, + { + "epoch": 0.14439, + "grad_norm": 0.7788124716110962, + "learning_rate": 0.003, + "loss": 4.068, + "step": 14439 + }, + { + "epoch": 0.1444, + "grad_norm": 0.8882533346037006, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 14440 + }, + { + "epoch": 0.14441, + "grad_norm": 0.9637094760046029, + "learning_rate": 0.003, + "loss": 4.063, + "step": 14441 + }, + { + "epoch": 0.14442, + "grad_norm": 1.0542485151682868, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 14442 + }, + { + "epoch": 0.14443, + "grad_norm": 1.0675152490569604, + "learning_rate": 0.003, + "loss": 4.015, + "step": 14443 + }, + { + "epoch": 0.14444, + "grad_norm": 1.068655760621334, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 14444 + }, + { + "epoch": 0.14445, + "grad_norm": 0.7473739132205487, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 14445 + }, + { + "epoch": 0.14446, + "grad_norm": 0.6281665809254701, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 14446 + }, + { + "epoch": 0.14447, + "grad_norm": 0.5541523715087511, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 14447 + }, + { + "epoch": 0.14448, + "grad_norm": 0.547401998532125, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 14448 + }, + { + "epoch": 0.14449, + "grad_norm": 0.6002790406695873, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 14449 + }, + { + "epoch": 0.1445, + "grad_norm": 0.5822464483767282, + "learning_rate": 0.003, + "loss": 4.036, + "step": 14450 + }, + { + "epoch": 0.14451, + "grad_norm": 0.5796443600528274, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 14451 + }, + { + "epoch": 0.14452, + "grad_norm": 0.5763053929414035, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14452 + }, + { + "epoch": 0.14453, + "grad_norm": 0.5815588975756844, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 14453 + }, + { + "epoch": 0.14454, + "grad_norm": 0.5085654755920457, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 14454 + }, + { + "epoch": 0.14455, + "grad_norm": 0.45677890465485554, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 14455 + }, + { + "epoch": 0.14456, + "grad_norm": 0.4780626498860529, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 14456 + }, + { + "epoch": 0.14457, + "grad_norm": 0.4184026414065751, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 14457 + }, + { + "epoch": 0.14458, + "grad_norm": 0.38242252005883726, + "learning_rate": 0.003, + "loss": 3.994, + "step": 14458 + }, + { + "epoch": 0.14459, + "grad_norm": 0.3746142478434623, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 14459 + }, + { + "epoch": 0.1446, + "grad_norm": 0.396789272505137, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 14460 + }, + { + "epoch": 0.14461, + "grad_norm": 0.4522992170395448, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 14461 + }, + { + "epoch": 0.14462, + "grad_norm": 0.5335685882449777, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 14462 + }, + { + "epoch": 0.14463, + "grad_norm": 0.6592331809225305, + "learning_rate": 0.003, + "loss": 4.019, + "step": 14463 + }, + { + "epoch": 0.14464, + "grad_norm": 0.7858158749220078, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14464 + }, + { + "epoch": 0.14465, + "grad_norm": 0.9490084678688924, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 14465 + }, + { + "epoch": 0.14466, + "grad_norm": 1.0089569122057602, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14466 + }, + { + "epoch": 0.14467, + "grad_norm": 0.8384511228636733, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 14467 + }, + { + "epoch": 0.14468, + "grad_norm": 0.6275001101265367, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 14468 + }, + { + "epoch": 0.14469, + "grad_norm": 0.5510035437066796, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 14469 + }, + { + "epoch": 0.1447, + "grad_norm": 0.6339698269974869, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 14470 + }, + { + "epoch": 0.14471, + "grad_norm": 0.6415961079856902, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 14471 + }, + { + "epoch": 0.14472, + "grad_norm": 0.7329414673111437, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 14472 + }, + { + "epoch": 0.14473, + "grad_norm": 0.8289034323840451, + "learning_rate": 0.003, + "loss": 3.989, + "step": 14473 + }, + { + "epoch": 0.14474, + "grad_norm": 0.7900061172548047, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 14474 + }, + { + "epoch": 0.14475, + "grad_norm": 0.746534806100544, + "learning_rate": 0.003, + "loss": 4.041, + "step": 14475 + }, + { + "epoch": 0.14476, + "grad_norm": 0.7329569695474576, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 14476 + }, + { + "epoch": 0.14477, + "grad_norm": 0.7883929120682559, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 14477 + }, + { + "epoch": 0.14478, + "grad_norm": 0.7224199049083543, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 14478 + }, + { + "epoch": 0.14479, + "grad_norm": 0.689996140826465, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 14479 + }, + { + "epoch": 0.1448, + "grad_norm": 0.8504765366693361, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 14480 + }, + { + "epoch": 0.14481, + "grad_norm": 0.8185369527015324, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 14481 + }, + { + "epoch": 0.14482, + "grad_norm": 0.9079231424570736, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 14482 + }, + { + "epoch": 0.14483, + "grad_norm": 1.017576401185957, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 14483 + }, + { + "epoch": 0.14484, + "grad_norm": 0.8681745047956826, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 14484 + }, + { + "epoch": 0.14485, + "grad_norm": 0.7725501384937365, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 14485 + }, + { + "epoch": 0.14486, + "grad_norm": 0.7134432127979615, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 14486 + }, + { + "epoch": 0.14487, + "grad_norm": 0.706076873467523, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 14487 + }, + { + "epoch": 0.14488, + "grad_norm": 0.6918729230486471, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 14488 + }, + { + "epoch": 0.14489, + "grad_norm": 0.6989334521260058, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14489 + }, + { + "epoch": 0.1449, + "grad_norm": 0.7293389174766947, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 14490 + }, + { + "epoch": 0.14491, + "grad_norm": 0.8706084011174724, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 14491 + }, + { + "epoch": 0.14492, + "grad_norm": 0.9845208860723057, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 14492 + }, + { + "epoch": 0.14493, + "grad_norm": 1.051094919873643, + "learning_rate": 0.003, + "loss": 4.034, + "step": 14493 + }, + { + "epoch": 0.14494, + "grad_norm": 0.9705335938835851, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 14494 + }, + { + "epoch": 0.14495, + "grad_norm": 0.8927162337066793, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 14495 + }, + { + "epoch": 0.14496, + "grad_norm": 0.906339289596155, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 14496 + }, + { + "epoch": 0.14497, + "grad_norm": 0.7568510905226626, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 14497 + }, + { + "epoch": 0.14498, + "grad_norm": 0.719201592366577, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 14498 + }, + { + "epoch": 0.14499, + "grad_norm": 0.6314546141732514, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 14499 + }, + { + "epoch": 0.145, + "grad_norm": 0.6298408439133971, + "learning_rate": 0.003, + "loss": 4.0527, + "step": 14500 + }, + { + "epoch": 0.14501, + "grad_norm": 0.7531812135593454, + "learning_rate": 0.003, + "loss": 4.033, + "step": 14501 + }, + { + "epoch": 0.14502, + "grad_norm": 0.8425310404138221, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 14502 + }, + { + "epoch": 0.14503, + "grad_norm": 0.9070770739941331, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 14503 + }, + { + "epoch": 0.14504, + "grad_norm": 0.9093770151005623, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 14504 + }, + { + "epoch": 0.14505, + "grad_norm": 0.853336451846276, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 14505 + }, + { + "epoch": 0.14506, + "grad_norm": 0.7464505286153793, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 14506 + }, + { + "epoch": 0.14507, + "grad_norm": 0.6548605193827478, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 14507 + }, + { + "epoch": 0.14508, + "grad_norm": 0.6840532562126073, + "learning_rate": 0.003, + "loss": 4.004, + "step": 14508 + }, + { + "epoch": 0.14509, + "grad_norm": 0.6292278019937274, + "learning_rate": 0.003, + "loss": 4.033, + "step": 14509 + }, + { + "epoch": 0.1451, + "grad_norm": 0.5669775565766129, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 14510 + }, + { + "epoch": 0.14511, + "grad_norm": 0.5226732514936051, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 14511 + }, + { + "epoch": 0.14512, + "grad_norm": 0.519556217015763, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 14512 + }, + { + "epoch": 0.14513, + "grad_norm": 0.5787908426295746, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 14513 + }, + { + "epoch": 0.14514, + "grad_norm": 0.6559924305894286, + "learning_rate": 0.003, + "loss": 4.032, + "step": 14514 + }, + { + "epoch": 0.14515, + "grad_norm": 0.7847792278933329, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14515 + }, + { + "epoch": 0.14516, + "grad_norm": 0.8520177895194592, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 14516 + }, + { + "epoch": 0.14517, + "grad_norm": 0.7655481920756856, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 14517 + }, + { + "epoch": 0.14518, + "grad_norm": 0.8045638011990464, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 14518 + }, + { + "epoch": 0.14519, + "grad_norm": 0.8308273005966306, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 14519 + }, + { + "epoch": 0.1452, + "grad_norm": 0.8662653326127868, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 14520 + }, + { + "epoch": 0.14521, + "grad_norm": 0.8409089757765718, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 14521 + }, + { + "epoch": 0.14522, + "grad_norm": 0.7962354286725256, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 14522 + }, + { + "epoch": 0.14523, + "grad_norm": 0.7010851299474119, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 14523 + }, + { + "epoch": 0.14524, + "grad_norm": 0.6774573345014829, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 14524 + }, + { + "epoch": 0.14525, + "grad_norm": 0.6341566828657795, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 14525 + }, + { + "epoch": 0.14526, + "grad_norm": 0.6241642686980894, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 14526 + }, + { + "epoch": 0.14527, + "grad_norm": 0.6883551067367291, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 14527 + }, + { + "epoch": 0.14528, + "grad_norm": 0.8153423029838588, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14528 + }, + { + "epoch": 0.14529, + "grad_norm": 0.813945658729641, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 14529 + }, + { + "epoch": 0.1453, + "grad_norm": 0.7667769718454905, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 14530 + }, + { + "epoch": 0.14531, + "grad_norm": 0.6772286505990486, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 14531 + }, + { + "epoch": 0.14532, + "grad_norm": 0.5715075147243073, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 14532 + }, + { + "epoch": 0.14533, + "grad_norm": 0.7382754240898729, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 14533 + }, + { + "epoch": 0.14534, + "grad_norm": 0.7606483828493937, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 14534 + }, + { + "epoch": 0.14535, + "grad_norm": 0.6292245809162564, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 14535 + }, + { + "epoch": 0.14536, + "grad_norm": 0.5970634739074995, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 14536 + }, + { + "epoch": 0.14537, + "grad_norm": 0.6166299284204556, + "learning_rate": 0.003, + "loss": 3.991, + "step": 14537 + }, + { + "epoch": 0.14538, + "grad_norm": 0.546485161895953, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 14538 + }, + { + "epoch": 0.14539, + "grad_norm": 0.5762396846118211, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 14539 + }, + { + "epoch": 0.1454, + "grad_norm": 0.5321189787143538, + "learning_rate": 0.003, + "loss": 4.009, + "step": 14540 + }, + { + "epoch": 0.14541, + "grad_norm": 0.5265260644049988, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 14541 + }, + { + "epoch": 0.14542, + "grad_norm": 0.5412175593502493, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 14542 + }, + { + "epoch": 0.14543, + "grad_norm": 0.6046926387308197, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 14543 + }, + { + "epoch": 0.14544, + "grad_norm": 0.7270719502953031, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 14544 + }, + { + "epoch": 0.14545, + "grad_norm": 0.7114121716909256, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 14545 + }, + { + "epoch": 0.14546, + "grad_norm": 0.6522727485640463, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 14546 + }, + { + "epoch": 0.14547, + "grad_norm": 0.7147542119346554, + "learning_rate": 0.003, + "loss": 4.029, + "step": 14547 + }, + { + "epoch": 0.14548, + "grad_norm": 0.7717636891407744, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 14548 + }, + { + "epoch": 0.14549, + "grad_norm": 0.6980244272988703, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 14549 + }, + { + "epoch": 0.1455, + "grad_norm": 0.7505929371530453, + "learning_rate": 0.003, + "loss": 4.023, + "step": 14550 + }, + { + "epoch": 0.14551, + "grad_norm": 0.7438930296581573, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14551 + }, + { + "epoch": 0.14552, + "grad_norm": 0.7865486136617482, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 14552 + }, + { + "epoch": 0.14553, + "grad_norm": 0.8208998613118857, + "learning_rate": 0.003, + "loss": 4.037, + "step": 14553 + }, + { + "epoch": 0.14554, + "grad_norm": 0.8323802405075481, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 14554 + }, + { + "epoch": 0.14555, + "grad_norm": 0.8276240613175501, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 14555 + }, + { + "epoch": 0.14556, + "grad_norm": 0.753606852065734, + "learning_rate": 0.003, + "loss": 4.01, + "step": 14556 + }, + { + "epoch": 0.14557, + "grad_norm": 0.9589473886404751, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 14557 + }, + { + "epoch": 0.14558, + "grad_norm": 1.261213115194116, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 14558 + }, + { + "epoch": 0.14559, + "grad_norm": 0.7007563833297489, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 14559 + }, + { + "epoch": 0.1456, + "grad_norm": 0.7769417727001159, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 14560 + }, + { + "epoch": 0.14561, + "grad_norm": 0.8814119945244198, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 14561 + }, + { + "epoch": 0.14562, + "grad_norm": 0.9149846692563315, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 14562 + }, + { + "epoch": 0.14563, + "grad_norm": 0.9947925369139463, + "learning_rate": 0.003, + "loss": 4.043, + "step": 14563 + }, + { + "epoch": 0.14564, + "grad_norm": 1.00187007579209, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 14564 + }, + { + "epoch": 0.14565, + "grad_norm": 0.8612708863368196, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 14565 + }, + { + "epoch": 0.14566, + "grad_norm": 0.7921822820463634, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 14566 + }, + { + "epoch": 0.14567, + "grad_norm": 0.6977192960290127, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 14567 + }, + { + "epoch": 0.14568, + "grad_norm": 0.7390116074872014, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 14568 + }, + { + "epoch": 0.14569, + "grad_norm": 0.8278200352268891, + "learning_rate": 0.003, + "loss": 4.0554, + "step": 14569 + }, + { + "epoch": 0.1457, + "grad_norm": 0.8628144528607142, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 14570 + }, + { + "epoch": 0.14571, + "grad_norm": 0.8121354695177575, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 14571 + }, + { + "epoch": 0.14572, + "grad_norm": 0.7553706049868479, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14572 + }, + { + "epoch": 0.14573, + "grad_norm": 0.8163149894883702, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14573 + }, + { + "epoch": 0.14574, + "grad_norm": 0.6990740010086758, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 14574 + }, + { + "epoch": 0.14575, + "grad_norm": 0.6761581069244099, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 14575 + }, + { + "epoch": 0.14576, + "grad_norm": 0.6931269619482177, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14576 + }, + { + "epoch": 0.14577, + "grad_norm": 0.7130479758794105, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 14577 + }, + { + "epoch": 0.14578, + "grad_norm": 0.8292987750958368, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 14578 + }, + { + "epoch": 0.14579, + "grad_norm": 0.700892865328089, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 14579 + }, + { + "epoch": 0.1458, + "grad_norm": 0.6040010605422317, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 14580 + }, + { + "epoch": 0.14581, + "grad_norm": 0.6760437193841092, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 14581 + }, + { + "epoch": 0.14582, + "grad_norm": 0.694017083969666, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 14582 + }, + { + "epoch": 0.14583, + "grad_norm": 0.7798280690096578, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 14583 + }, + { + "epoch": 0.14584, + "grad_norm": 0.660684760024419, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 14584 + }, + { + "epoch": 0.14585, + "grad_norm": 0.6464208485328045, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 14585 + }, + { + "epoch": 0.14586, + "grad_norm": 0.6846914329185203, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 14586 + }, + { + "epoch": 0.14587, + "grad_norm": 0.5431233898954613, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14587 + }, + { + "epoch": 0.14588, + "grad_norm": 0.5058386674889934, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 14588 + }, + { + "epoch": 0.14589, + "grad_norm": 0.5123323146918984, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 14589 + }, + { + "epoch": 0.1459, + "grad_norm": 0.5678953009690293, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 14590 + }, + { + "epoch": 0.14591, + "grad_norm": 0.6095834014136225, + "learning_rate": 0.003, + "loss": 3.988, + "step": 14591 + }, + { + "epoch": 0.14592, + "grad_norm": 0.6930441674125132, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 14592 + }, + { + "epoch": 0.14593, + "grad_norm": 0.7396120208274483, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 14593 + }, + { + "epoch": 0.14594, + "grad_norm": 0.7263625030289407, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14594 + }, + { + "epoch": 0.14595, + "grad_norm": 0.6809265531955385, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 14595 + }, + { + "epoch": 0.14596, + "grad_norm": 0.7262097818121623, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 14596 + }, + { + "epoch": 0.14597, + "grad_norm": 0.7385299796593401, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 14597 + }, + { + "epoch": 0.14598, + "grad_norm": 0.8138619415905863, + "learning_rate": 0.003, + "loss": 4.064, + "step": 14598 + }, + { + "epoch": 0.14599, + "grad_norm": 0.825765109168549, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 14599 + }, + { + "epoch": 0.146, + "grad_norm": 0.7462415844894413, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 14600 + }, + { + "epoch": 0.14601, + "grad_norm": 0.794040197171896, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 14601 + }, + { + "epoch": 0.14602, + "grad_norm": 0.7657212720363907, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 14602 + }, + { + "epoch": 0.14603, + "grad_norm": 0.7080848560348811, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 14603 + }, + { + "epoch": 0.14604, + "grad_norm": 0.8409333197958356, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14604 + }, + { + "epoch": 0.14605, + "grad_norm": 0.691456694193018, + "learning_rate": 0.003, + "loss": 4.003, + "step": 14605 + }, + { + "epoch": 0.14606, + "grad_norm": 0.6500020966128263, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 14606 + }, + { + "epoch": 0.14607, + "grad_norm": 0.7028399385603951, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 14607 + }, + { + "epoch": 0.14608, + "grad_norm": 0.7298369275649219, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 14608 + }, + { + "epoch": 0.14609, + "grad_norm": 0.7530205265267652, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 14609 + }, + { + "epoch": 0.1461, + "grad_norm": 0.8018372387535869, + "learning_rate": 0.003, + "loss": 4.018, + "step": 14610 + }, + { + "epoch": 0.14611, + "grad_norm": 0.8497369978303065, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 14611 + }, + { + "epoch": 0.14612, + "grad_norm": 0.9905289452282907, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 14612 + }, + { + "epoch": 0.14613, + "grad_norm": 1.0767119999468018, + "learning_rate": 0.003, + "loss": 4.017, + "step": 14613 + }, + { + "epoch": 0.14614, + "grad_norm": 0.7820230460280234, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 14614 + }, + { + "epoch": 0.14615, + "grad_norm": 0.640262783839326, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 14615 + }, + { + "epoch": 0.14616, + "grad_norm": 0.7031109203702952, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 14616 + }, + { + "epoch": 0.14617, + "grad_norm": 0.8262353955164161, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 14617 + }, + { + "epoch": 0.14618, + "grad_norm": 0.6875765998760236, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 14618 + }, + { + "epoch": 0.14619, + "grad_norm": 0.5094982717710068, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 14619 + }, + { + "epoch": 0.1462, + "grad_norm": 0.48340116105091663, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 14620 + }, + { + "epoch": 0.14621, + "grad_norm": 0.48154198077058746, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 14621 + }, + { + "epoch": 0.14622, + "grad_norm": 0.49488391744513655, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 14622 + }, + { + "epoch": 0.14623, + "grad_norm": 0.5438297077969995, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 14623 + }, + { + "epoch": 0.14624, + "grad_norm": 0.6720799437382363, + "learning_rate": 0.003, + "loss": 4.027, + "step": 14624 + }, + { + "epoch": 0.14625, + "grad_norm": 0.8348351507580628, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 14625 + }, + { + "epoch": 0.14626, + "grad_norm": 0.8422392326542737, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 14626 + }, + { + "epoch": 0.14627, + "grad_norm": 0.6897880663844621, + "learning_rate": 0.003, + "loss": 4.021, + "step": 14627 + }, + { + "epoch": 0.14628, + "grad_norm": 0.7368569836150679, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 14628 + }, + { + "epoch": 0.14629, + "grad_norm": 0.7316171008820822, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 14629 + }, + { + "epoch": 0.1463, + "grad_norm": 0.7392503054584164, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 14630 + }, + { + "epoch": 0.14631, + "grad_norm": 0.7024486394661432, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 14631 + }, + { + "epoch": 0.14632, + "grad_norm": 0.62294890167663, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 14632 + }, + { + "epoch": 0.14633, + "grad_norm": 0.750526745252686, + "learning_rate": 0.003, + "loss": 4.01, + "step": 14633 + }, + { + "epoch": 0.14634, + "grad_norm": 0.9090769655544715, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 14634 + }, + { + "epoch": 0.14635, + "grad_norm": 0.9601488777945507, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 14635 + }, + { + "epoch": 0.14636, + "grad_norm": 0.982134670953479, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 14636 + }, + { + "epoch": 0.14637, + "grad_norm": 1.003749839160603, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 14637 + }, + { + "epoch": 0.14638, + "grad_norm": 0.853632172218639, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 14638 + }, + { + "epoch": 0.14639, + "grad_norm": 0.7335431150887074, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 14639 + }, + { + "epoch": 0.1464, + "grad_norm": 0.7015711885484931, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 14640 + }, + { + "epoch": 0.14641, + "grad_norm": 0.6136319784475244, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 14641 + }, + { + "epoch": 0.14642, + "grad_norm": 0.5870844231378712, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 14642 + }, + { + "epoch": 0.14643, + "grad_norm": 0.5347712697244325, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 14643 + }, + { + "epoch": 0.14644, + "grad_norm": 0.575670526832582, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 14644 + }, + { + "epoch": 0.14645, + "grad_norm": 0.6043680867010641, + "learning_rate": 0.003, + "loss": 4.026, + "step": 14645 + }, + { + "epoch": 0.14646, + "grad_norm": 0.6018478784035529, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 14646 + }, + { + "epoch": 0.14647, + "grad_norm": 0.5297445595759771, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 14647 + }, + { + "epoch": 0.14648, + "grad_norm": 0.5498427012090263, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 14648 + }, + { + "epoch": 0.14649, + "grad_norm": 0.686398400991011, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14649 + }, + { + "epoch": 0.1465, + "grad_norm": 0.7346111377826815, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 14650 + }, + { + "epoch": 0.14651, + "grad_norm": 0.6938357952140218, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 14651 + }, + { + "epoch": 0.14652, + "grad_norm": 0.6784750616876584, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 14652 + }, + { + "epoch": 0.14653, + "grad_norm": 0.6494020995617737, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 14653 + }, + { + "epoch": 0.14654, + "grad_norm": 0.6921255324399969, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 14654 + }, + { + "epoch": 0.14655, + "grad_norm": 0.6882255415544667, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 14655 + }, + { + "epoch": 0.14656, + "grad_norm": 0.6172170545846556, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 14656 + }, + { + "epoch": 0.14657, + "grad_norm": 0.5562253255102746, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 14657 + }, + { + "epoch": 0.14658, + "grad_norm": 0.47968835738454196, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 14658 + }, + { + "epoch": 0.14659, + "grad_norm": 0.5268654279351678, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 14659 + }, + { + "epoch": 0.1466, + "grad_norm": 0.5690260322163797, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 14660 + }, + { + "epoch": 0.14661, + "grad_norm": 0.6814589252472805, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 14661 + }, + { + "epoch": 0.14662, + "grad_norm": 0.7349117651824639, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 14662 + }, + { + "epoch": 0.14663, + "grad_norm": 0.7017825769418486, + "learning_rate": 0.003, + "loss": 4.033, + "step": 14663 + }, + { + "epoch": 0.14664, + "grad_norm": 0.7426110688115322, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 14664 + }, + { + "epoch": 0.14665, + "grad_norm": 0.839532574381832, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 14665 + }, + { + "epoch": 0.14666, + "grad_norm": 0.8615948942724577, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 14666 + }, + { + "epoch": 0.14667, + "grad_norm": 1.0105199499217659, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 14667 + }, + { + "epoch": 0.14668, + "grad_norm": 1.1385865092345768, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 14668 + }, + { + "epoch": 0.14669, + "grad_norm": 0.9351625962474065, + "learning_rate": 0.003, + "loss": 4.04, + "step": 14669 + }, + { + "epoch": 0.1467, + "grad_norm": 0.8531353961779137, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14670 + }, + { + "epoch": 0.14671, + "grad_norm": 0.8383697772015284, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 14671 + }, + { + "epoch": 0.14672, + "grad_norm": 0.8892525418406511, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 14672 + }, + { + "epoch": 0.14673, + "grad_norm": 0.8288779048353828, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 14673 + }, + { + "epoch": 0.14674, + "grad_norm": 0.7743496794995989, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 14674 + }, + { + "epoch": 0.14675, + "grad_norm": 0.7504268130382062, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 14675 + }, + { + "epoch": 0.14676, + "grad_norm": 0.7895774250282772, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 14676 + }, + { + "epoch": 0.14677, + "grad_norm": 0.8139481206192979, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 14677 + }, + { + "epoch": 0.14678, + "grad_norm": 0.8805556358380788, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 14678 + }, + { + "epoch": 0.14679, + "grad_norm": 0.8977728113966148, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 14679 + }, + { + "epoch": 0.1468, + "grad_norm": 0.9379091534155573, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 14680 + }, + { + "epoch": 0.14681, + "grad_norm": 0.8858554399969494, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 14681 + }, + { + "epoch": 0.14682, + "grad_norm": 0.8849426906350751, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14682 + }, + { + "epoch": 0.14683, + "grad_norm": 0.9263532621859001, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 14683 + }, + { + "epoch": 0.14684, + "grad_norm": 0.7924064098769555, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 14684 + }, + { + "epoch": 0.14685, + "grad_norm": 0.7460852211374167, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 14685 + }, + { + "epoch": 0.14686, + "grad_norm": 0.660306584090183, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 14686 + }, + { + "epoch": 0.14687, + "grad_norm": 0.6483109269406181, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 14687 + }, + { + "epoch": 0.14688, + "grad_norm": 0.6057433984883117, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 14688 + }, + { + "epoch": 0.14689, + "grad_norm": 0.6353633999537076, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 14689 + }, + { + "epoch": 0.1469, + "grad_norm": 0.7296735839540702, + "learning_rate": 0.003, + "loss": 4.011, + "step": 14690 + }, + { + "epoch": 0.14691, + "grad_norm": 0.6940209114268016, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 14691 + }, + { + "epoch": 0.14692, + "grad_norm": 0.674537455667571, + "learning_rate": 0.003, + "loss": 4.015, + "step": 14692 + }, + { + "epoch": 0.14693, + "grad_norm": 0.6494749632653558, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 14693 + }, + { + "epoch": 0.14694, + "grad_norm": 0.6407686892732671, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 14694 + }, + { + "epoch": 0.14695, + "grad_norm": 0.6711373047260024, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 14695 + }, + { + "epoch": 0.14696, + "grad_norm": 0.8858097640454092, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 14696 + }, + { + "epoch": 0.14697, + "grad_norm": 0.9060781976884374, + "learning_rate": 0.003, + "loss": 4.018, + "step": 14697 + }, + { + "epoch": 0.14698, + "grad_norm": 0.8354630635483112, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 14698 + }, + { + "epoch": 0.14699, + "grad_norm": 0.6913118268156139, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 14699 + }, + { + "epoch": 0.147, + "grad_norm": 0.5855691711169272, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 14700 + }, + { + "epoch": 0.14701, + "grad_norm": 0.6277310910139257, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 14701 + }, + { + "epoch": 0.14702, + "grad_norm": 0.7248766917565059, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 14702 + }, + { + "epoch": 0.14703, + "grad_norm": 0.7855527712450379, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 14703 + }, + { + "epoch": 0.14704, + "grad_norm": 0.7658209437192152, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14704 + }, + { + "epoch": 0.14705, + "grad_norm": 0.8042642262603081, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 14705 + }, + { + "epoch": 0.14706, + "grad_norm": 0.780338946335645, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 14706 + }, + { + "epoch": 0.14707, + "grad_norm": 0.7843382902381029, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 14707 + }, + { + "epoch": 0.14708, + "grad_norm": 0.7024228270885627, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 14708 + }, + { + "epoch": 0.14709, + "grad_norm": 0.7337732148288765, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 14709 + }, + { + "epoch": 0.1471, + "grad_norm": 0.8033622872259374, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 14710 + }, + { + "epoch": 0.14711, + "grad_norm": 0.791281735652731, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 14711 + }, + { + "epoch": 0.14712, + "grad_norm": 0.6579427300126074, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 14712 + }, + { + "epoch": 0.14713, + "grad_norm": 0.6936763960550125, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 14713 + }, + { + "epoch": 0.14714, + "grad_norm": 0.6938972319356259, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 14714 + }, + { + "epoch": 0.14715, + "grad_norm": 0.7567928161518429, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 14715 + }, + { + "epoch": 0.14716, + "grad_norm": 0.8758316356275139, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 14716 + }, + { + "epoch": 0.14717, + "grad_norm": 0.9666082641614469, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 14717 + }, + { + "epoch": 0.14718, + "grad_norm": 0.8495274712510803, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 14718 + }, + { + "epoch": 0.14719, + "grad_norm": 0.7855324460965826, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 14719 + }, + { + "epoch": 0.1472, + "grad_norm": 0.715496405051123, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14720 + }, + { + "epoch": 0.14721, + "grad_norm": 0.686295161951173, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 14721 + }, + { + "epoch": 0.14722, + "grad_norm": 0.6975891473309932, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 14722 + }, + { + "epoch": 0.14723, + "grad_norm": 0.5964116622061237, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 14723 + }, + { + "epoch": 0.14724, + "grad_norm": 0.5293998561513408, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 14724 + }, + { + "epoch": 0.14725, + "grad_norm": 0.5515442095182653, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 14725 + }, + { + "epoch": 0.14726, + "grad_norm": 0.5290955502573769, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 14726 + }, + { + "epoch": 0.14727, + "grad_norm": 0.493148066766284, + "learning_rate": 0.003, + "loss": 4.007, + "step": 14727 + }, + { + "epoch": 0.14728, + "grad_norm": 0.48677326845947233, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 14728 + }, + { + "epoch": 0.14729, + "grad_norm": 0.46148560655995363, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 14729 + }, + { + "epoch": 0.1473, + "grad_norm": 0.502596595832462, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 14730 + }, + { + "epoch": 0.14731, + "grad_norm": 0.5742665778871303, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 14731 + }, + { + "epoch": 0.14732, + "grad_norm": 0.5971027423632758, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 14732 + }, + { + "epoch": 0.14733, + "grad_norm": 0.5828842525642582, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 14733 + }, + { + "epoch": 0.14734, + "grad_norm": 0.5445733334304773, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14734 + }, + { + "epoch": 0.14735, + "grad_norm": 0.5694085368497307, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14735 + }, + { + "epoch": 0.14736, + "grad_norm": 0.6047158991557761, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 14736 + }, + { + "epoch": 0.14737, + "grad_norm": 0.7043837789861571, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 14737 + }, + { + "epoch": 0.14738, + "grad_norm": 0.9009982609310595, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 14738 + }, + { + "epoch": 0.14739, + "grad_norm": 1.1311989582636257, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 14739 + }, + { + "epoch": 0.1474, + "grad_norm": 0.7873108049275206, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 14740 + }, + { + "epoch": 0.14741, + "grad_norm": 0.7628914506362686, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 14741 + }, + { + "epoch": 0.14742, + "grad_norm": 0.8205863874805673, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 14742 + }, + { + "epoch": 0.14743, + "grad_norm": 0.8752967899655629, + "learning_rate": 0.003, + "loss": 4.033, + "step": 14743 + }, + { + "epoch": 0.14744, + "grad_norm": 0.957854685216169, + "learning_rate": 0.003, + "loss": 4.01, + "step": 14744 + }, + { + "epoch": 0.14745, + "grad_norm": 1.0298511446775493, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 14745 + }, + { + "epoch": 0.14746, + "grad_norm": 0.9990636943344798, + "learning_rate": 0.003, + "loss": 3.994, + "step": 14746 + }, + { + "epoch": 0.14747, + "grad_norm": 0.95234807368851, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 14747 + }, + { + "epoch": 0.14748, + "grad_norm": 0.8011268709279428, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 14748 + }, + { + "epoch": 0.14749, + "grad_norm": 0.8658137394194376, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 14749 + }, + { + "epoch": 0.1475, + "grad_norm": 0.8438604506751153, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 14750 + }, + { + "epoch": 0.14751, + "grad_norm": 0.7888826611183828, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 14751 + }, + { + "epoch": 0.14752, + "grad_norm": 0.8281842884563885, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 14752 + }, + { + "epoch": 0.14753, + "grad_norm": 0.9342547599783473, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14753 + }, + { + "epoch": 0.14754, + "grad_norm": 0.8880846644632576, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 14754 + }, + { + "epoch": 0.14755, + "grad_norm": 0.8918082210293028, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 14755 + }, + { + "epoch": 0.14756, + "grad_norm": 0.7456189836093348, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14756 + }, + { + "epoch": 0.14757, + "grad_norm": 0.696120822201115, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 14757 + }, + { + "epoch": 0.14758, + "grad_norm": 0.7872812088151739, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 14758 + }, + { + "epoch": 0.14759, + "grad_norm": 0.8308017605265626, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 14759 + }, + { + "epoch": 0.1476, + "grad_norm": 0.8066029853207918, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 14760 + }, + { + "epoch": 0.14761, + "grad_norm": 0.7075390838270097, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 14761 + }, + { + "epoch": 0.14762, + "grad_norm": 0.7469377101858142, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 14762 + }, + { + "epoch": 0.14763, + "grad_norm": 0.7261029908391988, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14763 + }, + { + "epoch": 0.14764, + "grad_norm": 0.6508689464088501, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 14764 + }, + { + "epoch": 0.14765, + "grad_norm": 0.6544343529895242, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 14765 + }, + { + "epoch": 0.14766, + "grad_norm": 0.677430108681501, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 14766 + }, + { + "epoch": 0.14767, + "grad_norm": 0.6939798491048794, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 14767 + }, + { + "epoch": 0.14768, + "grad_norm": 0.5969509571010395, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 14768 + }, + { + "epoch": 0.14769, + "grad_norm": 0.5972482409547665, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 14769 + }, + { + "epoch": 0.1477, + "grad_norm": 0.6776476674626442, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 14770 + }, + { + "epoch": 0.14771, + "grad_norm": 0.7098933644515899, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 14771 + }, + { + "epoch": 0.14772, + "grad_norm": 0.7715967684608087, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 14772 + }, + { + "epoch": 0.14773, + "grad_norm": 0.8315730895167478, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 14773 + }, + { + "epoch": 0.14774, + "grad_norm": 0.8336936277861199, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 14774 + }, + { + "epoch": 0.14775, + "grad_norm": 0.8015383060465553, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 14775 + }, + { + "epoch": 0.14776, + "grad_norm": 0.7192355233350791, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 14776 + }, + { + "epoch": 0.14777, + "grad_norm": 0.6242459154516727, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 14777 + }, + { + "epoch": 0.14778, + "grad_norm": 0.651871173855858, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 14778 + }, + { + "epoch": 0.14779, + "grad_norm": 0.6459418376109122, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 14779 + }, + { + "epoch": 0.1478, + "grad_norm": 0.6101236941012168, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14780 + }, + { + "epoch": 0.14781, + "grad_norm": 0.648559109365841, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 14781 + }, + { + "epoch": 0.14782, + "grad_norm": 0.6919980791115065, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 14782 + }, + { + "epoch": 0.14783, + "grad_norm": 0.699216773125641, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 14783 + }, + { + "epoch": 0.14784, + "grad_norm": 0.7183091541734864, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 14784 + }, + { + "epoch": 0.14785, + "grad_norm": 0.7445528861624208, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 14785 + }, + { + "epoch": 0.14786, + "grad_norm": 0.7654501528311922, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 14786 + }, + { + "epoch": 0.14787, + "grad_norm": 0.8172182938442426, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 14787 + }, + { + "epoch": 0.14788, + "grad_norm": 0.9489428365124956, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 14788 + }, + { + "epoch": 0.14789, + "grad_norm": 1.071016835630401, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 14789 + }, + { + "epoch": 0.1479, + "grad_norm": 0.9225081936129478, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 14790 + }, + { + "epoch": 0.14791, + "grad_norm": 0.8381620090225637, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 14791 + }, + { + "epoch": 0.14792, + "grad_norm": 0.7852016192892289, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 14792 + }, + { + "epoch": 0.14793, + "grad_norm": 0.7137331914278024, + "learning_rate": 0.003, + "loss": 4.033, + "step": 14793 + }, + { + "epoch": 0.14794, + "grad_norm": 0.6756733283576574, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 14794 + }, + { + "epoch": 0.14795, + "grad_norm": 0.6784933378219624, + "learning_rate": 0.003, + "loss": 4.0655, + "step": 14795 + }, + { + "epoch": 0.14796, + "grad_norm": 0.7532574750632449, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 14796 + }, + { + "epoch": 0.14797, + "grad_norm": 0.7682122232693501, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 14797 + }, + { + "epoch": 0.14798, + "grad_norm": 0.6919803801545071, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 14798 + }, + { + "epoch": 0.14799, + "grad_norm": 0.6894506353070324, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 14799 + }, + { + "epoch": 0.148, + "grad_norm": 0.6506534158349215, + "learning_rate": 0.003, + "loss": 4.002, + "step": 14800 + }, + { + "epoch": 0.14801, + "grad_norm": 0.5953742305434083, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 14801 + }, + { + "epoch": 0.14802, + "grad_norm": 0.5352091480199368, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 14802 + }, + { + "epoch": 0.14803, + "grad_norm": 0.497694625743089, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 14803 + }, + { + "epoch": 0.14804, + "grad_norm": 0.5012231635870196, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 14804 + }, + { + "epoch": 0.14805, + "grad_norm": 0.490654026874642, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 14805 + }, + { + "epoch": 0.14806, + "grad_norm": 0.48406193253881996, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 14806 + }, + { + "epoch": 0.14807, + "grad_norm": 0.5153014787910969, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 14807 + }, + { + "epoch": 0.14808, + "grad_norm": 0.5766652968169343, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 14808 + }, + { + "epoch": 0.14809, + "grad_norm": 0.6325258071228433, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 14809 + }, + { + "epoch": 0.1481, + "grad_norm": 0.8312116108288012, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14810 + }, + { + "epoch": 0.14811, + "grad_norm": 1.207091916270145, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 14811 + }, + { + "epoch": 0.14812, + "grad_norm": 0.773341678738452, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 14812 + }, + { + "epoch": 0.14813, + "grad_norm": 0.6230346691298881, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 14813 + }, + { + "epoch": 0.14814, + "grad_norm": 0.6551781455655734, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 14814 + }, + { + "epoch": 0.14815, + "grad_norm": 0.6735167322299169, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 14815 + }, + { + "epoch": 0.14816, + "grad_norm": 0.6718500417056529, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 14816 + }, + { + "epoch": 0.14817, + "grad_norm": 0.6624184066370538, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 14817 + }, + { + "epoch": 0.14818, + "grad_norm": 0.6120303300140975, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 14818 + }, + { + "epoch": 0.14819, + "grad_norm": 0.5992405650077496, + "learning_rate": 0.003, + "loss": 4.004, + "step": 14819 + }, + { + "epoch": 0.1482, + "grad_norm": 0.5973908843817647, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 14820 + }, + { + "epoch": 0.14821, + "grad_norm": 0.6663794429622332, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 14821 + }, + { + "epoch": 0.14822, + "grad_norm": 0.7440480806457982, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 14822 + }, + { + "epoch": 0.14823, + "grad_norm": 0.7191763202591483, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 14823 + }, + { + "epoch": 0.14824, + "grad_norm": 0.766152165148274, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 14824 + }, + { + "epoch": 0.14825, + "grad_norm": 0.8618416710149183, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 14825 + }, + { + "epoch": 0.14826, + "grad_norm": 0.7679580760568375, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14826 + }, + { + "epoch": 0.14827, + "grad_norm": 0.7402681248819744, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 14827 + }, + { + "epoch": 0.14828, + "grad_norm": 0.6633063866625318, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 14828 + }, + { + "epoch": 0.14829, + "grad_norm": 0.6797758880979279, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 14829 + }, + { + "epoch": 0.1483, + "grad_norm": 0.6867094326257899, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 14830 + }, + { + "epoch": 0.14831, + "grad_norm": 0.7980846945458534, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 14831 + }, + { + "epoch": 0.14832, + "grad_norm": 0.9088786771535408, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 14832 + }, + { + "epoch": 0.14833, + "grad_norm": 0.9046178033965966, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 14833 + }, + { + "epoch": 0.14834, + "grad_norm": 0.7814873143779848, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 14834 + }, + { + "epoch": 0.14835, + "grad_norm": 0.7968760599407801, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 14835 + }, + { + "epoch": 0.14836, + "grad_norm": 0.7504846081615631, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 14836 + }, + { + "epoch": 0.14837, + "grad_norm": 0.7475880462831856, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14837 + }, + { + "epoch": 0.14838, + "grad_norm": 0.7980851169565474, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 14838 + }, + { + "epoch": 0.14839, + "grad_norm": 0.870922006427334, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 14839 + }, + { + "epoch": 0.1484, + "grad_norm": 0.8692164081738153, + "learning_rate": 0.003, + "loss": 4.007, + "step": 14840 + }, + { + "epoch": 0.14841, + "grad_norm": 0.8197399170432321, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 14841 + }, + { + "epoch": 0.14842, + "grad_norm": 0.8180276532926982, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 14842 + }, + { + "epoch": 0.14843, + "grad_norm": 0.8486310737294599, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 14843 + }, + { + "epoch": 0.14844, + "grad_norm": 0.8327232357195816, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 14844 + }, + { + "epoch": 0.14845, + "grad_norm": 0.7343153132486305, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 14845 + }, + { + "epoch": 0.14846, + "grad_norm": 0.804418558483046, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 14846 + }, + { + "epoch": 0.14847, + "grad_norm": 0.9285499292084929, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 14847 + }, + { + "epoch": 0.14848, + "grad_norm": 1.0625822225350008, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 14848 + }, + { + "epoch": 0.14849, + "grad_norm": 1.0071426816720745, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 14849 + }, + { + "epoch": 0.1485, + "grad_norm": 0.901559706042572, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 14850 + }, + { + "epoch": 0.14851, + "grad_norm": 0.8324251416843788, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 14851 + }, + { + "epoch": 0.14852, + "grad_norm": 0.6955829347824211, + "learning_rate": 0.003, + "loss": 4.013, + "step": 14852 + }, + { + "epoch": 0.14853, + "grad_norm": 0.7076782042328994, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 14853 + }, + { + "epoch": 0.14854, + "grad_norm": 0.6431139279079517, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 14854 + }, + { + "epoch": 0.14855, + "grad_norm": 0.6357390069388968, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 14855 + }, + { + "epoch": 0.14856, + "grad_norm": 0.6482908424146612, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 14856 + }, + { + "epoch": 0.14857, + "grad_norm": 0.5838413020441089, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 14857 + }, + { + "epoch": 0.14858, + "grad_norm": 0.5189693150991993, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 14858 + }, + { + "epoch": 0.14859, + "grad_norm": 0.5382045516705722, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 14859 + }, + { + "epoch": 0.1486, + "grad_norm": 0.5435059375746106, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 14860 + }, + { + "epoch": 0.14861, + "grad_norm": 0.5312187085666117, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 14861 + }, + { + "epoch": 0.14862, + "grad_norm": 0.5687179581708006, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 14862 + }, + { + "epoch": 0.14863, + "grad_norm": 0.5734181228441037, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 14863 + }, + { + "epoch": 0.14864, + "grad_norm": 0.6380741244891152, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14864 + }, + { + "epoch": 0.14865, + "grad_norm": 0.7524318252021959, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 14865 + }, + { + "epoch": 0.14866, + "grad_norm": 0.6895500825229726, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 14866 + }, + { + "epoch": 0.14867, + "grad_norm": 0.7004319754716466, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 14867 + }, + { + "epoch": 0.14868, + "grad_norm": 0.7241743547360892, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 14868 + }, + { + "epoch": 0.14869, + "grad_norm": 0.6867339130506787, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 14869 + }, + { + "epoch": 0.1487, + "grad_norm": 0.7308384919302151, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 14870 + }, + { + "epoch": 0.14871, + "grad_norm": 0.7793260356922631, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 14871 + }, + { + "epoch": 0.14872, + "grad_norm": 0.7891373350241717, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 14872 + }, + { + "epoch": 0.14873, + "grad_norm": 0.9387926298274875, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 14873 + }, + { + "epoch": 0.14874, + "grad_norm": 0.9503331854684032, + "learning_rate": 0.003, + "loss": 3.99, + "step": 14874 + }, + { + "epoch": 0.14875, + "grad_norm": 0.9154074399211979, + "learning_rate": 0.003, + "loss": 4.024, + "step": 14875 + }, + { + "epoch": 0.14876, + "grad_norm": 0.7614377871275438, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 14876 + }, + { + "epoch": 0.14877, + "grad_norm": 0.7031979616437721, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 14877 + }, + { + "epoch": 0.14878, + "grad_norm": 0.6913557185476851, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 14878 + }, + { + "epoch": 0.14879, + "grad_norm": 0.7444615654488437, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 14879 + }, + { + "epoch": 0.1488, + "grad_norm": 0.7572704397655801, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 14880 + }, + { + "epoch": 0.14881, + "grad_norm": 0.9172723335006604, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 14881 + }, + { + "epoch": 0.14882, + "grad_norm": 0.9982167309854223, + "learning_rate": 0.003, + "loss": 4.043, + "step": 14882 + }, + { + "epoch": 0.14883, + "grad_norm": 0.9453378146551381, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 14883 + }, + { + "epoch": 0.14884, + "grad_norm": 0.9129920843377315, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 14884 + }, + { + "epoch": 0.14885, + "grad_norm": 0.812205609728804, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 14885 + }, + { + "epoch": 0.14886, + "grad_norm": 0.7307749768641048, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 14886 + }, + { + "epoch": 0.14887, + "grad_norm": 0.7133474021735441, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 14887 + }, + { + "epoch": 0.14888, + "grad_norm": 0.720807148026326, + "learning_rate": 0.003, + "loss": 4.0565, + "step": 14888 + }, + { + "epoch": 0.14889, + "grad_norm": 0.9192626678065914, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 14889 + }, + { + "epoch": 0.1489, + "grad_norm": 0.8843333317263382, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14890 + }, + { + "epoch": 0.14891, + "grad_norm": 0.686598584546183, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 14891 + }, + { + "epoch": 0.14892, + "grad_norm": 0.698028707590579, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 14892 + }, + { + "epoch": 0.14893, + "grad_norm": 0.6931346624063444, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 14893 + }, + { + "epoch": 0.14894, + "grad_norm": 0.6862494510732242, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 14894 + }, + { + "epoch": 0.14895, + "grad_norm": 0.7350876665093794, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 14895 + }, + { + "epoch": 0.14896, + "grad_norm": 0.6778548926408298, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14896 + }, + { + "epoch": 0.14897, + "grad_norm": 0.7297708020480338, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 14897 + }, + { + "epoch": 0.14898, + "grad_norm": 0.6987911507859321, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 14898 + }, + { + "epoch": 0.14899, + "grad_norm": 0.7239140883366258, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 14899 + }, + { + "epoch": 0.149, + "grad_norm": 0.7917347704435576, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 14900 + }, + { + "epoch": 0.14901, + "grad_norm": 1.0629128554278506, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 14901 + }, + { + "epoch": 0.14902, + "grad_norm": 1.0336011137555334, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 14902 + }, + { + "epoch": 0.14903, + "grad_norm": 0.8088030025154945, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 14903 + }, + { + "epoch": 0.14904, + "grad_norm": 0.6937947196546024, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 14904 + }, + { + "epoch": 0.14905, + "grad_norm": 0.7384751068615351, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 14905 + }, + { + "epoch": 0.14906, + "grad_norm": 0.7910976130867444, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 14906 + }, + { + "epoch": 0.14907, + "grad_norm": 0.6262387475655553, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 14907 + }, + { + "epoch": 0.14908, + "grad_norm": 0.6027657807653753, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 14908 + }, + { + "epoch": 0.14909, + "grad_norm": 0.5924563838050411, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 14909 + }, + { + "epoch": 0.1491, + "grad_norm": 0.685251223076166, + "learning_rate": 0.003, + "loss": 4.019, + "step": 14910 + }, + { + "epoch": 0.14911, + "grad_norm": 0.6696551339474495, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 14911 + }, + { + "epoch": 0.14912, + "grad_norm": 0.6096943902927998, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 14912 + }, + { + "epoch": 0.14913, + "grad_norm": 0.5210546258100551, + "learning_rate": 0.003, + "loss": 3.992, + "step": 14913 + }, + { + "epoch": 0.14914, + "grad_norm": 0.4501967801820687, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 14914 + }, + { + "epoch": 0.14915, + "grad_norm": 0.4793775360915418, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 14915 + }, + { + "epoch": 0.14916, + "grad_norm": 0.4906318115567648, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14916 + }, + { + "epoch": 0.14917, + "grad_norm": 0.3953187050765534, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 14917 + }, + { + "epoch": 0.14918, + "grad_norm": 0.4568199550431618, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 14918 + }, + { + "epoch": 0.14919, + "grad_norm": 0.43737017774960246, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 14919 + }, + { + "epoch": 0.1492, + "grad_norm": 0.49696498917415327, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 14920 + }, + { + "epoch": 0.14921, + "grad_norm": 0.5266288632342483, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 14921 + }, + { + "epoch": 0.14922, + "grad_norm": 0.6094993659116348, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 14922 + }, + { + "epoch": 0.14923, + "grad_norm": 0.5625663459295522, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 14923 + }, + { + "epoch": 0.14924, + "grad_norm": 0.5757058242569297, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 14924 + }, + { + "epoch": 0.14925, + "grad_norm": 0.6009739103299622, + "learning_rate": 0.003, + "loss": 3.998, + "step": 14925 + }, + { + "epoch": 0.14926, + "grad_norm": 0.5476532599696575, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 14926 + }, + { + "epoch": 0.14927, + "grad_norm": 0.6212069261543305, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 14927 + }, + { + "epoch": 0.14928, + "grad_norm": 0.7212064576723622, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 14928 + }, + { + "epoch": 0.14929, + "grad_norm": 0.772322137710829, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 14929 + }, + { + "epoch": 0.1493, + "grad_norm": 0.7843035160805176, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 14930 + }, + { + "epoch": 0.14931, + "grad_norm": 0.7808549066626157, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 14931 + }, + { + "epoch": 0.14932, + "grad_norm": 0.8305072075081507, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 14932 + }, + { + "epoch": 0.14933, + "grad_norm": 1.0284872494848272, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 14933 + }, + { + "epoch": 0.14934, + "grad_norm": 1.3728366341570264, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 14934 + }, + { + "epoch": 0.14935, + "grad_norm": 0.5938919358164987, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 14935 + }, + { + "epoch": 0.14936, + "grad_norm": 0.9041129139649393, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 14936 + }, + { + "epoch": 0.14937, + "grad_norm": 1.1290817736241172, + "learning_rate": 0.003, + "loss": 4.024, + "step": 14937 + }, + { + "epoch": 0.14938, + "grad_norm": 0.9462357753336604, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 14938 + }, + { + "epoch": 0.14939, + "grad_norm": 0.9213858091151335, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 14939 + }, + { + "epoch": 0.1494, + "grad_norm": 1.0068601942957645, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 14940 + }, + { + "epoch": 0.14941, + "grad_norm": 1.1131036793848033, + "learning_rate": 0.003, + "loss": 4.0773, + "step": 14941 + }, + { + "epoch": 0.14942, + "grad_norm": 0.8424404212162646, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 14942 + }, + { + "epoch": 0.14943, + "grad_norm": 0.8316800477904414, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 14943 + }, + { + "epoch": 0.14944, + "grad_norm": 0.7297842979772468, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 14944 + }, + { + "epoch": 0.14945, + "grad_norm": 0.7476074853660047, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 14945 + }, + { + "epoch": 0.14946, + "grad_norm": 0.7170138409315941, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 14946 + }, + { + "epoch": 0.14947, + "grad_norm": 0.718447311489337, + "learning_rate": 0.003, + "loss": 4.024, + "step": 14947 + }, + { + "epoch": 0.14948, + "grad_norm": 0.6967861956028377, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 14948 + }, + { + "epoch": 0.14949, + "grad_norm": 0.6700473312481864, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 14949 + }, + { + "epoch": 0.1495, + "grad_norm": 0.7558236884736979, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 14950 + }, + { + "epoch": 0.14951, + "grad_norm": 0.8064749354299661, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 14951 + }, + { + "epoch": 0.14952, + "grad_norm": 0.82103060734683, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 14952 + }, + { + "epoch": 0.14953, + "grad_norm": 0.7359274167019315, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 14953 + }, + { + "epoch": 0.14954, + "grad_norm": 0.6094831212381653, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 14954 + }, + { + "epoch": 0.14955, + "grad_norm": 0.5689097477038582, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 14955 + }, + { + "epoch": 0.14956, + "grad_norm": 0.5985738445075873, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 14956 + }, + { + "epoch": 0.14957, + "grad_norm": 0.5990697461052491, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 14957 + }, + { + "epoch": 0.14958, + "grad_norm": 0.5790665740565895, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 14958 + }, + { + "epoch": 0.14959, + "grad_norm": 0.5908230518602763, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 14959 + }, + { + "epoch": 0.1496, + "grad_norm": 0.5812306979497697, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 14960 + }, + { + "epoch": 0.14961, + "grad_norm": 0.712099666910256, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 14961 + }, + { + "epoch": 0.14962, + "grad_norm": 0.8392571897203948, + "learning_rate": 0.003, + "loss": 4.007, + "step": 14962 + }, + { + "epoch": 0.14963, + "grad_norm": 1.001145268968051, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 14963 + }, + { + "epoch": 0.14964, + "grad_norm": 1.0446537105735392, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 14964 + }, + { + "epoch": 0.14965, + "grad_norm": 0.7976048795092633, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 14965 + }, + { + "epoch": 0.14966, + "grad_norm": 0.7096146610274412, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 14966 + }, + { + "epoch": 0.14967, + "grad_norm": 0.8200616451452598, + "learning_rate": 0.003, + "loss": 4.0475, + "step": 14967 + }, + { + "epoch": 0.14968, + "grad_norm": 0.8920051100510048, + "learning_rate": 0.003, + "loss": 4.024, + "step": 14968 + }, + { + "epoch": 0.14969, + "grad_norm": 0.7911931919663526, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 14969 + }, + { + "epoch": 0.1497, + "grad_norm": 0.621790069739747, + "learning_rate": 0.003, + "loss": 4.031, + "step": 14970 + }, + { + "epoch": 0.14971, + "grad_norm": 0.5328575537390182, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 14971 + }, + { + "epoch": 0.14972, + "grad_norm": 0.6065037137076532, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 14972 + }, + { + "epoch": 0.14973, + "grad_norm": 0.6690089274430213, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 14973 + }, + { + "epoch": 0.14974, + "grad_norm": 0.7521722455132389, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 14974 + }, + { + "epoch": 0.14975, + "grad_norm": 0.786262158420773, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 14975 + }, + { + "epoch": 0.14976, + "grad_norm": 0.7589377382938809, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 14976 + }, + { + "epoch": 0.14977, + "grad_norm": 0.7680271697115066, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 14977 + }, + { + "epoch": 0.14978, + "grad_norm": 0.8002892511450407, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 14978 + }, + { + "epoch": 0.14979, + "grad_norm": 0.7940588001424898, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 14979 + }, + { + "epoch": 0.1498, + "grad_norm": 0.8393763901080196, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 14980 + }, + { + "epoch": 0.14981, + "grad_norm": 0.9632015250837859, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 14981 + }, + { + "epoch": 0.14982, + "grad_norm": 0.9751967742446861, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 14982 + }, + { + "epoch": 0.14983, + "grad_norm": 0.83179894953928, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 14983 + }, + { + "epoch": 0.14984, + "grad_norm": 0.6958810950239727, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 14984 + }, + { + "epoch": 0.14985, + "grad_norm": 0.6712195172983859, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 14985 + }, + { + "epoch": 0.14986, + "grad_norm": 0.6927470035806449, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 14986 + }, + { + "epoch": 0.14987, + "grad_norm": 0.7697054081416388, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 14987 + }, + { + "epoch": 0.14988, + "grad_norm": 0.6897407380475598, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 14988 + }, + { + "epoch": 0.14989, + "grad_norm": 0.7095696912696359, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 14989 + }, + { + "epoch": 0.1499, + "grad_norm": 0.8662905097259082, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 14990 + }, + { + "epoch": 0.14991, + "grad_norm": 0.928249058766288, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 14991 + }, + { + "epoch": 0.14992, + "grad_norm": 0.8938469044770959, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 14992 + }, + { + "epoch": 0.14993, + "grad_norm": 0.6654717063371433, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 14993 + }, + { + "epoch": 0.14994, + "grad_norm": 0.5336269721381617, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 14994 + }, + { + "epoch": 0.14995, + "grad_norm": 0.5760207683101092, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 14995 + }, + { + "epoch": 0.14996, + "grad_norm": 0.5147809209405754, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 14996 + }, + { + "epoch": 0.14997, + "grad_norm": 0.4572533330595954, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 14997 + }, + { + "epoch": 0.14998, + "grad_norm": 0.4129970310686191, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 14998 + }, + { + "epoch": 0.14999, + "grad_norm": 0.4384218089104503, + "learning_rate": 0.003, + "loss": 4.028, + "step": 14999 + }, + { + "epoch": 0.15, + "grad_norm": 0.43849415845769896, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 15000 + }, + { + "epoch": 0.15001, + "grad_norm": 0.4625227354635212, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 15001 + }, + { + "epoch": 0.15002, + "grad_norm": 0.5017578875949327, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 15002 + }, + { + "epoch": 0.15003, + "grad_norm": 0.6244777633522479, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 15003 + }, + { + "epoch": 0.15004, + "grad_norm": 0.6905903421682339, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 15004 + }, + { + "epoch": 0.15005, + "grad_norm": 0.6823253446521532, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 15005 + }, + { + "epoch": 0.15006, + "grad_norm": 0.6790786077228567, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 15006 + }, + { + "epoch": 0.15007, + "grad_norm": 0.5964155318523472, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 15007 + }, + { + "epoch": 0.15008, + "grad_norm": 0.602472722609136, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 15008 + }, + { + "epoch": 0.15009, + "grad_norm": 0.6498184948949767, + "learning_rate": 0.003, + "loss": 3.984, + "step": 15009 + }, + { + "epoch": 0.1501, + "grad_norm": 0.6824098892016357, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 15010 + }, + { + "epoch": 0.15011, + "grad_norm": 0.762082826960762, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 15011 + }, + { + "epoch": 0.15012, + "grad_norm": 0.8534928266875933, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 15012 + }, + { + "epoch": 0.15013, + "grad_norm": 1.079821902959634, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 15013 + }, + { + "epoch": 0.15014, + "grad_norm": 0.947981984258979, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 15014 + }, + { + "epoch": 0.15015, + "grad_norm": 0.8831934632220668, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15015 + }, + { + "epoch": 0.15016, + "grad_norm": 0.7777072028928107, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 15016 + }, + { + "epoch": 0.15017, + "grad_norm": 0.8315095610823321, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 15017 + }, + { + "epoch": 0.15018, + "grad_norm": 0.8966909147006316, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 15018 + }, + { + "epoch": 0.15019, + "grad_norm": 0.9371979411580385, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 15019 + }, + { + "epoch": 0.1502, + "grad_norm": 0.9449750753541227, + "learning_rate": 0.003, + "loss": 4.02, + "step": 15020 + }, + { + "epoch": 0.15021, + "grad_norm": 0.9600694765232334, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 15021 + }, + { + "epoch": 0.15022, + "grad_norm": 0.961416264815989, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 15022 + }, + { + "epoch": 0.15023, + "grad_norm": 0.7507181704587175, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 15023 + }, + { + "epoch": 0.15024, + "grad_norm": 0.6104796465936684, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 15024 + }, + { + "epoch": 0.15025, + "grad_norm": 0.5697063181966483, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 15025 + }, + { + "epoch": 0.15026, + "grad_norm": 0.5475909034296759, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 15026 + }, + { + "epoch": 0.15027, + "grad_norm": 0.5496992761410575, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 15027 + }, + { + "epoch": 0.15028, + "grad_norm": 0.524749156188949, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 15028 + }, + { + "epoch": 0.15029, + "grad_norm": 0.6020759442996997, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 15029 + }, + { + "epoch": 0.1503, + "grad_norm": 0.6428883174588407, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 15030 + }, + { + "epoch": 0.15031, + "grad_norm": 0.6288282803297456, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 15031 + }, + { + "epoch": 0.15032, + "grad_norm": 0.5941688026487564, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 15032 + }, + { + "epoch": 0.15033, + "grad_norm": 0.5627484686845755, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 15033 + }, + { + "epoch": 0.15034, + "grad_norm": 0.7121923917260269, + "learning_rate": 0.003, + "loss": 4.022, + "step": 15034 + }, + { + "epoch": 0.15035, + "grad_norm": 0.6989424402905254, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 15035 + }, + { + "epoch": 0.15036, + "grad_norm": 0.6624351745844684, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 15036 + }, + { + "epoch": 0.15037, + "grad_norm": 0.6505710827572576, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 15037 + }, + { + "epoch": 0.15038, + "grad_norm": 0.6723598892752009, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 15038 + }, + { + "epoch": 0.15039, + "grad_norm": 0.6102080867086398, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 15039 + }, + { + "epoch": 0.1504, + "grad_norm": 0.6040853212372258, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 15040 + }, + { + "epoch": 0.15041, + "grad_norm": 0.6648271723350402, + "learning_rate": 0.003, + "loss": 4.0533, + "step": 15041 + }, + { + "epoch": 0.15042, + "grad_norm": 0.7847198619036705, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 15042 + }, + { + "epoch": 0.15043, + "grad_norm": 0.8206059674630864, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 15043 + }, + { + "epoch": 0.15044, + "grad_norm": 0.7837810077717581, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 15044 + }, + { + "epoch": 0.15045, + "grad_norm": 0.8681640530827219, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 15045 + }, + { + "epoch": 0.15046, + "grad_norm": 0.8363186699326943, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 15046 + }, + { + "epoch": 0.15047, + "grad_norm": 0.7217784926599973, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 15047 + }, + { + "epoch": 0.15048, + "grad_norm": 0.6328816714681581, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 15048 + }, + { + "epoch": 0.15049, + "grad_norm": 0.5248575445439408, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 15049 + }, + { + "epoch": 0.1505, + "grad_norm": 0.646393445226273, + "learning_rate": 0.003, + "loss": 4.021, + "step": 15050 + }, + { + "epoch": 0.15051, + "grad_norm": 0.8657777357028381, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 15051 + }, + { + "epoch": 0.15052, + "grad_norm": 1.0055071905729553, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 15052 + }, + { + "epoch": 0.15053, + "grad_norm": 0.9182051360042865, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 15053 + }, + { + "epoch": 0.15054, + "grad_norm": 0.82481131204497, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 15054 + }, + { + "epoch": 0.15055, + "grad_norm": 0.7652944363819487, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 15055 + }, + { + "epoch": 0.15056, + "grad_norm": 0.8721707419962583, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 15056 + }, + { + "epoch": 0.15057, + "grad_norm": 0.7776054419007827, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15057 + }, + { + "epoch": 0.15058, + "grad_norm": 0.7554474477262264, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 15058 + }, + { + "epoch": 0.15059, + "grad_norm": 0.9295553841477092, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 15059 + }, + { + "epoch": 0.1506, + "grad_norm": 0.9666617563650143, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15060 + }, + { + "epoch": 0.15061, + "grad_norm": 0.8668653416588769, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 15061 + }, + { + "epoch": 0.15062, + "grad_norm": 0.8393627515060834, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 15062 + }, + { + "epoch": 0.15063, + "grad_norm": 0.9022607697591591, + "learning_rate": 0.003, + "loss": 4.0445, + "step": 15063 + }, + { + "epoch": 0.15064, + "grad_norm": 0.8487369484574422, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 15064 + }, + { + "epoch": 0.15065, + "grad_norm": 0.7495603858507854, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 15065 + }, + { + "epoch": 0.15066, + "grad_norm": 0.7236534949339661, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 15066 + }, + { + "epoch": 0.15067, + "grad_norm": 0.7757582290022547, + "learning_rate": 0.003, + "loss": 4.0577, + "step": 15067 + }, + { + "epoch": 0.15068, + "grad_norm": 0.9969893855641154, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 15068 + }, + { + "epoch": 0.15069, + "grad_norm": 1.0272059756716876, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 15069 + }, + { + "epoch": 0.1507, + "grad_norm": 0.8342754587420498, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15070 + }, + { + "epoch": 0.15071, + "grad_norm": 0.72296282138692, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 15071 + }, + { + "epoch": 0.15072, + "grad_norm": 0.6272046821646781, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 15072 + }, + { + "epoch": 0.15073, + "grad_norm": 0.6364337729857572, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 15073 + }, + { + "epoch": 0.15074, + "grad_norm": 0.7519710604342456, + "learning_rate": 0.003, + "loss": 4.029, + "step": 15074 + }, + { + "epoch": 0.15075, + "grad_norm": 0.8018463513116556, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 15075 + }, + { + "epoch": 0.15076, + "grad_norm": 0.8203769577078823, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 15076 + }, + { + "epoch": 0.15077, + "grad_norm": 0.7484980821435413, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 15077 + }, + { + "epoch": 0.15078, + "grad_norm": 0.7346424046023494, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 15078 + }, + { + "epoch": 0.15079, + "grad_norm": 0.813321117910262, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 15079 + }, + { + "epoch": 0.1508, + "grad_norm": 0.724181078995489, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 15080 + }, + { + "epoch": 0.15081, + "grad_norm": 0.7115608593310115, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 15081 + }, + { + "epoch": 0.15082, + "grad_norm": 0.586508063155618, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 15082 + }, + { + "epoch": 0.15083, + "grad_norm": 0.5769685897289706, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 15083 + }, + { + "epoch": 0.15084, + "grad_norm": 0.6348398294788625, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 15084 + }, + { + "epoch": 0.15085, + "grad_norm": 0.6568454752724143, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 15085 + }, + { + "epoch": 0.15086, + "grad_norm": 0.5913716218102637, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 15086 + }, + { + "epoch": 0.15087, + "grad_norm": 0.5491872653058648, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 15087 + }, + { + "epoch": 0.15088, + "grad_norm": 0.5203266101201471, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 15088 + }, + { + "epoch": 0.15089, + "grad_norm": 0.6043756593188749, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 15089 + }, + { + "epoch": 0.1509, + "grad_norm": 0.6841911673658455, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 15090 + }, + { + "epoch": 0.15091, + "grad_norm": 0.6432505508579005, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 15091 + }, + { + "epoch": 0.15092, + "grad_norm": 0.6724209982841587, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 15092 + }, + { + "epoch": 0.15093, + "grad_norm": 0.7343062539770523, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 15093 + }, + { + "epoch": 0.15094, + "grad_norm": 0.6897345645845774, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 15094 + }, + { + "epoch": 0.15095, + "grad_norm": 0.6800588015928665, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 15095 + }, + { + "epoch": 0.15096, + "grad_norm": 0.6634958807000957, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 15096 + }, + { + "epoch": 0.15097, + "grad_norm": 0.9489500579675485, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 15097 + }, + { + "epoch": 0.15098, + "grad_norm": 1.130851429772493, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 15098 + }, + { + "epoch": 0.15099, + "grad_norm": 0.9767545405433613, + "learning_rate": 0.003, + "loss": 4.014, + "step": 15099 + }, + { + "epoch": 0.151, + "grad_norm": 0.7301100160538779, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 15100 + }, + { + "epoch": 0.15101, + "grad_norm": 0.8494732417330505, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 15101 + }, + { + "epoch": 0.15102, + "grad_norm": 0.8776002154854353, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 15102 + }, + { + "epoch": 0.15103, + "grad_norm": 0.8174802183334039, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 15103 + }, + { + "epoch": 0.15104, + "grad_norm": 0.9464278280314267, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 15104 + }, + { + "epoch": 0.15105, + "grad_norm": 0.9426470226591377, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 15105 + }, + { + "epoch": 0.15106, + "grad_norm": 0.890983646234427, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 15106 + }, + { + "epoch": 0.15107, + "grad_norm": 0.8751518632660679, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 15107 + }, + { + "epoch": 0.15108, + "grad_norm": 0.9558563090824278, + "learning_rate": 0.003, + "loss": 4.044, + "step": 15108 + }, + { + "epoch": 0.15109, + "grad_norm": 1.0629250021605672, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 15109 + }, + { + "epoch": 0.1511, + "grad_norm": 1.106763209839843, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15110 + }, + { + "epoch": 0.15111, + "grad_norm": 1.23472434709432, + "learning_rate": 0.003, + "loss": 4.0478, + "step": 15111 + }, + { + "epoch": 0.15112, + "grad_norm": 1.0607199796002629, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 15112 + }, + { + "epoch": 0.15113, + "grad_norm": 1.0774946192952357, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 15113 + }, + { + "epoch": 0.15114, + "grad_norm": 0.9422564115459301, + "learning_rate": 0.003, + "loss": 4.024, + "step": 15114 + }, + { + "epoch": 0.15115, + "grad_norm": 0.8118722476118688, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 15115 + }, + { + "epoch": 0.15116, + "grad_norm": 0.7802455119782956, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 15116 + }, + { + "epoch": 0.15117, + "grad_norm": 0.7484492916833859, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 15117 + }, + { + "epoch": 0.15118, + "grad_norm": 1.1063637930086452, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 15118 + }, + { + "epoch": 0.15119, + "grad_norm": 1.3163328526374443, + "learning_rate": 0.003, + "loss": 4.0588, + "step": 15119 + }, + { + "epoch": 0.1512, + "grad_norm": 0.6497006272424084, + "learning_rate": 0.003, + "loss": 4.0602, + "step": 15120 + }, + { + "epoch": 0.15121, + "grad_norm": 0.7809071802054189, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 15121 + }, + { + "epoch": 0.15122, + "grad_norm": 0.9493727114996644, + "learning_rate": 0.003, + "loss": 4.034, + "step": 15122 + }, + { + "epoch": 0.15123, + "grad_norm": 0.9146225400856951, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 15123 + }, + { + "epoch": 0.15124, + "grad_norm": 0.7736704658704119, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 15124 + }, + { + "epoch": 0.15125, + "grad_norm": 0.8056507775023118, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 15125 + }, + { + "epoch": 0.15126, + "grad_norm": 0.8095563321947759, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 15126 + }, + { + "epoch": 0.15127, + "grad_norm": 0.8576167836762469, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 15127 + }, + { + "epoch": 0.15128, + "grad_norm": 0.9154507252853304, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 15128 + }, + { + "epoch": 0.15129, + "grad_norm": 0.8099262404113706, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 15129 + }, + { + "epoch": 0.1513, + "grad_norm": 0.6367615908205472, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 15130 + }, + { + "epoch": 0.15131, + "grad_norm": 0.640533273762432, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15131 + }, + { + "epoch": 0.15132, + "grad_norm": 0.6163276454109838, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 15132 + }, + { + "epoch": 0.15133, + "grad_norm": 0.719900627549059, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 15133 + }, + { + "epoch": 0.15134, + "grad_norm": 0.8271705467680855, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15134 + }, + { + "epoch": 0.15135, + "grad_norm": 0.8030090253325086, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 15135 + }, + { + "epoch": 0.15136, + "grad_norm": 0.7087596963024795, + "learning_rate": 0.003, + "loss": 3.982, + "step": 15136 + }, + { + "epoch": 0.15137, + "grad_norm": 0.5913261587333299, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 15137 + }, + { + "epoch": 0.15138, + "grad_norm": 0.6582041124999534, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 15138 + }, + { + "epoch": 0.15139, + "grad_norm": 0.8020938396475599, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15139 + }, + { + "epoch": 0.1514, + "grad_norm": 0.8328231348505806, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 15140 + }, + { + "epoch": 0.15141, + "grad_norm": 0.8789671962577075, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15141 + }, + { + "epoch": 0.15142, + "grad_norm": 0.9375251702512206, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 15142 + }, + { + "epoch": 0.15143, + "grad_norm": 0.9452439835336452, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 15143 + }, + { + "epoch": 0.15144, + "grad_norm": 0.7582854261133937, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 15144 + }, + { + "epoch": 0.15145, + "grad_norm": 0.7252237418968446, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 15145 + }, + { + "epoch": 0.15146, + "grad_norm": 0.6452971517261687, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 15146 + }, + { + "epoch": 0.15147, + "grad_norm": 0.6870777146109712, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 15147 + }, + { + "epoch": 0.15148, + "grad_norm": 0.663763526693065, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 15148 + }, + { + "epoch": 0.15149, + "grad_norm": 0.7233420335503309, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 15149 + }, + { + "epoch": 0.1515, + "grad_norm": 0.6778911070278002, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 15150 + }, + { + "epoch": 0.15151, + "grad_norm": 0.7401669644849158, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 15151 + }, + { + "epoch": 0.15152, + "grad_norm": 0.7560043827099342, + "learning_rate": 0.003, + "loss": 4.0599, + "step": 15152 + }, + { + "epoch": 0.15153, + "grad_norm": 0.6551508428575531, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 15153 + }, + { + "epoch": 0.15154, + "grad_norm": 0.6786826382790307, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 15154 + }, + { + "epoch": 0.15155, + "grad_norm": 0.5871800420684692, + "learning_rate": 0.003, + "loss": 4.022, + "step": 15155 + }, + { + "epoch": 0.15156, + "grad_norm": 0.6237249032404997, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 15156 + }, + { + "epoch": 0.15157, + "grad_norm": 0.5779654175797776, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 15157 + }, + { + "epoch": 0.15158, + "grad_norm": 0.5247510459903141, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 15158 + }, + { + "epoch": 0.15159, + "grad_norm": 0.4992457127166718, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 15159 + }, + { + "epoch": 0.1516, + "grad_norm": 0.4946031965332622, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 15160 + }, + { + "epoch": 0.15161, + "grad_norm": 0.49255031065897426, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 15161 + }, + { + "epoch": 0.15162, + "grad_norm": 0.5020861633046483, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 15162 + }, + { + "epoch": 0.15163, + "grad_norm": 0.6511415514511834, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 15163 + }, + { + "epoch": 0.15164, + "grad_norm": 0.9213131524216522, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 15164 + }, + { + "epoch": 0.15165, + "grad_norm": 1.189420437222522, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 15165 + }, + { + "epoch": 0.15166, + "grad_norm": 0.8021515163168167, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 15166 + }, + { + "epoch": 0.15167, + "grad_norm": 0.6124817280156382, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 15167 + }, + { + "epoch": 0.15168, + "grad_norm": 0.6374703983876506, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 15168 + }, + { + "epoch": 0.15169, + "grad_norm": 0.6468483751561573, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 15169 + }, + { + "epoch": 0.1517, + "grad_norm": 0.7009922670505879, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 15170 + }, + { + "epoch": 0.15171, + "grad_norm": 0.7164695313266437, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 15171 + }, + { + "epoch": 0.15172, + "grad_norm": 0.7577772926064306, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 15172 + }, + { + "epoch": 0.15173, + "grad_norm": 0.7864425686883857, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 15173 + }, + { + "epoch": 0.15174, + "grad_norm": 0.837240643029611, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 15174 + }, + { + "epoch": 0.15175, + "grad_norm": 0.8580491132734692, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 15175 + }, + { + "epoch": 0.15176, + "grad_norm": 0.8138122495070191, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 15176 + }, + { + "epoch": 0.15177, + "grad_norm": 0.843466556500434, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 15177 + }, + { + "epoch": 0.15178, + "grad_norm": 0.6929303008172377, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 15178 + }, + { + "epoch": 0.15179, + "grad_norm": 0.7115946370284995, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 15179 + }, + { + "epoch": 0.1518, + "grad_norm": 0.7434961837459984, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 15180 + }, + { + "epoch": 0.15181, + "grad_norm": 0.9315616487701691, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 15181 + }, + { + "epoch": 0.15182, + "grad_norm": 1.4095435489602892, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 15182 + }, + { + "epoch": 0.15183, + "grad_norm": 0.801821088804249, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 15183 + }, + { + "epoch": 0.15184, + "grad_norm": 0.8620576544975488, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 15184 + }, + { + "epoch": 0.15185, + "grad_norm": 0.9808157358623367, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 15185 + }, + { + "epoch": 0.15186, + "grad_norm": 1.1231599511802468, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 15186 + }, + { + "epoch": 0.15187, + "grad_norm": 1.016733422402779, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 15187 + }, + { + "epoch": 0.15188, + "grad_norm": 1.034337989559245, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 15188 + }, + { + "epoch": 0.15189, + "grad_norm": 1.0001458241725552, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 15189 + }, + { + "epoch": 0.1519, + "grad_norm": 0.9289118672312923, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 15190 + }, + { + "epoch": 0.15191, + "grad_norm": 0.8867299476551963, + "learning_rate": 0.003, + "loss": 4.0646, + "step": 15191 + }, + { + "epoch": 0.15192, + "grad_norm": 0.983146708162778, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 15192 + }, + { + "epoch": 0.15193, + "grad_norm": 1.0643450068484763, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 15193 + }, + { + "epoch": 0.15194, + "grad_norm": 1.0794758916559124, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 15194 + }, + { + "epoch": 0.15195, + "grad_norm": 1.0745630196748042, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 15195 + }, + { + "epoch": 0.15196, + "grad_norm": 1.0280824228848064, + "learning_rate": 0.003, + "loss": 4.0542, + "step": 15196 + }, + { + "epoch": 0.15197, + "grad_norm": 0.8287982905706978, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 15197 + }, + { + "epoch": 0.15198, + "grad_norm": 0.7980194198164525, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 15198 + }, + { + "epoch": 0.15199, + "grad_norm": 0.8536088270740431, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 15199 + }, + { + "epoch": 0.152, + "grad_norm": 0.8027801273560791, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 15200 + }, + { + "epoch": 0.15201, + "grad_norm": 0.6344109692594445, + "learning_rate": 0.003, + "loss": 4.029, + "step": 15201 + }, + { + "epoch": 0.15202, + "grad_norm": 0.6190132301174052, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 15202 + }, + { + "epoch": 0.15203, + "grad_norm": 0.5765399569449291, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 15203 + }, + { + "epoch": 0.15204, + "grad_norm": 0.5161402277059763, + "learning_rate": 0.003, + "loss": 4.049, + "step": 15204 + }, + { + "epoch": 0.15205, + "grad_norm": 0.5368377855292451, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 15205 + }, + { + "epoch": 0.15206, + "grad_norm": 0.6428741038844351, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 15206 + }, + { + "epoch": 0.15207, + "grad_norm": 0.6658547786906518, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 15207 + }, + { + "epoch": 0.15208, + "grad_norm": 0.7173461569592662, + "learning_rate": 0.003, + "loss": 3.983, + "step": 15208 + }, + { + "epoch": 0.15209, + "grad_norm": 0.7056020387574599, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 15209 + }, + { + "epoch": 0.1521, + "grad_norm": 0.7410444071478915, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 15210 + }, + { + "epoch": 0.15211, + "grad_norm": 0.7775259843292164, + "learning_rate": 0.003, + "loss": 4.033, + "step": 15211 + }, + { + "epoch": 0.15212, + "grad_norm": 0.802249574303423, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 15212 + }, + { + "epoch": 0.15213, + "grad_norm": 0.7463368165069671, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 15213 + }, + { + "epoch": 0.15214, + "grad_norm": 0.6684514494042204, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 15214 + }, + { + "epoch": 0.15215, + "grad_norm": 0.6256418224355852, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 15215 + }, + { + "epoch": 0.15216, + "grad_norm": 0.6053072946052895, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 15216 + }, + { + "epoch": 0.15217, + "grad_norm": 0.5841422333245886, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15217 + }, + { + "epoch": 0.15218, + "grad_norm": 0.5272461870322407, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 15218 + }, + { + "epoch": 0.15219, + "grad_norm": 0.5649243905819423, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15219 + }, + { + "epoch": 0.1522, + "grad_norm": 0.5580943296223243, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 15220 + }, + { + "epoch": 0.15221, + "grad_norm": 0.5302877751518691, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 15221 + }, + { + "epoch": 0.15222, + "grad_norm": 0.5318505697784107, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 15222 + }, + { + "epoch": 0.15223, + "grad_norm": 0.5219506699187867, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 15223 + }, + { + "epoch": 0.15224, + "grad_norm": 0.5569416944945081, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 15224 + }, + { + "epoch": 0.15225, + "grad_norm": 0.5811361862283355, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 15225 + }, + { + "epoch": 0.15226, + "grad_norm": 0.6112711300140231, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 15226 + }, + { + "epoch": 0.15227, + "grad_norm": 0.5425053397772428, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 15227 + }, + { + "epoch": 0.15228, + "grad_norm": 0.6584523148157674, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 15228 + }, + { + "epoch": 0.15229, + "grad_norm": 0.7122393716831691, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 15229 + }, + { + "epoch": 0.1523, + "grad_norm": 0.5943442641708316, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 15230 + }, + { + "epoch": 0.15231, + "grad_norm": 0.5865286076725247, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 15231 + }, + { + "epoch": 0.15232, + "grad_norm": 0.7938964829181663, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 15232 + }, + { + "epoch": 0.15233, + "grad_norm": 1.1553880713156224, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 15233 + }, + { + "epoch": 0.15234, + "grad_norm": 0.905944539427331, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 15234 + }, + { + "epoch": 0.15235, + "grad_norm": 0.7343899628924413, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 15235 + }, + { + "epoch": 0.15236, + "grad_norm": 0.7138053658577701, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 15236 + }, + { + "epoch": 0.15237, + "grad_norm": 0.7450088517820894, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 15237 + }, + { + "epoch": 0.15238, + "grad_norm": 0.8447929990535474, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 15238 + }, + { + "epoch": 0.15239, + "grad_norm": 0.8215677220170566, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 15239 + }, + { + "epoch": 0.1524, + "grad_norm": 0.741829460222774, + "learning_rate": 0.003, + "loss": 4.009, + "step": 15240 + }, + { + "epoch": 0.15241, + "grad_norm": 0.6382503992649025, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 15241 + }, + { + "epoch": 0.15242, + "grad_norm": 0.6919796312411944, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 15242 + }, + { + "epoch": 0.15243, + "grad_norm": 0.6805701473730174, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 15243 + }, + { + "epoch": 0.15244, + "grad_norm": 0.6632898497538519, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 15244 + }, + { + "epoch": 0.15245, + "grad_norm": 0.6301514947893682, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15245 + }, + { + "epoch": 0.15246, + "grad_norm": 0.638323100382206, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 15246 + }, + { + "epoch": 0.15247, + "grad_norm": 0.6484066178037714, + "learning_rate": 0.003, + "loss": 4.003, + "step": 15247 + }, + { + "epoch": 0.15248, + "grad_norm": 0.7253989964036154, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 15248 + }, + { + "epoch": 0.15249, + "grad_norm": 0.8564218867436894, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 15249 + }, + { + "epoch": 0.1525, + "grad_norm": 0.9867069338790672, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 15250 + }, + { + "epoch": 0.15251, + "grad_norm": 1.029116225523085, + "learning_rate": 0.003, + "loss": 4.0574, + "step": 15251 + }, + { + "epoch": 0.15252, + "grad_norm": 0.9531720101762685, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 15252 + }, + { + "epoch": 0.15253, + "grad_norm": 0.7650031980090297, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 15253 + }, + { + "epoch": 0.15254, + "grad_norm": 0.7455914640324973, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15254 + }, + { + "epoch": 0.15255, + "grad_norm": 0.8082264965610977, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 15255 + }, + { + "epoch": 0.15256, + "grad_norm": 0.7253810469338458, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15256 + }, + { + "epoch": 0.15257, + "grad_norm": 0.7038740091617554, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 15257 + }, + { + "epoch": 0.15258, + "grad_norm": 0.7354549601890636, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 15258 + }, + { + "epoch": 0.15259, + "grad_norm": 0.7205672926250963, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 15259 + }, + { + "epoch": 0.1526, + "grad_norm": 0.6898993388562645, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 15260 + }, + { + "epoch": 0.15261, + "grad_norm": 0.6536684054964241, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 15261 + }, + { + "epoch": 0.15262, + "grad_norm": 0.5153500656569336, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15262 + }, + { + "epoch": 0.15263, + "grad_norm": 0.5244336164078591, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 15263 + }, + { + "epoch": 0.15264, + "grad_norm": 0.505081529206954, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 15264 + }, + { + "epoch": 0.15265, + "grad_norm": 0.5698793105052802, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 15265 + }, + { + "epoch": 0.15266, + "grad_norm": 0.6897454799803584, + "learning_rate": 0.003, + "loss": 3.984, + "step": 15266 + }, + { + "epoch": 0.15267, + "grad_norm": 0.8462637599729078, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 15267 + }, + { + "epoch": 0.15268, + "grad_norm": 0.9229954358184341, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15268 + }, + { + "epoch": 0.15269, + "grad_norm": 0.8993021329227955, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 15269 + }, + { + "epoch": 0.1527, + "grad_norm": 0.7745762815451938, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 15270 + }, + { + "epoch": 0.15271, + "grad_norm": 0.9015333643279616, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 15271 + }, + { + "epoch": 0.15272, + "grad_norm": 1.057311829641025, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 15272 + }, + { + "epoch": 0.15273, + "grad_norm": 0.8509038695848292, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 15273 + }, + { + "epoch": 0.15274, + "grad_norm": 0.6676502081914841, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 15274 + }, + { + "epoch": 0.15275, + "grad_norm": 0.6119554863771784, + "learning_rate": 0.003, + "loss": 4.014, + "step": 15275 + }, + { + "epoch": 0.15276, + "grad_norm": 0.6498895972974201, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 15276 + }, + { + "epoch": 0.15277, + "grad_norm": 0.6897529066855737, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 15277 + }, + { + "epoch": 0.15278, + "grad_norm": 0.7134782833601273, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 15278 + }, + { + "epoch": 0.15279, + "grad_norm": 0.7412962538306852, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 15279 + }, + { + "epoch": 0.1528, + "grad_norm": 0.825421710549414, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 15280 + }, + { + "epoch": 0.15281, + "grad_norm": 0.886082970710489, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 15281 + }, + { + "epoch": 0.15282, + "grad_norm": 0.8947999491013672, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 15282 + }, + { + "epoch": 0.15283, + "grad_norm": 0.9265135393517352, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 15283 + }, + { + "epoch": 0.15284, + "grad_norm": 0.8853288884888935, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 15284 + }, + { + "epoch": 0.15285, + "grad_norm": 0.8005871164625958, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 15285 + }, + { + "epoch": 0.15286, + "grad_norm": 0.82596890267072, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 15286 + }, + { + "epoch": 0.15287, + "grad_norm": 0.8074543413189929, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 15287 + }, + { + "epoch": 0.15288, + "grad_norm": 0.7004858762601054, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 15288 + }, + { + "epoch": 0.15289, + "grad_norm": 0.6077715032621099, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 15289 + }, + { + "epoch": 0.1529, + "grad_norm": 0.5989581351900294, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 15290 + }, + { + "epoch": 0.15291, + "grad_norm": 0.5794144872454734, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 15291 + }, + { + "epoch": 0.15292, + "grad_norm": 0.5870203094275899, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 15292 + }, + { + "epoch": 0.15293, + "grad_norm": 0.5854949685748848, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 15293 + }, + { + "epoch": 0.15294, + "grad_norm": 0.62444999797568, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 15294 + }, + { + "epoch": 0.15295, + "grad_norm": 0.6905051278271871, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 15295 + }, + { + "epoch": 0.15296, + "grad_norm": 0.7040553213359015, + "learning_rate": 0.003, + "loss": 4.035, + "step": 15296 + }, + { + "epoch": 0.15297, + "grad_norm": 0.6832879859974013, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 15297 + }, + { + "epoch": 0.15298, + "grad_norm": 0.7189607219099731, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 15298 + }, + { + "epoch": 0.15299, + "grad_norm": 0.740121270491045, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 15299 + }, + { + "epoch": 0.153, + "grad_norm": 0.65572424179827, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 15300 + }, + { + "epoch": 0.15301, + "grad_norm": 0.6488893208083502, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 15301 + }, + { + "epoch": 0.15302, + "grad_norm": 0.7437616389065304, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 15302 + }, + { + "epoch": 0.15303, + "grad_norm": 0.8500053505687012, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 15303 + }, + { + "epoch": 0.15304, + "grad_norm": 0.9927679009649721, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 15304 + }, + { + "epoch": 0.15305, + "grad_norm": 1.0866008800254616, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 15305 + }, + { + "epoch": 0.15306, + "grad_norm": 0.7263969705507841, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 15306 + }, + { + "epoch": 0.15307, + "grad_norm": 0.7000871915223934, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 15307 + }, + { + "epoch": 0.15308, + "grad_norm": 0.6567365319225736, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 15308 + }, + { + "epoch": 0.15309, + "grad_norm": 0.7304477553347791, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 15309 + }, + { + "epoch": 0.1531, + "grad_norm": 0.6949574643412535, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 15310 + }, + { + "epoch": 0.15311, + "grad_norm": 0.6299099903989768, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 15311 + }, + { + "epoch": 0.15312, + "grad_norm": 0.601278909334528, + "learning_rate": 0.003, + "loss": 4.009, + "step": 15312 + }, + { + "epoch": 0.15313, + "grad_norm": 0.5868559846907907, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 15313 + }, + { + "epoch": 0.15314, + "grad_norm": 0.5842895790188793, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 15314 + }, + { + "epoch": 0.15315, + "grad_norm": 0.6389563967547562, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 15315 + }, + { + "epoch": 0.15316, + "grad_norm": 0.7017034815421501, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 15316 + }, + { + "epoch": 0.15317, + "grad_norm": 0.7279992263141329, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 15317 + }, + { + "epoch": 0.15318, + "grad_norm": 0.655026369587428, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 15318 + }, + { + "epoch": 0.15319, + "grad_norm": 0.6593184729996399, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 15319 + }, + { + "epoch": 0.1532, + "grad_norm": 0.7105761931008014, + "learning_rate": 0.003, + "loss": 4.007, + "step": 15320 + }, + { + "epoch": 0.15321, + "grad_norm": 0.7924328601071415, + "learning_rate": 0.003, + "loss": 4.015, + "step": 15321 + }, + { + "epoch": 0.15322, + "grad_norm": 0.7850925486621507, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 15322 + }, + { + "epoch": 0.15323, + "grad_norm": 0.8106813895453915, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 15323 + }, + { + "epoch": 0.15324, + "grad_norm": 0.7723266916300019, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 15324 + }, + { + "epoch": 0.15325, + "grad_norm": 0.8328265149636992, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 15325 + }, + { + "epoch": 0.15326, + "grad_norm": 0.682560193722491, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 15326 + }, + { + "epoch": 0.15327, + "grad_norm": 0.7226028949055369, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 15327 + }, + { + "epoch": 0.15328, + "grad_norm": 0.7026774225150242, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 15328 + }, + { + "epoch": 0.15329, + "grad_norm": 0.7657495916568188, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 15329 + }, + { + "epoch": 0.1533, + "grad_norm": 0.7637842016293398, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 15330 + }, + { + "epoch": 0.15331, + "grad_norm": 0.8200691921295185, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 15331 + }, + { + "epoch": 0.15332, + "grad_norm": 0.9855454331377194, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 15332 + }, + { + "epoch": 0.15333, + "grad_norm": 1.0658194690355363, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 15333 + }, + { + "epoch": 0.15334, + "grad_norm": 0.816007050283126, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 15334 + }, + { + "epoch": 0.15335, + "grad_norm": 0.7972615108144802, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 15335 + }, + { + "epoch": 0.15336, + "grad_norm": 0.674209335985942, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 15336 + }, + { + "epoch": 0.15337, + "grad_norm": 0.7280734558569729, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 15337 + }, + { + "epoch": 0.15338, + "grad_norm": 0.7883329997298748, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 15338 + }, + { + "epoch": 0.15339, + "grad_norm": 0.7925766342032173, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 15339 + }, + { + "epoch": 0.1534, + "grad_norm": 0.8062953036573143, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 15340 + }, + { + "epoch": 0.15341, + "grad_norm": 0.9458232497572975, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 15341 + }, + { + "epoch": 0.15342, + "grad_norm": 1.0751994901963307, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 15342 + }, + { + "epoch": 0.15343, + "grad_norm": 0.82465335193746, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15343 + }, + { + "epoch": 0.15344, + "grad_norm": 0.7240939409628286, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 15344 + }, + { + "epoch": 0.15345, + "grad_norm": 0.7354291899368567, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 15345 + }, + { + "epoch": 0.15346, + "grad_norm": 0.7030218576669397, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 15346 + }, + { + "epoch": 0.15347, + "grad_norm": 0.7287815505026385, + "learning_rate": 0.003, + "loss": 4.001, + "step": 15347 + }, + { + "epoch": 0.15348, + "grad_norm": 0.7722922796783326, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 15348 + }, + { + "epoch": 0.15349, + "grad_norm": 0.8327830747760604, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 15349 + }, + { + "epoch": 0.1535, + "grad_norm": 0.7295508852305, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 15350 + }, + { + "epoch": 0.15351, + "grad_norm": 0.6307739894086412, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 15351 + }, + { + "epoch": 0.15352, + "grad_norm": 0.5649470981192235, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 15352 + }, + { + "epoch": 0.15353, + "grad_norm": 0.5163164284898917, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 15353 + }, + { + "epoch": 0.15354, + "grad_norm": 0.5581505447269689, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 15354 + }, + { + "epoch": 0.15355, + "grad_norm": 0.5677968405450405, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 15355 + }, + { + "epoch": 0.15356, + "grad_norm": 0.598810935121117, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 15356 + }, + { + "epoch": 0.15357, + "grad_norm": 0.5694783211080593, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 15357 + }, + { + "epoch": 0.15358, + "grad_norm": 0.6195740174308718, + "learning_rate": 0.003, + "loss": 3.999, + "step": 15358 + }, + { + "epoch": 0.15359, + "grad_norm": 0.5655552821923536, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 15359 + }, + { + "epoch": 0.1536, + "grad_norm": 0.5331590170089439, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 15360 + }, + { + "epoch": 0.15361, + "grad_norm": 0.48124670130964226, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 15361 + }, + { + "epoch": 0.15362, + "grad_norm": 0.588817098573973, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 15362 + }, + { + "epoch": 0.15363, + "grad_norm": 0.7714486766391149, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 15363 + }, + { + "epoch": 0.15364, + "grad_norm": 1.0157582673431647, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 15364 + }, + { + "epoch": 0.15365, + "grad_norm": 1.2080815666320524, + "learning_rate": 0.003, + "loss": 4.044, + "step": 15365 + }, + { + "epoch": 0.15366, + "grad_norm": 0.5913285234266148, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 15366 + }, + { + "epoch": 0.15367, + "grad_norm": 0.6794852154022021, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 15367 + }, + { + "epoch": 0.15368, + "grad_norm": 0.8184536845699646, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 15368 + }, + { + "epoch": 0.15369, + "grad_norm": 0.9643990254545532, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 15369 + }, + { + "epoch": 0.1537, + "grad_norm": 1.061437433374094, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 15370 + }, + { + "epoch": 0.15371, + "grad_norm": 0.8111625231347493, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 15371 + }, + { + "epoch": 0.15372, + "grad_norm": 0.7610913542958062, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 15372 + }, + { + "epoch": 0.15373, + "grad_norm": 0.7066614494032745, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 15373 + }, + { + "epoch": 0.15374, + "grad_norm": 0.6462381740507261, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 15374 + }, + { + "epoch": 0.15375, + "grad_norm": 0.6002781992910454, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 15375 + }, + { + "epoch": 0.15376, + "grad_norm": 0.6308157484835002, + "learning_rate": 0.003, + "loss": 4.026, + "step": 15376 + }, + { + "epoch": 0.15377, + "grad_norm": 0.591818653811065, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 15377 + }, + { + "epoch": 0.15378, + "grad_norm": 0.6344841990965829, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 15378 + }, + { + "epoch": 0.15379, + "grad_norm": 0.6191008176859728, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 15379 + }, + { + "epoch": 0.1538, + "grad_norm": 0.671042109693734, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 15380 + }, + { + "epoch": 0.15381, + "grad_norm": 0.7325212713155278, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 15381 + }, + { + "epoch": 0.15382, + "grad_norm": 0.6349128611615833, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 15382 + }, + { + "epoch": 0.15383, + "grad_norm": 0.5508065476488367, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 15383 + }, + { + "epoch": 0.15384, + "grad_norm": 0.5307510688837791, + "learning_rate": 0.003, + "loss": 4.005, + "step": 15384 + }, + { + "epoch": 0.15385, + "grad_norm": 0.6013524426025216, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 15385 + }, + { + "epoch": 0.15386, + "grad_norm": 0.7310203278758454, + "learning_rate": 0.003, + "loss": 3.981, + "step": 15386 + }, + { + "epoch": 0.15387, + "grad_norm": 0.8904027414974, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 15387 + }, + { + "epoch": 0.15388, + "grad_norm": 0.9850954202841145, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 15388 + }, + { + "epoch": 0.15389, + "grad_norm": 1.0721880633000358, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 15389 + }, + { + "epoch": 0.1539, + "grad_norm": 0.7978715613426326, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 15390 + }, + { + "epoch": 0.15391, + "grad_norm": 0.678826703296583, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 15391 + }, + { + "epoch": 0.15392, + "grad_norm": 0.627895073167145, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15392 + }, + { + "epoch": 0.15393, + "grad_norm": 0.6564841180293643, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 15393 + }, + { + "epoch": 0.15394, + "grad_norm": 0.6116069421196719, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 15394 + }, + { + "epoch": 0.15395, + "grad_norm": 0.5517679598262513, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 15395 + }, + { + "epoch": 0.15396, + "grad_norm": 0.5911917839889207, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 15396 + }, + { + "epoch": 0.15397, + "grad_norm": 0.5863640616724395, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 15397 + }, + { + "epoch": 0.15398, + "grad_norm": 0.5727585626253856, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 15398 + }, + { + "epoch": 0.15399, + "grad_norm": 0.510968328480277, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 15399 + }, + { + "epoch": 0.154, + "grad_norm": 0.541629472454527, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 15400 + }, + { + "epoch": 0.15401, + "grad_norm": 0.6791602455063686, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 15401 + }, + { + "epoch": 0.15402, + "grad_norm": 0.866377742277208, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 15402 + }, + { + "epoch": 0.15403, + "grad_norm": 0.9555404179707458, + "learning_rate": 0.003, + "loss": 3.994, + "step": 15403 + }, + { + "epoch": 0.15404, + "grad_norm": 0.9493374086670127, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 15404 + }, + { + "epoch": 0.15405, + "grad_norm": 0.9496814057891808, + "learning_rate": 0.003, + "loss": 4.0567, + "step": 15405 + }, + { + "epoch": 0.15406, + "grad_norm": 0.9121919406690933, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15406 + }, + { + "epoch": 0.15407, + "grad_norm": 0.7573670836032641, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 15407 + }, + { + "epoch": 0.15408, + "grad_norm": 0.755892955856889, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 15408 + }, + { + "epoch": 0.15409, + "grad_norm": 0.7743898254217108, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 15409 + }, + { + "epoch": 0.1541, + "grad_norm": 0.7986861727165735, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 15410 + }, + { + "epoch": 0.15411, + "grad_norm": 0.8071568617283438, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 15411 + }, + { + "epoch": 0.15412, + "grad_norm": 0.8286469729399101, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15412 + }, + { + "epoch": 0.15413, + "grad_norm": 0.8892172169738111, + "learning_rate": 0.003, + "loss": 4.0607, + "step": 15413 + }, + { + "epoch": 0.15414, + "grad_norm": 0.9841807401898698, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 15414 + }, + { + "epoch": 0.15415, + "grad_norm": 0.8827652836281675, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 15415 + }, + { + "epoch": 0.15416, + "grad_norm": 0.8152060597104859, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 15416 + }, + { + "epoch": 0.15417, + "grad_norm": 0.8889084260325876, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 15417 + }, + { + "epoch": 0.15418, + "grad_norm": 1.1560663896668033, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 15418 + }, + { + "epoch": 0.15419, + "grad_norm": 0.9297082235269862, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 15419 + }, + { + "epoch": 0.1542, + "grad_norm": 0.8624625922855407, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 15420 + }, + { + "epoch": 0.15421, + "grad_norm": 0.8689782984838949, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 15421 + }, + { + "epoch": 0.15422, + "grad_norm": 0.824494046509237, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15422 + }, + { + "epoch": 0.15423, + "grad_norm": 0.9033547595891516, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 15423 + }, + { + "epoch": 0.15424, + "grad_norm": 0.8487702158951896, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 15424 + }, + { + "epoch": 0.15425, + "grad_norm": 0.9002333381880879, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15425 + }, + { + "epoch": 0.15426, + "grad_norm": 0.8635246010963689, + "learning_rate": 0.003, + "loss": 4.007, + "step": 15426 + }, + { + "epoch": 0.15427, + "grad_norm": 0.6887618859803827, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 15427 + }, + { + "epoch": 0.15428, + "grad_norm": 0.7489132440411961, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 15428 + }, + { + "epoch": 0.15429, + "grad_norm": 0.6924049741651879, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 15429 + }, + { + "epoch": 0.1543, + "grad_norm": 0.6526640206012081, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 15430 + }, + { + "epoch": 0.15431, + "grad_norm": 0.5637289725717021, + "learning_rate": 0.003, + "loss": 4.042, + "step": 15431 + }, + { + "epoch": 0.15432, + "grad_norm": 0.5380505601070791, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 15432 + }, + { + "epoch": 0.15433, + "grad_norm": 0.54993468244848, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 15433 + }, + { + "epoch": 0.15434, + "grad_norm": 0.5009217125845753, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 15434 + }, + { + "epoch": 0.15435, + "grad_norm": 0.5165559632601738, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 15435 + }, + { + "epoch": 0.15436, + "grad_norm": 0.5100327648265632, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 15436 + }, + { + "epoch": 0.15437, + "grad_norm": 0.5325383830673746, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 15437 + }, + { + "epoch": 0.15438, + "grad_norm": 0.5874977953624729, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 15438 + }, + { + "epoch": 0.15439, + "grad_norm": 0.7342590069300173, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 15439 + }, + { + "epoch": 0.1544, + "grad_norm": 0.9056012694537432, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 15440 + }, + { + "epoch": 0.15441, + "grad_norm": 0.9751630349222324, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 15441 + }, + { + "epoch": 0.15442, + "grad_norm": 0.7668223000131096, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 15442 + }, + { + "epoch": 0.15443, + "grad_norm": 0.6420887132079744, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 15443 + }, + { + "epoch": 0.15444, + "grad_norm": 0.72289835502226, + "learning_rate": 0.003, + "loss": 4.004, + "step": 15444 + }, + { + "epoch": 0.15445, + "grad_norm": 0.7002610152384341, + "learning_rate": 0.003, + "loss": 3.969, + "step": 15445 + }, + { + "epoch": 0.15446, + "grad_norm": 0.6347760696384761, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 15446 + }, + { + "epoch": 0.15447, + "grad_norm": 0.5421844441898771, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 15447 + }, + { + "epoch": 0.15448, + "grad_norm": 0.5623365172051468, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 15448 + }, + { + "epoch": 0.15449, + "grad_norm": 0.6089907240486931, + "learning_rate": 0.003, + "loss": 3.993, + "step": 15449 + }, + { + "epoch": 0.1545, + "grad_norm": 0.5852111704663608, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 15450 + }, + { + "epoch": 0.15451, + "grad_norm": 0.613234716648087, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 15451 + }, + { + "epoch": 0.15452, + "grad_norm": 0.5879351753933841, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 15452 + }, + { + "epoch": 0.15453, + "grad_norm": 0.4899041589720017, + "learning_rate": 0.003, + "loss": 3.99, + "step": 15453 + }, + { + "epoch": 0.15454, + "grad_norm": 0.5058735467383264, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 15454 + }, + { + "epoch": 0.15455, + "grad_norm": 0.42214705346315806, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 15455 + }, + { + "epoch": 0.15456, + "grad_norm": 0.43464594008546864, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 15456 + }, + { + "epoch": 0.15457, + "grad_norm": 0.49892823653749724, + "learning_rate": 0.003, + "loss": 4.007, + "step": 15457 + }, + { + "epoch": 0.15458, + "grad_norm": 0.6351817095674852, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15458 + }, + { + "epoch": 0.15459, + "grad_norm": 0.8294508086861385, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 15459 + }, + { + "epoch": 0.1546, + "grad_norm": 0.9144930843378102, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 15460 + }, + { + "epoch": 0.15461, + "grad_norm": 0.7338717018728959, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 15461 + }, + { + "epoch": 0.15462, + "grad_norm": 0.751184692835154, + "learning_rate": 0.003, + "loss": 3.992, + "step": 15462 + }, + { + "epoch": 0.15463, + "grad_norm": 0.8154314491374894, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 15463 + }, + { + "epoch": 0.15464, + "grad_norm": 0.7300752196315508, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 15464 + }, + { + "epoch": 0.15465, + "grad_norm": 0.7239281624222859, + "learning_rate": 0.003, + "loss": 3.993, + "step": 15465 + }, + { + "epoch": 0.15466, + "grad_norm": 0.7637940734192566, + "learning_rate": 0.003, + "loss": 4.02, + "step": 15466 + }, + { + "epoch": 0.15467, + "grad_norm": 0.7282862626908271, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 15467 + }, + { + "epoch": 0.15468, + "grad_norm": 0.8659888685079891, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 15468 + }, + { + "epoch": 0.15469, + "grad_norm": 1.0532446466264087, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 15469 + }, + { + "epoch": 0.1547, + "grad_norm": 0.8322569754645831, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 15470 + }, + { + "epoch": 0.15471, + "grad_norm": 0.7129918460509422, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 15471 + }, + { + "epoch": 0.15472, + "grad_norm": 0.6625779706728259, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 15472 + }, + { + "epoch": 0.15473, + "grad_norm": 0.6864529574539354, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 15473 + }, + { + "epoch": 0.15474, + "grad_norm": 0.7094412683000448, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 15474 + }, + { + "epoch": 0.15475, + "grad_norm": 0.7644808387686064, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 15475 + }, + { + "epoch": 0.15476, + "grad_norm": 0.7970507768159413, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 15476 + }, + { + "epoch": 0.15477, + "grad_norm": 0.8096017378990372, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 15477 + }, + { + "epoch": 0.15478, + "grad_norm": 0.7386666706273246, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 15478 + }, + { + "epoch": 0.15479, + "grad_norm": 0.6633049474350242, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 15479 + }, + { + "epoch": 0.1548, + "grad_norm": 0.6678599322763658, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 15480 + }, + { + "epoch": 0.15481, + "grad_norm": 0.6328592459092854, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 15481 + }, + { + "epoch": 0.15482, + "grad_norm": 0.6159242897626688, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 15482 + }, + { + "epoch": 0.15483, + "grad_norm": 0.7092590306203305, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 15483 + }, + { + "epoch": 0.15484, + "grad_norm": 0.860979338565048, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 15484 + }, + { + "epoch": 0.15485, + "grad_norm": 0.9110974211473755, + "learning_rate": 0.003, + "loss": 4.0437, + "step": 15485 + }, + { + "epoch": 0.15486, + "grad_norm": 0.7995215308235661, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 15486 + }, + { + "epoch": 0.15487, + "grad_norm": 0.7572599876403941, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 15487 + }, + { + "epoch": 0.15488, + "grad_norm": 0.7876043175043531, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 15488 + }, + { + "epoch": 0.15489, + "grad_norm": 0.7744435049715426, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 15489 + }, + { + "epoch": 0.1549, + "grad_norm": 0.8902141170678104, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 15490 + }, + { + "epoch": 0.15491, + "grad_norm": 1.027198894243121, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 15491 + }, + { + "epoch": 0.15492, + "grad_norm": 0.933100825910686, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 15492 + }, + { + "epoch": 0.15493, + "grad_norm": 0.9925318734742423, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 15493 + }, + { + "epoch": 0.15494, + "grad_norm": 1.0623096174728075, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 15494 + }, + { + "epoch": 0.15495, + "grad_norm": 1.0068324341341486, + "learning_rate": 0.003, + "loss": 4.0601, + "step": 15495 + }, + { + "epoch": 0.15496, + "grad_norm": 0.8855063467047851, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 15496 + }, + { + "epoch": 0.15497, + "grad_norm": 0.9807195764184613, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 15497 + }, + { + "epoch": 0.15498, + "grad_norm": 0.8299377546453467, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 15498 + }, + { + "epoch": 0.15499, + "grad_norm": 0.7987939337613517, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 15499 + }, + { + "epoch": 0.155, + "grad_norm": 0.838956932834113, + "learning_rate": 0.003, + "loss": 4.034, + "step": 15500 + }, + { + "epoch": 0.15501, + "grad_norm": 0.8643720924213201, + "learning_rate": 0.003, + "loss": 4.025, + "step": 15501 + }, + { + "epoch": 0.15502, + "grad_norm": 0.7767517199061892, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 15502 + }, + { + "epoch": 0.15503, + "grad_norm": 0.6936068122786285, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 15503 + }, + { + "epoch": 0.15504, + "grad_norm": 0.6091911526222404, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 15504 + }, + { + "epoch": 0.15505, + "grad_norm": 0.5486952110411002, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 15505 + }, + { + "epoch": 0.15506, + "grad_norm": 0.5399886330955185, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 15506 + }, + { + "epoch": 0.15507, + "grad_norm": 0.5423984945447168, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 15507 + }, + { + "epoch": 0.15508, + "grad_norm": 0.6092483327575672, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 15508 + }, + { + "epoch": 0.15509, + "grad_norm": 0.7416695892907649, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 15509 + }, + { + "epoch": 0.1551, + "grad_norm": 0.8928959946405204, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 15510 + }, + { + "epoch": 0.15511, + "grad_norm": 0.9457258821501877, + "learning_rate": 0.003, + "loss": 4.053, + "step": 15511 + }, + { + "epoch": 0.15512, + "grad_norm": 0.825183325082099, + "learning_rate": 0.003, + "loss": 4.048, + "step": 15512 + }, + { + "epoch": 0.15513, + "grad_norm": 0.6496122269694347, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 15513 + }, + { + "epoch": 0.15514, + "grad_norm": 0.7047667919224528, + "learning_rate": 0.003, + "loss": 4.045, + "step": 15514 + }, + { + "epoch": 0.15515, + "grad_norm": 0.8627742097184382, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 15515 + }, + { + "epoch": 0.15516, + "grad_norm": 0.7579852907056911, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 15516 + }, + { + "epoch": 0.15517, + "grad_norm": 0.7718410137895048, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 15517 + }, + { + "epoch": 0.15518, + "grad_norm": 0.7198112219989408, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 15518 + }, + { + "epoch": 0.15519, + "grad_norm": 0.651173206242662, + "learning_rate": 0.003, + "loss": 3.999, + "step": 15519 + }, + { + "epoch": 0.1552, + "grad_norm": 0.5127531583063887, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 15520 + }, + { + "epoch": 0.15521, + "grad_norm": 0.5998245317190984, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 15521 + }, + { + "epoch": 0.15522, + "grad_norm": 0.6113473958004474, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 15522 + }, + { + "epoch": 0.15523, + "grad_norm": 0.6517424197163805, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 15523 + }, + { + "epoch": 0.15524, + "grad_norm": 0.6463335899757282, + "learning_rate": 0.003, + "loss": 4.022, + "step": 15524 + }, + { + "epoch": 0.15525, + "grad_norm": 0.6388109002356205, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 15525 + }, + { + "epoch": 0.15526, + "grad_norm": 0.6829255537096299, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 15526 + }, + { + "epoch": 0.15527, + "grad_norm": 0.7159334070955671, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 15527 + }, + { + "epoch": 0.15528, + "grad_norm": 0.6812173657150831, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 15528 + }, + { + "epoch": 0.15529, + "grad_norm": 0.6309782120561529, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 15529 + }, + { + "epoch": 0.1553, + "grad_norm": 0.591381848034642, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 15530 + }, + { + "epoch": 0.15531, + "grad_norm": 0.635929769803623, + "learning_rate": 0.003, + "loss": 4.0, + "step": 15531 + }, + { + "epoch": 0.15532, + "grad_norm": 0.5749094141896738, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 15532 + }, + { + "epoch": 0.15533, + "grad_norm": 0.591514849973928, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 15533 + }, + { + "epoch": 0.15534, + "grad_norm": 0.6898780954120327, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 15534 + }, + { + "epoch": 0.15535, + "grad_norm": 0.7015211705932873, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 15535 + }, + { + "epoch": 0.15536, + "grad_norm": 0.7579445711473217, + "learning_rate": 0.003, + "loss": 4.001, + "step": 15536 + }, + { + "epoch": 0.15537, + "grad_norm": 0.799438416352439, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 15537 + }, + { + "epoch": 0.15538, + "grad_norm": 0.7306019519419089, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 15538 + }, + { + "epoch": 0.15539, + "grad_norm": 0.6005455677613533, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 15539 + }, + { + "epoch": 0.1554, + "grad_norm": 0.5215439517940001, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 15540 + }, + { + "epoch": 0.15541, + "grad_norm": 0.6216855213239662, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 15541 + }, + { + "epoch": 0.15542, + "grad_norm": 0.6149979988277356, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 15542 + }, + { + "epoch": 0.15543, + "grad_norm": 0.6421483162267317, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 15543 + }, + { + "epoch": 0.15544, + "grad_norm": 0.6374570715610997, + "learning_rate": 0.003, + "loss": 4.0499, + "step": 15544 + }, + { + "epoch": 0.15545, + "grad_norm": 0.6273435752500629, + "learning_rate": 0.003, + "loss": 3.983, + "step": 15545 + }, + { + "epoch": 0.15546, + "grad_norm": 0.5911646958669942, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 15546 + }, + { + "epoch": 0.15547, + "grad_norm": 0.6007388393437674, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 15547 + }, + { + "epoch": 0.15548, + "grad_norm": 0.6815153380493059, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 15548 + }, + { + "epoch": 0.15549, + "grad_norm": 0.8029654507374164, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 15549 + }, + { + "epoch": 0.1555, + "grad_norm": 0.8958200154022117, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 15550 + }, + { + "epoch": 0.15551, + "grad_norm": 0.9900227981435769, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 15551 + }, + { + "epoch": 0.15552, + "grad_norm": 1.0635350490573978, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 15552 + }, + { + "epoch": 0.15553, + "grad_norm": 0.9683598823839643, + "learning_rate": 0.003, + "loss": 3.986, + "step": 15553 + }, + { + "epoch": 0.15554, + "grad_norm": 1.0099701253520676, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 15554 + }, + { + "epoch": 0.15555, + "grad_norm": 1.0096896487427125, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 15555 + }, + { + "epoch": 0.15556, + "grad_norm": 1.016860024076801, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 15556 + }, + { + "epoch": 0.15557, + "grad_norm": 0.9347144092566355, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 15557 + }, + { + "epoch": 0.15558, + "grad_norm": 0.9177331202748121, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 15558 + }, + { + "epoch": 0.15559, + "grad_norm": 1.0070143198263801, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 15559 + }, + { + "epoch": 0.1556, + "grad_norm": 0.9160951397727032, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 15560 + }, + { + "epoch": 0.15561, + "grad_norm": 0.8591710771341173, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 15561 + }, + { + "epoch": 0.15562, + "grad_norm": 0.8395202996308428, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 15562 + }, + { + "epoch": 0.15563, + "grad_norm": 0.7685152662475938, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 15563 + }, + { + "epoch": 0.15564, + "grad_norm": 0.72536957410754, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 15564 + }, + { + "epoch": 0.15565, + "grad_norm": 0.8021868679695753, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 15565 + }, + { + "epoch": 0.15566, + "grad_norm": 0.8490069311757922, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 15566 + }, + { + "epoch": 0.15567, + "grad_norm": 0.950489344848763, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 15567 + }, + { + "epoch": 0.15568, + "grad_norm": 0.9791285668787945, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 15568 + }, + { + "epoch": 0.15569, + "grad_norm": 0.8852036202315171, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 15569 + }, + { + "epoch": 0.1557, + "grad_norm": 0.8848909439726236, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 15570 + }, + { + "epoch": 0.15571, + "grad_norm": 0.8082349754716159, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15571 + }, + { + "epoch": 0.15572, + "grad_norm": 0.7048529567301629, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 15572 + }, + { + "epoch": 0.15573, + "grad_norm": 0.608411859186664, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 15573 + }, + { + "epoch": 0.15574, + "grad_norm": 0.5715889740422453, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 15574 + }, + { + "epoch": 0.15575, + "grad_norm": 0.6352082082363288, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 15575 + }, + { + "epoch": 0.15576, + "grad_norm": 0.7985091345450283, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 15576 + }, + { + "epoch": 0.15577, + "grad_norm": 0.8760130061051367, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 15577 + }, + { + "epoch": 0.15578, + "grad_norm": 0.7490012265012769, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 15578 + }, + { + "epoch": 0.15579, + "grad_norm": 0.6254712935745266, + "learning_rate": 0.003, + "loss": 4.008, + "step": 15579 + }, + { + "epoch": 0.1558, + "grad_norm": 0.5958567332601931, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 15580 + }, + { + "epoch": 0.15581, + "grad_norm": 0.6529370818519394, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 15581 + }, + { + "epoch": 0.15582, + "grad_norm": 0.6855083691487892, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 15582 + }, + { + "epoch": 0.15583, + "grad_norm": 0.5894124702006397, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 15583 + }, + { + "epoch": 0.15584, + "grad_norm": 0.4969076205500627, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 15584 + }, + { + "epoch": 0.15585, + "grad_norm": 0.46159180218614954, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 15585 + }, + { + "epoch": 0.15586, + "grad_norm": 0.44235137002572317, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 15586 + }, + { + "epoch": 0.15587, + "grad_norm": 0.41533428985234544, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 15587 + }, + { + "epoch": 0.15588, + "grad_norm": 0.4088917744565381, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 15588 + }, + { + "epoch": 0.15589, + "grad_norm": 0.4100632959885097, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 15589 + }, + { + "epoch": 0.1559, + "grad_norm": 0.4367071841537676, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 15590 + }, + { + "epoch": 0.15591, + "grad_norm": 0.50672853986763, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 15591 + }, + { + "epoch": 0.15592, + "grad_norm": 0.5723984946200392, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 15592 + }, + { + "epoch": 0.15593, + "grad_norm": 0.725928260722575, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 15593 + }, + { + "epoch": 0.15594, + "grad_norm": 0.9709073310807451, + "learning_rate": 0.003, + "loss": 3.998, + "step": 15594 + }, + { + "epoch": 0.15595, + "grad_norm": 1.0011701696088906, + "learning_rate": 0.003, + "loss": 3.987, + "step": 15595 + }, + { + "epoch": 0.15596, + "grad_norm": 0.7838244762693315, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 15596 + }, + { + "epoch": 0.15597, + "grad_norm": 0.7197879273432198, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 15597 + }, + { + "epoch": 0.15598, + "grad_norm": 0.7031343248207099, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 15598 + }, + { + "epoch": 0.15599, + "grad_norm": 0.6607925493299033, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 15599 + }, + { + "epoch": 0.156, + "grad_norm": 0.7215238616192661, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 15600 + }, + { + "epoch": 0.15601, + "grad_norm": 0.7600933810973404, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 15601 + }, + { + "epoch": 0.15602, + "grad_norm": 0.7924677418226641, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 15602 + }, + { + "epoch": 0.15603, + "grad_norm": 0.865985780333317, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 15603 + }, + { + "epoch": 0.15604, + "grad_norm": 0.7881436792170692, + "learning_rate": 0.003, + "loss": 4.014, + "step": 15604 + }, + { + "epoch": 0.15605, + "grad_norm": 0.8425472499220391, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 15605 + }, + { + "epoch": 0.15606, + "grad_norm": 0.7961620885023344, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 15606 + }, + { + "epoch": 0.15607, + "grad_norm": 0.8690706507704246, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 15607 + }, + { + "epoch": 0.15608, + "grad_norm": 0.9255052421963075, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 15608 + }, + { + "epoch": 0.15609, + "grad_norm": 0.9725565073393386, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 15609 + }, + { + "epoch": 0.1561, + "grad_norm": 0.9077002156258192, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 15610 + }, + { + "epoch": 0.15611, + "grad_norm": 0.881456267279888, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 15611 + }, + { + "epoch": 0.15612, + "grad_norm": 0.8231983763471588, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 15612 + }, + { + "epoch": 0.15613, + "grad_norm": 0.8155604963245934, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 15613 + }, + { + "epoch": 0.15614, + "grad_norm": 0.8069519247808924, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 15614 + }, + { + "epoch": 0.15615, + "grad_norm": 0.7993709967477285, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 15615 + }, + { + "epoch": 0.15616, + "grad_norm": 0.8090949292786541, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 15616 + }, + { + "epoch": 0.15617, + "grad_norm": 0.9666392410905393, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 15617 + }, + { + "epoch": 0.15618, + "grad_norm": 0.9937513674476818, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 15618 + }, + { + "epoch": 0.15619, + "grad_norm": 0.875233426498004, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 15619 + }, + { + "epoch": 0.1562, + "grad_norm": 0.7912488251743891, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 15620 + }, + { + "epoch": 0.15621, + "grad_norm": 0.6143764544045114, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 15621 + }, + { + "epoch": 0.15622, + "grad_norm": 0.5135939656153742, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 15622 + }, + { + "epoch": 0.15623, + "grad_norm": 0.5030373087684754, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 15623 + }, + { + "epoch": 0.15624, + "grad_norm": 0.5563217182619539, + "learning_rate": 0.003, + "loss": 4.038, + "step": 15624 + }, + { + "epoch": 0.15625, + "grad_norm": 0.582122502360409, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 15625 + }, + { + "epoch": 0.15626, + "grad_norm": 0.649067292618393, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15626 + }, + { + "epoch": 0.15627, + "grad_norm": 0.8290505835784656, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 15627 + }, + { + "epoch": 0.15628, + "grad_norm": 0.9431034496755974, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 15628 + }, + { + "epoch": 0.15629, + "grad_norm": 0.8990682470417183, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 15629 + }, + { + "epoch": 0.1563, + "grad_norm": 0.7396651791873612, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 15630 + }, + { + "epoch": 0.15631, + "grad_norm": 0.680764442717998, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 15631 + }, + { + "epoch": 0.15632, + "grad_norm": 0.640063551806679, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 15632 + }, + { + "epoch": 0.15633, + "grad_norm": 0.6060373619304447, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 15633 + }, + { + "epoch": 0.15634, + "grad_norm": 0.6053353008584665, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 15634 + }, + { + "epoch": 0.15635, + "grad_norm": 0.6330018076142412, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 15635 + }, + { + "epoch": 0.15636, + "grad_norm": 0.6182348118943397, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 15636 + }, + { + "epoch": 0.15637, + "grad_norm": 0.5701824182083595, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 15637 + }, + { + "epoch": 0.15638, + "grad_norm": 0.5925582470567741, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 15638 + }, + { + "epoch": 0.15639, + "grad_norm": 0.673840575212614, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 15639 + }, + { + "epoch": 0.1564, + "grad_norm": 0.6268865700667507, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 15640 + }, + { + "epoch": 0.15641, + "grad_norm": 0.5786207948271682, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 15641 + }, + { + "epoch": 0.15642, + "grad_norm": 0.559344998064897, + "learning_rate": 0.003, + "loss": 4.016, + "step": 15642 + }, + { + "epoch": 0.15643, + "grad_norm": 0.5254444826155936, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 15643 + }, + { + "epoch": 0.15644, + "grad_norm": 0.49801410934319224, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 15644 + }, + { + "epoch": 0.15645, + "grad_norm": 0.5937272295703302, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 15645 + }, + { + "epoch": 0.15646, + "grad_norm": 0.6486996095072027, + "learning_rate": 0.003, + "loss": 3.972, + "step": 15646 + }, + { + "epoch": 0.15647, + "grad_norm": 0.7324643141900073, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 15647 + }, + { + "epoch": 0.15648, + "grad_norm": 0.8574707311225193, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 15648 + }, + { + "epoch": 0.15649, + "grad_norm": 0.874006807805858, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 15649 + }, + { + "epoch": 0.1565, + "grad_norm": 0.7941086958268307, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 15650 + }, + { + "epoch": 0.15651, + "grad_norm": 0.8026691326367495, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 15651 + }, + { + "epoch": 0.15652, + "grad_norm": 0.7453533592920598, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 15652 + }, + { + "epoch": 0.15653, + "grad_norm": 0.7758764272708664, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 15653 + }, + { + "epoch": 0.15654, + "grad_norm": 0.853080865024767, + "learning_rate": 0.003, + "loss": 4.013, + "step": 15654 + }, + { + "epoch": 0.15655, + "grad_norm": 0.8726175628721387, + "learning_rate": 0.003, + "loss": 4.0644, + "step": 15655 + }, + { + "epoch": 0.15656, + "grad_norm": 0.7142036393764818, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 15656 + }, + { + "epoch": 0.15657, + "grad_norm": 0.806271707032899, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 15657 + }, + { + "epoch": 0.15658, + "grad_norm": 0.8676655093219628, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 15658 + }, + { + "epoch": 0.15659, + "grad_norm": 0.8636804737342687, + "learning_rate": 0.003, + "loss": 4.04, + "step": 15659 + }, + { + "epoch": 0.1566, + "grad_norm": 0.819718417948259, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 15660 + }, + { + "epoch": 0.15661, + "grad_norm": 0.7993739719826846, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 15661 + }, + { + "epoch": 0.15662, + "grad_norm": 0.774352378168353, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 15662 + }, + { + "epoch": 0.15663, + "grad_norm": 0.7395932375760476, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15663 + }, + { + "epoch": 0.15664, + "grad_norm": 0.8853351225314133, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 15664 + }, + { + "epoch": 0.15665, + "grad_norm": 0.9375212903705973, + "learning_rate": 0.003, + "loss": 4.03, + "step": 15665 + }, + { + "epoch": 0.15666, + "grad_norm": 0.906899327620966, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 15666 + }, + { + "epoch": 0.15667, + "grad_norm": 0.846919561150316, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 15667 + }, + { + "epoch": 0.15668, + "grad_norm": 0.7822289355264589, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 15668 + }, + { + "epoch": 0.15669, + "grad_norm": 0.8613726199531282, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 15669 + }, + { + "epoch": 0.1567, + "grad_norm": 0.9532589654061366, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 15670 + }, + { + "epoch": 0.15671, + "grad_norm": 0.8739601195186565, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 15671 + }, + { + "epoch": 0.15672, + "grad_norm": 0.7130663473119927, + "learning_rate": 0.003, + "loss": 4.013, + "step": 15672 + }, + { + "epoch": 0.15673, + "grad_norm": 0.715211092112699, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 15673 + }, + { + "epoch": 0.15674, + "grad_norm": 0.7739383836647019, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 15674 + }, + { + "epoch": 0.15675, + "grad_norm": 0.7605449708924737, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 15675 + }, + { + "epoch": 0.15676, + "grad_norm": 0.6320183200925991, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 15676 + }, + { + "epoch": 0.15677, + "grad_norm": 0.6272117457669908, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 15677 + }, + { + "epoch": 0.15678, + "grad_norm": 0.6472969753231188, + "learning_rate": 0.003, + "loss": 4.02, + "step": 15678 + }, + { + "epoch": 0.15679, + "grad_norm": 0.4987504681983765, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 15679 + }, + { + "epoch": 0.1568, + "grad_norm": 0.43226541019936277, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 15680 + }, + { + "epoch": 0.15681, + "grad_norm": 0.4323812897568941, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 15681 + }, + { + "epoch": 0.15682, + "grad_norm": 0.4063113854927988, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 15682 + }, + { + "epoch": 0.15683, + "grad_norm": 0.40764038851259127, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 15683 + }, + { + "epoch": 0.15684, + "grad_norm": 0.4487153243473453, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 15684 + }, + { + "epoch": 0.15685, + "grad_norm": 0.5334537015454429, + "learning_rate": 0.003, + "loss": 4.007, + "step": 15685 + }, + { + "epoch": 0.15686, + "grad_norm": 0.708197053096469, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 15686 + }, + { + "epoch": 0.15687, + "grad_norm": 0.9235732779495441, + "learning_rate": 0.003, + "loss": 4.03, + "step": 15687 + }, + { + "epoch": 0.15688, + "grad_norm": 1.1504146989607653, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 15688 + }, + { + "epoch": 0.15689, + "grad_norm": 0.8579374039750891, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 15689 + }, + { + "epoch": 0.1569, + "grad_norm": 0.767361687584915, + "learning_rate": 0.003, + "loss": 4.001, + "step": 15690 + }, + { + "epoch": 0.15691, + "grad_norm": 0.8122957622617395, + "learning_rate": 0.003, + "loss": 4.015, + "step": 15691 + }, + { + "epoch": 0.15692, + "grad_norm": 1.0284225307735397, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 15692 + }, + { + "epoch": 0.15693, + "grad_norm": 0.9494031189043398, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 15693 + }, + { + "epoch": 0.15694, + "grad_norm": 0.9344866663048261, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 15694 + }, + { + "epoch": 0.15695, + "grad_norm": 0.8912978156520955, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 15695 + }, + { + "epoch": 0.15696, + "grad_norm": 0.7857841244837427, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 15696 + }, + { + "epoch": 0.15697, + "grad_norm": 0.8848781941757603, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15697 + }, + { + "epoch": 0.15698, + "grad_norm": 0.8701638381928962, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 15698 + }, + { + "epoch": 0.15699, + "grad_norm": 0.8371969370557966, + "learning_rate": 0.003, + "loss": 4.0637, + "step": 15699 + }, + { + "epoch": 0.157, + "grad_norm": 0.8041067647140022, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 15700 + }, + { + "epoch": 0.15701, + "grad_norm": 0.6962325033159212, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 15701 + }, + { + "epoch": 0.15702, + "grad_norm": 0.7657720106751605, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 15702 + }, + { + "epoch": 0.15703, + "grad_norm": 0.741663827093284, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 15703 + }, + { + "epoch": 0.15704, + "grad_norm": 0.7804209005746866, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 15704 + }, + { + "epoch": 0.15705, + "grad_norm": 0.7437921325923658, + "learning_rate": 0.003, + "loss": 4.008, + "step": 15705 + }, + { + "epoch": 0.15706, + "grad_norm": 0.6668536213085892, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 15706 + }, + { + "epoch": 0.15707, + "grad_norm": 0.6280854722400809, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 15707 + }, + { + "epoch": 0.15708, + "grad_norm": 0.5718534274390086, + "learning_rate": 0.003, + "loss": 4.019, + "step": 15708 + }, + { + "epoch": 0.15709, + "grad_norm": 0.5541573547398432, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 15709 + }, + { + "epoch": 0.1571, + "grad_norm": 0.5574464378046001, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 15710 + }, + { + "epoch": 0.15711, + "grad_norm": 0.5917733915187963, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 15711 + }, + { + "epoch": 0.15712, + "grad_norm": 0.6010617354091231, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 15712 + }, + { + "epoch": 0.15713, + "grad_norm": 0.6172667115438685, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 15713 + }, + { + "epoch": 0.15714, + "grad_norm": 0.6683783132626372, + "learning_rate": 0.003, + "loss": 4.0, + "step": 15714 + }, + { + "epoch": 0.15715, + "grad_norm": 0.6896838237181508, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 15715 + }, + { + "epoch": 0.15716, + "grad_norm": 0.7480196862421763, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 15716 + }, + { + "epoch": 0.15717, + "grad_norm": 0.8509821418800899, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15717 + }, + { + "epoch": 0.15718, + "grad_norm": 0.7436380999724544, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 15718 + }, + { + "epoch": 0.15719, + "grad_norm": 0.6744709205713507, + "learning_rate": 0.003, + "loss": 4.013, + "step": 15719 + }, + { + "epoch": 0.1572, + "grad_norm": 0.7691532953850261, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 15720 + }, + { + "epoch": 0.15721, + "grad_norm": 0.6966736553085414, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 15721 + }, + { + "epoch": 0.15722, + "grad_norm": 0.6674618078571818, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 15722 + }, + { + "epoch": 0.15723, + "grad_norm": 0.6125386810749728, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 15723 + }, + { + "epoch": 0.15724, + "grad_norm": 0.702089847448902, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 15724 + }, + { + "epoch": 0.15725, + "grad_norm": 0.8077876139122758, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 15725 + }, + { + "epoch": 0.15726, + "grad_norm": 0.8599335000896925, + "learning_rate": 0.003, + "loss": 3.991, + "step": 15726 + }, + { + "epoch": 0.15727, + "grad_norm": 0.7962520651524224, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 15727 + }, + { + "epoch": 0.15728, + "grad_norm": 0.746975124874623, + "learning_rate": 0.003, + "loss": 4.021, + "step": 15728 + }, + { + "epoch": 0.15729, + "grad_norm": 0.7686145693255922, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 15729 + }, + { + "epoch": 0.1573, + "grad_norm": 0.7945959288992333, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 15730 + }, + { + "epoch": 0.15731, + "grad_norm": 0.8029578638305038, + "learning_rate": 0.003, + "loss": 3.995, + "step": 15731 + }, + { + "epoch": 0.15732, + "grad_norm": 0.7721015014199698, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 15732 + }, + { + "epoch": 0.15733, + "grad_norm": 0.7808365786898782, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 15733 + }, + { + "epoch": 0.15734, + "grad_norm": 0.7316800750696419, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 15734 + }, + { + "epoch": 0.15735, + "grad_norm": 0.6572533760773999, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 15735 + }, + { + "epoch": 0.15736, + "grad_norm": 0.7813809603970673, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 15736 + }, + { + "epoch": 0.15737, + "grad_norm": 0.874900977894494, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 15737 + }, + { + "epoch": 0.15738, + "grad_norm": 0.9063258262451599, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 15738 + }, + { + "epoch": 0.15739, + "grad_norm": 0.7503375450907923, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 15739 + }, + { + "epoch": 0.1574, + "grad_norm": 0.654127711583441, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 15740 + }, + { + "epoch": 0.15741, + "grad_norm": 0.5810132095232045, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 15741 + }, + { + "epoch": 0.15742, + "grad_norm": 0.5809750026281869, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 15742 + }, + { + "epoch": 0.15743, + "grad_norm": 0.6153875209449347, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 15743 + }, + { + "epoch": 0.15744, + "grad_norm": 0.5773012206048198, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 15744 + }, + { + "epoch": 0.15745, + "grad_norm": 0.5579097731773733, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 15745 + }, + { + "epoch": 0.15746, + "grad_norm": 0.6058041590856887, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 15746 + }, + { + "epoch": 0.15747, + "grad_norm": 0.7141918010116859, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 15747 + }, + { + "epoch": 0.15748, + "grad_norm": 0.7768591853106724, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 15748 + }, + { + "epoch": 0.15749, + "grad_norm": 0.9275107075978863, + "learning_rate": 0.003, + "loss": 4.008, + "step": 15749 + }, + { + "epoch": 0.1575, + "grad_norm": 0.9438427659528532, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 15750 + }, + { + "epoch": 0.15751, + "grad_norm": 0.826055822634805, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 15751 + }, + { + "epoch": 0.15752, + "grad_norm": 0.7910304998050793, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 15752 + }, + { + "epoch": 0.15753, + "grad_norm": 0.7226884585117523, + "learning_rate": 0.003, + "loss": 4.029, + "step": 15753 + }, + { + "epoch": 0.15754, + "grad_norm": 0.7329789101050534, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15754 + }, + { + "epoch": 0.15755, + "grad_norm": 0.6692239082631861, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 15755 + }, + { + "epoch": 0.15756, + "grad_norm": 0.5789302622151473, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 15756 + }, + { + "epoch": 0.15757, + "grad_norm": 0.5625001519389233, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 15757 + }, + { + "epoch": 0.15758, + "grad_norm": 0.5116994804188493, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 15758 + }, + { + "epoch": 0.15759, + "grad_norm": 0.48783400366439555, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 15759 + }, + { + "epoch": 0.1576, + "grad_norm": 0.45377109584336045, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 15760 + }, + { + "epoch": 0.15761, + "grad_norm": 0.5049755781592149, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 15761 + }, + { + "epoch": 0.15762, + "grad_norm": 0.5402888979868046, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 15762 + }, + { + "epoch": 0.15763, + "grad_norm": 0.6606794165750385, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 15763 + }, + { + "epoch": 0.15764, + "grad_norm": 0.8866583189361682, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 15764 + }, + { + "epoch": 0.15765, + "grad_norm": 1.074754394249539, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 15765 + }, + { + "epoch": 0.15766, + "grad_norm": 0.917468210327863, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 15766 + }, + { + "epoch": 0.15767, + "grad_norm": 0.9244104422185474, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 15767 + }, + { + "epoch": 0.15768, + "grad_norm": 1.0424453105335632, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 15768 + }, + { + "epoch": 0.15769, + "grad_norm": 0.9789309127716987, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 15769 + }, + { + "epoch": 0.1577, + "grad_norm": 0.8588737797474159, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 15770 + }, + { + "epoch": 0.15771, + "grad_norm": 0.839897094454728, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 15771 + }, + { + "epoch": 0.15772, + "grad_norm": 0.7359451800928299, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 15772 + }, + { + "epoch": 0.15773, + "grad_norm": 0.760370215168736, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 15773 + }, + { + "epoch": 0.15774, + "grad_norm": 1.0921065942546064, + "learning_rate": 0.003, + "loss": 4.03, + "step": 15774 + }, + { + "epoch": 0.15775, + "grad_norm": 1.0858665661349218, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 15775 + }, + { + "epoch": 0.15776, + "grad_norm": 0.805110722567032, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 15776 + }, + { + "epoch": 0.15777, + "grad_norm": 0.6080593501809872, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 15777 + }, + { + "epoch": 0.15778, + "grad_norm": 0.6002295673503456, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 15778 + }, + { + "epoch": 0.15779, + "grad_norm": 0.6323196033634219, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15779 + }, + { + "epoch": 0.1578, + "grad_norm": 0.6663188977507563, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 15780 + }, + { + "epoch": 0.15781, + "grad_norm": 0.677146280894016, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 15781 + }, + { + "epoch": 0.15782, + "grad_norm": 0.5831426543661401, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 15782 + }, + { + "epoch": 0.15783, + "grad_norm": 0.5363488606227224, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 15783 + }, + { + "epoch": 0.15784, + "grad_norm": 0.640857795001523, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 15784 + }, + { + "epoch": 0.15785, + "grad_norm": 0.7140908924248571, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 15785 + }, + { + "epoch": 0.15786, + "grad_norm": 0.7596665504403666, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 15786 + }, + { + "epoch": 0.15787, + "grad_norm": 0.8436684764304795, + "learning_rate": 0.003, + "loss": 4.027, + "step": 15787 + }, + { + "epoch": 0.15788, + "grad_norm": 0.8104213818036485, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 15788 + }, + { + "epoch": 0.15789, + "grad_norm": 0.714297174246003, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 15789 + }, + { + "epoch": 0.1579, + "grad_norm": 0.674515283598818, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 15790 + }, + { + "epoch": 0.15791, + "grad_norm": 0.6513778632425948, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 15791 + }, + { + "epoch": 0.15792, + "grad_norm": 0.6194197255041204, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 15792 + }, + { + "epoch": 0.15793, + "grad_norm": 0.6591871483696258, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 15793 + }, + { + "epoch": 0.15794, + "grad_norm": 0.7464843619662981, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 15794 + }, + { + "epoch": 0.15795, + "grad_norm": 0.8408995226515188, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 15795 + }, + { + "epoch": 0.15796, + "grad_norm": 0.8227748982081021, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 15796 + }, + { + "epoch": 0.15797, + "grad_norm": 0.7120545039432742, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 15797 + }, + { + "epoch": 0.15798, + "grad_norm": 0.7824143468180024, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 15798 + }, + { + "epoch": 0.15799, + "grad_norm": 0.7487145406671606, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 15799 + }, + { + "epoch": 0.158, + "grad_norm": 0.7145962453853103, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 15800 + }, + { + "epoch": 0.15801, + "grad_norm": 0.7836012571057879, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 15801 + }, + { + "epoch": 0.15802, + "grad_norm": 0.7179820708222092, + "learning_rate": 0.003, + "loss": 4.012, + "step": 15802 + }, + { + "epoch": 0.15803, + "grad_norm": 0.6739962125999319, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 15803 + }, + { + "epoch": 0.15804, + "grad_norm": 0.6019475628565186, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 15804 + }, + { + "epoch": 0.15805, + "grad_norm": 0.6269637013533936, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 15805 + }, + { + "epoch": 0.15806, + "grad_norm": 0.7004576412501466, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 15806 + }, + { + "epoch": 0.15807, + "grad_norm": 0.6839894283807149, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 15807 + }, + { + "epoch": 0.15808, + "grad_norm": 0.661040920828652, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 15808 + }, + { + "epoch": 0.15809, + "grad_norm": 0.7174004882181979, + "learning_rate": 0.003, + "loss": 4.009, + "step": 15809 + }, + { + "epoch": 0.1581, + "grad_norm": 0.7354364699135215, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 15810 + }, + { + "epoch": 0.15811, + "grad_norm": 0.7950537819138425, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 15811 + }, + { + "epoch": 0.15812, + "grad_norm": 1.0065594305297532, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 15812 + }, + { + "epoch": 0.15813, + "grad_norm": 1.0045527594485757, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 15813 + }, + { + "epoch": 0.15814, + "grad_norm": 0.9838357132040904, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 15814 + }, + { + "epoch": 0.15815, + "grad_norm": 1.0253942444071018, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 15815 + }, + { + "epoch": 0.15816, + "grad_norm": 1.0627044835569048, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 15816 + }, + { + "epoch": 0.15817, + "grad_norm": 0.9923658270204568, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 15817 + }, + { + "epoch": 0.15818, + "grad_norm": 0.9190045107595943, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 15818 + }, + { + "epoch": 0.15819, + "grad_norm": 0.8860236045467113, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 15819 + }, + { + "epoch": 0.1582, + "grad_norm": 0.7696683752568235, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 15820 + }, + { + "epoch": 0.15821, + "grad_norm": 0.7359999673158344, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 15821 + }, + { + "epoch": 0.15822, + "grad_norm": 0.7564984970266696, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 15822 + }, + { + "epoch": 0.15823, + "grad_norm": 0.7114143619452966, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 15823 + }, + { + "epoch": 0.15824, + "grad_norm": 0.6475973385841104, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 15824 + }, + { + "epoch": 0.15825, + "grad_norm": 0.5597767523007879, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 15825 + }, + { + "epoch": 0.15826, + "grad_norm": 0.62573565185223, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15826 + }, + { + "epoch": 0.15827, + "grad_norm": 0.6521554160673176, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 15827 + }, + { + "epoch": 0.15828, + "grad_norm": 0.5547679771082523, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 15828 + }, + { + "epoch": 0.15829, + "grad_norm": 0.49559783966173276, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 15829 + }, + { + "epoch": 0.1583, + "grad_norm": 0.4876124968651738, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 15830 + }, + { + "epoch": 0.15831, + "grad_norm": 0.510795123347659, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 15831 + }, + { + "epoch": 0.15832, + "grad_norm": 0.5479296095157911, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 15832 + }, + { + "epoch": 0.15833, + "grad_norm": 0.5828875133637585, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 15833 + }, + { + "epoch": 0.15834, + "grad_norm": 0.5716278033176444, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 15834 + }, + { + "epoch": 0.15835, + "grad_norm": 0.5753877038516331, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 15835 + }, + { + "epoch": 0.15836, + "grad_norm": 0.5927606551705119, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 15836 + }, + { + "epoch": 0.15837, + "grad_norm": 0.6406940457242933, + "learning_rate": 0.003, + "loss": 3.999, + "step": 15837 + }, + { + "epoch": 0.15838, + "grad_norm": 0.7514693307936401, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 15838 + }, + { + "epoch": 0.15839, + "grad_norm": 0.8750621023457512, + "learning_rate": 0.003, + "loss": 3.996, + "step": 15839 + }, + { + "epoch": 0.1584, + "grad_norm": 0.8834083559055544, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 15840 + }, + { + "epoch": 0.15841, + "grad_norm": 0.837482675269418, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 15841 + }, + { + "epoch": 0.15842, + "grad_norm": 0.8962853296952716, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 15842 + }, + { + "epoch": 0.15843, + "grad_norm": 0.8826870423186285, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 15843 + }, + { + "epoch": 0.15844, + "grad_norm": 0.902240569003256, + "learning_rate": 0.003, + "loss": 4.032, + "step": 15844 + }, + { + "epoch": 0.15845, + "grad_norm": 0.853859476316182, + "learning_rate": 0.003, + "loss": 4.0501, + "step": 15845 + }, + { + "epoch": 0.15846, + "grad_norm": 1.0069976213217864, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 15846 + }, + { + "epoch": 0.15847, + "grad_norm": 1.1399109390749675, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 15847 + }, + { + "epoch": 0.15848, + "grad_norm": 0.9559100060618182, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 15848 + }, + { + "epoch": 0.15849, + "grad_norm": 0.9643022941896654, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 15849 + }, + { + "epoch": 0.1585, + "grad_norm": 0.9146487914385231, + "learning_rate": 0.003, + "loss": 4.0604, + "step": 15850 + }, + { + "epoch": 0.15851, + "grad_norm": 0.900950221523909, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 15851 + }, + { + "epoch": 0.15852, + "grad_norm": 0.7626787872881088, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 15852 + }, + { + "epoch": 0.15853, + "grad_norm": 0.7538738062036032, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 15853 + }, + { + "epoch": 0.15854, + "grad_norm": 0.7455920728884269, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 15854 + }, + { + "epoch": 0.15855, + "grad_norm": 0.9163529642712943, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 15855 + }, + { + "epoch": 0.15856, + "grad_norm": 1.036401362086216, + "learning_rate": 0.003, + "loss": 4.057, + "step": 15856 + }, + { + "epoch": 0.15857, + "grad_norm": 0.8043634394537015, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 15857 + }, + { + "epoch": 0.15858, + "grad_norm": 0.6952689988903984, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 15858 + }, + { + "epoch": 0.15859, + "grad_norm": 0.6980672751033222, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 15859 + }, + { + "epoch": 0.1586, + "grad_norm": 0.7574531090029488, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 15860 + }, + { + "epoch": 0.15861, + "grad_norm": 0.7421652327431558, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 15861 + }, + { + "epoch": 0.15862, + "grad_norm": 0.6872117969760646, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 15862 + }, + { + "epoch": 0.15863, + "grad_norm": 0.6011277858232327, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 15863 + }, + { + "epoch": 0.15864, + "grad_norm": 0.48782767351254686, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 15864 + }, + { + "epoch": 0.15865, + "grad_norm": 0.4949472035319615, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 15865 + }, + { + "epoch": 0.15866, + "grad_norm": 0.5397842993627814, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 15866 + }, + { + "epoch": 0.15867, + "grad_norm": 0.60605653468352, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 15867 + }, + { + "epoch": 0.15868, + "grad_norm": 0.6175539402787972, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 15868 + }, + { + "epoch": 0.15869, + "grad_norm": 0.5851672770417058, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 15869 + }, + { + "epoch": 0.1587, + "grad_norm": 0.5486684413215428, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 15870 + }, + { + "epoch": 0.15871, + "grad_norm": 0.598179471335982, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 15871 + }, + { + "epoch": 0.15872, + "grad_norm": 0.6565995199004536, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 15872 + }, + { + "epoch": 0.15873, + "grad_norm": 0.7238126670305417, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 15873 + }, + { + "epoch": 0.15874, + "grad_norm": 0.8666198578575829, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 15874 + }, + { + "epoch": 0.15875, + "grad_norm": 0.9357088867870735, + "learning_rate": 0.003, + "loss": 3.992, + "step": 15875 + }, + { + "epoch": 0.15876, + "grad_norm": 0.8303480379812397, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 15876 + }, + { + "epoch": 0.15877, + "grad_norm": 0.8018933442891364, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 15877 + }, + { + "epoch": 0.15878, + "grad_norm": 0.7564559445328369, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 15878 + }, + { + "epoch": 0.15879, + "grad_norm": 0.7353891714735834, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 15879 + }, + { + "epoch": 0.1588, + "grad_norm": 0.6912163282088741, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 15880 + }, + { + "epoch": 0.15881, + "grad_norm": 0.6819822915057494, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 15881 + }, + { + "epoch": 0.15882, + "grad_norm": 0.6718010503658044, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 15882 + }, + { + "epoch": 0.15883, + "grad_norm": 0.6185092573008172, + "learning_rate": 0.003, + "loss": 3.987, + "step": 15883 + }, + { + "epoch": 0.15884, + "grad_norm": 0.5776010864440924, + "learning_rate": 0.003, + "loss": 4.015, + "step": 15884 + }, + { + "epoch": 0.15885, + "grad_norm": 0.488649197151387, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 15885 + }, + { + "epoch": 0.15886, + "grad_norm": 0.4709595399500116, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 15886 + }, + { + "epoch": 0.15887, + "grad_norm": 0.5086535431853145, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 15887 + }, + { + "epoch": 0.15888, + "grad_norm": 0.5347554305988607, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 15888 + }, + { + "epoch": 0.15889, + "grad_norm": 0.5970482038450456, + "learning_rate": 0.003, + "loss": 4.029, + "step": 15889 + }, + { + "epoch": 0.1589, + "grad_norm": 0.748932410726409, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 15890 + }, + { + "epoch": 0.15891, + "grad_norm": 0.9061514676560063, + "learning_rate": 0.003, + "loss": 4.042, + "step": 15891 + }, + { + "epoch": 0.15892, + "grad_norm": 0.913470496069581, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 15892 + }, + { + "epoch": 0.15893, + "grad_norm": 0.8422476727752833, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 15893 + }, + { + "epoch": 0.15894, + "grad_norm": 0.8168190625070573, + "learning_rate": 0.003, + "loss": 4.002, + "step": 15894 + }, + { + "epoch": 0.15895, + "grad_norm": 0.7317314067172604, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 15895 + }, + { + "epoch": 0.15896, + "grad_norm": 0.7691133452491093, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 15896 + }, + { + "epoch": 0.15897, + "grad_norm": 0.6929738107225618, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 15897 + }, + { + "epoch": 0.15898, + "grad_norm": 0.6536972833997907, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 15898 + }, + { + "epoch": 0.15899, + "grad_norm": 0.7236252507394011, + "learning_rate": 0.003, + "loss": 3.978, + "step": 15899 + }, + { + "epoch": 0.159, + "grad_norm": 0.7240268234470876, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 15900 + }, + { + "epoch": 0.15901, + "grad_norm": 0.7678228942116273, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15901 + }, + { + "epoch": 0.15902, + "grad_norm": 0.9177516827621967, + "learning_rate": 0.003, + "loss": 4.0589, + "step": 15902 + }, + { + "epoch": 0.15903, + "grad_norm": 0.899546118950815, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 15903 + }, + { + "epoch": 0.15904, + "grad_norm": 0.8250907128282055, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 15904 + }, + { + "epoch": 0.15905, + "grad_norm": 0.8363854226384185, + "learning_rate": 0.003, + "loss": 4.015, + "step": 15905 + }, + { + "epoch": 0.15906, + "grad_norm": 0.9352723520734928, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 15906 + }, + { + "epoch": 0.15907, + "grad_norm": 1.0286741580585013, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 15907 + }, + { + "epoch": 0.15908, + "grad_norm": 1.1260565820069743, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 15908 + }, + { + "epoch": 0.15909, + "grad_norm": 0.949159359192028, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 15909 + }, + { + "epoch": 0.1591, + "grad_norm": 0.8973304742399385, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 15910 + }, + { + "epoch": 0.15911, + "grad_norm": 0.7972489505420549, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 15911 + }, + { + "epoch": 0.15912, + "grad_norm": 0.7652758854087552, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 15912 + }, + { + "epoch": 0.15913, + "grad_norm": 0.8100989927144087, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 15913 + }, + { + "epoch": 0.15914, + "grad_norm": 0.8535930323451651, + "learning_rate": 0.003, + "loss": 4.0524, + "step": 15914 + }, + { + "epoch": 0.15915, + "grad_norm": 0.8130914350997571, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 15915 + }, + { + "epoch": 0.15916, + "grad_norm": 0.7487228469121741, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 15916 + }, + { + "epoch": 0.15917, + "grad_norm": 0.6731792542473658, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 15917 + }, + { + "epoch": 0.15918, + "grad_norm": 0.6753075199628372, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 15918 + }, + { + "epoch": 0.15919, + "grad_norm": 0.6314567703146373, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 15919 + }, + { + "epoch": 0.1592, + "grad_norm": 0.6317311831408045, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 15920 + }, + { + "epoch": 0.15921, + "grad_norm": 0.6532419731707999, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 15921 + }, + { + "epoch": 0.15922, + "grad_norm": 0.6758435668922811, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 15922 + }, + { + "epoch": 0.15923, + "grad_norm": 0.6899584778881791, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 15923 + }, + { + "epoch": 0.15924, + "grad_norm": 0.6478239680296207, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 15924 + }, + { + "epoch": 0.15925, + "grad_norm": 0.6627966462467095, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 15925 + }, + { + "epoch": 0.15926, + "grad_norm": 0.5839030360767814, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 15926 + }, + { + "epoch": 0.15927, + "grad_norm": 0.4700623752575492, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 15927 + }, + { + "epoch": 0.15928, + "grad_norm": 0.43851432942319574, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 15928 + }, + { + "epoch": 0.15929, + "grad_norm": 0.449293734137095, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 15929 + }, + { + "epoch": 0.1593, + "grad_norm": 0.5111176175774331, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 15930 + }, + { + "epoch": 0.15931, + "grad_norm": 0.5253348903135279, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 15931 + }, + { + "epoch": 0.15932, + "grad_norm": 0.6365599050391122, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 15932 + }, + { + "epoch": 0.15933, + "grad_norm": 0.7472219409090677, + "learning_rate": 0.003, + "loss": 3.988, + "step": 15933 + }, + { + "epoch": 0.15934, + "grad_norm": 0.842653594597801, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 15934 + }, + { + "epoch": 0.15935, + "grad_norm": 0.7663140987960378, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 15935 + }, + { + "epoch": 0.15936, + "grad_norm": 0.6528471285137005, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 15936 + }, + { + "epoch": 0.15937, + "grad_norm": 0.5877960166108128, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 15937 + }, + { + "epoch": 0.15938, + "grad_norm": 0.62909772924451, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 15938 + }, + { + "epoch": 0.15939, + "grad_norm": 0.6551033547541886, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 15939 + }, + { + "epoch": 0.1594, + "grad_norm": 0.5955218317578371, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 15940 + }, + { + "epoch": 0.15941, + "grad_norm": 0.643086339364613, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 15941 + }, + { + "epoch": 0.15942, + "grad_norm": 0.6258572967673671, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 15942 + }, + { + "epoch": 0.15943, + "grad_norm": 0.6757678129844892, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 15943 + }, + { + "epoch": 0.15944, + "grad_norm": 0.6648753759952436, + "learning_rate": 0.003, + "loss": 4.019, + "step": 15944 + }, + { + "epoch": 0.15945, + "grad_norm": 0.6819579264640631, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 15945 + }, + { + "epoch": 0.15946, + "grad_norm": 0.8130824703480173, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 15946 + }, + { + "epoch": 0.15947, + "grad_norm": 0.9489880001596879, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 15947 + }, + { + "epoch": 0.15948, + "grad_norm": 0.9328696857096722, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 15948 + }, + { + "epoch": 0.15949, + "grad_norm": 1.0404847052840045, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 15949 + }, + { + "epoch": 0.1595, + "grad_norm": 0.9419816679161276, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 15950 + }, + { + "epoch": 0.15951, + "grad_norm": 1.043122904345796, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 15951 + }, + { + "epoch": 0.15952, + "grad_norm": 1.1169464249855507, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 15952 + }, + { + "epoch": 0.15953, + "grad_norm": 0.8953205428836682, + "learning_rate": 0.003, + "loss": 4.0583, + "step": 15953 + }, + { + "epoch": 0.15954, + "grad_norm": 0.9722476588912641, + "learning_rate": 0.003, + "loss": 4.0815, + "step": 15954 + }, + { + "epoch": 0.15955, + "grad_norm": 1.0355585432719587, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 15955 + }, + { + "epoch": 0.15956, + "grad_norm": 1.1685443953699555, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 15956 + }, + { + "epoch": 0.15957, + "grad_norm": 1.1736696502200186, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 15957 + }, + { + "epoch": 0.15958, + "grad_norm": 0.8952538671974775, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 15958 + }, + { + "epoch": 0.15959, + "grad_norm": 0.7136720058383998, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 15959 + }, + { + "epoch": 0.1596, + "grad_norm": 0.8839988256286213, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 15960 + }, + { + "epoch": 0.15961, + "grad_norm": 1.0582564526043372, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 15961 + }, + { + "epoch": 0.15962, + "grad_norm": 1.0367386093439788, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 15962 + }, + { + "epoch": 0.15963, + "grad_norm": 0.9547289027844985, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 15963 + }, + { + "epoch": 0.15964, + "grad_norm": 1.1649316405780181, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 15964 + }, + { + "epoch": 0.15965, + "grad_norm": 0.8448196066069004, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 15965 + }, + { + "epoch": 0.15966, + "grad_norm": 0.769603933601852, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 15966 + }, + { + "epoch": 0.15967, + "grad_norm": 0.7818567546883032, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 15967 + }, + { + "epoch": 0.15968, + "grad_norm": 0.6294502238926833, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 15968 + }, + { + "epoch": 0.15969, + "grad_norm": 0.6129016677609104, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 15969 + }, + { + "epoch": 0.1597, + "grad_norm": 0.5956236389039562, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 15970 + }, + { + "epoch": 0.15971, + "grad_norm": 0.7407286066730498, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 15971 + }, + { + "epoch": 0.15972, + "grad_norm": 0.9021365056935701, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 15972 + }, + { + "epoch": 0.15973, + "grad_norm": 0.9653683965582521, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 15973 + }, + { + "epoch": 0.15974, + "grad_norm": 0.8027573086370479, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 15974 + }, + { + "epoch": 0.15975, + "grad_norm": 0.6091043224841364, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 15975 + }, + { + "epoch": 0.15976, + "grad_norm": 0.5472784805662756, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 15976 + }, + { + "epoch": 0.15977, + "grad_norm": 0.5521116912686626, + "learning_rate": 0.003, + "loss": 4.023, + "step": 15977 + }, + { + "epoch": 0.15978, + "grad_norm": 0.5348489183882431, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 15978 + }, + { + "epoch": 0.15979, + "grad_norm": 0.4956933930986142, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 15979 + }, + { + "epoch": 0.1598, + "grad_norm": 0.5686057576489544, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 15980 + }, + { + "epoch": 0.15981, + "grad_norm": 0.5640262618357024, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 15981 + }, + { + "epoch": 0.15982, + "grad_norm": 0.5675738834177462, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 15982 + }, + { + "epoch": 0.15983, + "grad_norm": 0.6692628155970604, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 15983 + }, + { + "epoch": 0.15984, + "grad_norm": 0.7697688310257875, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 15984 + }, + { + "epoch": 0.15985, + "grad_norm": 0.8268716348975755, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 15985 + }, + { + "epoch": 0.15986, + "grad_norm": 0.8080041327465158, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 15986 + }, + { + "epoch": 0.15987, + "grad_norm": 0.7147998844974516, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 15987 + }, + { + "epoch": 0.15988, + "grad_norm": 0.5527585869339717, + "learning_rate": 0.003, + "loss": 4.022, + "step": 15988 + }, + { + "epoch": 0.15989, + "grad_norm": 0.515865312651708, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 15989 + }, + { + "epoch": 0.1599, + "grad_norm": 0.50269264077406, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 15990 + }, + { + "epoch": 0.15991, + "grad_norm": 0.5613111553193959, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 15991 + }, + { + "epoch": 0.15992, + "grad_norm": 0.5544158988948062, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 15992 + }, + { + "epoch": 0.15993, + "grad_norm": 0.5623165486775694, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 15993 + }, + { + "epoch": 0.15994, + "grad_norm": 0.5653789857500269, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 15994 + }, + { + "epoch": 0.15995, + "grad_norm": 0.6530783484433561, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 15995 + }, + { + "epoch": 0.15996, + "grad_norm": 0.7760665631109803, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 15996 + }, + { + "epoch": 0.15997, + "grad_norm": 0.8656794452428339, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 15997 + }, + { + "epoch": 0.15998, + "grad_norm": 0.8613095275551708, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 15998 + }, + { + "epoch": 0.15999, + "grad_norm": 0.841542518821868, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 15999 + }, + { + "epoch": 0.16, + "grad_norm": 0.7328913522486569, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 16000 + }, + { + "epoch": 0.16001, + "grad_norm": 0.7203247539335782, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 16001 + }, + { + "epoch": 0.16002, + "grad_norm": 0.874038325626642, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 16002 + }, + { + "epoch": 0.16003, + "grad_norm": 0.8209922558718113, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 16003 + }, + { + "epoch": 0.16004, + "grad_norm": 0.7704090220518899, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 16004 + }, + { + "epoch": 0.16005, + "grad_norm": 0.7811643164086096, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 16005 + }, + { + "epoch": 0.16006, + "grad_norm": 0.7288154312654078, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 16006 + }, + { + "epoch": 0.16007, + "grad_norm": 0.7393611927967718, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 16007 + }, + { + "epoch": 0.16008, + "grad_norm": 0.7399363685218638, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 16008 + }, + { + "epoch": 0.16009, + "grad_norm": 0.6825377944697435, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 16009 + }, + { + "epoch": 0.1601, + "grad_norm": 0.6637496935287176, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 16010 + }, + { + "epoch": 0.16011, + "grad_norm": 0.6635873798034125, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 16011 + }, + { + "epoch": 0.16012, + "grad_norm": 0.7136995753756666, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 16012 + }, + { + "epoch": 0.16013, + "grad_norm": 0.680236926187956, + "learning_rate": 0.003, + "loss": 4.044, + "step": 16013 + }, + { + "epoch": 0.16014, + "grad_norm": 0.6744582074246789, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 16014 + }, + { + "epoch": 0.16015, + "grad_norm": 0.7300955784950756, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 16015 + }, + { + "epoch": 0.16016, + "grad_norm": 0.8070908074082678, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 16016 + }, + { + "epoch": 0.16017, + "grad_norm": 0.8550827034765707, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16017 + }, + { + "epoch": 0.16018, + "grad_norm": 0.7375161648760126, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 16018 + }, + { + "epoch": 0.16019, + "grad_norm": 0.8508755660569645, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 16019 + }, + { + "epoch": 0.1602, + "grad_norm": 0.9384797300177965, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 16020 + }, + { + "epoch": 0.16021, + "grad_norm": 0.8705488160629888, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 16021 + }, + { + "epoch": 0.16022, + "grad_norm": 0.8429107779414039, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 16022 + }, + { + "epoch": 0.16023, + "grad_norm": 0.8073276859701234, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16023 + }, + { + "epoch": 0.16024, + "grad_norm": 0.8502679264883467, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 16024 + }, + { + "epoch": 0.16025, + "grad_norm": 0.786510230790806, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 16025 + }, + { + "epoch": 0.16026, + "grad_norm": 0.8144963692905746, + "learning_rate": 0.003, + "loss": 3.994, + "step": 16026 + }, + { + "epoch": 0.16027, + "grad_norm": 0.7682359510250966, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 16027 + }, + { + "epoch": 0.16028, + "grad_norm": 0.6587177880714806, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 16028 + }, + { + "epoch": 0.16029, + "grad_norm": 0.6866032214420894, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 16029 + }, + { + "epoch": 0.1603, + "grad_norm": 0.63102793574837, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 16030 + }, + { + "epoch": 0.16031, + "grad_norm": 0.642548515827938, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 16031 + }, + { + "epoch": 0.16032, + "grad_norm": 0.6945429230773069, + "learning_rate": 0.003, + "loss": 4.025, + "step": 16032 + }, + { + "epoch": 0.16033, + "grad_norm": 0.7802967831059502, + "learning_rate": 0.003, + "loss": 4.013, + "step": 16033 + }, + { + "epoch": 0.16034, + "grad_norm": 0.7938313258046359, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 16034 + }, + { + "epoch": 0.16035, + "grad_norm": 0.7206171101361393, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 16035 + }, + { + "epoch": 0.16036, + "grad_norm": 0.972674729662394, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 16036 + }, + { + "epoch": 0.16037, + "grad_norm": 1.2910587220676204, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 16037 + }, + { + "epoch": 0.16038, + "grad_norm": 0.5618389020666137, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 16038 + }, + { + "epoch": 0.16039, + "grad_norm": 0.6460745102104615, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 16039 + }, + { + "epoch": 0.1604, + "grad_norm": 0.754267678457873, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16040 + }, + { + "epoch": 0.16041, + "grad_norm": 0.7380205884277299, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 16041 + }, + { + "epoch": 0.16042, + "grad_norm": 0.7776856148609154, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 16042 + }, + { + "epoch": 0.16043, + "grad_norm": 0.8016311871404913, + "learning_rate": 0.003, + "loss": 4.009, + "step": 16043 + }, + { + "epoch": 0.16044, + "grad_norm": 0.836067534147884, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 16044 + }, + { + "epoch": 0.16045, + "grad_norm": 0.9499535691667454, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 16045 + }, + { + "epoch": 0.16046, + "grad_norm": 0.9188899626444613, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 16046 + }, + { + "epoch": 0.16047, + "grad_norm": 0.8043384485311449, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 16047 + }, + { + "epoch": 0.16048, + "grad_norm": 0.8162358964412746, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 16048 + }, + { + "epoch": 0.16049, + "grad_norm": 0.7372277399830667, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 16049 + }, + { + "epoch": 0.1605, + "grad_norm": 0.6969497688230643, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16050 + }, + { + "epoch": 0.16051, + "grad_norm": 0.5898244581161662, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 16051 + }, + { + "epoch": 0.16052, + "grad_norm": 0.568450975935238, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 16052 + }, + { + "epoch": 0.16053, + "grad_norm": 0.48606799379296795, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 16053 + }, + { + "epoch": 0.16054, + "grad_norm": 0.47440333307146715, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 16054 + }, + { + "epoch": 0.16055, + "grad_norm": 0.5071485027197873, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 16055 + }, + { + "epoch": 0.16056, + "grad_norm": 0.5234393954193303, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 16056 + }, + { + "epoch": 0.16057, + "grad_norm": 0.5527499734762888, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16057 + }, + { + "epoch": 0.16058, + "grad_norm": 0.6223880745611952, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 16058 + }, + { + "epoch": 0.16059, + "grad_norm": 0.7387028616913739, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 16059 + }, + { + "epoch": 0.1606, + "grad_norm": 0.7035142758880868, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 16060 + }, + { + "epoch": 0.16061, + "grad_norm": 0.5867275337488125, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 16061 + }, + { + "epoch": 0.16062, + "grad_norm": 0.5914888215594999, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 16062 + }, + { + "epoch": 0.16063, + "grad_norm": 0.6314955785751368, + "learning_rate": 0.003, + "loss": 4.014, + "step": 16063 + }, + { + "epoch": 0.16064, + "grad_norm": 0.7259243629957435, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 16064 + }, + { + "epoch": 0.16065, + "grad_norm": 0.8207181856580842, + "learning_rate": 0.003, + "loss": 4.009, + "step": 16065 + }, + { + "epoch": 0.16066, + "grad_norm": 0.7926497502657461, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 16066 + }, + { + "epoch": 0.16067, + "grad_norm": 0.6870090156957772, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 16067 + }, + { + "epoch": 0.16068, + "grad_norm": 0.7905201874031024, + "learning_rate": 0.003, + "loss": 3.995, + "step": 16068 + }, + { + "epoch": 0.16069, + "grad_norm": 0.7498910016871344, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 16069 + }, + { + "epoch": 0.1607, + "grad_norm": 0.6507953992813502, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 16070 + }, + { + "epoch": 0.16071, + "grad_norm": 0.6363587236802856, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 16071 + }, + { + "epoch": 0.16072, + "grad_norm": 0.6808746507037334, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 16072 + }, + { + "epoch": 0.16073, + "grad_norm": 0.7070629892183206, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 16073 + }, + { + "epoch": 0.16074, + "grad_norm": 0.8187989926341253, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16074 + }, + { + "epoch": 0.16075, + "grad_norm": 0.8247181493453092, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 16075 + }, + { + "epoch": 0.16076, + "grad_norm": 0.7785424311164701, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 16076 + }, + { + "epoch": 0.16077, + "grad_norm": 0.802653619060269, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 16077 + }, + { + "epoch": 0.16078, + "grad_norm": 0.9354265866944814, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 16078 + }, + { + "epoch": 0.16079, + "grad_norm": 1.1180016884315818, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 16079 + }, + { + "epoch": 0.1608, + "grad_norm": 1.0264755079664163, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 16080 + }, + { + "epoch": 0.16081, + "grad_norm": 0.8990063287169537, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 16081 + }, + { + "epoch": 0.16082, + "grad_norm": 0.8285899174628206, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 16082 + }, + { + "epoch": 0.16083, + "grad_norm": 0.8128543448127217, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 16083 + }, + { + "epoch": 0.16084, + "grad_norm": 0.7354044849259688, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 16084 + }, + { + "epoch": 0.16085, + "grad_norm": 0.7480544273771731, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 16085 + }, + { + "epoch": 0.16086, + "grad_norm": 0.7560874676273818, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 16086 + }, + { + "epoch": 0.16087, + "grad_norm": 0.8155819217335142, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 16087 + }, + { + "epoch": 0.16088, + "grad_norm": 0.8051281695147977, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16088 + }, + { + "epoch": 0.16089, + "grad_norm": 0.7114352942425747, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 16089 + }, + { + "epoch": 0.1609, + "grad_norm": 0.5709791049100779, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 16090 + }, + { + "epoch": 0.16091, + "grad_norm": 0.659447951888532, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 16091 + }, + { + "epoch": 0.16092, + "grad_norm": 0.5850652350010963, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 16092 + }, + { + "epoch": 0.16093, + "grad_norm": 0.6133683820710318, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 16093 + }, + { + "epoch": 0.16094, + "grad_norm": 0.5906108623254351, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 16094 + }, + { + "epoch": 0.16095, + "grad_norm": 0.6507974636829058, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 16095 + }, + { + "epoch": 0.16096, + "grad_norm": 0.6912451480183617, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 16096 + }, + { + "epoch": 0.16097, + "grad_norm": 0.6574589184634025, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 16097 + }, + { + "epoch": 0.16098, + "grad_norm": 0.5467397832890231, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16098 + }, + { + "epoch": 0.16099, + "grad_norm": 0.508585633633795, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16099 + }, + { + "epoch": 0.161, + "grad_norm": 0.48599642924559777, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 16100 + }, + { + "epoch": 0.16101, + "grad_norm": 0.490443231114694, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 16101 + }, + { + "epoch": 0.16102, + "grad_norm": 0.565098862715796, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 16102 + }, + { + "epoch": 0.16103, + "grad_norm": 0.6590595741164376, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 16103 + }, + { + "epoch": 0.16104, + "grad_norm": 0.674582684774745, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 16104 + }, + { + "epoch": 0.16105, + "grad_norm": 0.5988268381470454, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 16105 + }, + { + "epoch": 0.16106, + "grad_norm": 0.6437399785590763, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 16106 + }, + { + "epoch": 0.16107, + "grad_norm": 0.8541603273243035, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 16107 + }, + { + "epoch": 0.16108, + "grad_norm": 1.097273910851997, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 16108 + }, + { + "epoch": 0.16109, + "grad_norm": 0.9609693183864391, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 16109 + }, + { + "epoch": 0.1611, + "grad_norm": 0.9239330857199889, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 16110 + }, + { + "epoch": 0.16111, + "grad_norm": 0.7998163550119576, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 16111 + }, + { + "epoch": 0.16112, + "grad_norm": 0.9416418634299187, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 16112 + }, + { + "epoch": 0.16113, + "grad_norm": 0.9113022998995228, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 16113 + }, + { + "epoch": 0.16114, + "grad_norm": 0.7912387800209134, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16114 + }, + { + "epoch": 0.16115, + "grad_norm": 0.8909357268613136, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 16115 + }, + { + "epoch": 0.16116, + "grad_norm": 0.7663943701559953, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 16116 + }, + { + "epoch": 0.16117, + "grad_norm": 0.7528703596258809, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 16117 + }, + { + "epoch": 0.16118, + "grad_norm": 0.8108097918338969, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 16118 + }, + { + "epoch": 0.16119, + "grad_norm": 0.8989951700723642, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 16119 + }, + { + "epoch": 0.1612, + "grad_norm": 0.7545408986130517, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 16120 + }, + { + "epoch": 0.16121, + "grad_norm": 0.605762420493875, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 16121 + }, + { + "epoch": 0.16122, + "grad_norm": 0.6589709748931336, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 16122 + }, + { + "epoch": 0.16123, + "grad_norm": 0.7630162511765206, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 16123 + }, + { + "epoch": 0.16124, + "grad_norm": 0.7958687953247566, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16124 + }, + { + "epoch": 0.16125, + "grad_norm": 0.9046180803092724, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 16125 + }, + { + "epoch": 0.16126, + "grad_norm": 0.9990068565365005, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 16126 + }, + { + "epoch": 0.16127, + "grad_norm": 0.8952294506951165, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 16127 + }, + { + "epoch": 0.16128, + "grad_norm": 0.6976266217678653, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 16128 + }, + { + "epoch": 0.16129, + "grad_norm": 0.6833411773282388, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 16129 + }, + { + "epoch": 0.1613, + "grad_norm": 0.6790075379594158, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 16130 + }, + { + "epoch": 0.16131, + "grad_norm": 0.5940614623172595, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 16131 + }, + { + "epoch": 0.16132, + "grad_norm": 0.6214550853285304, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 16132 + }, + { + "epoch": 0.16133, + "grad_norm": 0.6246946968573867, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 16133 + }, + { + "epoch": 0.16134, + "grad_norm": 0.6304352856303006, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 16134 + }, + { + "epoch": 0.16135, + "grad_norm": 0.5878584366615266, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 16135 + }, + { + "epoch": 0.16136, + "grad_norm": 0.49648187454864207, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 16136 + }, + { + "epoch": 0.16137, + "grad_norm": 0.5303085537685186, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 16137 + }, + { + "epoch": 0.16138, + "grad_norm": 0.6296936463741298, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 16138 + }, + { + "epoch": 0.16139, + "grad_norm": 0.6521582477896798, + "learning_rate": 0.003, + "loss": 4.02, + "step": 16139 + }, + { + "epoch": 0.1614, + "grad_norm": 0.6374880881181769, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 16140 + }, + { + "epoch": 0.16141, + "grad_norm": 0.6702355865901751, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 16141 + }, + { + "epoch": 0.16142, + "grad_norm": 0.9475081306866651, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 16142 + }, + { + "epoch": 0.16143, + "grad_norm": 1.0875682457795606, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 16143 + }, + { + "epoch": 0.16144, + "grad_norm": 0.8878779606889708, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 16144 + }, + { + "epoch": 0.16145, + "grad_norm": 0.8752096928318694, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16145 + }, + { + "epoch": 0.16146, + "grad_norm": 0.9084094160561826, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 16146 + }, + { + "epoch": 0.16147, + "grad_norm": 0.8563838628019046, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 16147 + }, + { + "epoch": 0.16148, + "grad_norm": 0.7498523468931342, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 16148 + }, + { + "epoch": 0.16149, + "grad_norm": 0.7825743651117434, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 16149 + }, + { + "epoch": 0.1615, + "grad_norm": 0.8023581350429726, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 16150 + }, + { + "epoch": 0.16151, + "grad_norm": 0.7260174953871584, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 16151 + }, + { + "epoch": 0.16152, + "grad_norm": 0.6897676984494139, + "learning_rate": 0.003, + "loss": 4.014, + "step": 16152 + }, + { + "epoch": 0.16153, + "grad_norm": 0.7034284857735298, + "learning_rate": 0.003, + "loss": 4.013, + "step": 16153 + }, + { + "epoch": 0.16154, + "grad_norm": 0.7507020994827392, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 16154 + }, + { + "epoch": 0.16155, + "grad_norm": 0.887547839520277, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 16155 + }, + { + "epoch": 0.16156, + "grad_norm": 1.1048397804938632, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 16156 + }, + { + "epoch": 0.16157, + "grad_norm": 1.0783829545633832, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 16157 + }, + { + "epoch": 0.16158, + "grad_norm": 0.744974921393579, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 16158 + }, + { + "epoch": 0.16159, + "grad_norm": 0.6109099836484049, + "learning_rate": 0.003, + "loss": 4.04, + "step": 16159 + }, + { + "epoch": 0.1616, + "grad_norm": 0.5764480231561824, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 16160 + }, + { + "epoch": 0.16161, + "grad_norm": 0.5933424736847807, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 16161 + }, + { + "epoch": 0.16162, + "grad_norm": 0.5953032039897719, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 16162 + }, + { + "epoch": 0.16163, + "grad_norm": 0.5807359098236737, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 16163 + }, + { + "epoch": 0.16164, + "grad_norm": 0.5283101500771322, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 16164 + }, + { + "epoch": 0.16165, + "grad_norm": 0.6053224932578491, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 16165 + }, + { + "epoch": 0.16166, + "grad_norm": 0.5774848345138113, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 16166 + }, + { + "epoch": 0.16167, + "grad_norm": 0.610819440784283, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 16167 + }, + { + "epoch": 0.16168, + "grad_norm": 0.6341551224138574, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 16168 + }, + { + "epoch": 0.16169, + "grad_norm": 0.6657034788727697, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 16169 + }, + { + "epoch": 0.1617, + "grad_norm": 0.7289660375645008, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 16170 + }, + { + "epoch": 0.16171, + "grad_norm": 0.804616387658045, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16171 + }, + { + "epoch": 0.16172, + "grad_norm": 0.7334452697200826, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16172 + }, + { + "epoch": 0.16173, + "grad_norm": 0.6599918824804044, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 16173 + }, + { + "epoch": 0.16174, + "grad_norm": 0.9002703557209406, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 16174 + }, + { + "epoch": 0.16175, + "grad_norm": 1.078591855871073, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 16175 + }, + { + "epoch": 0.16176, + "grad_norm": 1.067601159426037, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 16176 + }, + { + "epoch": 0.16177, + "grad_norm": 0.8216405102544481, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 16177 + }, + { + "epoch": 0.16178, + "grad_norm": 0.7472200305871399, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 16178 + }, + { + "epoch": 0.16179, + "grad_norm": 0.7248981737187594, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 16179 + }, + { + "epoch": 0.1618, + "grad_norm": 0.7445379440669236, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 16180 + }, + { + "epoch": 0.16181, + "grad_norm": 0.7129831288351488, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 16181 + }, + { + "epoch": 0.16182, + "grad_norm": 0.8638720812874842, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 16182 + }, + { + "epoch": 0.16183, + "grad_norm": 0.8844774135762427, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 16183 + }, + { + "epoch": 0.16184, + "grad_norm": 0.8815461591345237, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 16184 + }, + { + "epoch": 0.16185, + "grad_norm": 0.8998320569059928, + "learning_rate": 0.003, + "loss": 4.0517, + "step": 16185 + }, + { + "epoch": 0.16186, + "grad_norm": 0.8259068628632481, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 16186 + }, + { + "epoch": 0.16187, + "grad_norm": 0.734129101842798, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 16187 + }, + { + "epoch": 0.16188, + "grad_norm": 0.65896509055404, + "learning_rate": 0.003, + "loss": 4.032, + "step": 16188 + }, + { + "epoch": 0.16189, + "grad_norm": 0.6753498756051349, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 16189 + }, + { + "epoch": 0.1619, + "grad_norm": 0.6032331582089555, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 16190 + }, + { + "epoch": 0.16191, + "grad_norm": 0.47880274231086173, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 16191 + }, + { + "epoch": 0.16192, + "grad_norm": 0.4997830616020925, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 16192 + }, + { + "epoch": 0.16193, + "grad_norm": 0.5318040396927591, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 16193 + }, + { + "epoch": 0.16194, + "grad_norm": 0.5632949192163702, + "learning_rate": 0.003, + "loss": 3.994, + "step": 16194 + }, + { + "epoch": 0.16195, + "grad_norm": 0.6058575772082412, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16195 + }, + { + "epoch": 0.16196, + "grad_norm": 0.6484135134508132, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 16196 + }, + { + "epoch": 0.16197, + "grad_norm": 0.7808893129872457, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 16197 + }, + { + "epoch": 0.16198, + "grad_norm": 1.0984714710232961, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 16198 + }, + { + "epoch": 0.16199, + "grad_norm": 0.9149974585036751, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 16199 + }, + { + "epoch": 0.162, + "grad_norm": 0.7851508469769148, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 16200 + }, + { + "epoch": 0.16201, + "grad_norm": 0.5758409008390454, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 16201 + }, + { + "epoch": 0.16202, + "grad_norm": 0.5853888670606042, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 16202 + }, + { + "epoch": 0.16203, + "grad_norm": 0.5454840121783263, + "learning_rate": 0.003, + "loss": 3.994, + "step": 16203 + }, + { + "epoch": 0.16204, + "grad_norm": 0.5324545838272559, + "learning_rate": 0.003, + "loss": 4.033, + "step": 16204 + }, + { + "epoch": 0.16205, + "grad_norm": 0.5009860329213156, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 16205 + }, + { + "epoch": 0.16206, + "grad_norm": 0.5389285573947773, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 16206 + }, + { + "epoch": 0.16207, + "grad_norm": 0.5670979539676142, + "learning_rate": 0.003, + "loss": 4.015, + "step": 16207 + }, + { + "epoch": 0.16208, + "grad_norm": 0.6519958336632837, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 16208 + }, + { + "epoch": 0.16209, + "grad_norm": 0.6979847436255846, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 16209 + }, + { + "epoch": 0.1621, + "grad_norm": 0.7919191550921083, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 16210 + }, + { + "epoch": 0.16211, + "grad_norm": 0.8334947945405766, + "learning_rate": 0.003, + "loss": 4.002, + "step": 16211 + }, + { + "epoch": 0.16212, + "grad_norm": 0.8657092367668201, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 16212 + }, + { + "epoch": 0.16213, + "grad_norm": 0.8708515184373422, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 16213 + }, + { + "epoch": 0.16214, + "grad_norm": 0.8799585039440236, + "learning_rate": 0.003, + "loss": 4.041, + "step": 16214 + }, + { + "epoch": 0.16215, + "grad_norm": 1.0667641925676892, + "learning_rate": 0.003, + "loss": 4.0551, + "step": 16215 + }, + { + "epoch": 0.16216, + "grad_norm": 0.953930487544628, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 16216 + }, + { + "epoch": 0.16217, + "grad_norm": 0.9809809229724272, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 16217 + }, + { + "epoch": 0.16218, + "grad_norm": 1.0345277747142525, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 16218 + }, + { + "epoch": 0.16219, + "grad_norm": 0.9870946071472233, + "learning_rate": 0.003, + "loss": 4.029, + "step": 16219 + }, + { + "epoch": 0.1622, + "grad_norm": 0.9997293492588936, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 16220 + }, + { + "epoch": 0.16221, + "grad_norm": 0.872005108035573, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 16221 + }, + { + "epoch": 0.16222, + "grad_norm": 0.9092185885943878, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 16222 + }, + { + "epoch": 0.16223, + "grad_norm": 0.872799867134176, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 16223 + }, + { + "epoch": 0.16224, + "grad_norm": 0.8436704397699154, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 16224 + }, + { + "epoch": 0.16225, + "grad_norm": 0.6765726254424981, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 16225 + }, + { + "epoch": 0.16226, + "grad_norm": 0.788509951211089, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 16226 + }, + { + "epoch": 0.16227, + "grad_norm": 0.8939237703095619, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 16227 + }, + { + "epoch": 0.16228, + "grad_norm": 0.9099774742305652, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 16228 + }, + { + "epoch": 0.16229, + "grad_norm": 0.802085068595386, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 16229 + }, + { + "epoch": 0.1623, + "grad_norm": 0.734457046569413, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 16230 + }, + { + "epoch": 0.16231, + "grad_norm": 0.702724913575687, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 16231 + }, + { + "epoch": 0.16232, + "grad_norm": 0.6452982666936272, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 16232 + }, + { + "epoch": 0.16233, + "grad_norm": 0.6538409124173807, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 16233 + }, + { + "epoch": 0.16234, + "grad_norm": 0.7040975495812338, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 16234 + }, + { + "epoch": 0.16235, + "grad_norm": 0.7088208550867093, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 16235 + }, + { + "epoch": 0.16236, + "grad_norm": 0.6952724608499704, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 16236 + }, + { + "epoch": 0.16237, + "grad_norm": 0.7336803780352178, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 16237 + }, + { + "epoch": 0.16238, + "grad_norm": 0.6866333530332546, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 16238 + }, + { + "epoch": 0.16239, + "grad_norm": 0.6025927572764416, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 16239 + }, + { + "epoch": 0.1624, + "grad_norm": 0.5969180132833037, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16240 + }, + { + "epoch": 0.16241, + "grad_norm": 0.572593310607837, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 16241 + }, + { + "epoch": 0.16242, + "grad_norm": 0.6271330325351604, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16242 + }, + { + "epoch": 0.16243, + "grad_norm": 0.6579351654808658, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 16243 + }, + { + "epoch": 0.16244, + "grad_norm": 0.6939535707279298, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 16244 + }, + { + "epoch": 0.16245, + "grad_norm": 0.6848241603299615, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 16245 + }, + { + "epoch": 0.16246, + "grad_norm": 0.6365872833871931, + "learning_rate": 0.003, + "loss": 4.007, + "step": 16246 + }, + { + "epoch": 0.16247, + "grad_norm": 0.6316716074665942, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 16247 + }, + { + "epoch": 0.16248, + "grad_norm": 0.5025711449802966, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 16248 + }, + { + "epoch": 0.16249, + "grad_norm": 0.5668093355303251, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16249 + }, + { + "epoch": 0.1625, + "grad_norm": 0.5686863063847305, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 16250 + }, + { + "epoch": 0.16251, + "grad_norm": 0.5385976050244601, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 16251 + }, + { + "epoch": 0.16252, + "grad_norm": 0.4551452909637282, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 16252 + }, + { + "epoch": 0.16253, + "grad_norm": 0.46408726452996973, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 16253 + }, + { + "epoch": 0.16254, + "grad_norm": 0.47828005757273456, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 16254 + }, + { + "epoch": 0.16255, + "grad_norm": 0.5889145991740289, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 16255 + }, + { + "epoch": 0.16256, + "grad_norm": 0.7201355962524655, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 16256 + }, + { + "epoch": 0.16257, + "grad_norm": 0.7202815398124883, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 16257 + }, + { + "epoch": 0.16258, + "grad_norm": 0.6972445596748231, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 16258 + }, + { + "epoch": 0.16259, + "grad_norm": 0.7266683562942571, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 16259 + }, + { + "epoch": 0.1626, + "grad_norm": 0.6706879230151489, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 16260 + }, + { + "epoch": 0.16261, + "grad_norm": 0.5380161277758212, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 16261 + }, + { + "epoch": 0.16262, + "grad_norm": 0.6189664054883972, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16262 + }, + { + "epoch": 0.16263, + "grad_norm": 0.6743196345306963, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 16263 + }, + { + "epoch": 0.16264, + "grad_norm": 0.7430257664689909, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 16264 + }, + { + "epoch": 0.16265, + "grad_norm": 0.7156882311046834, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 16265 + }, + { + "epoch": 0.16266, + "grad_norm": 0.6795102294312222, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16266 + }, + { + "epoch": 0.16267, + "grad_norm": 0.7924924229077378, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 16267 + }, + { + "epoch": 0.16268, + "grad_norm": 0.7997339526196382, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 16268 + }, + { + "epoch": 0.16269, + "grad_norm": 0.8307902642456197, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 16269 + }, + { + "epoch": 0.1627, + "grad_norm": 0.9155892121559324, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 16270 + }, + { + "epoch": 0.16271, + "grad_norm": 0.9218896462064676, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 16271 + }, + { + "epoch": 0.16272, + "grad_norm": 1.0487724514624892, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 16272 + }, + { + "epoch": 0.16273, + "grad_norm": 1.116289266781017, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 16273 + }, + { + "epoch": 0.16274, + "grad_norm": 1.1020193957481552, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 16274 + }, + { + "epoch": 0.16275, + "grad_norm": 0.8216444441883671, + "learning_rate": 0.003, + "loss": 4.0552, + "step": 16275 + }, + { + "epoch": 0.16276, + "grad_norm": 0.7485287911430606, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 16276 + }, + { + "epoch": 0.16277, + "grad_norm": 0.7568613666068665, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 16277 + }, + { + "epoch": 0.16278, + "grad_norm": 0.7184223825418345, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 16278 + }, + { + "epoch": 0.16279, + "grad_norm": 0.7143724997391959, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16279 + }, + { + "epoch": 0.1628, + "grad_norm": 0.665659022350711, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 16280 + }, + { + "epoch": 0.16281, + "grad_norm": 0.6858205177492304, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 16281 + }, + { + "epoch": 0.16282, + "grad_norm": 0.7475526102744618, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 16282 + }, + { + "epoch": 0.16283, + "grad_norm": 0.8695527096296636, + "learning_rate": 0.003, + "loss": 3.981, + "step": 16283 + }, + { + "epoch": 0.16284, + "grad_norm": 1.036011690825396, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 16284 + }, + { + "epoch": 0.16285, + "grad_norm": 1.0375382957218422, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 16285 + }, + { + "epoch": 0.16286, + "grad_norm": 0.7322428899208343, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 16286 + }, + { + "epoch": 0.16287, + "grad_norm": 0.7477557813921698, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 16287 + }, + { + "epoch": 0.16288, + "grad_norm": 0.6979929926408653, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 16288 + }, + { + "epoch": 0.16289, + "grad_norm": 0.6649174538842552, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 16289 + }, + { + "epoch": 0.1629, + "grad_norm": 0.6948971581768276, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 16290 + }, + { + "epoch": 0.16291, + "grad_norm": 0.6352227750020072, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 16291 + }, + { + "epoch": 0.16292, + "grad_norm": 0.633154059904313, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 16292 + }, + { + "epoch": 0.16293, + "grad_norm": 0.5768506430250397, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 16293 + }, + { + "epoch": 0.16294, + "grad_norm": 0.60571653408596, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 16294 + }, + { + "epoch": 0.16295, + "grad_norm": 0.621393457298354, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 16295 + }, + { + "epoch": 0.16296, + "grad_norm": 0.5871115357821157, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 16296 + }, + { + "epoch": 0.16297, + "grad_norm": 0.5363282464982695, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 16297 + }, + { + "epoch": 0.16298, + "grad_norm": 0.6325808902171773, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 16298 + }, + { + "epoch": 0.16299, + "grad_norm": 0.7206593321481302, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 16299 + }, + { + "epoch": 0.163, + "grad_norm": 0.8553579478010428, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16300 + }, + { + "epoch": 0.16301, + "grad_norm": 1.076745870167687, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 16301 + }, + { + "epoch": 0.16302, + "grad_norm": 0.9465477291493395, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 16302 + }, + { + "epoch": 0.16303, + "grad_norm": 0.8711387462008231, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 16303 + }, + { + "epoch": 0.16304, + "grad_norm": 0.7289841081920717, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 16304 + }, + { + "epoch": 0.16305, + "grad_norm": 0.7006599680368707, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 16305 + }, + { + "epoch": 0.16306, + "grad_norm": 0.7194253284976674, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 16306 + }, + { + "epoch": 0.16307, + "grad_norm": 0.7063118216896344, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 16307 + }, + { + "epoch": 0.16308, + "grad_norm": 0.6425186257830108, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 16308 + }, + { + "epoch": 0.16309, + "grad_norm": 0.6169836365279108, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 16309 + }, + { + "epoch": 0.1631, + "grad_norm": 0.7019979322817804, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 16310 + }, + { + "epoch": 0.16311, + "grad_norm": 0.7189132318239398, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16311 + }, + { + "epoch": 0.16312, + "grad_norm": 0.7519758551222371, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 16312 + }, + { + "epoch": 0.16313, + "grad_norm": 0.7124391015512757, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 16313 + }, + { + "epoch": 0.16314, + "grad_norm": 0.6737926903167968, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 16314 + }, + { + "epoch": 0.16315, + "grad_norm": 0.7124793999649871, + "learning_rate": 0.003, + "loss": 3.988, + "step": 16315 + }, + { + "epoch": 0.16316, + "grad_norm": 0.765544321221971, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 16316 + }, + { + "epoch": 0.16317, + "grad_norm": 0.8086978339186814, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 16317 + }, + { + "epoch": 0.16318, + "grad_norm": 0.769274064186725, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 16318 + }, + { + "epoch": 0.16319, + "grad_norm": 0.7589679817493752, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 16319 + }, + { + "epoch": 0.1632, + "grad_norm": 0.765620153628993, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 16320 + }, + { + "epoch": 0.16321, + "grad_norm": 0.8486099594126554, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 16321 + }, + { + "epoch": 0.16322, + "grad_norm": 0.9616804029049413, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 16322 + }, + { + "epoch": 0.16323, + "grad_norm": 1.1384661219325642, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 16323 + }, + { + "epoch": 0.16324, + "grad_norm": 0.8340010615186295, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 16324 + }, + { + "epoch": 0.16325, + "grad_norm": 0.755199460639956, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 16325 + }, + { + "epoch": 0.16326, + "grad_norm": 0.751337112271056, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 16326 + }, + { + "epoch": 0.16327, + "grad_norm": 0.8639381717376634, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 16327 + }, + { + "epoch": 0.16328, + "grad_norm": 0.7844440993209423, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 16328 + }, + { + "epoch": 0.16329, + "grad_norm": 0.6983863270264371, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 16329 + }, + { + "epoch": 0.1633, + "grad_norm": 0.7013503618087201, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 16330 + }, + { + "epoch": 0.16331, + "grad_norm": 0.6787715551991773, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 16331 + }, + { + "epoch": 0.16332, + "grad_norm": 0.759908559446759, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 16332 + }, + { + "epoch": 0.16333, + "grad_norm": 0.7251999914038219, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 16333 + }, + { + "epoch": 0.16334, + "grad_norm": 0.7516532538440844, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 16334 + }, + { + "epoch": 0.16335, + "grad_norm": 0.7839828244770668, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 16335 + }, + { + "epoch": 0.16336, + "grad_norm": 0.7389410595991239, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 16336 + }, + { + "epoch": 0.16337, + "grad_norm": 0.6853234865769776, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16337 + }, + { + "epoch": 0.16338, + "grad_norm": 0.5722459442139302, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 16338 + }, + { + "epoch": 0.16339, + "grad_norm": 0.5529686623998846, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 16339 + }, + { + "epoch": 0.1634, + "grad_norm": 0.5518303502169559, + "learning_rate": 0.003, + "loss": 3.998, + "step": 16340 + }, + { + "epoch": 0.16341, + "grad_norm": 0.549411996949786, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 16341 + }, + { + "epoch": 0.16342, + "grad_norm": 0.5396355144022832, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 16342 + }, + { + "epoch": 0.16343, + "grad_norm": 0.5342421406793538, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 16343 + }, + { + "epoch": 0.16344, + "grad_norm": 0.5512441326919023, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 16344 + }, + { + "epoch": 0.16345, + "grad_norm": 0.6228111869492626, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 16345 + }, + { + "epoch": 0.16346, + "grad_norm": 0.7127444162969343, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 16346 + }, + { + "epoch": 0.16347, + "grad_norm": 0.7180734287131072, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 16347 + }, + { + "epoch": 0.16348, + "grad_norm": 0.8352210294620825, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 16348 + }, + { + "epoch": 0.16349, + "grad_norm": 0.9588001740572473, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 16349 + }, + { + "epoch": 0.1635, + "grad_norm": 1.0904789706131348, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 16350 + }, + { + "epoch": 0.16351, + "grad_norm": 0.9526242244579395, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16351 + }, + { + "epoch": 0.16352, + "grad_norm": 0.948085908561441, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 16352 + }, + { + "epoch": 0.16353, + "grad_norm": 0.8819566654480747, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 16353 + }, + { + "epoch": 0.16354, + "grad_norm": 0.7678065847987564, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 16354 + }, + { + "epoch": 0.16355, + "grad_norm": 0.6909468647644972, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 16355 + }, + { + "epoch": 0.16356, + "grad_norm": 0.6701128608535649, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 16356 + }, + { + "epoch": 0.16357, + "grad_norm": 0.6369516210517373, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16357 + }, + { + "epoch": 0.16358, + "grad_norm": 0.6732213786221984, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 16358 + }, + { + "epoch": 0.16359, + "grad_norm": 0.7834635147853903, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 16359 + }, + { + "epoch": 0.1636, + "grad_norm": 0.8188549499420716, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 16360 + }, + { + "epoch": 0.16361, + "grad_norm": 0.7579720774548057, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 16361 + }, + { + "epoch": 0.16362, + "grad_norm": 0.7081779600811833, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 16362 + }, + { + "epoch": 0.16363, + "grad_norm": 0.6772555640213418, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16363 + }, + { + "epoch": 0.16364, + "grad_norm": 0.7191750412797707, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 16364 + }, + { + "epoch": 0.16365, + "grad_norm": 0.785481923630699, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 16365 + }, + { + "epoch": 0.16366, + "grad_norm": 0.7689297426243351, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 16366 + }, + { + "epoch": 0.16367, + "grad_norm": 0.7358276567597489, + "learning_rate": 0.003, + "loss": 3.995, + "step": 16367 + }, + { + "epoch": 0.16368, + "grad_norm": 0.7395026643218053, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 16368 + }, + { + "epoch": 0.16369, + "grad_norm": 0.7930609900941302, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 16369 + }, + { + "epoch": 0.1637, + "grad_norm": 0.878393063755209, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 16370 + }, + { + "epoch": 0.16371, + "grad_norm": 0.9244332067570696, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 16371 + }, + { + "epoch": 0.16372, + "grad_norm": 0.9566689172479622, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 16372 + }, + { + "epoch": 0.16373, + "grad_norm": 0.8388356877731539, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 16373 + }, + { + "epoch": 0.16374, + "grad_norm": 0.7370318333178116, + "learning_rate": 0.003, + "loss": 4.013, + "step": 16374 + }, + { + "epoch": 0.16375, + "grad_norm": 0.6943733594081891, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 16375 + }, + { + "epoch": 0.16376, + "grad_norm": 0.7518910493924712, + "learning_rate": 0.003, + "loss": 4.009, + "step": 16376 + }, + { + "epoch": 0.16377, + "grad_norm": 0.9613842629641994, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 16377 + }, + { + "epoch": 0.16378, + "grad_norm": 1.0872594793142445, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 16378 + }, + { + "epoch": 0.16379, + "grad_norm": 0.7907909216701896, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 16379 + }, + { + "epoch": 0.1638, + "grad_norm": 0.6094247540259998, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 16380 + }, + { + "epoch": 0.16381, + "grad_norm": 0.625499374594474, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 16381 + }, + { + "epoch": 0.16382, + "grad_norm": 0.690187868947564, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 16382 + }, + { + "epoch": 0.16383, + "grad_norm": 0.6060456482247091, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 16383 + }, + { + "epoch": 0.16384, + "grad_norm": 0.5134089306334945, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 16384 + }, + { + "epoch": 0.16385, + "grad_norm": 0.5324007332052176, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16385 + }, + { + "epoch": 0.16386, + "grad_norm": 0.5486830993463355, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 16386 + }, + { + "epoch": 0.16387, + "grad_norm": 0.48576435925575684, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 16387 + }, + { + "epoch": 0.16388, + "grad_norm": 0.47274998914073607, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 16388 + }, + { + "epoch": 0.16389, + "grad_norm": 0.49428747696802877, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 16389 + }, + { + "epoch": 0.1639, + "grad_norm": 0.5695940078575901, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 16390 + }, + { + "epoch": 0.16391, + "grad_norm": 0.6216006033396068, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16391 + }, + { + "epoch": 0.16392, + "grad_norm": 0.7175028222608615, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 16392 + }, + { + "epoch": 0.16393, + "grad_norm": 0.9764802233160966, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 16393 + }, + { + "epoch": 0.16394, + "grad_norm": 1.0702479944492942, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 16394 + }, + { + "epoch": 0.16395, + "grad_norm": 0.8777996682590724, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 16395 + }, + { + "epoch": 0.16396, + "grad_norm": 0.7155818550068563, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 16396 + }, + { + "epoch": 0.16397, + "grad_norm": 0.5666721499627739, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 16397 + }, + { + "epoch": 0.16398, + "grad_norm": 0.5682908224909041, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 16398 + }, + { + "epoch": 0.16399, + "grad_norm": 0.555761340419061, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 16399 + }, + { + "epoch": 0.164, + "grad_norm": 0.6268500187264583, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 16400 + }, + { + "epoch": 0.16401, + "grad_norm": 0.7989067750953144, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 16401 + }, + { + "epoch": 0.16402, + "grad_norm": 0.8308146899980249, + "learning_rate": 0.003, + "loss": 4.0495, + "step": 16402 + }, + { + "epoch": 0.16403, + "grad_norm": 0.8230695870433359, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 16403 + }, + { + "epoch": 0.16404, + "grad_norm": 0.8758201146437213, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 16404 + }, + { + "epoch": 0.16405, + "grad_norm": 0.7344539185044385, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 16405 + }, + { + "epoch": 0.16406, + "grad_norm": 0.7134930985824104, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 16406 + }, + { + "epoch": 0.16407, + "grad_norm": 0.7363233558172064, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 16407 + }, + { + "epoch": 0.16408, + "grad_norm": 0.8833839529986364, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16408 + }, + { + "epoch": 0.16409, + "grad_norm": 1.0355934624412118, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 16409 + }, + { + "epoch": 0.1641, + "grad_norm": 0.965255890769591, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 16410 + }, + { + "epoch": 0.16411, + "grad_norm": 0.9774698065198513, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 16411 + }, + { + "epoch": 0.16412, + "grad_norm": 0.8417000020541207, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 16412 + }, + { + "epoch": 0.16413, + "grad_norm": 0.7863405871424867, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 16413 + }, + { + "epoch": 0.16414, + "grad_norm": 0.6451336526510499, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 16414 + }, + { + "epoch": 0.16415, + "grad_norm": 0.623838395579373, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 16415 + }, + { + "epoch": 0.16416, + "grad_norm": 0.5937779829657698, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 16416 + }, + { + "epoch": 0.16417, + "grad_norm": 0.6193328358230892, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 16417 + }, + { + "epoch": 0.16418, + "grad_norm": 0.6554149913204125, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 16418 + }, + { + "epoch": 0.16419, + "grad_norm": 0.6101665308331341, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 16419 + }, + { + "epoch": 0.1642, + "grad_norm": 0.5701165193904526, + "learning_rate": 0.003, + "loss": 4.0519, + "step": 16420 + }, + { + "epoch": 0.16421, + "grad_norm": 0.6452338591440953, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16421 + }, + { + "epoch": 0.16422, + "grad_norm": 0.8031660149796181, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 16422 + }, + { + "epoch": 0.16423, + "grad_norm": 0.9491918663346045, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 16423 + }, + { + "epoch": 0.16424, + "grad_norm": 1.0335877395685158, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 16424 + }, + { + "epoch": 0.16425, + "grad_norm": 0.8298612613291083, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16425 + }, + { + "epoch": 0.16426, + "grad_norm": 0.8582090494805862, + "learning_rate": 0.003, + "loss": 4.045, + "step": 16426 + }, + { + "epoch": 0.16427, + "grad_norm": 0.8049478330110112, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16427 + }, + { + "epoch": 0.16428, + "grad_norm": 0.7081441641589571, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16428 + }, + { + "epoch": 0.16429, + "grad_norm": 0.7012940790636004, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 16429 + }, + { + "epoch": 0.1643, + "grad_norm": 0.5416775807969733, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 16430 + }, + { + "epoch": 0.16431, + "grad_norm": 0.5826579003985968, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 16431 + }, + { + "epoch": 0.16432, + "grad_norm": 0.6026830665292304, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 16432 + }, + { + "epoch": 0.16433, + "grad_norm": 0.6130162397220662, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 16433 + }, + { + "epoch": 0.16434, + "grad_norm": 0.6110976682406473, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 16434 + }, + { + "epoch": 0.16435, + "grad_norm": 0.6228558564679885, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 16435 + }, + { + "epoch": 0.16436, + "grad_norm": 0.6838090927932065, + "learning_rate": 0.003, + "loss": 4.007, + "step": 16436 + }, + { + "epoch": 0.16437, + "grad_norm": 0.6996980789733964, + "learning_rate": 0.003, + "loss": 4.03, + "step": 16437 + }, + { + "epoch": 0.16438, + "grad_norm": 0.6751407662658467, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 16438 + }, + { + "epoch": 0.16439, + "grad_norm": 0.6261333618808353, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 16439 + }, + { + "epoch": 0.1644, + "grad_norm": 0.5348366864511399, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 16440 + }, + { + "epoch": 0.16441, + "grad_norm": 0.5234800913644054, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 16441 + }, + { + "epoch": 0.16442, + "grad_norm": 0.49088270718253285, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 16442 + }, + { + "epoch": 0.16443, + "grad_norm": 0.4606029937945308, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 16443 + }, + { + "epoch": 0.16444, + "grad_norm": 0.484358388287165, + "learning_rate": 0.003, + "loss": 4.012, + "step": 16444 + }, + { + "epoch": 0.16445, + "grad_norm": 0.6014883026714138, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 16445 + }, + { + "epoch": 0.16446, + "grad_norm": 0.7251880621026916, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 16446 + }, + { + "epoch": 0.16447, + "grad_norm": 0.7579477406269334, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 16447 + }, + { + "epoch": 0.16448, + "grad_norm": 0.8400459812339308, + "learning_rate": 0.003, + "loss": 4.028, + "step": 16448 + }, + { + "epoch": 0.16449, + "grad_norm": 0.8584508772803555, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 16449 + }, + { + "epoch": 0.1645, + "grad_norm": 0.8376433502633845, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 16450 + }, + { + "epoch": 0.16451, + "grad_norm": 0.878260924992957, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 16451 + }, + { + "epoch": 0.16452, + "grad_norm": 0.8708676184721396, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 16452 + }, + { + "epoch": 0.16453, + "grad_norm": 0.833399373349953, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 16453 + }, + { + "epoch": 0.16454, + "grad_norm": 0.7485977280130607, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 16454 + }, + { + "epoch": 0.16455, + "grad_norm": 0.857585708548467, + "learning_rate": 0.003, + "loss": 4.039, + "step": 16455 + }, + { + "epoch": 0.16456, + "grad_norm": 0.8825035438361902, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 16456 + }, + { + "epoch": 0.16457, + "grad_norm": 0.9288411622256061, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 16457 + }, + { + "epoch": 0.16458, + "grad_norm": 1.149589601421984, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16458 + }, + { + "epoch": 0.16459, + "grad_norm": 1.066196652932666, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 16459 + }, + { + "epoch": 0.1646, + "grad_norm": 0.7680245119854924, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 16460 + }, + { + "epoch": 0.16461, + "grad_norm": 0.6642299562861197, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 16461 + }, + { + "epoch": 0.16462, + "grad_norm": 0.6844478747370674, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 16462 + }, + { + "epoch": 0.16463, + "grad_norm": 0.7665512621344426, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 16463 + }, + { + "epoch": 0.16464, + "grad_norm": 0.8607080892818392, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 16464 + }, + { + "epoch": 0.16465, + "grad_norm": 0.8452272293199399, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 16465 + }, + { + "epoch": 0.16466, + "grad_norm": 0.7375206157582019, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 16466 + }, + { + "epoch": 0.16467, + "grad_norm": 0.7971169672650823, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 16467 + }, + { + "epoch": 0.16468, + "grad_norm": 0.8075774634039125, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 16468 + }, + { + "epoch": 0.16469, + "grad_norm": 0.7983662231627953, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 16469 + }, + { + "epoch": 0.1647, + "grad_norm": 0.8019743462899563, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 16470 + }, + { + "epoch": 0.16471, + "grad_norm": 0.8493446062391746, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 16471 + }, + { + "epoch": 0.16472, + "grad_norm": 0.735023649322399, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 16472 + }, + { + "epoch": 0.16473, + "grad_norm": 0.7187804018831155, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 16473 + }, + { + "epoch": 0.16474, + "grad_norm": 0.7464700058684623, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 16474 + }, + { + "epoch": 0.16475, + "grad_norm": 0.7125643694773537, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 16475 + }, + { + "epoch": 0.16476, + "grad_norm": 0.8122211566386365, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 16476 + }, + { + "epoch": 0.16477, + "grad_norm": 1.013561338009026, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 16477 + }, + { + "epoch": 0.16478, + "grad_norm": 1.0523394196648967, + "learning_rate": 0.003, + "loss": 4.0571, + "step": 16478 + }, + { + "epoch": 0.16479, + "grad_norm": 0.9961628164478313, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 16479 + }, + { + "epoch": 0.1648, + "grad_norm": 0.8801933397245033, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16480 + }, + { + "epoch": 0.16481, + "grad_norm": 0.7104424189539608, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16481 + }, + { + "epoch": 0.16482, + "grad_norm": 0.6532205943373002, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16482 + }, + { + "epoch": 0.16483, + "grad_norm": 0.670010609585426, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 16483 + }, + { + "epoch": 0.16484, + "grad_norm": 0.7303092175408049, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 16484 + }, + { + "epoch": 0.16485, + "grad_norm": 0.6849760481748425, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 16485 + }, + { + "epoch": 0.16486, + "grad_norm": 0.6254766269924723, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 16486 + }, + { + "epoch": 0.16487, + "grad_norm": 0.5420751443970653, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 16487 + }, + { + "epoch": 0.16488, + "grad_norm": 0.5376818712322721, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 16488 + }, + { + "epoch": 0.16489, + "grad_norm": 0.44406899828957674, + "learning_rate": 0.003, + "loss": 4.012, + "step": 16489 + }, + { + "epoch": 0.1649, + "grad_norm": 0.46680516781620096, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 16490 + }, + { + "epoch": 0.16491, + "grad_norm": 0.5056397407883515, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 16491 + }, + { + "epoch": 0.16492, + "grad_norm": 0.582395016371069, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 16492 + }, + { + "epoch": 0.16493, + "grad_norm": 0.7463657898903787, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 16493 + }, + { + "epoch": 0.16494, + "grad_norm": 0.8830468295027808, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 16494 + }, + { + "epoch": 0.16495, + "grad_norm": 0.7593767188183723, + "learning_rate": 0.003, + "loss": 4.039, + "step": 16495 + }, + { + "epoch": 0.16496, + "grad_norm": 0.5838212493668908, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 16496 + }, + { + "epoch": 0.16497, + "grad_norm": 0.6434167522590665, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 16497 + }, + { + "epoch": 0.16498, + "grad_norm": 0.7032207165750831, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 16498 + }, + { + "epoch": 0.16499, + "grad_norm": 0.6953824910567883, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16499 + }, + { + "epoch": 0.165, + "grad_norm": 0.7098328871841306, + "learning_rate": 0.003, + "loss": 3.989, + "step": 16500 + }, + { + "epoch": 0.16501, + "grad_norm": 0.7061954467716096, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 16501 + }, + { + "epoch": 0.16502, + "grad_norm": 0.6338287728435836, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 16502 + }, + { + "epoch": 0.16503, + "grad_norm": 0.5730800814635354, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 16503 + }, + { + "epoch": 0.16504, + "grad_norm": 0.45257824913421824, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 16504 + }, + { + "epoch": 0.16505, + "grad_norm": 0.4676988614818606, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 16505 + }, + { + "epoch": 0.16506, + "grad_norm": 0.5048201268352747, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 16506 + }, + { + "epoch": 0.16507, + "grad_norm": 0.5363348276835753, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 16507 + }, + { + "epoch": 0.16508, + "grad_norm": 0.6223193511147211, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 16508 + }, + { + "epoch": 0.16509, + "grad_norm": 0.6628025625226366, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 16509 + }, + { + "epoch": 0.1651, + "grad_norm": 0.7231224204906437, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 16510 + }, + { + "epoch": 0.16511, + "grad_norm": 0.8363616049492013, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 16511 + }, + { + "epoch": 0.16512, + "grad_norm": 0.9656787780962515, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 16512 + }, + { + "epoch": 0.16513, + "grad_norm": 1.0312904910358542, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 16513 + }, + { + "epoch": 0.16514, + "grad_norm": 0.9576512945740546, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 16514 + }, + { + "epoch": 0.16515, + "grad_norm": 1.028215133511749, + "learning_rate": 0.003, + "loss": 4.026, + "step": 16515 + }, + { + "epoch": 0.16516, + "grad_norm": 1.1317288811541488, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 16516 + }, + { + "epoch": 0.16517, + "grad_norm": 0.9184951947242392, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 16517 + }, + { + "epoch": 0.16518, + "grad_norm": 0.8228872476269836, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 16518 + }, + { + "epoch": 0.16519, + "grad_norm": 0.894348447124691, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 16519 + }, + { + "epoch": 0.1652, + "grad_norm": 1.0649263400339763, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16520 + }, + { + "epoch": 0.16521, + "grad_norm": 0.9556713558571924, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 16521 + }, + { + "epoch": 0.16522, + "grad_norm": 0.8155354850150442, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 16522 + }, + { + "epoch": 0.16523, + "grad_norm": 0.6923996036687416, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 16523 + }, + { + "epoch": 0.16524, + "grad_norm": 0.6692984730813865, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 16524 + }, + { + "epoch": 0.16525, + "grad_norm": 0.6425488469448529, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 16525 + }, + { + "epoch": 0.16526, + "grad_norm": 0.5700480099481833, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 16526 + }, + { + "epoch": 0.16527, + "grad_norm": 0.5434421628674331, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 16527 + }, + { + "epoch": 0.16528, + "grad_norm": 0.5868787803396613, + "learning_rate": 0.003, + "loss": 3.99, + "step": 16528 + }, + { + "epoch": 0.16529, + "grad_norm": 0.6201473508240588, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 16529 + }, + { + "epoch": 0.1653, + "grad_norm": 0.6062242436333888, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 16530 + }, + { + "epoch": 0.16531, + "grad_norm": 0.5753237808188633, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16531 + }, + { + "epoch": 0.16532, + "grad_norm": 0.5803208159671768, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 16532 + }, + { + "epoch": 0.16533, + "grad_norm": 0.6021238547330601, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 16533 + }, + { + "epoch": 0.16534, + "grad_norm": 0.7256867469784206, + "learning_rate": 0.003, + "loss": 4.026, + "step": 16534 + }, + { + "epoch": 0.16535, + "grad_norm": 0.796027962308231, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 16535 + }, + { + "epoch": 0.16536, + "grad_norm": 0.7642171937523711, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 16536 + }, + { + "epoch": 0.16537, + "grad_norm": 0.8038896299032112, + "learning_rate": 0.003, + "loss": 4.03, + "step": 16537 + }, + { + "epoch": 0.16538, + "grad_norm": 0.8146244391921398, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 16538 + }, + { + "epoch": 0.16539, + "grad_norm": 0.8804562236498741, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 16539 + }, + { + "epoch": 0.1654, + "grad_norm": 0.873315547648944, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16540 + }, + { + "epoch": 0.16541, + "grad_norm": 0.8177297139128609, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 16541 + }, + { + "epoch": 0.16542, + "grad_norm": 0.7237785474040892, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16542 + }, + { + "epoch": 0.16543, + "grad_norm": 0.6746408331136379, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 16543 + }, + { + "epoch": 0.16544, + "grad_norm": 0.6389986309248177, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 16544 + }, + { + "epoch": 0.16545, + "grad_norm": 0.6220239479323114, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 16545 + }, + { + "epoch": 0.16546, + "grad_norm": 0.6239266109036372, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 16546 + }, + { + "epoch": 0.16547, + "grad_norm": 0.6693923178275285, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 16547 + }, + { + "epoch": 0.16548, + "grad_norm": 0.5766345907089254, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 16548 + }, + { + "epoch": 0.16549, + "grad_norm": 0.6069880830877322, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 16549 + }, + { + "epoch": 0.1655, + "grad_norm": 0.6635385903989455, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 16550 + }, + { + "epoch": 0.16551, + "grad_norm": 0.6964533591742508, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 16551 + }, + { + "epoch": 0.16552, + "grad_norm": 0.64613626777268, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 16552 + }, + { + "epoch": 0.16553, + "grad_norm": 0.545911093812616, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16553 + }, + { + "epoch": 0.16554, + "grad_norm": 0.5573538003915022, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 16554 + }, + { + "epoch": 0.16555, + "grad_norm": 0.5842997198604437, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 16555 + }, + { + "epoch": 0.16556, + "grad_norm": 0.660152332466579, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 16556 + }, + { + "epoch": 0.16557, + "grad_norm": 0.7610553568741445, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 16557 + }, + { + "epoch": 0.16558, + "grad_norm": 0.8942034729009333, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 16558 + }, + { + "epoch": 0.16559, + "grad_norm": 0.9697428366666488, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 16559 + }, + { + "epoch": 0.1656, + "grad_norm": 1.0473214113933182, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 16560 + }, + { + "epoch": 0.16561, + "grad_norm": 0.8610206946661342, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 16561 + }, + { + "epoch": 0.16562, + "grad_norm": 0.6308301406426157, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 16562 + }, + { + "epoch": 0.16563, + "grad_norm": 0.6624493928615727, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16563 + }, + { + "epoch": 0.16564, + "grad_norm": 0.6413298742534803, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 16564 + }, + { + "epoch": 0.16565, + "grad_norm": 0.6194382500310178, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 16565 + }, + { + "epoch": 0.16566, + "grad_norm": 0.6078141114801726, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 16566 + }, + { + "epoch": 0.16567, + "grad_norm": 0.5257363739558373, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 16567 + }, + { + "epoch": 0.16568, + "grad_norm": 0.5712562809132455, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 16568 + }, + { + "epoch": 0.16569, + "grad_norm": 0.5992294445038069, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 16569 + }, + { + "epoch": 0.1657, + "grad_norm": 0.7077951666407302, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 16570 + }, + { + "epoch": 0.16571, + "grad_norm": 0.7727476877039821, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 16571 + }, + { + "epoch": 0.16572, + "grad_norm": 0.7374954796317202, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 16572 + }, + { + "epoch": 0.16573, + "grad_norm": 0.7676109691346559, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 16573 + }, + { + "epoch": 0.16574, + "grad_norm": 0.7183427348826723, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 16574 + }, + { + "epoch": 0.16575, + "grad_norm": 0.6830484369482456, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 16575 + }, + { + "epoch": 0.16576, + "grad_norm": 0.7131490213739717, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 16576 + }, + { + "epoch": 0.16577, + "grad_norm": 0.7029946163686476, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 16577 + }, + { + "epoch": 0.16578, + "grad_norm": 0.6354240061585735, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 16578 + }, + { + "epoch": 0.16579, + "grad_norm": 0.5297546780249613, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 16579 + }, + { + "epoch": 0.1658, + "grad_norm": 0.5348034964453966, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 16580 + }, + { + "epoch": 0.16581, + "grad_norm": 0.47108538814367157, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 16581 + }, + { + "epoch": 0.16582, + "grad_norm": 0.5212505492363512, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 16582 + }, + { + "epoch": 0.16583, + "grad_norm": 0.6000944173637724, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 16583 + }, + { + "epoch": 0.16584, + "grad_norm": 0.6792806520795382, + "learning_rate": 0.003, + "loss": 3.999, + "step": 16584 + }, + { + "epoch": 0.16585, + "grad_norm": 0.893553983060512, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 16585 + }, + { + "epoch": 0.16586, + "grad_norm": 1.0657342511555685, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 16586 + }, + { + "epoch": 0.16587, + "grad_norm": 1.0615600764218098, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 16587 + }, + { + "epoch": 0.16588, + "grad_norm": 0.9596993485664361, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 16588 + }, + { + "epoch": 0.16589, + "grad_norm": 0.9652244093366414, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 16589 + }, + { + "epoch": 0.1659, + "grad_norm": 0.8915985880688833, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 16590 + }, + { + "epoch": 0.16591, + "grad_norm": 0.8909160817235098, + "learning_rate": 0.003, + "loss": 4.0563, + "step": 16591 + }, + { + "epoch": 0.16592, + "grad_norm": 1.03524228678834, + "learning_rate": 0.003, + "loss": 4.038, + "step": 16592 + }, + { + "epoch": 0.16593, + "grad_norm": 0.9386113051134701, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 16593 + }, + { + "epoch": 0.16594, + "grad_norm": 1.048529978153803, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 16594 + }, + { + "epoch": 0.16595, + "grad_norm": 1.0794076962928847, + "learning_rate": 0.003, + "loss": 4.0645, + "step": 16595 + }, + { + "epoch": 0.16596, + "grad_norm": 0.8868750212393541, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 16596 + }, + { + "epoch": 0.16597, + "grad_norm": 0.8522044835098456, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 16597 + }, + { + "epoch": 0.16598, + "grad_norm": 0.9436455399391949, + "learning_rate": 0.003, + "loss": 4.001, + "step": 16598 + }, + { + "epoch": 0.16599, + "grad_norm": 1.1107872976867337, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 16599 + }, + { + "epoch": 0.166, + "grad_norm": 0.9792273305236139, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 16600 + }, + { + "epoch": 0.16601, + "grad_norm": 0.8372395782881267, + "learning_rate": 0.003, + "loss": 4.067, + "step": 16601 + }, + { + "epoch": 0.16602, + "grad_norm": 0.7646032971191477, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 16602 + }, + { + "epoch": 0.16603, + "grad_norm": 0.8028017843869113, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 16603 + }, + { + "epoch": 0.16604, + "grad_norm": 0.9279757762790702, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 16604 + }, + { + "epoch": 0.16605, + "grad_norm": 0.9627432564846211, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 16605 + }, + { + "epoch": 0.16606, + "grad_norm": 0.8583188352173863, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 16606 + }, + { + "epoch": 0.16607, + "grad_norm": 0.7036843659580695, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 16607 + }, + { + "epoch": 0.16608, + "grad_norm": 0.6524598429544792, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 16608 + }, + { + "epoch": 0.16609, + "grad_norm": 0.6244286741960031, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 16609 + }, + { + "epoch": 0.1661, + "grad_norm": 0.5586733644994009, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 16610 + }, + { + "epoch": 0.16611, + "grad_norm": 0.5812382384052477, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 16611 + }, + { + "epoch": 0.16612, + "grad_norm": 0.6037017733360739, + "learning_rate": 0.003, + "loss": 4.009, + "step": 16612 + }, + { + "epoch": 0.16613, + "grad_norm": 0.64352966203228, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 16613 + }, + { + "epoch": 0.16614, + "grad_norm": 0.6523900922510829, + "learning_rate": 0.003, + "loss": 4.028, + "step": 16614 + }, + { + "epoch": 0.16615, + "grad_norm": 0.6733260776837747, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 16615 + }, + { + "epoch": 0.16616, + "grad_norm": 0.6885057686983833, + "learning_rate": 0.003, + "loss": 4.0627, + "step": 16616 + }, + { + "epoch": 0.16617, + "grad_norm": 0.6578363938270875, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 16617 + }, + { + "epoch": 0.16618, + "grad_norm": 0.5117687646826794, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 16618 + }, + { + "epoch": 0.16619, + "grad_norm": 0.5060450359303601, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 16619 + }, + { + "epoch": 0.1662, + "grad_norm": 0.5156616672797957, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 16620 + }, + { + "epoch": 0.16621, + "grad_norm": 0.5026937610775707, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 16621 + }, + { + "epoch": 0.16622, + "grad_norm": 0.4775696829462301, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 16622 + }, + { + "epoch": 0.16623, + "grad_norm": 0.4598893676976056, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 16623 + }, + { + "epoch": 0.16624, + "grad_norm": 0.4802907761785501, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 16624 + }, + { + "epoch": 0.16625, + "grad_norm": 0.6363494824752323, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 16625 + }, + { + "epoch": 0.16626, + "grad_norm": 0.8976520170400485, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 16626 + }, + { + "epoch": 0.16627, + "grad_norm": 1.1653702407814712, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 16627 + }, + { + "epoch": 0.16628, + "grad_norm": 0.593774301280545, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 16628 + }, + { + "epoch": 0.16629, + "grad_norm": 0.7468795437825465, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 16629 + }, + { + "epoch": 0.1663, + "grad_norm": 1.1022392135862265, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 16630 + }, + { + "epoch": 0.16631, + "grad_norm": 0.8842233857661201, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 16631 + }, + { + "epoch": 0.16632, + "grad_norm": 0.7594948664885439, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 16632 + }, + { + "epoch": 0.16633, + "grad_norm": 0.7840774794608127, + "learning_rate": 0.003, + "loss": 4.026, + "step": 16633 + }, + { + "epoch": 0.16634, + "grad_norm": 0.644774700786173, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 16634 + }, + { + "epoch": 0.16635, + "grad_norm": 0.7235567099470066, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 16635 + }, + { + "epoch": 0.16636, + "grad_norm": 0.811120442429958, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 16636 + }, + { + "epoch": 0.16637, + "grad_norm": 0.8101378313575973, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 16637 + }, + { + "epoch": 0.16638, + "grad_norm": 0.7659825138445999, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 16638 + }, + { + "epoch": 0.16639, + "grad_norm": 0.7299647883275843, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 16639 + }, + { + "epoch": 0.1664, + "grad_norm": 0.741231239562297, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 16640 + }, + { + "epoch": 0.16641, + "grad_norm": 0.7569383412400844, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 16641 + }, + { + "epoch": 0.16642, + "grad_norm": 0.8824003287125595, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 16642 + }, + { + "epoch": 0.16643, + "grad_norm": 0.8401022707765351, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 16643 + }, + { + "epoch": 0.16644, + "grad_norm": 0.8411312891313598, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 16644 + }, + { + "epoch": 0.16645, + "grad_norm": 0.8601514461763959, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 16645 + }, + { + "epoch": 0.16646, + "grad_norm": 0.8241228413121059, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16646 + }, + { + "epoch": 0.16647, + "grad_norm": 0.7306910097383713, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 16647 + }, + { + "epoch": 0.16648, + "grad_norm": 0.7123672549025807, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 16648 + }, + { + "epoch": 0.16649, + "grad_norm": 0.6650785906138053, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 16649 + }, + { + "epoch": 0.1665, + "grad_norm": 0.8041409216722969, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 16650 + }, + { + "epoch": 0.16651, + "grad_norm": 0.7242744981621208, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 16651 + }, + { + "epoch": 0.16652, + "grad_norm": 0.7525163319178036, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 16652 + }, + { + "epoch": 0.16653, + "grad_norm": 0.8286989080538483, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 16653 + }, + { + "epoch": 0.16654, + "grad_norm": 0.7619532497884564, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 16654 + }, + { + "epoch": 0.16655, + "grad_norm": 0.7851128823161653, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 16655 + }, + { + "epoch": 0.16656, + "grad_norm": 0.7396768014754642, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 16656 + }, + { + "epoch": 0.16657, + "grad_norm": 0.7030460133203666, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 16657 + }, + { + "epoch": 0.16658, + "grad_norm": 0.7733620550905135, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 16658 + }, + { + "epoch": 0.16659, + "grad_norm": 0.8097166598147543, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 16659 + }, + { + "epoch": 0.1666, + "grad_norm": 0.8347156142983782, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 16660 + }, + { + "epoch": 0.16661, + "grad_norm": 0.8715704542322701, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 16661 + }, + { + "epoch": 0.16662, + "grad_norm": 0.9024117377594998, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 16662 + }, + { + "epoch": 0.16663, + "grad_norm": 0.7602906925775678, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 16663 + }, + { + "epoch": 0.16664, + "grad_norm": 0.6650816529658774, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 16664 + }, + { + "epoch": 0.16665, + "grad_norm": 0.6105322893294864, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 16665 + }, + { + "epoch": 0.16666, + "grad_norm": 0.588510620791206, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 16666 + }, + { + "epoch": 0.16667, + "grad_norm": 0.5630924088452839, + "learning_rate": 0.003, + "loss": 4.003, + "step": 16667 + }, + { + "epoch": 0.16668, + "grad_norm": 0.5903964402249022, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 16668 + }, + { + "epoch": 0.16669, + "grad_norm": 0.7069933302344071, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 16669 + }, + { + "epoch": 0.1667, + "grad_norm": 0.8791067888111274, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 16670 + }, + { + "epoch": 0.16671, + "grad_norm": 0.8742420689541519, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 16671 + }, + { + "epoch": 0.16672, + "grad_norm": 0.7876662742812034, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 16672 + }, + { + "epoch": 0.16673, + "grad_norm": 0.9827649804891835, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 16673 + }, + { + "epoch": 0.16674, + "grad_norm": 1.0376375213641038, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 16674 + }, + { + "epoch": 0.16675, + "grad_norm": 0.9037954996999619, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 16675 + }, + { + "epoch": 0.16676, + "grad_norm": 0.811889932396361, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 16676 + }, + { + "epoch": 0.16677, + "grad_norm": 0.7858416873185846, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 16677 + }, + { + "epoch": 0.16678, + "grad_norm": 0.7186202839571197, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 16678 + }, + { + "epoch": 0.16679, + "grad_norm": 0.7416643534962624, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 16679 + }, + { + "epoch": 0.1668, + "grad_norm": 0.7189457942462366, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 16680 + }, + { + "epoch": 0.16681, + "grad_norm": 0.6674559582954873, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 16681 + }, + { + "epoch": 0.16682, + "grad_norm": 0.6314509907385235, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 16682 + }, + { + "epoch": 0.16683, + "grad_norm": 0.5913854227738622, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 16683 + }, + { + "epoch": 0.16684, + "grad_norm": 0.6351017341768526, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 16684 + }, + { + "epoch": 0.16685, + "grad_norm": 0.6761164394583006, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 16685 + }, + { + "epoch": 0.16686, + "grad_norm": 0.7723075442170572, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 16686 + }, + { + "epoch": 0.16687, + "grad_norm": 0.8419766154858084, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 16687 + }, + { + "epoch": 0.16688, + "grad_norm": 0.8436707322016122, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 16688 + }, + { + "epoch": 0.16689, + "grad_norm": 0.7119563090251725, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 16689 + }, + { + "epoch": 0.1669, + "grad_norm": 0.654681828430589, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 16690 + }, + { + "epoch": 0.16691, + "grad_norm": 0.5697939845476346, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 16691 + }, + { + "epoch": 0.16692, + "grad_norm": 0.5463554891589016, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 16692 + }, + { + "epoch": 0.16693, + "grad_norm": 0.5864305563871623, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 16693 + }, + { + "epoch": 0.16694, + "grad_norm": 0.5635010910046606, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 16694 + }, + { + "epoch": 0.16695, + "grad_norm": 0.5106454112966375, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 16695 + }, + { + "epoch": 0.16696, + "grad_norm": 0.5607818333633833, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 16696 + }, + { + "epoch": 0.16697, + "grad_norm": 0.6518375652717548, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 16697 + }, + { + "epoch": 0.16698, + "grad_norm": 0.6755317952783747, + "learning_rate": 0.003, + "loss": 4.006, + "step": 16698 + }, + { + "epoch": 0.16699, + "grad_norm": 0.7973110319928953, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 16699 + }, + { + "epoch": 0.167, + "grad_norm": 0.9173216538892576, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 16700 + }, + { + "epoch": 0.16701, + "grad_norm": 0.8028931574925393, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 16701 + }, + { + "epoch": 0.16702, + "grad_norm": 0.9366465468145423, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 16702 + }, + { + "epoch": 0.16703, + "grad_norm": 1.0397658257613431, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 16703 + }, + { + "epoch": 0.16704, + "grad_norm": 0.9647412052649074, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 16704 + }, + { + "epoch": 0.16705, + "grad_norm": 0.8615606653113577, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 16705 + }, + { + "epoch": 0.16706, + "grad_norm": 0.804122838704138, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 16706 + }, + { + "epoch": 0.16707, + "grad_norm": 0.7722543363064113, + "learning_rate": 0.003, + "loss": 4.012, + "step": 16707 + }, + { + "epoch": 0.16708, + "grad_norm": 0.7910931652718152, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 16708 + }, + { + "epoch": 0.16709, + "grad_norm": 0.8002273061322099, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 16709 + }, + { + "epoch": 0.1671, + "grad_norm": 0.7897750795074798, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 16710 + }, + { + "epoch": 0.16711, + "grad_norm": 0.8725024527913912, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 16711 + }, + { + "epoch": 0.16712, + "grad_norm": 0.8454031320755834, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 16712 + }, + { + "epoch": 0.16713, + "grad_norm": 0.7364556913260923, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16713 + }, + { + "epoch": 0.16714, + "grad_norm": 0.6398530334992498, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 16714 + }, + { + "epoch": 0.16715, + "grad_norm": 0.7114671198193028, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 16715 + }, + { + "epoch": 0.16716, + "grad_norm": 0.7115196228800367, + "learning_rate": 0.003, + "loss": 4.021, + "step": 16716 + }, + { + "epoch": 0.16717, + "grad_norm": 0.6451231789408886, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 16717 + }, + { + "epoch": 0.16718, + "grad_norm": 0.6584247807451084, + "learning_rate": 0.003, + "loss": 4.003, + "step": 16718 + }, + { + "epoch": 0.16719, + "grad_norm": 0.6249796722084001, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 16719 + }, + { + "epoch": 0.1672, + "grad_norm": 0.5383604432025493, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 16720 + }, + { + "epoch": 0.16721, + "grad_norm": 0.6058184683521908, + "learning_rate": 0.003, + "loss": 4.006, + "step": 16721 + }, + { + "epoch": 0.16722, + "grad_norm": 0.59347748877088, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 16722 + }, + { + "epoch": 0.16723, + "grad_norm": 0.6142979708255785, + "learning_rate": 0.003, + "loss": 3.992, + "step": 16723 + }, + { + "epoch": 0.16724, + "grad_norm": 0.5840631370223596, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 16724 + }, + { + "epoch": 0.16725, + "grad_norm": 0.5367925508446715, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 16725 + }, + { + "epoch": 0.16726, + "grad_norm": 0.48890010964714997, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 16726 + }, + { + "epoch": 0.16727, + "grad_norm": 0.5672254518949784, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 16727 + }, + { + "epoch": 0.16728, + "grad_norm": 0.6266122831741209, + "learning_rate": 0.003, + "loss": 4.011, + "step": 16728 + }, + { + "epoch": 0.16729, + "grad_norm": 0.7491806114185904, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 16729 + }, + { + "epoch": 0.1673, + "grad_norm": 0.8275689035634474, + "learning_rate": 0.003, + "loss": 4.0545, + "step": 16730 + }, + { + "epoch": 0.16731, + "grad_norm": 0.7690222834853109, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 16731 + }, + { + "epoch": 0.16732, + "grad_norm": 0.591075265268295, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 16732 + }, + { + "epoch": 0.16733, + "grad_norm": 0.6564854185743794, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 16733 + }, + { + "epoch": 0.16734, + "grad_norm": 0.5947707098146063, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 16734 + }, + { + "epoch": 0.16735, + "grad_norm": 0.6563516119367269, + "learning_rate": 0.003, + "loss": 4.034, + "step": 16735 + }, + { + "epoch": 0.16736, + "grad_norm": 0.7261527775324446, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 16736 + }, + { + "epoch": 0.16737, + "grad_norm": 0.6936945597889562, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 16737 + }, + { + "epoch": 0.16738, + "grad_norm": 0.7075168163436032, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 16738 + }, + { + "epoch": 0.16739, + "grad_norm": 0.7556956393523185, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 16739 + }, + { + "epoch": 0.1674, + "grad_norm": 0.8490249952953227, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 16740 + }, + { + "epoch": 0.16741, + "grad_norm": 0.9243400148146629, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 16741 + }, + { + "epoch": 0.16742, + "grad_norm": 0.8933983465452434, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 16742 + }, + { + "epoch": 0.16743, + "grad_norm": 0.8475593969628112, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 16743 + }, + { + "epoch": 0.16744, + "grad_norm": 0.8432738385217966, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 16744 + }, + { + "epoch": 0.16745, + "grad_norm": 0.8122289806506147, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 16745 + }, + { + "epoch": 0.16746, + "grad_norm": 0.7230136474631398, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 16746 + }, + { + "epoch": 0.16747, + "grad_norm": 0.6705237117111017, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 16747 + }, + { + "epoch": 0.16748, + "grad_norm": 0.6775589825591243, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 16748 + }, + { + "epoch": 0.16749, + "grad_norm": 0.7264421554824158, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 16749 + }, + { + "epoch": 0.1675, + "grad_norm": 0.8552460384509942, + "learning_rate": 0.003, + "loss": 4.004, + "step": 16750 + }, + { + "epoch": 0.16751, + "grad_norm": 0.8671152367770061, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 16751 + }, + { + "epoch": 0.16752, + "grad_norm": 0.9309298891794899, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 16752 + }, + { + "epoch": 0.16753, + "grad_norm": 0.8361208365507218, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 16753 + }, + { + "epoch": 0.16754, + "grad_norm": 0.7647484598481045, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 16754 + }, + { + "epoch": 0.16755, + "grad_norm": 0.8589998183947432, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 16755 + }, + { + "epoch": 0.16756, + "grad_norm": 0.8924377110307843, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 16756 + }, + { + "epoch": 0.16757, + "grad_norm": 0.7680231580155855, + "learning_rate": 0.003, + "loss": 4.023, + "step": 16757 + }, + { + "epoch": 0.16758, + "grad_norm": 0.7560084542605441, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 16758 + }, + { + "epoch": 0.16759, + "grad_norm": 0.740996353244455, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 16759 + }, + { + "epoch": 0.1676, + "grad_norm": 0.7831210759823999, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 16760 + }, + { + "epoch": 0.16761, + "grad_norm": 0.7735722916103702, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 16761 + }, + { + "epoch": 0.16762, + "grad_norm": 0.7751703015000184, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 16762 + }, + { + "epoch": 0.16763, + "grad_norm": 0.8198367031222362, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 16763 + }, + { + "epoch": 0.16764, + "grad_norm": 0.7284877909282639, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 16764 + }, + { + "epoch": 0.16765, + "grad_norm": 0.6450231234714806, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 16765 + }, + { + "epoch": 0.16766, + "grad_norm": 0.6620247013720122, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 16766 + }, + { + "epoch": 0.16767, + "grad_norm": 0.6482779719723644, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 16767 + }, + { + "epoch": 0.16768, + "grad_norm": 0.7326699909064435, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 16768 + }, + { + "epoch": 0.16769, + "grad_norm": 0.8239256793417282, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 16769 + }, + { + "epoch": 0.1677, + "grad_norm": 0.8878881853869682, + "learning_rate": 0.003, + "loss": 4.016, + "step": 16770 + }, + { + "epoch": 0.16771, + "grad_norm": 0.91585863679991, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 16771 + }, + { + "epoch": 0.16772, + "grad_norm": 0.9687999832102712, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 16772 + }, + { + "epoch": 0.16773, + "grad_norm": 0.8954591918216721, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 16773 + }, + { + "epoch": 0.16774, + "grad_norm": 0.7930032892718559, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 16774 + }, + { + "epoch": 0.16775, + "grad_norm": 0.9055443829725194, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 16775 + }, + { + "epoch": 0.16776, + "grad_norm": 0.9086146212994963, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 16776 + }, + { + "epoch": 0.16777, + "grad_norm": 0.8793903466579152, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 16777 + }, + { + "epoch": 0.16778, + "grad_norm": 0.8859953294356973, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 16778 + }, + { + "epoch": 0.16779, + "grad_norm": 0.8879314411546951, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 16779 + }, + { + "epoch": 0.1678, + "grad_norm": 0.7699862216637151, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 16780 + }, + { + "epoch": 0.16781, + "grad_norm": 0.7357890056601423, + "learning_rate": 0.003, + "loss": 4.0639, + "step": 16781 + }, + { + "epoch": 0.16782, + "grad_norm": 0.844258636822252, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 16782 + }, + { + "epoch": 0.16783, + "grad_norm": 0.8802860474280216, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 16783 + }, + { + "epoch": 0.16784, + "grad_norm": 0.7949918295837642, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 16784 + }, + { + "epoch": 0.16785, + "grad_norm": 0.6872862802052109, + "learning_rate": 0.003, + "loss": 4.034, + "step": 16785 + }, + { + "epoch": 0.16786, + "grad_norm": 0.6656420189355488, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 16786 + }, + { + "epoch": 0.16787, + "grad_norm": 0.5398548236649521, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 16787 + }, + { + "epoch": 0.16788, + "grad_norm": 0.49422516873126654, + "learning_rate": 0.003, + "loss": 4.04, + "step": 16788 + }, + { + "epoch": 0.16789, + "grad_norm": 0.5238738627388017, + "learning_rate": 0.003, + "loss": 4.019, + "step": 16789 + }, + { + "epoch": 0.1679, + "grad_norm": 0.5179973901228337, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 16790 + }, + { + "epoch": 0.16791, + "grad_norm": 0.5386634213773159, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 16791 + }, + { + "epoch": 0.16792, + "grad_norm": 0.5667131732366432, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 16792 + }, + { + "epoch": 0.16793, + "grad_norm": 0.5838665154627402, + "learning_rate": 0.003, + "loss": 3.994, + "step": 16793 + }, + { + "epoch": 0.16794, + "grad_norm": 0.618691361279987, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 16794 + }, + { + "epoch": 0.16795, + "grad_norm": 0.6379281668548806, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 16795 + }, + { + "epoch": 0.16796, + "grad_norm": 0.6496360161670717, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 16796 + }, + { + "epoch": 0.16797, + "grad_norm": 0.6382194191720955, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16797 + }, + { + "epoch": 0.16798, + "grad_norm": 0.6435134632830762, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16798 + }, + { + "epoch": 0.16799, + "grad_norm": 0.6745659407373322, + "learning_rate": 0.003, + "loss": 4.018, + "step": 16799 + }, + { + "epoch": 0.168, + "grad_norm": 0.7312583401572542, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 16800 + }, + { + "epoch": 0.16801, + "grad_norm": 0.8070789387858536, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 16801 + }, + { + "epoch": 0.16802, + "grad_norm": 0.7600302151673037, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 16802 + }, + { + "epoch": 0.16803, + "grad_norm": 0.683277498240685, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 16803 + }, + { + "epoch": 0.16804, + "grad_norm": 0.7543477407464146, + "learning_rate": 0.003, + "loss": 4.02, + "step": 16804 + }, + { + "epoch": 0.16805, + "grad_norm": 0.7784385909025363, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 16805 + }, + { + "epoch": 0.16806, + "grad_norm": 0.8313076128532343, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 16806 + }, + { + "epoch": 0.16807, + "grad_norm": 0.9240498082039174, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 16807 + }, + { + "epoch": 0.16808, + "grad_norm": 0.9008672481060601, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 16808 + }, + { + "epoch": 0.16809, + "grad_norm": 0.767978936217339, + "learning_rate": 0.003, + "loss": 4.019, + "step": 16809 + }, + { + "epoch": 0.1681, + "grad_norm": 0.5685551581065619, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 16810 + }, + { + "epoch": 0.16811, + "grad_norm": 0.6604902166888691, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 16811 + }, + { + "epoch": 0.16812, + "grad_norm": 0.7652538626111153, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 16812 + }, + { + "epoch": 0.16813, + "grad_norm": 0.7622971490308715, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16813 + }, + { + "epoch": 0.16814, + "grad_norm": 0.7255490620423525, + "learning_rate": 0.003, + "loss": 3.989, + "step": 16814 + }, + { + "epoch": 0.16815, + "grad_norm": 0.726258897210262, + "learning_rate": 0.003, + "loss": 4.0535, + "step": 16815 + }, + { + "epoch": 0.16816, + "grad_norm": 0.7798262849599433, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 16816 + }, + { + "epoch": 0.16817, + "grad_norm": 0.7349356525472343, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 16817 + }, + { + "epoch": 0.16818, + "grad_norm": 0.8099059991281846, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 16818 + }, + { + "epoch": 0.16819, + "grad_norm": 0.7132782316212541, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 16819 + }, + { + "epoch": 0.1682, + "grad_norm": 0.6792238050270083, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 16820 + }, + { + "epoch": 0.16821, + "grad_norm": 0.7475653875895177, + "learning_rate": 0.003, + "loss": 3.981, + "step": 16821 + }, + { + "epoch": 0.16822, + "grad_norm": 0.8150592232575337, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 16822 + }, + { + "epoch": 0.16823, + "grad_norm": 0.797119544266411, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 16823 + }, + { + "epoch": 0.16824, + "grad_norm": 0.8114778333249525, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 16824 + }, + { + "epoch": 0.16825, + "grad_norm": 0.8514512654990984, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 16825 + }, + { + "epoch": 0.16826, + "grad_norm": 0.8620320777178825, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 16826 + }, + { + "epoch": 0.16827, + "grad_norm": 0.8488029996004607, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 16827 + }, + { + "epoch": 0.16828, + "grad_norm": 0.8941131030394607, + "learning_rate": 0.003, + "loss": 4.022, + "step": 16828 + }, + { + "epoch": 0.16829, + "grad_norm": 0.9926874940667274, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 16829 + }, + { + "epoch": 0.1683, + "grad_norm": 0.951882518090118, + "learning_rate": 0.003, + "loss": 4.025, + "step": 16830 + }, + { + "epoch": 0.16831, + "grad_norm": 0.839025979844505, + "learning_rate": 0.003, + "loss": 3.999, + "step": 16831 + }, + { + "epoch": 0.16832, + "grad_norm": 0.8045343323029652, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 16832 + }, + { + "epoch": 0.16833, + "grad_norm": 0.7528059063619884, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 16833 + }, + { + "epoch": 0.16834, + "grad_norm": 0.8131074477951502, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 16834 + }, + { + "epoch": 0.16835, + "grad_norm": 0.8868890235139484, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 16835 + }, + { + "epoch": 0.16836, + "grad_norm": 0.7596994910590674, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 16836 + }, + { + "epoch": 0.16837, + "grad_norm": 0.7278987242445316, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 16837 + }, + { + "epoch": 0.16838, + "grad_norm": 0.6823660983926129, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 16838 + }, + { + "epoch": 0.16839, + "grad_norm": 0.5900695314929626, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 16839 + }, + { + "epoch": 0.1684, + "grad_norm": 0.6351232250423694, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 16840 + }, + { + "epoch": 0.16841, + "grad_norm": 0.6593507304516766, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 16841 + }, + { + "epoch": 0.16842, + "grad_norm": 0.7709707675316848, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16842 + }, + { + "epoch": 0.16843, + "grad_norm": 0.8813189151541583, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 16843 + }, + { + "epoch": 0.16844, + "grad_norm": 0.7412506674571124, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 16844 + }, + { + "epoch": 0.16845, + "grad_norm": 0.5865193924459494, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 16845 + }, + { + "epoch": 0.16846, + "grad_norm": 0.6227814324662859, + "learning_rate": 0.003, + "loss": 4.032, + "step": 16846 + }, + { + "epoch": 0.16847, + "grad_norm": 0.6617255530732737, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 16847 + }, + { + "epoch": 0.16848, + "grad_norm": 0.6272498486594033, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 16848 + }, + { + "epoch": 0.16849, + "grad_norm": 0.5716076057927707, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 16849 + }, + { + "epoch": 0.1685, + "grad_norm": 0.5627584538178395, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 16850 + }, + { + "epoch": 0.16851, + "grad_norm": 0.5135026273189454, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 16851 + }, + { + "epoch": 0.16852, + "grad_norm": 0.5279716847304485, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 16852 + }, + { + "epoch": 0.16853, + "grad_norm": 0.5245548094107838, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16853 + }, + { + "epoch": 0.16854, + "grad_norm": 0.5599026522268685, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 16854 + }, + { + "epoch": 0.16855, + "grad_norm": 0.6639630128768249, + "learning_rate": 0.003, + "loss": 3.999, + "step": 16855 + }, + { + "epoch": 0.16856, + "grad_norm": 0.7007213500952891, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 16856 + }, + { + "epoch": 0.16857, + "grad_norm": 0.7470466622371494, + "learning_rate": 0.003, + "loss": 3.964, + "step": 16857 + }, + { + "epoch": 0.16858, + "grad_norm": 0.8156204401615365, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 16858 + }, + { + "epoch": 0.16859, + "grad_norm": 0.9153029992107364, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 16859 + }, + { + "epoch": 0.1686, + "grad_norm": 0.898679626653985, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 16860 + }, + { + "epoch": 0.16861, + "grad_norm": 0.87610737199239, + "learning_rate": 0.003, + "loss": 4.019, + "step": 16861 + }, + { + "epoch": 0.16862, + "grad_norm": 0.884955282175233, + "learning_rate": 0.003, + "loss": 4.014, + "step": 16862 + }, + { + "epoch": 0.16863, + "grad_norm": 0.8601842220998611, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 16863 + }, + { + "epoch": 0.16864, + "grad_norm": 0.8405760509991281, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 16864 + }, + { + "epoch": 0.16865, + "grad_norm": 0.7399290365966521, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 16865 + }, + { + "epoch": 0.16866, + "grad_norm": 0.6625537814016639, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16866 + }, + { + "epoch": 0.16867, + "grad_norm": 0.6077992203341109, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 16867 + }, + { + "epoch": 0.16868, + "grad_norm": 0.5934373288246633, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 16868 + }, + { + "epoch": 0.16869, + "grad_norm": 0.6849338513113239, + "learning_rate": 0.003, + "loss": 4.002, + "step": 16869 + }, + { + "epoch": 0.1687, + "grad_norm": 0.7510214477087447, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 16870 + }, + { + "epoch": 0.16871, + "grad_norm": 0.782257610084173, + "learning_rate": 0.003, + "loss": 4.027, + "step": 16871 + }, + { + "epoch": 0.16872, + "grad_norm": 0.8731637596955213, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 16872 + }, + { + "epoch": 0.16873, + "grad_norm": 1.004534618176648, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 16873 + }, + { + "epoch": 0.16874, + "grad_norm": 0.9898607091801839, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 16874 + }, + { + "epoch": 0.16875, + "grad_norm": 0.8079107238887434, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 16875 + }, + { + "epoch": 0.16876, + "grad_norm": 0.7852542339274141, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 16876 + }, + { + "epoch": 0.16877, + "grad_norm": 0.7050533156219742, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 16877 + }, + { + "epoch": 0.16878, + "grad_norm": 0.7970136766246021, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 16878 + }, + { + "epoch": 0.16879, + "grad_norm": 0.7632273446179336, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 16879 + }, + { + "epoch": 0.1688, + "grad_norm": 0.7337621151007924, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 16880 + }, + { + "epoch": 0.16881, + "grad_norm": 0.7043838295093796, + "learning_rate": 0.003, + "loss": 4.0, + "step": 16881 + }, + { + "epoch": 0.16882, + "grad_norm": 0.738882177788936, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 16882 + }, + { + "epoch": 0.16883, + "grad_norm": 0.8334859095171616, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 16883 + }, + { + "epoch": 0.16884, + "grad_norm": 0.8495116352071703, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 16884 + }, + { + "epoch": 0.16885, + "grad_norm": 0.7928743654884041, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 16885 + }, + { + "epoch": 0.16886, + "grad_norm": 0.6624612847624027, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 16886 + }, + { + "epoch": 0.16887, + "grad_norm": 0.5894351779334055, + "learning_rate": 0.003, + "loss": 4.033, + "step": 16887 + }, + { + "epoch": 0.16888, + "grad_norm": 0.5889291469171579, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 16888 + }, + { + "epoch": 0.16889, + "grad_norm": 0.6599218805700745, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 16889 + }, + { + "epoch": 0.1689, + "grad_norm": 0.66022598183978, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 16890 + }, + { + "epoch": 0.16891, + "grad_norm": 0.6674568540095174, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 16891 + }, + { + "epoch": 0.16892, + "grad_norm": 0.7183661183710668, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 16892 + }, + { + "epoch": 0.16893, + "grad_norm": 0.7540287771697072, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 16893 + }, + { + "epoch": 0.16894, + "grad_norm": 0.7809037810616588, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 16894 + }, + { + "epoch": 0.16895, + "grad_norm": 0.648181326262796, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 16895 + }, + { + "epoch": 0.16896, + "grad_norm": 0.5506814079433512, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 16896 + }, + { + "epoch": 0.16897, + "grad_norm": 0.5561973696860486, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 16897 + }, + { + "epoch": 0.16898, + "grad_norm": 0.5569219196441896, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 16898 + }, + { + "epoch": 0.16899, + "grad_norm": 0.5494656330266635, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 16899 + }, + { + "epoch": 0.169, + "grad_norm": 0.5846868324904451, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 16900 + }, + { + "epoch": 0.16901, + "grad_norm": 0.6998568733741364, + "learning_rate": 0.003, + "loss": 3.998, + "step": 16901 + }, + { + "epoch": 0.16902, + "grad_norm": 0.8611084575445206, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 16902 + }, + { + "epoch": 0.16903, + "grad_norm": 0.9349673229326824, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 16903 + }, + { + "epoch": 0.16904, + "grad_norm": 0.8623620143455528, + "learning_rate": 0.003, + "loss": 4.0481, + "step": 16904 + }, + { + "epoch": 0.16905, + "grad_norm": 0.8477323441416168, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 16905 + }, + { + "epoch": 0.16906, + "grad_norm": 0.8553168740457334, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16906 + }, + { + "epoch": 0.16907, + "grad_norm": 0.8476110032834518, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 16907 + }, + { + "epoch": 0.16908, + "grad_norm": 0.7990787924987449, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 16908 + }, + { + "epoch": 0.16909, + "grad_norm": 0.7339814070748067, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 16909 + }, + { + "epoch": 0.1691, + "grad_norm": 0.716124863180639, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 16910 + }, + { + "epoch": 0.16911, + "grad_norm": 0.6718598641702379, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 16911 + }, + { + "epoch": 0.16912, + "grad_norm": 0.6611914480900384, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 16912 + }, + { + "epoch": 0.16913, + "grad_norm": 0.6371893872494692, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 16913 + }, + { + "epoch": 0.16914, + "grad_norm": 0.6888263751190603, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 16914 + }, + { + "epoch": 0.16915, + "grad_norm": 0.7606485412925441, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 16915 + }, + { + "epoch": 0.16916, + "grad_norm": 0.7396804108369526, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 16916 + }, + { + "epoch": 0.16917, + "grad_norm": 0.7774871030954023, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 16917 + }, + { + "epoch": 0.16918, + "grad_norm": 0.9095527514966636, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 16918 + }, + { + "epoch": 0.16919, + "grad_norm": 0.9862445661490731, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 16919 + }, + { + "epoch": 0.1692, + "grad_norm": 1.031832216636907, + "learning_rate": 0.003, + "loss": 3.993, + "step": 16920 + }, + { + "epoch": 0.16921, + "grad_norm": 0.8881869969535184, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 16921 + }, + { + "epoch": 0.16922, + "grad_norm": 0.8880463835565062, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 16922 + }, + { + "epoch": 0.16923, + "grad_norm": 0.8649003171155096, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 16923 + }, + { + "epoch": 0.16924, + "grad_norm": 0.8828340346907788, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 16924 + }, + { + "epoch": 0.16925, + "grad_norm": 0.9483788224510387, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 16925 + }, + { + "epoch": 0.16926, + "grad_norm": 1.0465106332890544, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 16926 + }, + { + "epoch": 0.16927, + "grad_norm": 0.9389344312122384, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 16927 + }, + { + "epoch": 0.16928, + "grad_norm": 0.8617979210102559, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 16928 + }, + { + "epoch": 0.16929, + "grad_norm": 0.8152511110455027, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 16929 + }, + { + "epoch": 0.1693, + "grad_norm": 0.7291877297285105, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 16930 + }, + { + "epoch": 0.16931, + "grad_norm": 0.6776667662680718, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 16931 + }, + { + "epoch": 0.16932, + "grad_norm": 0.6281069750538274, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 16932 + }, + { + "epoch": 0.16933, + "grad_norm": 0.6111053668365116, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 16933 + }, + { + "epoch": 0.16934, + "grad_norm": 0.6127988709288319, + "learning_rate": 0.003, + "loss": 4.031, + "step": 16934 + }, + { + "epoch": 0.16935, + "grad_norm": 0.6797020496567745, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 16935 + }, + { + "epoch": 0.16936, + "grad_norm": 0.7339402670949243, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 16936 + }, + { + "epoch": 0.16937, + "grad_norm": 0.8387084481047926, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 16937 + }, + { + "epoch": 0.16938, + "grad_norm": 0.7351077790119371, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 16938 + }, + { + "epoch": 0.16939, + "grad_norm": 0.6692223841930796, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 16939 + }, + { + "epoch": 0.1694, + "grad_norm": 0.6763580631289592, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 16940 + }, + { + "epoch": 0.16941, + "grad_norm": 0.7575155616982289, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 16941 + }, + { + "epoch": 0.16942, + "grad_norm": 0.7119022618618516, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 16942 + }, + { + "epoch": 0.16943, + "grad_norm": 0.6926487456099352, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 16943 + }, + { + "epoch": 0.16944, + "grad_norm": 0.6375838402814343, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 16944 + }, + { + "epoch": 0.16945, + "grad_norm": 0.6030133827257046, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 16945 + }, + { + "epoch": 0.16946, + "grad_norm": 0.551258650593797, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 16946 + }, + { + "epoch": 0.16947, + "grad_norm": 0.5132657448547289, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 16947 + }, + { + "epoch": 0.16948, + "grad_norm": 0.48516608498604774, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 16948 + }, + { + "epoch": 0.16949, + "grad_norm": 0.4594570227558095, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 16949 + }, + { + "epoch": 0.1695, + "grad_norm": 0.469179996465847, + "learning_rate": 0.003, + "loss": 4.008, + "step": 16950 + }, + { + "epoch": 0.16951, + "grad_norm": 0.46630118580100116, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 16951 + }, + { + "epoch": 0.16952, + "grad_norm": 0.4620744426215333, + "learning_rate": 0.003, + "loss": 3.99, + "step": 16952 + }, + { + "epoch": 0.16953, + "grad_norm": 0.51274523386973, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 16953 + }, + { + "epoch": 0.16954, + "grad_norm": 0.6543526074332281, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 16954 + }, + { + "epoch": 0.16955, + "grad_norm": 0.773239597438771, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 16955 + }, + { + "epoch": 0.16956, + "grad_norm": 0.839932114677449, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 16956 + }, + { + "epoch": 0.16957, + "grad_norm": 0.8812247992451209, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 16957 + }, + { + "epoch": 0.16958, + "grad_norm": 0.9710649399248793, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 16958 + }, + { + "epoch": 0.16959, + "grad_norm": 0.8515425920795181, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 16959 + }, + { + "epoch": 0.1696, + "grad_norm": 1.096248211187238, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 16960 + }, + { + "epoch": 0.16961, + "grad_norm": 0.9804473658182166, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 16961 + }, + { + "epoch": 0.16962, + "grad_norm": 1.092408438991426, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 16962 + }, + { + "epoch": 0.16963, + "grad_norm": 0.8872405917354457, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 16963 + }, + { + "epoch": 0.16964, + "grad_norm": 0.8581214028592054, + "learning_rate": 0.003, + "loss": 4.046, + "step": 16964 + }, + { + "epoch": 0.16965, + "grad_norm": 0.7737583304055957, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 16965 + }, + { + "epoch": 0.16966, + "grad_norm": 0.7919887396519223, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 16966 + }, + { + "epoch": 0.16967, + "grad_norm": 0.7743737516042823, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 16967 + }, + { + "epoch": 0.16968, + "grad_norm": 0.8501711416412858, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 16968 + }, + { + "epoch": 0.16969, + "grad_norm": 1.0351780304926104, + "learning_rate": 0.003, + "loss": 4.037, + "step": 16969 + }, + { + "epoch": 0.1697, + "grad_norm": 0.919605561210908, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 16970 + }, + { + "epoch": 0.16971, + "grad_norm": 0.9206960961098278, + "learning_rate": 0.003, + "loss": 4.036, + "step": 16971 + }, + { + "epoch": 0.16972, + "grad_norm": 0.8607780909332682, + "learning_rate": 0.003, + "loss": 4.031, + "step": 16972 + }, + { + "epoch": 0.16973, + "grad_norm": 0.7160068249900132, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 16973 + }, + { + "epoch": 0.16974, + "grad_norm": 0.6679762403378336, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 16974 + }, + { + "epoch": 0.16975, + "grad_norm": 0.7880974761793975, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 16975 + }, + { + "epoch": 0.16976, + "grad_norm": 0.7624998319458187, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 16976 + }, + { + "epoch": 0.16977, + "grad_norm": 0.7209071558149863, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 16977 + }, + { + "epoch": 0.16978, + "grad_norm": 0.6767451351627289, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 16978 + }, + { + "epoch": 0.16979, + "grad_norm": 0.7793066529994811, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 16979 + }, + { + "epoch": 0.1698, + "grad_norm": 0.8423426909106505, + "learning_rate": 0.003, + "loss": 4.008, + "step": 16980 + }, + { + "epoch": 0.16981, + "grad_norm": 0.761568482253797, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 16981 + }, + { + "epoch": 0.16982, + "grad_norm": 0.6253971508055719, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 16982 + }, + { + "epoch": 0.16983, + "grad_norm": 0.6718818076862838, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 16983 + }, + { + "epoch": 0.16984, + "grad_norm": 0.6630387556917772, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 16984 + }, + { + "epoch": 0.16985, + "grad_norm": 0.5988763313765193, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 16985 + }, + { + "epoch": 0.16986, + "grad_norm": 0.5554399995848343, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 16986 + }, + { + "epoch": 0.16987, + "grad_norm": 0.4626548126775766, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 16987 + }, + { + "epoch": 0.16988, + "grad_norm": 0.4931026442435748, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 16988 + }, + { + "epoch": 0.16989, + "grad_norm": 0.4503909167126027, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 16989 + }, + { + "epoch": 0.1699, + "grad_norm": 0.5216321807035523, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 16990 + }, + { + "epoch": 0.16991, + "grad_norm": 0.557276785807371, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 16991 + }, + { + "epoch": 0.16992, + "grad_norm": 0.5735520980217623, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 16992 + }, + { + "epoch": 0.16993, + "grad_norm": 0.559116874616211, + "learning_rate": 0.003, + "loss": 4.013, + "step": 16993 + }, + { + "epoch": 0.16994, + "grad_norm": 0.5593148949519718, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 16994 + }, + { + "epoch": 0.16995, + "grad_norm": 0.5086634464113681, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 16995 + }, + { + "epoch": 0.16996, + "grad_norm": 0.4671755424443961, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 16996 + }, + { + "epoch": 0.16997, + "grad_norm": 0.5803392632205588, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 16997 + }, + { + "epoch": 0.16998, + "grad_norm": 0.73249413437869, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 16998 + }, + { + "epoch": 0.16999, + "grad_norm": 0.8167205869381062, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 16999 + }, + { + "epoch": 0.17, + "grad_norm": 0.8836493914914115, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 17000 + }, + { + "epoch": 0.17001, + "grad_norm": 0.9542663327431696, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 17001 + }, + { + "epoch": 0.17002, + "grad_norm": 1.1057516724718057, + "learning_rate": 0.003, + "loss": 4.035, + "step": 17002 + }, + { + "epoch": 0.17003, + "grad_norm": 0.8076940903584046, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 17003 + }, + { + "epoch": 0.17004, + "grad_norm": 0.6623979568395343, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17004 + }, + { + "epoch": 0.17005, + "grad_norm": 0.682317432212222, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 17005 + }, + { + "epoch": 0.17006, + "grad_norm": 0.6871264128217303, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 17006 + }, + { + "epoch": 0.17007, + "grad_norm": 0.7662487269428042, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 17007 + }, + { + "epoch": 0.17008, + "grad_norm": 0.9066666479977625, + "learning_rate": 0.003, + "loss": 4.043, + "step": 17008 + }, + { + "epoch": 0.17009, + "grad_norm": 0.9734083485278109, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 17009 + }, + { + "epoch": 0.1701, + "grad_norm": 0.9057153350812239, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 17010 + }, + { + "epoch": 0.17011, + "grad_norm": 0.8368153812206399, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 17011 + }, + { + "epoch": 0.17012, + "grad_norm": 0.8200700562335327, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 17012 + }, + { + "epoch": 0.17013, + "grad_norm": 0.8693302597860023, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 17013 + }, + { + "epoch": 0.17014, + "grad_norm": 0.8411504939260394, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 17014 + }, + { + "epoch": 0.17015, + "grad_norm": 0.8376385070618715, + "learning_rate": 0.003, + "loss": 4.011, + "step": 17015 + }, + { + "epoch": 0.17016, + "grad_norm": 0.7430874043614377, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 17016 + }, + { + "epoch": 0.17017, + "grad_norm": 0.6756767725838364, + "learning_rate": 0.003, + "loss": 4.014, + "step": 17017 + }, + { + "epoch": 0.17018, + "grad_norm": 0.6826424106294728, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 17018 + }, + { + "epoch": 0.17019, + "grad_norm": 0.7677174372221247, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 17019 + }, + { + "epoch": 0.1702, + "grad_norm": 0.7301427903205723, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 17020 + }, + { + "epoch": 0.17021, + "grad_norm": 0.6529015043699053, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 17021 + }, + { + "epoch": 0.17022, + "grad_norm": 0.7769853001392928, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 17022 + }, + { + "epoch": 0.17023, + "grad_norm": 0.8019010215925481, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 17023 + }, + { + "epoch": 0.17024, + "grad_norm": 0.8173399625585499, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 17024 + }, + { + "epoch": 0.17025, + "grad_norm": 0.7320622040264967, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 17025 + }, + { + "epoch": 0.17026, + "grad_norm": 0.8370243472563126, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 17026 + }, + { + "epoch": 0.17027, + "grad_norm": 0.8405844442025142, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 17027 + }, + { + "epoch": 0.17028, + "grad_norm": 0.9203922754963085, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 17028 + }, + { + "epoch": 0.17029, + "grad_norm": 1.0965406922371823, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 17029 + }, + { + "epoch": 0.1703, + "grad_norm": 0.7095281872044997, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 17030 + }, + { + "epoch": 0.17031, + "grad_norm": 0.6336486593776273, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 17031 + }, + { + "epoch": 0.17032, + "grad_norm": 0.7843414483971201, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17032 + }, + { + "epoch": 0.17033, + "grad_norm": 0.8205045472128958, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 17033 + }, + { + "epoch": 0.17034, + "grad_norm": 0.8290237227704815, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 17034 + }, + { + "epoch": 0.17035, + "grad_norm": 0.8002354067993146, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 17035 + }, + { + "epoch": 0.17036, + "grad_norm": 0.6699724107534464, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 17036 + }, + { + "epoch": 0.17037, + "grad_norm": 0.5790342197148493, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 17037 + }, + { + "epoch": 0.17038, + "grad_norm": 0.5490089774962841, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17038 + }, + { + "epoch": 0.17039, + "grad_norm": 0.5874575790207482, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 17039 + }, + { + "epoch": 0.1704, + "grad_norm": 0.6562435982420084, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 17040 + }, + { + "epoch": 0.17041, + "grad_norm": 0.6764785873949335, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 17041 + }, + { + "epoch": 0.17042, + "grad_norm": 0.6763977490078303, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 17042 + }, + { + "epoch": 0.17043, + "grad_norm": 0.7183931251006868, + "learning_rate": 0.003, + "loss": 3.992, + "step": 17043 + }, + { + "epoch": 0.17044, + "grad_norm": 0.6466704686220157, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 17044 + }, + { + "epoch": 0.17045, + "grad_norm": 0.6176437079395135, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 17045 + }, + { + "epoch": 0.17046, + "grad_norm": 0.6615467564894798, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17046 + }, + { + "epoch": 0.17047, + "grad_norm": 0.70873318248042, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17047 + }, + { + "epoch": 0.17048, + "grad_norm": 0.6700715624403683, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 17048 + }, + { + "epoch": 0.17049, + "grad_norm": 0.5368187247327539, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 17049 + }, + { + "epoch": 0.1705, + "grad_norm": 0.5441585261132242, + "learning_rate": 0.003, + "loss": 3.988, + "step": 17050 + }, + { + "epoch": 0.17051, + "grad_norm": 0.600913060622111, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 17051 + }, + { + "epoch": 0.17052, + "grad_norm": 0.6684782985382182, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 17052 + }, + { + "epoch": 0.17053, + "grad_norm": 0.6831990626272514, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 17053 + }, + { + "epoch": 0.17054, + "grad_norm": 0.7815684328959158, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 17054 + }, + { + "epoch": 0.17055, + "grad_norm": 0.9704018151971032, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 17055 + }, + { + "epoch": 0.17056, + "grad_norm": 1.094038472038823, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 17056 + }, + { + "epoch": 0.17057, + "grad_norm": 0.7874618441816014, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 17057 + }, + { + "epoch": 0.17058, + "grad_norm": 0.6905755197201305, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 17058 + }, + { + "epoch": 0.17059, + "grad_norm": 0.8513619951833328, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 17059 + }, + { + "epoch": 0.1706, + "grad_norm": 0.7450638982647823, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 17060 + }, + { + "epoch": 0.17061, + "grad_norm": 0.6488580809376148, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 17061 + }, + { + "epoch": 0.17062, + "grad_norm": 0.6081792322695182, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 17062 + }, + { + "epoch": 0.17063, + "grad_norm": 0.6633797947134488, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 17063 + }, + { + "epoch": 0.17064, + "grad_norm": 0.7404783201396364, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 17064 + }, + { + "epoch": 0.17065, + "grad_norm": 0.721790418749709, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 17065 + }, + { + "epoch": 0.17066, + "grad_norm": 0.7934964071795458, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 17066 + }, + { + "epoch": 0.17067, + "grad_norm": 0.7964963287686474, + "learning_rate": 0.003, + "loss": 4.034, + "step": 17067 + }, + { + "epoch": 0.17068, + "grad_norm": 0.8720746505709076, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 17068 + }, + { + "epoch": 0.17069, + "grad_norm": 0.88477573023967, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 17069 + }, + { + "epoch": 0.1707, + "grad_norm": 0.8653673688027542, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 17070 + }, + { + "epoch": 0.17071, + "grad_norm": 0.8827050757246879, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 17071 + }, + { + "epoch": 0.17072, + "grad_norm": 0.8832745864882469, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 17072 + }, + { + "epoch": 0.17073, + "grad_norm": 0.8280272513126309, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 17073 + }, + { + "epoch": 0.17074, + "grad_norm": 0.7528853678109788, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 17074 + }, + { + "epoch": 0.17075, + "grad_norm": 0.8041741336784379, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 17075 + }, + { + "epoch": 0.17076, + "grad_norm": 0.7982671469452584, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17076 + }, + { + "epoch": 0.17077, + "grad_norm": 0.820122799250735, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 17077 + }, + { + "epoch": 0.17078, + "grad_norm": 0.9224797852070756, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 17078 + }, + { + "epoch": 0.17079, + "grad_norm": 0.9064525071827488, + "learning_rate": 0.003, + "loss": 4.028, + "step": 17079 + }, + { + "epoch": 0.1708, + "grad_norm": 0.8402424588024713, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 17080 + }, + { + "epoch": 0.17081, + "grad_norm": 0.8629486123225542, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 17081 + }, + { + "epoch": 0.17082, + "grad_norm": 0.7656959630560087, + "learning_rate": 0.003, + "loss": 4.0582, + "step": 17082 + }, + { + "epoch": 0.17083, + "grad_norm": 0.8184874893455825, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 17083 + }, + { + "epoch": 0.17084, + "grad_norm": 0.7367886049217272, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 17084 + }, + { + "epoch": 0.17085, + "grad_norm": 0.7938352415085668, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 17085 + }, + { + "epoch": 0.17086, + "grad_norm": 0.870302086094702, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 17086 + }, + { + "epoch": 0.17087, + "grad_norm": 0.8563920854711037, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 17087 + }, + { + "epoch": 0.17088, + "grad_norm": 0.7790612645080579, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 17088 + }, + { + "epoch": 0.17089, + "grad_norm": 0.8155196413458076, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17089 + }, + { + "epoch": 0.1709, + "grad_norm": 0.6870233413276922, + "learning_rate": 0.003, + "loss": 4.01, + "step": 17090 + }, + { + "epoch": 0.17091, + "grad_norm": 0.7185784955175307, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 17091 + }, + { + "epoch": 0.17092, + "grad_norm": 0.6952714825699388, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 17092 + }, + { + "epoch": 0.17093, + "grad_norm": 0.6329192671346364, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 17093 + }, + { + "epoch": 0.17094, + "grad_norm": 0.5903137928979486, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17094 + }, + { + "epoch": 0.17095, + "grad_norm": 0.5672875936508815, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 17095 + }, + { + "epoch": 0.17096, + "grad_norm": 0.5491490283710623, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 17096 + }, + { + "epoch": 0.17097, + "grad_norm": 0.5427707288344307, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 17097 + }, + { + "epoch": 0.17098, + "grad_norm": 0.48636541490229224, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 17098 + }, + { + "epoch": 0.17099, + "grad_norm": 0.4557664203881534, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 17099 + }, + { + "epoch": 0.171, + "grad_norm": 0.4781626426268215, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 17100 + }, + { + "epoch": 0.17101, + "grad_norm": 0.5223480173844036, + "learning_rate": 0.003, + "loss": 3.99, + "step": 17101 + }, + { + "epoch": 0.17102, + "grad_norm": 0.6091213839788104, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 17102 + }, + { + "epoch": 0.17103, + "grad_norm": 0.7080999080549218, + "learning_rate": 0.003, + "loss": 4.004, + "step": 17103 + }, + { + "epoch": 0.17104, + "grad_norm": 0.7141054981161705, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 17104 + }, + { + "epoch": 0.17105, + "grad_norm": 0.5568884767350075, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 17105 + }, + { + "epoch": 0.17106, + "grad_norm": 0.5130460013127117, + "learning_rate": 0.003, + "loss": 3.998, + "step": 17106 + }, + { + "epoch": 0.17107, + "grad_norm": 0.5573282010270056, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17107 + }, + { + "epoch": 0.17108, + "grad_norm": 0.615906667828662, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 17108 + }, + { + "epoch": 0.17109, + "grad_norm": 0.6807300883589703, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 17109 + }, + { + "epoch": 0.1711, + "grad_norm": 0.7283268195398905, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 17110 + }, + { + "epoch": 0.17111, + "grad_norm": 0.8222387554491175, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 17111 + }, + { + "epoch": 0.17112, + "grad_norm": 0.8800237890434398, + "learning_rate": 0.003, + "loss": 3.99, + "step": 17112 + }, + { + "epoch": 0.17113, + "grad_norm": 0.8064671683206728, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 17113 + }, + { + "epoch": 0.17114, + "grad_norm": 0.6726728111830116, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 17114 + }, + { + "epoch": 0.17115, + "grad_norm": 0.5915848799380349, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 17115 + }, + { + "epoch": 0.17116, + "grad_norm": 0.5924313260269027, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 17116 + }, + { + "epoch": 0.17117, + "grad_norm": 0.6328504893661724, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 17117 + }, + { + "epoch": 0.17118, + "grad_norm": 0.7204589050458863, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 17118 + }, + { + "epoch": 0.17119, + "grad_norm": 0.851016902090998, + "learning_rate": 0.003, + "loss": 4.017, + "step": 17119 + }, + { + "epoch": 0.1712, + "grad_norm": 0.8483436628527788, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 17120 + }, + { + "epoch": 0.17121, + "grad_norm": 0.8635094519133777, + "learning_rate": 0.003, + "loss": 4.008, + "step": 17121 + }, + { + "epoch": 0.17122, + "grad_norm": 0.9372514011026325, + "learning_rate": 0.003, + "loss": 4.0, + "step": 17122 + }, + { + "epoch": 0.17123, + "grad_norm": 0.9570031089192139, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17123 + }, + { + "epoch": 0.17124, + "grad_norm": 1.0105236067932282, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 17124 + }, + { + "epoch": 0.17125, + "grad_norm": 0.9570756025960214, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 17125 + }, + { + "epoch": 0.17126, + "grad_norm": 0.7552632855115113, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 17126 + }, + { + "epoch": 0.17127, + "grad_norm": 0.6893675800600579, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 17127 + }, + { + "epoch": 0.17128, + "grad_norm": 0.7735801611576151, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 17128 + }, + { + "epoch": 0.17129, + "grad_norm": 0.9842513440930503, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 17129 + }, + { + "epoch": 0.1713, + "grad_norm": 1.026283687577696, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 17130 + }, + { + "epoch": 0.17131, + "grad_norm": 0.8560009989407565, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 17131 + }, + { + "epoch": 0.17132, + "grad_norm": 0.7042923831252912, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 17132 + }, + { + "epoch": 0.17133, + "grad_norm": 0.8151371434903343, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 17133 + }, + { + "epoch": 0.17134, + "grad_norm": 0.7762840729689812, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 17134 + }, + { + "epoch": 0.17135, + "grad_norm": 0.7859754874831201, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 17135 + }, + { + "epoch": 0.17136, + "grad_norm": 0.836563060671449, + "learning_rate": 0.003, + "loss": 4.008, + "step": 17136 + }, + { + "epoch": 0.17137, + "grad_norm": 0.8974281875473841, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 17137 + }, + { + "epoch": 0.17138, + "grad_norm": 0.8294026624955068, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 17138 + }, + { + "epoch": 0.17139, + "grad_norm": 0.9136124912127187, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 17139 + }, + { + "epoch": 0.1714, + "grad_norm": 1.0730555422143557, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 17140 + }, + { + "epoch": 0.17141, + "grad_norm": 0.9521211227348815, + "learning_rate": 0.003, + "loss": 4.027, + "step": 17141 + }, + { + "epoch": 0.17142, + "grad_norm": 0.9024514651609155, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 17142 + }, + { + "epoch": 0.17143, + "grad_norm": 0.8400309143275912, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 17143 + }, + { + "epoch": 0.17144, + "grad_norm": 0.6986977239596697, + "learning_rate": 0.003, + "loss": 4.017, + "step": 17144 + }, + { + "epoch": 0.17145, + "grad_norm": 0.7155100771248984, + "learning_rate": 0.003, + "loss": 4.004, + "step": 17145 + }, + { + "epoch": 0.17146, + "grad_norm": 0.6976085352212389, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 17146 + }, + { + "epoch": 0.17147, + "grad_norm": 0.7359115290629179, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 17147 + }, + { + "epoch": 0.17148, + "grad_norm": 0.7658308499325717, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 17148 + }, + { + "epoch": 0.17149, + "grad_norm": 0.7989599516270722, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 17149 + }, + { + "epoch": 0.1715, + "grad_norm": 0.8507968589111446, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 17150 + }, + { + "epoch": 0.17151, + "grad_norm": 0.6783746295949759, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 17151 + }, + { + "epoch": 0.17152, + "grad_norm": 0.5627501109519509, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 17152 + }, + { + "epoch": 0.17153, + "grad_norm": 0.5704108582911749, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 17153 + }, + { + "epoch": 0.17154, + "grad_norm": 0.5147904613294461, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17154 + }, + { + "epoch": 0.17155, + "grad_norm": 0.48249163575573123, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 17155 + }, + { + "epoch": 0.17156, + "grad_norm": 0.463974938278209, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 17156 + }, + { + "epoch": 0.17157, + "grad_norm": 0.4526681479262139, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17157 + }, + { + "epoch": 0.17158, + "grad_norm": 0.4682335216431008, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 17158 + }, + { + "epoch": 0.17159, + "grad_norm": 0.49723718532761124, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17159 + }, + { + "epoch": 0.1716, + "grad_norm": 0.5765540346966052, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 17160 + }, + { + "epoch": 0.17161, + "grad_norm": 0.6547271517156172, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 17161 + }, + { + "epoch": 0.17162, + "grad_norm": 0.7486899415028669, + "learning_rate": 0.003, + "loss": 4.033, + "step": 17162 + }, + { + "epoch": 0.17163, + "grad_norm": 0.8450393777805939, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 17163 + }, + { + "epoch": 0.17164, + "grad_norm": 0.8366061294038614, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 17164 + }, + { + "epoch": 0.17165, + "grad_norm": 0.7778930988600739, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 17165 + }, + { + "epoch": 0.17166, + "grad_norm": 0.8402100721873467, + "learning_rate": 0.003, + "loss": 4.013, + "step": 17166 + }, + { + "epoch": 0.17167, + "grad_norm": 0.8834496787665719, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 17167 + }, + { + "epoch": 0.17168, + "grad_norm": 0.8475598650863471, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 17168 + }, + { + "epoch": 0.17169, + "grad_norm": 0.7722976229802555, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 17169 + }, + { + "epoch": 0.1717, + "grad_norm": 0.7079412049272558, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 17170 + }, + { + "epoch": 0.17171, + "grad_norm": 0.6889957183263626, + "learning_rate": 0.003, + "loss": 4.012, + "step": 17171 + }, + { + "epoch": 0.17172, + "grad_norm": 0.7110572144291701, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 17172 + }, + { + "epoch": 0.17173, + "grad_norm": 0.6688989938002743, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 17173 + }, + { + "epoch": 0.17174, + "grad_norm": 0.741488844343974, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 17174 + }, + { + "epoch": 0.17175, + "grad_norm": 0.7828060254322132, + "learning_rate": 0.003, + "loss": 4.01, + "step": 17175 + }, + { + "epoch": 0.17176, + "grad_norm": 0.7842563471043481, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 17176 + }, + { + "epoch": 0.17177, + "grad_norm": 0.6824338973894194, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 17177 + }, + { + "epoch": 0.17178, + "grad_norm": 0.7364629657218406, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 17178 + }, + { + "epoch": 0.17179, + "grad_norm": 0.7457972604265986, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 17179 + }, + { + "epoch": 0.1718, + "grad_norm": 0.7262323757539157, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 17180 + }, + { + "epoch": 0.17181, + "grad_norm": 0.7157704691590209, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 17181 + }, + { + "epoch": 0.17182, + "grad_norm": 0.7241343578686508, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 17182 + }, + { + "epoch": 0.17183, + "grad_norm": 0.7288985426738073, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17183 + }, + { + "epoch": 0.17184, + "grad_norm": 0.7013294956213563, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 17184 + }, + { + "epoch": 0.17185, + "grad_norm": 0.7069261133585735, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 17185 + }, + { + "epoch": 0.17186, + "grad_norm": 0.7600361596049441, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 17186 + }, + { + "epoch": 0.17187, + "grad_norm": 0.8497343319582621, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 17187 + }, + { + "epoch": 0.17188, + "grad_norm": 0.8650061062830031, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 17188 + }, + { + "epoch": 0.17189, + "grad_norm": 0.8958194168450007, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 17189 + }, + { + "epoch": 0.1719, + "grad_norm": 0.9483144940390243, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 17190 + }, + { + "epoch": 0.17191, + "grad_norm": 0.9706061040955183, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 17191 + }, + { + "epoch": 0.17192, + "grad_norm": 0.9132717452704538, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 17192 + }, + { + "epoch": 0.17193, + "grad_norm": 0.8182599228553088, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17193 + }, + { + "epoch": 0.17194, + "grad_norm": 0.8430017021211981, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 17194 + }, + { + "epoch": 0.17195, + "grad_norm": 0.7352615257172569, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 17195 + }, + { + "epoch": 0.17196, + "grad_norm": 0.7527260792387832, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 17196 + }, + { + "epoch": 0.17197, + "grad_norm": 0.7738577022852078, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17197 + }, + { + "epoch": 0.17198, + "grad_norm": 0.7231025784319881, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 17198 + }, + { + "epoch": 0.17199, + "grad_norm": 0.6791634160000043, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 17199 + }, + { + "epoch": 0.172, + "grad_norm": 0.5860904375504952, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 17200 + }, + { + "epoch": 0.17201, + "grad_norm": 0.5301360072617385, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 17201 + }, + { + "epoch": 0.17202, + "grad_norm": 0.5117601621667984, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 17202 + }, + { + "epoch": 0.17203, + "grad_norm": 0.54620661027039, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 17203 + }, + { + "epoch": 0.17204, + "grad_norm": 0.6170242658029927, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 17204 + }, + { + "epoch": 0.17205, + "grad_norm": 0.7342835666313463, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 17205 + }, + { + "epoch": 0.17206, + "grad_norm": 0.8721126627908405, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 17206 + }, + { + "epoch": 0.17207, + "grad_norm": 0.856240671368542, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 17207 + }, + { + "epoch": 0.17208, + "grad_norm": 0.7769863678580673, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 17208 + }, + { + "epoch": 0.17209, + "grad_norm": 0.7570203383218392, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 17209 + }, + { + "epoch": 0.1721, + "grad_norm": 0.7179311331210521, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 17210 + }, + { + "epoch": 0.17211, + "grad_norm": 0.6365547829587447, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 17211 + }, + { + "epoch": 0.17212, + "grad_norm": 0.6684733662486948, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 17212 + }, + { + "epoch": 0.17213, + "grad_norm": 0.5958094136668693, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 17213 + }, + { + "epoch": 0.17214, + "grad_norm": 0.5209059403859658, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 17214 + }, + { + "epoch": 0.17215, + "grad_norm": 0.5782940091838192, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 17215 + }, + { + "epoch": 0.17216, + "grad_norm": 0.5616892096559163, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 17216 + }, + { + "epoch": 0.17217, + "grad_norm": 0.5392997850614464, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 17217 + }, + { + "epoch": 0.17218, + "grad_norm": 0.5838709100091114, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 17218 + }, + { + "epoch": 0.17219, + "grad_norm": 0.6120115202066513, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 17219 + }, + { + "epoch": 0.1722, + "grad_norm": 0.5966082038161974, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 17220 + }, + { + "epoch": 0.17221, + "grad_norm": 0.5382626335304387, + "learning_rate": 0.003, + "loss": 4.001, + "step": 17221 + }, + { + "epoch": 0.17222, + "grad_norm": 0.61367787972728, + "learning_rate": 0.003, + "loss": 4.003, + "step": 17222 + }, + { + "epoch": 0.17223, + "grad_norm": 0.7442342185951052, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 17223 + }, + { + "epoch": 0.17224, + "grad_norm": 0.9141777623175399, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 17224 + }, + { + "epoch": 0.17225, + "grad_norm": 1.1580774068643465, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 17225 + }, + { + "epoch": 0.17226, + "grad_norm": 0.8576062622582994, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 17226 + }, + { + "epoch": 0.17227, + "grad_norm": 0.7290036120549149, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 17227 + }, + { + "epoch": 0.17228, + "grad_norm": 0.6501537257212758, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 17228 + }, + { + "epoch": 0.17229, + "grad_norm": 0.7245028096396091, + "learning_rate": 0.003, + "loss": 4.015, + "step": 17229 + }, + { + "epoch": 0.1723, + "grad_norm": 0.8554471026755649, + "learning_rate": 0.003, + "loss": 4.018, + "step": 17230 + }, + { + "epoch": 0.17231, + "grad_norm": 0.880550570308038, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 17231 + }, + { + "epoch": 0.17232, + "grad_norm": 0.8396039037047295, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 17232 + }, + { + "epoch": 0.17233, + "grad_norm": 0.911819733687189, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 17233 + }, + { + "epoch": 0.17234, + "grad_norm": 0.9811859882989339, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 17234 + }, + { + "epoch": 0.17235, + "grad_norm": 1.0200583631652882, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 17235 + }, + { + "epoch": 0.17236, + "grad_norm": 0.8849979604774756, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 17236 + }, + { + "epoch": 0.17237, + "grad_norm": 0.8098535993772426, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 17237 + }, + { + "epoch": 0.17238, + "grad_norm": 0.8230015041694175, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 17238 + }, + { + "epoch": 0.17239, + "grad_norm": 0.9034729014127706, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 17239 + }, + { + "epoch": 0.1724, + "grad_norm": 0.9768744106099628, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 17240 + }, + { + "epoch": 0.17241, + "grad_norm": 0.8180593629819266, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 17241 + }, + { + "epoch": 0.17242, + "grad_norm": 0.8163884384810415, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17242 + }, + { + "epoch": 0.17243, + "grad_norm": 0.9664534475964014, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 17243 + }, + { + "epoch": 0.17244, + "grad_norm": 1.0032422570808268, + "learning_rate": 0.003, + "loss": 4.0659, + "step": 17244 + }, + { + "epoch": 0.17245, + "grad_norm": 0.8918531049472689, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 17245 + }, + { + "epoch": 0.17246, + "grad_norm": 0.8277744702758794, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 17246 + }, + { + "epoch": 0.17247, + "grad_norm": 0.8288137570951843, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 17247 + }, + { + "epoch": 0.17248, + "grad_norm": 0.879018548996737, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 17248 + }, + { + "epoch": 0.17249, + "grad_norm": 0.9178216110577104, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 17249 + }, + { + "epoch": 0.1725, + "grad_norm": 0.8250875692841292, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 17250 + }, + { + "epoch": 0.17251, + "grad_norm": 0.6478073355349894, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 17251 + }, + { + "epoch": 0.17252, + "grad_norm": 0.6635198643957008, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 17252 + }, + { + "epoch": 0.17253, + "grad_norm": 0.6242006808339946, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 17253 + }, + { + "epoch": 0.17254, + "grad_norm": 0.5948841915694275, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 17254 + }, + { + "epoch": 0.17255, + "grad_norm": 0.5737209038087729, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 17255 + }, + { + "epoch": 0.17256, + "grad_norm": 0.5285150849146768, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 17256 + }, + { + "epoch": 0.17257, + "grad_norm": 0.5806481854147068, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 17257 + }, + { + "epoch": 0.17258, + "grad_norm": 0.6560609200149118, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 17258 + }, + { + "epoch": 0.17259, + "grad_norm": 0.7188899768924045, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 17259 + }, + { + "epoch": 0.1726, + "grad_norm": 0.8175613106315833, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 17260 + }, + { + "epoch": 0.17261, + "grad_norm": 0.7405024764682759, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 17261 + }, + { + "epoch": 0.17262, + "grad_norm": 0.6224292113402982, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 17262 + }, + { + "epoch": 0.17263, + "grad_norm": 0.5955912470642905, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 17263 + }, + { + "epoch": 0.17264, + "grad_norm": 0.5768824766706631, + "learning_rate": 0.003, + "loss": 3.984, + "step": 17264 + }, + { + "epoch": 0.17265, + "grad_norm": 0.5124948833070845, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 17265 + }, + { + "epoch": 0.17266, + "grad_norm": 0.5669447473083066, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 17266 + }, + { + "epoch": 0.17267, + "grad_norm": 0.540922570706934, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 17267 + }, + { + "epoch": 0.17268, + "grad_norm": 0.484813462824433, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 17268 + }, + { + "epoch": 0.17269, + "grad_norm": 0.46556420783973895, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 17269 + }, + { + "epoch": 0.1727, + "grad_norm": 0.5351363779961413, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 17270 + }, + { + "epoch": 0.17271, + "grad_norm": 0.57817258746467, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 17271 + }, + { + "epoch": 0.17272, + "grad_norm": 0.7444938829894141, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 17272 + }, + { + "epoch": 0.17273, + "grad_norm": 0.9432961313665126, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 17273 + }, + { + "epoch": 0.17274, + "grad_norm": 1.0774766867881174, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 17274 + }, + { + "epoch": 0.17275, + "grad_norm": 0.5891086458165826, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 17275 + }, + { + "epoch": 0.17276, + "grad_norm": 0.6364512440882404, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 17276 + }, + { + "epoch": 0.17277, + "grad_norm": 0.823503517851109, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 17277 + }, + { + "epoch": 0.17278, + "grad_norm": 0.9344895635966044, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 17278 + }, + { + "epoch": 0.17279, + "grad_norm": 0.8886420430601134, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 17279 + }, + { + "epoch": 0.1728, + "grad_norm": 0.8760128696906585, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 17280 + }, + { + "epoch": 0.17281, + "grad_norm": 0.8658982971084332, + "learning_rate": 0.003, + "loss": 3.97, + "step": 17281 + }, + { + "epoch": 0.17282, + "grad_norm": 0.8556317286379161, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 17282 + }, + { + "epoch": 0.17283, + "grad_norm": 0.9889502767821967, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 17283 + }, + { + "epoch": 0.17284, + "grad_norm": 1.0472449367122711, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 17284 + }, + { + "epoch": 0.17285, + "grad_norm": 0.9660402018268783, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 17285 + }, + { + "epoch": 0.17286, + "grad_norm": 0.964694555188517, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 17286 + }, + { + "epoch": 0.17287, + "grad_norm": 0.9537508489437995, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 17287 + }, + { + "epoch": 0.17288, + "grad_norm": 1.0143809289208703, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 17288 + }, + { + "epoch": 0.17289, + "grad_norm": 0.9092400139301646, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 17289 + }, + { + "epoch": 0.1729, + "grad_norm": 0.9525845202267462, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 17290 + }, + { + "epoch": 0.17291, + "grad_norm": 0.9236858844658075, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 17291 + }, + { + "epoch": 0.17292, + "grad_norm": 0.8516059697355082, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 17292 + }, + { + "epoch": 0.17293, + "grad_norm": 0.8463611073430487, + "learning_rate": 0.003, + "loss": 4.035, + "step": 17293 + }, + { + "epoch": 0.17294, + "grad_norm": 0.642461351448897, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 17294 + }, + { + "epoch": 0.17295, + "grad_norm": 0.6533832385016205, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 17295 + }, + { + "epoch": 0.17296, + "grad_norm": 0.6290469077426999, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 17296 + }, + { + "epoch": 0.17297, + "grad_norm": 0.6700351287343538, + "learning_rate": 0.003, + "loss": 4.034, + "step": 17297 + }, + { + "epoch": 0.17298, + "grad_norm": 0.7573872286723606, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 17298 + }, + { + "epoch": 0.17299, + "grad_norm": 0.7077765596456906, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 17299 + }, + { + "epoch": 0.173, + "grad_norm": 0.6632730752477496, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 17300 + }, + { + "epoch": 0.17301, + "grad_norm": 0.6791214462065852, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 17301 + }, + { + "epoch": 0.17302, + "grad_norm": 0.737893077409774, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 17302 + }, + { + "epoch": 0.17303, + "grad_norm": 0.7365543023801143, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 17303 + }, + { + "epoch": 0.17304, + "grad_norm": 0.7239152435601959, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 17304 + }, + { + "epoch": 0.17305, + "grad_norm": 0.6864970510836362, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 17305 + }, + { + "epoch": 0.17306, + "grad_norm": 0.5821041688566785, + "learning_rate": 0.003, + "loss": 3.986, + "step": 17306 + }, + { + "epoch": 0.17307, + "grad_norm": 0.48534323234179044, + "learning_rate": 0.003, + "loss": 4.005, + "step": 17307 + }, + { + "epoch": 0.17308, + "grad_norm": 0.5743193588746298, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 17308 + }, + { + "epoch": 0.17309, + "grad_norm": 0.6242888591401119, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 17309 + }, + { + "epoch": 0.1731, + "grad_norm": 0.7648206589643427, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 17310 + }, + { + "epoch": 0.17311, + "grad_norm": 0.8363321542967938, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 17311 + }, + { + "epoch": 0.17312, + "grad_norm": 0.9592859862457663, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 17312 + }, + { + "epoch": 0.17313, + "grad_norm": 0.9037033563405308, + "learning_rate": 0.003, + "loss": 4.023, + "step": 17313 + }, + { + "epoch": 0.17314, + "grad_norm": 0.7166434249647443, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 17314 + }, + { + "epoch": 0.17315, + "grad_norm": 0.8231433673432491, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 17315 + }, + { + "epoch": 0.17316, + "grad_norm": 0.9356336368645922, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 17316 + }, + { + "epoch": 0.17317, + "grad_norm": 0.8572593896728062, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 17317 + }, + { + "epoch": 0.17318, + "grad_norm": 0.7027559521205037, + "learning_rate": 0.003, + "loss": 4.042, + "step": 17318 + }, + { + "epoch": 0.17319, + "grad_norm": 0.6454890766365828, + "learning_rate": 0.003, + "loss": 3.989, + "step": 17319 + }, + { + "epoch": 0.1732, + "grad_norm": 0.6928438432908273, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 17320 + }, + { + "epoch": 0.17321, + "grad_norm": 0.7188343798941086, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 17321 + }, + { + "epoch": 0.17322, + "grad_norm": 0.6544706123010271, + "learning_rate": 0.003, + "loss": 4.012, + "step": 17322 + }, + { + "epoch": 0.17323, + "grad_norm": 0.6216632692943628, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 17323 + }, + { + "epoch": 0.17324, + "grad_norm": 0.7155704366894524, + "learning_rate": 0.003, + "loss": 4.029, + "step": 17324 + }, + { + "epoch": 0.17325, + "grad_norm": 0.6832224672046081, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 17325 + }, + { + "epoch": 0.17326, + "grad_norm": 0.5591365121868614, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 17326 + }, + { + "epoch": 0.17327, + "grad_norm": 0.49555972671476006, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 17327 + }, + { + "epoch": 0.17328, + "grad_norm": 0.4498983602939436, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 17328 + }, + { + "epoch": 0.17329, + "grad_norm": 0.4333774222302085, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 17329 + }, + { + "epoch": 0.1733, + "grad_norm": 0.47850846848949224, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 17330 + }, + { + "epoch": 0.17331, + "grad_norm": 0.5216613755344367, + "learning_rate": 0.003, + "loss": 4.002, + "step": 17331 + }, + { + "epoch": 0.17332, + "grad_norm": 0.6461743097681549, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 17332 + }, + { + "epoch": 0.17333, + "grad_norm": 0.7675204212760159, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 17333 + }, + { + "epoch": 0.17334, + "grad_norm": 0.7590253264916148, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 17334 + }, + { + "epoch": 0.17335, + "grad_norm": 0.6768328478481178, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 17335 + }, + { + "epoch": 0.17336, + "grad_norm": 0.6007659025478136, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 17336 + }, + { + "epoch": 0.17337, + "grad_norm": 0.58678497759597, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17337 + }, + { + "epoch": 0.17338, + "grad_norm": 0.5523915843893515, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 17338 + }, + { + "epoch": 0.17339, + "grad_norm": 0.5894691272935666, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 17339 + }, + { + "epoch": 0.1734, + "grad_norm": 0.62054629713732, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 17340 + }, + { + "epoch": 0.17341, + "grad_norm": 0.7724447327020705, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 17341 + }, + { + "epoch": 0.17342, + "grad_norm": 0.9811129469080782, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 17342 + }, + { + "epoch": 0.17343, + "grad_norm": 1.2145773037105163, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 17343 + }, + { + "epoch": 0.17344, + "grad_norm": 0.6641174299954824, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 17344 + }, + { + "epoch": 0.17345, + "grad_norm": 0.7473125425382388, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 17345 + }, + { + "epoch": 0.17346, + "grad_norm": 0.8077341264714073, + "learning_rate": 0.003, + "loss": 4.017, + "step": 17346 + }, + { + "epoch": 0.17347, + "grad_norm": 0.7995328258631837, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 17347 + }, + { + "epoch": 0.17348, + "grad_norm": 0.7596285954845242, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 17348 + }, + { + "epoch": 0.17349, + "grad_norm": 0.7982518641949579, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17349 + }, + { + "epoch": 0.1735, + "grad_norm": 0.6901120822743435, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 17350 + }, + { + "epoch": 0.17351, + "grad_norm": 0.68649442558484, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17351 + }, + { + "epoch": 0.17352, + "grad_norm": 0.5685181998445901, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 17352 + }, + { + "epoch": 0.17353, + "grad_norm": 0.5567423659908116, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 17353 + }, + { + "epoch": 0.17354, + "grad_norm": 0.5828596142083001, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17354 + }, + { + "epoch": 0.17355, + "grad_norm": 0.6918408907916088, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 17355 + }, + { + "epoch": 0.17356, + "grad_norm": 0.7943251906494516, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 17356 + }, + { + "epoch": 0.17357, + "grad_norm": 0.7842436291833521, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 17357 + }, + { + "epoch": 0.17358, + "grad_norm": 0.8199167871264214, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 17358 + }, + { + "epoch": 0.17359, + "grad_norm": 0.7961921656898033, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 17359 + }, + { + "epoch": 0.1736, + "grad_norm": 0.801825935283319, + "learning_rate": 0.003, + "loss": 4.003, + "step": 17360 + }, + { + "epoch": 0.17361, + "grad_norm": 0.8717803089846826, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 17361 + }, + { + "epoch": 0.17362, + "grad_norm": 0.8249525310716522, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 17362 + }, + { + "epoch": 0.17363, + "grad_norm": 0.7136988841882079, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 17363 + }, + { + "epoch": 0.17364, + "grad_norm": 0.8865568619294606, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 17364 + }, + { + "epoch": 0.17365, + "grad_norm": 0.8799848673523597, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 17365 + }, + { + "epoch": 0.17366, + "grad_norm": 0.9084661150436764, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 17366 + }, + { + "epoch": 0.17367, + "grad_norm": 0.8479114577599507, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 17367 + }, + { + "epoch": 0.17368, + "grad_norm": 0.9914067579437349, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 17368 + }, + { + "epoch": 0.17369, + "grad_norm": 1.1397720003478693, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 17369 + }, + { + "epoch": 0.1737, + "grad_norm": 0.8125251187788938, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 17370 + }, + { + "epoch": 0.17371, + "grad_norm": 0.6689438821129744, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 17371 + }, + { + "epoch": 0.17372, + "grad_norm": 0.6846480274682645, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 17372 + }, + { + "epoch": 0.17373, + "grad_norm": 0.6862052881945565, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 17373 + }, + { + "epoch": 0.17374, + "grad_norm": 0.7772035591981205, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 17374 + }, + { + "epoch": 0.17375, + "grad_norm": 0.9448368819424854, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 17375 + }, + { + "epoch": 0.17376, + "grad_norm": 0.919455468486953, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 17376 + }, + { + "epoch": 0.17377, + "grad_norm": 0.8021782498719294, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 17377 + }, + { + "epoch": 0.17378, + "grad_norm": 0.8086116829630967, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 17378 + }, + { + "epoch": 0.17379, + "grad_norm": 0.7721847262022172, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 17379 + }, + { + "epoch": 0.1738, + "grad_norm": 0.6257721445098505, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 17380 + }, + { + "epoch": 0.17381, + "grad_norm": 0.6567759442693017, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 17381 + }, + { + "epoch": 0.17382, + "grad_norm": 0.6378073355118485, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 17382 + }, + { + "epoch": 0.17383, + "grad_norm": 0.6667647472179008, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 17383 + }, + { + "epoch": 0.17384, + "grad_norm": 0.5833443098276588, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 17384 + }, + { + "epoch": 0.17385, + "grad_norm": 0.6213330721741878, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 17385 + }, + { + "epoch": 0.17386, + "grad_norm": 0.6606426476961121, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 17386 + }, + { + "epoch": 0.17387, + "grad_norm": 0.7426681589146831, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17387 + }, + { + "epoch": 0.17388, + "grad_norm": 0.7818296396812834, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 17388 + }, + { + "epoch": 0.17389, + "grad_norm": 0.7061885312399905, + "learning_rate": 0.003, + "loss": 4.002, + "step": 17389 + }, + { + "epoch": 0.1739, + "grad_norm": 0.7669526565169381, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 17390 + }, + { + "epoch": 0.17391, + "grad_norm": 0.8200147824823788, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 17391 + }, + { + "epoch": 0.17392, + "grad_norm": 0.6974094902734338, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 17392 + }, + { + "epoch": 0.17393, + "grad_norm": 0.6319729399271037, + "learning_rate": 0.003, + "loss": 4.011, + "step": 17393 + }, + { + "epoch": 0.17394, + "grad_norm": 0.5664427424445811, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 17394 + }, + { + "epoch": 0.17395, + "grad_norm": 0.5670010577018194, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 17395 + }, + { + "epoch": 0.17396, + "grad_norm": 0.5785981844160106, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 17396 + }, + { + "epoch": 0.17397, + "grad_norm": 0.6837518859163503, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 17397 + }, + { + "epoch": 0.17398, + "grad_norm": 0.6792821085077131, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 17398 + }, + { + "epoch": 0.17399, + "grad_norm": 0.6008807461074882, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 17399 + }, + { + "epoch": 0.174, + "grad_norm": 0.5933358169137942, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17400 + }, + { + "epoch": 0.17401, + "grad_norm": 0.6441452816585017, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 17401 + }, + { + "epoch": 0.17402, + "grad_norm": 0.6956311319547491, + "learning_rate": 0.003, + "loss": 3.992, + "step": 17402 + }, + { + "epoch": 0.17403, + "grad_norm": 0.744284683929734, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 17403 + }, + { + "epoch": 0.17404, + "grad_norm": 0.7052138579913614, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 17404 + }, + { + "epoch": 0.17405, + "grad_norm": 0.6553948193601165, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 17405 + }, + { + "epoch": 0.17406, + "grad_norm": 0.7430020851926057, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 17406 + }, + { + "epoch": 0.17407, + "grad_norm": 0.9099518678117927, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 17407 + }, + { + "epoch": 0.17408, + "grad_norm": 1.0709322625675612, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 17408 + }, + { + "epoch": 0.17409, + "grad_norm": 0.9803215820026611, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 17409 + }, + { + "epoch": 0.1741, + "grad_norm": 0.8984529116541871, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 17410 + }, + { + "epoch": 0.17411, + "grad_norm": 0.7684695191996459, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17411 + }, + { + "epoch": 0.17412, + "grad_norm": 0.8278878508695315, + "learning_rate": 0.003, + "loss": 4.009, + "step": 17412 + }, + { + "epoch": 0.17413, + "grad_norm": 0.8538949364635449, + "learning_rate": 0.003, + "loss": 3.988, + "step": 17413 + }, + { + "epoch": 0.17414, + "grad_norm": 0.8546727355243328, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 17414 + }, + { + "epoch": 0.17415, + "grad_norm": 0.9226746552393756, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 17415 + }, + { + "epoch": 0.17416, + "grad_norm": 1.032752728283472, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 17416 + }, + { + "epoch": 0.17417, + "grad_norm": 0.9308387740774873, + "learning_rate": 0.003, + "loss": 4.022, + "step": 17417 + }, + { + "epoch": 0.17418, + "grad_norm": 0.9989926087791436, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 17418 + }, + { + "epoch": 0.17419, + "grad_norm": 1.0930269803427097, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 17419 + }, + { + "epoch": 0.1742, + "grad_norm": 0.9632950415089103, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 17420 + }, + { + "epoch": 0.17421, + "grad_norm": 0.8423738729183186, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 17421 + }, + { + "epoch": 0.17422, + "grad_norm": 0.6602972419628241, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 17422 + }, + { + "epoch": 0.17423, + "grad_norm": 0.601278775208128, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 17423 + }, + { + "epoch": 0.17424, + "grad_norm": 0.6730398514953155, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17424 + }, + { + "epoch": 0.17425, + "grad_norm": 0.5923091709888401, + "learning_rate": 0.003, + "loss": 4.025, + "step": 17425 + }, + { + "epoch": 0.17426, + "grad_norm": 0.6047230539476618, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 17426 + }, + { + "epoch": 0.17427, + "grad_norm": 0.5629660846883906, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 17427 + }, + { + "epoch": 0.17428, + "grad_norm": 0.5988016418271748, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 17428 + }, + { + "epoch": 0.17429, + "grad_norm": 0.7340014912685512, + "learning_rate": 0.003, + "loss": 4.0454, + "step": 17429 + }, + { + "epoch": 0.1743, + "grad_norm": 0.8097390366204548, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 17430 + }, + { + "epoch": 0.17431, + "grad_norm": 0.7596846975556525, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 17431 + }, + { + "epoch": 0.17432, + "grad_norm": 0.7395939244576796, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 17432 + }, + { + "epoch": 0.17433, + "grad_norm": 0.6649074469381537, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 17433 + }, + { + "epoch": 0.17434, + "grad_norm": 0.7005980165766259, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 17434 + }, + { + "epoch": 0.17435, + "grad_norm": 0.7674416751303273, + "learning_rate": 0.003, + "loss": 4.03, + "step": 17435 + }, + { + "epoch": 0.17436, + "grad_norm": 0.7843912784205122, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 17436 + }, + { + "epoch": 0.17437, + "grad_norm": 0.7975598638436348, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17437 + }, + { + "epoch": 0.17438, + "grad_norm": 0.876168183703435, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 17438 + }, + { + "epoch": 0.17439, + "grad_norm": 0.9677930263088985, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 17439 + }, + { + "epoch": 0.1744, + "grad_norm": 0.8899064501665144, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 17440 + }, + { + "epoch": 0.17441, + "grad_norm": 0.8023607090542267, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 17441 + }, + { + "epoch": 0.17442, + "grad_norm": 0.7513593691422762, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 17442 + }, + { + "epoch": 0.17443, + "grad_norm": 0.6882883050124783, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17443 + }, + { + "epoch": 0.17444, + "grad_norm": 0.6607864842062491, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 17444 + }, + { + "epoch": 0.17445, + "grad_norm": 0.6448136761604508, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 17445 + }, + { + "epoch": 0.17446, + "grad_norm": 0.7500022748038454, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 17446 + }, + { + "epoch": 0.17447, + "grad_norm": 0.9768968941769048, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 17447 + }, + { + "epoch": 0.17448, + "grad_norm": 0.9244606122337158, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 17448 + }, + { + "epoch": 0.17449, + "grad_norm": 0.6675362234199415, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 17449 + }, + { + "epoch": 0.1745, + "grad_norm": 0.5720935515174491, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 17450 + }, + { + "epoch": 0.17451, + "grad_norm": 0.58910505655655, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 17451 + }, + { + "epoch": 0.17452, + "grad_norm": 0.6532226743586879, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 17452 + }, + { + "epoch": 0.17453, + "grad_norm": 0.5721741200596072, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 17453 + }, + { + "epoch": 0.17454, + "grad_norm": 0.5288017631745136, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 17454 + }, + { + "epoch": 0.17455, + "grad_norm": 0.549777321148638, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 17455 + }, + { + "epoch": 0.17456, + "grad_norm": 0.5489115911335394, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 17456 + }, + { + "epoch": 0.17457, + "grad_norm": 0.5993528434988676, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 17457 + }, + { + "epoch": 0.17458, + "grad_norm": 0.656716513154675, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 17458 + }, + { + "epoch": 0.17459, + "grad_norm": 0.639996093183823, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 17459 + }, + { + "epoch": 0.1746, + "grad_norm": 0.7380618513629666, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 17460 + }, + { + "epoch": 0.17461, + "grad_norm": 0.7510182471654129, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 17461 + }, + { + "epoch": 0.17462, + "grad_norm": 0.7077478192580408, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 17462 + }, + { + "epoch": 0.17463, + "grad_norm": 0.7494222452218454, + "learning_rate": 0.003, + "loss": 3.977, + "step": 17463 + }, + { + "epoch": 0.17464, + "grad_norm": 0.8279634906229905, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 17464 + }, + { + "epoch": 0.17465, + "grad_norm": 0.8477740855546998, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 17465 + }, + { + "epoch": 0.17466, + "grad_norm": 0.8499387300448031, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 17466 + }, + { + "epoch": 0.17467, + "grad_norm": 0.7501860393701655, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 17467 + }, + { + "epoch": 0.17468, + "grad_norm": 0.7568320313968444, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 17468 + }, + { + "epoch": 0.17469, + "grad_norm": 0.7262318615352592, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 17469 + }, + { + "epoch": 0.1747, + "grad_norm": 0.8064252282389821, + "learning_rate": 0.003, + "loss": 4.032, + "step": 17470 + }, + { + "epoch": 0.17471, + "grad_norm": 0.9192852912794351, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 17471 + }, + { + "epoch": 0.17472, + "grad_norm": 0.9670662546703448, + "learning_rate": 0.003, + "loss": 4.031, + "step": 17472 + }, + { + "epoch": 0.17473, + "grad_norm": 0.8359271686771074, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 17473 + }, + { + "epoch": 0.17474, + "grad_norm": 0.7877209513341075, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 17474 + }, + { + "epoch": 0.17475, + "grad_norm": 0.8201877056043668, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 17475 + }, + { + "epoch": 0.17476, + "grad_norm": 0.8877581380862389, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 17476 + }, + { + "epoch": 0.17477, + "grad_norm": 0.8276931694647871, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 17477 + }, + { + "epoch": 0.17478, + "grad_norm": 0.7574627659324505, + "learning_rate": 0.003, + "loss": 3.994, + "step": 17478 + }, + { + "epoch": 0.17479, + "grad_norm": 0.7434927192302377, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 17479 + }, + { + "epoch": 0.1748, + "grad_norm": 0.869906023501737, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 17480 + }, + { + "epoch": 0.17481, + "grad_norm": 0.868035069156097, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 17481 + }, + { + "epoch": 0.17482, + "grad_norm": 0.8529127654188746, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 17482 + }, + { + "epoch": 0.17483, + "grad_norm": 0.8425282767977784, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 17483 + }, + { + "epoch": 0.17484, + "grad_norm": 0.7739914220434738, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 17484 + }, + { + "epoch": 0.17485, + "grad_norm": 0.7528259846326526, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 17485 + }, + { + "epoch": 0.17486, + "grad_norm": 0.6961229582041619, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 17486 + }, + { + "epoch": 0.17487, + "grad_norm": 0.7050156951504398, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 17487 + }, + { + "epoch": 0.17488, + "grad_norm": 0.713647523278865, + "learning_rate": 0.003, + "loss": 4.001, + "step": 17488 + }, + { + "epoch": 0.17489, + "grad_norm": 0.6505057818423013, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 17489 + }, + { + "epoch": 0.1749, + "grad_norm": 0.6399043974050708, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 17490 + }, + { + "epoch": 0.17491, + "grad_norm": 0.6318434964340484, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 17491 + }, + { + "epoch": 0.17492, + "grad_norm": 0.6981621417185642, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 17492 + }, + { + "epoch": 0.17493, + "grad_norm": 0.727799557704644, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 17493 + }, + { + "epoch": 0.17494, + "grad_norm": 0.7568614150253372, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 17494 + }, + { + "epoch": 0.17495, + "grad_norm": 0.8384287520693288, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 17495 + }, + { + "epoch": 0.17496, + "grad_norm": 0.8741632908523569, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 17496 + }, + { + "epoch": 0.17497, + "grad_norm": 0.8889450272344532, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17497 + }, + { + "epoch": 0.17498, + "grad_norm": 0.8886380290201331, + "learning_rate": 0.003, + "loss": 4.041, + "step": 17498 + }, + { + "epoch": 0.17499, + "grad_norm": 0.7454678092035932, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 17499 + }, + { + "epoch": 0.175, + "grad_norm": 0.6630474506626944, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 17500 + }, + { + "epoch": 0.17501, + "grad_norm": 0.6572908571386867, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 17501 + }, + { + "epoch": 0.17502, + "grad_norm": 0.7153133709156686, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 17502 + }, + { + "epoch": 0.17503, + "grad_norm": 0.7561785966207268, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 17503 + }, + { + "epoch": 0.17504, + "grad_norm": 0.6737752264375025, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 17504 + }, + { + "epoch": 0.17505, + "grad_norm": 0.6173925083837486, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 17505 + }, + { + "epoch": 0.17506, + "grad_norm": 0.6084377241223539, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 17506 + }, + { + "epoch": 0.17507, + "grad_norm": 0.5636125945932554, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 17507 + }, + { + "epoch": 0.17508, + "grad_norm": 0.6252100755634077, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 17508 + }, + { + "epoch": 0.17509, + "grad_norm": 0.6124295910069546, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 17509 + }, + { + "epoch": 0.1751, + "grad_norm": 0.6435919449975206, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 17510 + }, + { + "epoch": 0.17511, + "grad_norm": 0.6229596955055531, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 17511 + }, + { + "epoch": 0.17512, + "grad_norm": 0.5539193568523706, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 17512 + }, + { + "epoch": 0.17513, + "grad_norm": 0.5746611274414399, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 17513 + }, + { + "epoch": 0.17514, + "grad_norm": 0.6106335592712981, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 17514 + }, + { + "epoch": 0.17515, + "grad_norm": 0.7116396970268088, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 17515 + }, + { + "epoch": 0.17516, + "grad_norm": 0.8140291820089881, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 17516 + }, + { + "epoch": 0.17517, + "grad_norm": 0.7961122822328116, + "learning_rate": 0.003, + "loss": 4.022, + "step": 17517 + }, + { + "epoch": 0.17518, + "grad_norm": 0.9077986999661524, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 17518 + }, + { + "epoch": 0.17519, + "grad_norm": 0.9344324948895869, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 17519 + }, + { + "epoch": 0.1752, + "grad_norm": 0.8021325793245052, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 17520 + }, + { + "epoch": 0.17521, + "grad_norm": 0.6742139963887104, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 17521 + }, + { + "epoch": 0.17522, + "grad_norm": 0.6995223145618035, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 17522 + }, + { + "epoch": 0.17523, + "grad_norm": 0.7544820889537321, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 17523 + }, + { + "epoch": 0.17524, + "grad_norm": 0.7995030886534454, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 17524 + }, + { + "epoch": 0.17525, + "grad_norm": 0.927247501170024, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 17525 + }, + { + "epoch": 0.17526, + "grad_norm": 0.9727949377494464, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 17526 + }, + { + "epoch": 0.17527, + "grad_norm": 0.7889004865302941, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 17527 + }, + { + "epoch": 0.17528, + "grad_norm": 0.7856738908259283, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 17528 + }, + { + "epoch": 0.17529, + "grad_norm": 0.80800262760826, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 17529 + }, + { + "epoch": 0.1753, + "grad_norm": 0.7910749374817562, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 17530 + }, + { + "epoch": 0.17531, + "grad_norm": 0.7394811310959666, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 17531 + }, + { + "epoch": 0.17532, + "grad_norm": 0.7126792612797463, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 17532 + }, + { + "epoch": 0.17533, + "grad_norm": 0.7359329343880254, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 17533 + }, + { + "epoch": 0.17534, + "grad_norm": 0.6917705035332647, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 17534 + }, + { + "epoch": 0.17535, + "grad_norm": 0.5523437045780555, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 17535 + }, + { + "epoch": 0.17536, + "grad_norm": 0.5900948272236866, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 17536 + }, + { + "epoch": 0.17537, + "grad_norm": 0.7342716623322801, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 17537 + }, + { + "epoch": 0.17538, + "grad_norm": 0.8075841728178336, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 17538 + }, + { + "epoch": 0.17539, + "grad_norm": 0.7843027066640194, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 17539 + }, + { + "epoch": 0.1754, + "grad_norm": 0.7502279825701166, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 17540 + }, + { + "epoch": 0.17541, + "grad_norm": 0.6691769583199888, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 17541 + }, + { + "epoch": 0.17542, + "grad_norm": 0.6270678049639272, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 17542 + }, + { + "epoch": 0.17543, + "grad_norm": 0.628043738743877, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 17543 + }, + { + "epoch": 0.17544, + "grad_norm": 0.7212775188565475, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 17544 + }, + { + "epoch": 0.17545, + "grad_norm": 0.772836543281489, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 17545 + }, + { + "epoch": 0.17546, + "grad_norm": 0.8242580403010309, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17546 + }, + { + "epoch": 0.17547, + "grad_norm": 0.8303526615159839, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 17547 + }, + { + "epoch": 0.17548, + "grad_norm": 0.8252971496855028, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 17548 + }, + { + "epoch": 0.17549, + "grad_norm": 0.7856706508230953, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 17549 + }, + { + "epoch": 0.1755, + "grad_norm": 0.7361983306677732, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 17550 + }, + { + "epoch": 0.17551, + "grad_norm": 0.6336720581012585, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 17551 + }, + { + "epoch": 0.17552, + "grad_norm": 0.5831731581857059, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 17552 + }, + { + "epoch": 0.17553, + "grad_norm": 0.5768434724755112, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17553 + }, + { + "epoch": 0.17554, + "grad_norm": 0.5394666824031376, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 17554 + }, + { + "epoch": 0.17555, + "grad_norm": 0.6076478161325836, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 17555 + }, + { + "epoch": 0.17556, + "grad_norm": 0.6880854183174279, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 17556 + }, + { + "epoch": 0.17557, + "grad_norm": 0.7868849437633914, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 17557 + }, + { + "epoch": 0.17558, + "grad_norm": 0.8131067522243393, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 17558 + }, + { + "epoch": 0.17559, + "grad_norm": 0.8137370136021349, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 17559 + }, + { + "epoch": 0.1756, + "grad_norm": 0.8067708177915454, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 17560 + }, + { + "epoch": 0.17561, + "grad_norm": 0.8724858389617635, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 17561 + }, + { + "epoch": 0.17562, + "grad_norm": 0.8627092105556701, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 17562 + }, + { + "epoch": 0.17563, + "grad_norm": 0.7609551815118935, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 17563 + }, + { + "epoch": 0.17564, + "grad_norm": 0.7383727844128243, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17564 + }, + { + "epoch": 0.17565, + "grad_norm": 0.7831872149522712, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 17565 + }, + { + "epoch": 0.17566, + "grad_norm": 0.8409591890719151, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 17566 + }, + { + "epoch": 0.17567, + "grad_norm": 0.9153648361101778, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 17567 + }, + { + "epoch": 0.17568, + "grad_norm": 0.9942384708011097, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 17568 + }, + { + "epoch": 0.17569, + "grad_norm": 1.1151612328229246, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 17569 + }, + { + "epoch": 0.1757, + "grad_norm": 0.832982193471442, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 17570 + }, + { + "epoch": 0.17571, + "grad_norm": 0.7327520525550615, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 17571 + }, + { + "epoch": 0.17572, + "grad_norm": 0.7033248590903722, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 17572 + }, + { + "epoch": 0.17573, + "grad_norm": 0.8752464831313579, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 17573 + }, + { + "epoch": 0.17574, + "grad_norm": 1.0376223952774226, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17574 + }, + { + "epoch": 0.17575, + "grad_norm": 0.8837095791856532, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 17575 + }, + { + "epoch": 0.17576, + "grad_norm": 0.8168616442570847, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 17576 + }, + { + "epoch": 0.17577, + "grad_norm": 0.871945777933946, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 17577 + }, + { + "epoch": 0.17578, + "grad_norm": 0.8286759988658629, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 17578 + }, + { + "epoch": 0.17579, + "grad_norm": 0.7633038275059488, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 17579 + }, + { + "epoch": 0.1758, + "grad_norm": 0.7223108189989216, + "learning_rate": 0.003, + "loss": 4.02, + "step": 17580 + }, + { + "epoch": 0.17581, + "grad_norm": 0.6373083099129763, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 17581 + }, + { + "epoch": 0.17582, + "grad_norm": 0.6184811097915923, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 17582 + }, + { + "epoch": 0.17583, + "grad_norm": 0.5693952181757503, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17583 + }, + { + "epoch": 0.17584, + "grad_norm": 0.6791913795927627, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 17584 + }, + { + "epoch": 0.17585, + "grad_norm": 0.7311712981217636, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 17585 + }, + { + "epoch": 0.17586, + "grad_norm": 0.7749074204547705, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17586 + }, + { + "epoch": 0.17587, + "grad_norm": 0.864933202713721, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 17587 + }, + { + "epoch": 0.17588, + "grad_norm": 0.8718565110400126, + "learning_rate": 0.003, + "loss": 4.0584, + "step": 17588 + }, + { + "epoch": 0.17589, + "grad_norm": 0.7762310761209331, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 17589 + }, + { + "epoch": 0.1759, + "grad_norm": 0.6989736184249277, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 17590 + }, + { + "epoch": 0.17591, + "grad_norm": 0.7130168660461036, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 17591 + }, + { + "epoch": 0.17592, + "grad_norm": 0.7190540903593141, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 17592 + }, + { + "epoch": 0.17593, + "grad_norm": 0.6883262767044317, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 17593 + }, + { + "epoch": 0.17594, + "grad_norm": 0.7974217756220956, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 17594 + }, + { + "epoch": 0.17595, + "grad_norm": 0.8888594815861113, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 17595 + }, + { + "epoch": 0.17596, + "grad_norm": 0.9179065471102704, + "learning_rate": 0.003, + "loss": 4.021, + "step": 17596 + }, + { + "epoch": 0.17597, + "grad_norm": 0.7307831164517719, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 17597 + }, + { + "epoch": 0.17598, + "grad_norm": 0.5599747431597822, + "learning_rate": 0.003, + "loss": 4.035, + "step": 17598 + }, + { + "epoch": 0.17599, + "grad_norm": 0.5663331231423553, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 17599 + }, + { + "epoch": 0.176, + "grad_norm": 0.583911953924875, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 17600 + }, + { + "epoch": 0.17601, + "grad_norm": 0.5662202851452989, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 17601 + }, + { + "epoch": 0.17602, + "grad_norm": 0.5766049210734867, + "learning_rate": 0.003, + "loss": 3.993, + "step": 17602 + }, + { + "epoch": 0.17603, + "grad_norm": 0.4940600011241624, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 17603 + }, + { + "epoch": 0.17604, + "grad_norm": 0.4996281290825714, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 17604 + }, + { + "epoch": 0.17605, + "grad_norm": 0.5591100998433463, + "learning_rate": 0.003, + "loss": 4.019, + "step": 17605 + }, + { + "epoch": 0.17606, + "grad_norm": 0.5753290841745904, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 17606 + }, + { + "epoch": 0.17607, + "grad_norm": 0.6283656936792456, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 17607 + }, + { + "epoch": 0.17608, + "grad_norm": 0.6879395830472896, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 17608 + }, + { + "epoch": 0.17609, + "grad_norm": 0.7499764644096415, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 17609 + }, + { + "epoch": 0.1761, + "grad_norm": 0.7268825569659825, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 17610 + }, + { + "epoch": 0.17611, + "grad_norm": 0.6992552299151376, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 17611 + }, + { + "epoch": 0.17612, + "grad_norm": 0.7038271489225202, + "learning_rate": 0.003, + "loss": 3.9437, + "step": 17612 + }, + { + "epoch": 0.17613, + "grad_norm": 0.7225973949872105, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 17613 + }, + { + "epoch": 0.17614, + "grad_norm": 0.8064591429821653, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 17614 + }, + { + "epoch": 0.17615, + "grad_norm": 0.7687594433741777, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 17615 + }, + { + "epoch": 0.17616, + "grad_norm": 0.7773364873475865, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 17616 + }, + { + "epoch": 0.17617, + "grad_norm": 1.0112587918415696, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 17617 + }, + { + "epoch": 0.17618, + "grad_norm": 1.1486870957281123, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 17618 + }, + { + "epoch": 0.17619, + "grad_norm": 0.957599303787573, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 17619 + }, + { + "epoch": 0.1762, + "grad_norm": 0.869290822870177, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 17620 + }, + { + "epoch": 0.17621, + "grad_norm": 0.7454819273301404, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 17621 + }, + { + "epoch": 0.17622, + "grad_norm": 0.6884472762040221, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 17622 + }, + { + "epoch": 0.17623, + "grad_norm": 0.6202166377866123, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 17623 + }, + { + "epoch": 0.17624, + "grad_norm": 0.6731086951750674, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 17624 + }, + { + "epoch": 0.17625, + "grad_norm": 0.6149448592130394, + "learning_rate": 0.003, + "loss": 3.991, + "step": 17625 + }, + { + "epoch": 0.17626, + "grad_norm": 0.6425880143942744, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 17626 + }, + { + "epoch": 0.17627, + "grad_norm": 0.6451782323791839, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 17627 + }, + { + "epoch": 0.17628, + "grad_norm": 0.5922531677520573, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 17628 + }, + { + "epoch": 0.17629, + "grad_norm": 0.5700232038600105, + "learning_rate": 0.003, + "loss": 4.02, + "step": 17629 + }, + { + "epoch": 0.1763, + "grad_norm": 0.5641470245568329, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 17630 + }, + { + "epoch": 0.17631, + "grad_norm": 0.6463784528360764, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 17631 + }, + { + "epoch": 0.17632, + "grad_norm": 0.7187794913398393, + "learning_rate": 0.003, + "loss": 4.0528, + "step": 17632 + }, + { + "epoch": 0.17633, + "grad_norm": 0.7432416262053974, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 17633 + }, + { + "epoch": 0.17634, + "grad_norm": 0.7802715842631608, + "learning_rate": 0.003, + "loss": 3.991, + "step": 17634 + }, + { + "epoch": 0.17635, + "grad_norm": 0.895125351848697, + "learning_rate": 0.003, + "loss": 3.994, + "step": 17635 + }, + { + "epoch": 0.17636, + "grad_norm": 1.075119851837935, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 17636 + }, + { + "epoch": 0.17637, + "grad_norm": 0.8688171497305555, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 17637 + }, + { + "epoch": 0.17638, + "grad_norm": 0.7531148655333246, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 17638 + }, + { + "epoch": 0.17639, + "grad_norm": 0.8597867241411682, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 17639 + }, + { + "epoch": 0.1764, + "grad_norm": 0.9100308898168329, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 17640 + }, + { + "epoch": 0.17641, + "grad_norm": 0.8397647112581726, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 17641 + }, + { + "epoch": 0.17642, + "grad_norm": 0.7564149502658429, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 17642 + }, + { + "epoch": 0.17643, + "grad_norm": 0.6867891603138572, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 17643 + }, + { + "epoch": 0.17644, + "grad_norm": 0.644173364233107, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 17644 + }, + { + "epoch": 0.17645, + "grad_norm": 0.6025884313704164, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 17645 + }, + { + "epoch": 0.17646, + "grad_norm": 0.6035178403979408, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 17646 + }, + { + "epoch": 0.17647, + "grad_norm": 0.6942040972874673, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 17647 + }, + { + "epoch": 0.17648, + "grad_norm": 0.7343841642979171, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 17648 + }, + { + "epoch": 0.17649, + "grad_norm": 0.7932295622377227, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 17649 + }, + { + "epoch": 0.1765, + "grad_norm": 0.9444945334820432, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 17650 + }, + { + "epoch": 0.17651, + "grad_norm": 0.9596440546990449, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 17651 + }, + { + "epoch": 0.17652, + "grad_norm": 0.8070405190039496, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 17652 + }, + { + "epoch": 0.17653, + "grad_norm": 0.6423539207868153, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 17653 + }, + { + "epoch": 0.17654, + "grad_norm": 0.7346385893809405, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 17654 + }, + { + "epoch": 0.17655, + "grad_norm": 0.7712843050433327, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 17655 + }, + { + "epoch": 0.17656, + "grad_norm": 0.7861630476811571, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 17656 + }, + { + "epoch": 0.17657, + "grad_norm": 0.767944120234548, + "learning_rate": 0.003, + "loss": 3.97, + "step": 17657 + }, + { + "epoch": 0.17658, + "grad_norm": 0.6932978012935882, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 17658 + }, + { + "epoch": 0.17659, + "grad_norm": 0.6179731662368103, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 17659 + }, + { + "epoch": 0.1766, + "grad_norm": 0.6915303452965805, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 17660 + }, + { + "epoch": 0.17661, + "grad_norm": 0.705839952786067, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 17661 + }, + { + "epoch": 0.17662, + "grad_norm": 0.7206260163305941, + "learning_rate": 0.003, + "loss": 3.98, + "step": 17662 + }, + { + "epoch": 0.17663, + "grad_norm": 0.7489006122495757, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17663 + }, + { + "epoch": 0.17664, + "grad_norm": 0.7245636189301904, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 17664 + }, + { + "epoch": 0.17665, + "grad_norm": 0.6730886796032014, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 17665 + }, + { + "epoch": 0.17666, + "grad_norm": 0.6595460587962778, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 17666 + }, + { + "epoch": 0.17667, + "grad_norm": 0.6515964314597633, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 17667 + }, + { + "epoch": 0.17668, + "grad_norm": 0.7599795907146241, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 17668 + }, + { + "epoch": 0.17669, + "grad_norm": 0.8117631459677423, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 17669 + }, + { + "epoch": 0.1767, + "grad_norm": 0.8568799725588218, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 17670 + }, + { + "epoch": 0.17671, + "grad_norm": 0.934800365472657, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 17671 + }, + { + "epoch": 0.17672, + "grad_norm": 0.8992986575970445, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 17672 + }, + { + "epoch": 0.17673, + "grad_norm": 0.8579901120046466, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 17673 + }, + { + "epoch": 0.17674, + "grad_norm": 0.8540440988545649, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 17674 + }, + { + "epoch": 0.17675, + "grad_norm": 0.793414385300293, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 17675 + }, + { + "epoch": 0.17676, + "grad_norm": 0.7122281544856317, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 17676 + }, + { + "epoch": 0.17677, + "grad_norm": 0.6402837520529852, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 17677 + }, + { + "epoch": 0.17678, + "grad_norm": 0.6939918650497445, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 17678 + }, + { + "epoch": 0.17679, + "grad_norm": 0.7663249996023792, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 17679 + }, + { + "epoch": 0.1768, + "grad_norm": 0.7114100381179316, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 17680 + }, + { + "epoch": 0.17681, + "grad_norm": 0.7363688370782315, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 17681 + }, + { + "epoch": 0.17682, + "grad_norm": 0.6955151349028108, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 17682 + }, + { + "epoch": 0.17683, + "grad_norm": 0.6226680817629237, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 17683 + }, + { + "epoch": 0.17684, + "grad_norm": 0.6080833547357753, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 17684 + }, + { + "epoch": 0.17685, + "grad_norm": 0.6245259361840003, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 17685 + }, + { + "epoch": 0.17686, + "grad_norm": 0.6069127435029593, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 17686 + }, + { + "epoch": 0.17687, + "grad_norm": 0.5781184886919237, + "learning_rate": 0.003, + "loss": 3.985, + "step": 17687 + }, + { + "epoch": 0.17688, + "grad_norm": 0.6928700079295348, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 17688 + }, + { + "epoch": 0.17689, + "grad_norm": 0.8533464532107973, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 17689 + }, + { + "epoch": 0.1769, + "grad_norm": 1.0170898607503327, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 17690 + }, + { + "epoch": 0.17691, + "grad_norm": 1.063826494541355, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 17691 + }, + { + "epoch": 0.17692, + "grad_norm": 0.9178400556065082, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17692 + }, + { + "epoch": 0.17693, + "grad_norm": 0.7796637226766994, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 17693 + }, + { + "epoch": 0.17694, + "grad_norm": 0.7673234209800628, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 17694 + }, + { + "epoch": 0.17695, + "grad_norm": 0.7878036766220063, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 17695 + }, + { + "epoch": 0.17696, + "grad_norm": 0.8686530158599945, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17696 + }, + { + "epoch": 0.17697, + "grad_norm": 0.991063394784568, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 17697 + }, + { + "epoch": 0.17698, + "grad_norm": 0.9907765774479838, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 17698 + }, + { + "epoch": 0.17699, + "grad_norm": 0.9341945221197532, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 17699 + }, + { + "epoch": 0.177, + "grad_norm": 0.7788450448339939, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 17700 + }, + { + "epoch": 0.17701, + "grad_norm": 0.6663719305980664, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 17701 + }, + { + "epoch": 0.17702, + "grad_norm": 0.7496600975255501, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 17702 + }, + { + "epoch": 0.17703, + "grad_norm": 0.7620478724123704, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 17703 + }, + { + "epoch": 0.17704, + "grad_norm": 0.7318797140789409, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 17704 + }, + { + "epoch": 0.17705, + "grad_norm": 0.7214402000042405, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 17705 + }, + { + "epoch": 0.17706, + "grad_norm": 0.7210802210190488, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 17706 + }, + { + "epoch": 0.17707, + "grad_norm": 0.8650527239949911, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 17707 + }, + { + "epoch": 0.17708, + "grad_norm": 0.9221605064598223, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 17708 + }, + { + "epoch": 0.17709, + "grad_norm": 0.7924125583295596, + "learning_rate": 0.003, + "loss": 4.0462, + "step": 17709 + }, + { + "epoch": 0.1771, + "grad_norm": 0.8808191401509063, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 17710 + }, + { + "epoch": 0.17711, + "grad_norm": 0.8840348951238863, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 17711 + }, + { + "epoch": 0.17712, + "grad_norm": 0.9147133141470216, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 17712 + }, + { + "epoch": 0.17713, + "grad_norm": 0.8407336258998693, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 17713 + }, + { + "epoch": 0.17714, + "grad_norm": 0.7849602148614435, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 17714 + }, + { + "epoch": 0.17715, + "grad_norm": 0.7962378951222716, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17715 + }, + { + "epoch": 0.17716, + "grad_norm": 0.9249559533553178, + "learning_rate": 0.003, + "loss": 4.04, + "step": 17716 + }, + { + "epoch": 0.17717, + "grad_norm": 1.0403501274088574, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 17717 + }, + { + "epoch": 0.17718, + "grad_norm": 0.836752138987057, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 17718 + }, + { + "epoch": 0.17719, + "grad_norm": 0.7103200805265674, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 17719 + }, + { + "epoch": 0.1772, + "grad_norm": 0.6693065854881177, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 17720 + }, + { + "epoch": 0.17721, + "grad_norm": 0.6465054947263866, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 17721 + }, + { + "epoch": 0.17722, + "grad_norm": 0.6067492187421083, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 17722 + }, + { + "epoch": 0.17723, + "grad_norm": 0.5802086415975842, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 17723 + }, + { + "epoch": 0.17724, + "grad_norm": 0.5873653850724486, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 17724 + }, + { + "epoch": 0.17725, + "grad_norm": 0.6331400543007204, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17725 + }, + { + "epoch": 0.17726, + "grad_norm": 0.712801689314781, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 17726 + }, + { + "epoch": 0.17727, + "grad_norm": 0.9391904822180639, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 17727 + }, + { + "epoch": 0.17728, + "grad_norm": 1.1165205563796046, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 17728 + }, + { + "epoch": 0.17729, + "grad_norm": 0.8537321397874487, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 17729 + }, + { + "epoch": 0.1773, + "grad_norm": 0.6410784649056638, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 17730 + }, + { + "epoch": 0.17731, + "grad_norm": 0.6838987301255779, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 17731 + }, + { + "epoch": 0.17732, + "grad_norm": 0.7583656896170503, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 17732 + }, + { + "epoch": 0.17733, + "grad_norm": 0.7767266766041592, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 17733 + }, + { + "epoch": 0.17734, + "grad_norm": 0.691206118955507, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 17734 + }, + { + "epoch": 0.17735, + "grad_norm": 0.7134991820322427, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 17735 + }, + { + "epoch": 0.17736, + "grad_norm": 0.7425167661071269, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 17736 + }, + { + "epoch": 0.17737, + "grad_norm": 0.6897333174778817, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 17737 + }, + { + "epoch": 0.17738, + "grad_norm": 0.6599560983179622, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 17738 + }, + { + "epoch": 0.17739, + "grad_norm": 0.6646926760274924, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 17739 + }, + { + "epoch": 0.1774, + "grad_norm": 0.7311268774066235, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 17740 + }, + { + "epoch": 0.17741, + "grad_norm": 0.6277143196394622, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 17741 + }, + { + "epoch": 0.17742, + "grad_norm": 0.5266720472230978, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 17742 + }, + { + "epoch": 0.17743, + "grad_norm": 0.4986738848211866, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 17743 + }, + { + "epoch": 0.17744, + "grad_norm": 0.48922310871878266, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 17744 + }, + { + "epoch": 0.17745, + "grad_norm": 0.5400892682936996, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 17745 + }, + { + "epoch": 0.17746, + "grad_norm": 0.5810431567394441, + "learning_rate": 0.003, + "loss": 4.011, + "step": 17746 + }, + { + "epoch": 0.17747, + "grad_norm": 0.6213266034719422, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 17747 + }, + { + "epoch": 0.17748, + "grad_norm": 0.6033609581020897, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 17748 + }, + { + "epoch": 0.17749, + "grad_norm": 0.5007588027404936, + "learning_rate": 0.003, + "loss": 3.972, + "step": 17749 + }, + { + "epoch": 0.1775, + "grad_norm": 0.5193054677658865, + "learning_rate": 0.003, + "loss": 3.998, + "step": 17750 + }, + { + "epoch": 0.17751, + "grad_norm": 0.5513843578298679, + "learning_rate": 0.003, + "loss": 4.017, + "step": 17751 + }, + { + "epoch": 0.17752, + "grad_norm": 0.5973559925961731, + "learning_rate": 0.003, + "loss": 3.989, + "step": 17752 + }, + { + "epoch": 0.17753, + "grad_norm": 0.7408331596169623, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 17753 + }, + { + "epoch": 0.17754, + "grad_norm": 0.9670370638968752, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 17754 + }, + { + "epoch": 0.17755, + "grad_norm": 1.0815319573539943, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 17755 + }, + { + "epoch": 0.17756, + "grad_norm": 0.8509489988664808, + "learning_rate": 0.003, + "loss": 3.979, + "step": 17756 + }, + { + "epoch": 0.17757, + "grad_norm": 0.7447030898380376, + "learning_rate": 0.003, + "loss": 4.017, + "step": 17757 + }, + { + "epoch": 0.17758, + "grad_norm": 0.7260970168558156, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 17758 + }, + { + "epoch": 0.17759, + "grad_norm": 0.7762883540154504, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 17759 + }, + { + "epoch": 0.1776, + "grad_norm": 0.7053595377819348, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 17760 + }, + { + "epoch": 0.17761, + "grad_norm": 0.653238393632736, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 17761 + }, + { + "epoch": 0.17762, + "grad_norm": 0.7532454770878403, + "learning_rate": 0.003, + "loss": 4.018, + "step": 17762 + }, + { + "epoch": 0.17763, + "grad_norm": 0.868535797668387, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 17763 + }, + { + "epoch": 0.17764, + "grad_norm": 1.0331056114787867, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 17764 + }, + { + "epoch": 0.17765, + "grad_norm": 0.9189764037690792, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 17765 + }, + { + "epoch": 0.17766, + "grad_norm": 0.8524129838053207, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 17766 + }, + { + "epoch": 0.17767, + "grad_norm": 0.8410377931708058, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 17767 + }, + { + "epoch": 0.17768, + "grad_norm": 0.784735368523145, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 17768 + }, + { + "epoch": 0.17769, + "grad_norm": 0.7658325520339562, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 17769 + }, + { + "epoch": 0.1777, + "grad_norm": 0.8711294617911569, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 17770 + }, + { + "epoch": 0.17771, + "grad_norm": 1.005774937277879, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 17771 + }, + { + "epoch": 0.17772, + "grad_norm": 1.0633959851459704, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 17772 + }, + { + "epoch": 0.17773, + "grad_norm": 0.7610076293254614, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 17773 + }, + { + "epoch": 0.17774, + "grad_norm": 0.6409679372139958, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 17774 + }, + { + "epoch": 0.17775, + "grad_norm": 0.6654823552134196, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 17775 + }, + { + "epoch": 0.17776, + "grad_norm": 0.6458482283640213, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 17776 + }, + { + "epoch": 0.17777, + "grad_norm": 0.7357235070986475, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 17777 + }, + { + "epoch": 0.17778, + "grad_norm": 0.7353660813710062, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 17778 + }, + { + "epoch": 0.17779, + "grad_norm": 0.6865281050580523, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 17779 + }, + { + "epoch": 0.1778, + "grad_norm": 0.7662565629074398, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 17780 + }, + { + "epoch": 0.17781, + "grad_norm": 0.6742896749467833, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 17781 + }, + { + "epoch": 0.17782, + "grad_norm": 0.6036567156110828, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 17782 + }, + { + "epoch": 0.17783, + "grad_norm": 0.6205675381084828, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 17783 + }, + { + "epoch": 0.17784, + "grad_norm": 0.5749267697109011, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 17784 + }, + { + "epoch": 0.17785, + "grad_norm": 0.5868676557511001, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 17785 + }, + { + "epoch": 0.17786, + "grad_norm": 0.5923958282036329, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 17786 + }, + { + "epoch": 0.17787, + "grad_norm": 0.5888185594354658, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 17787 + }, + { + "epoch": 0.17788, + "grad_norm": 0.5924482195908057, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 17788 + }, + { + "epoch": 0.17789, + "grad_norm": 0.5228708489436498, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 17789 + }, + { + "epoch": 0.1779, + "grad_norm": 0.5362610272874899, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 17790 + }, + { + "epoch": 0.17791, + "grad_norm": 0.48920930407833646, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 17791 + }, + { + "epoch": 0.17792, + "grad_norm": 0.5080217498183867, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 17792 + }, + { + "epoch": 0.17793, + "grad_norm": 0.5381464506739798, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 17793 + }, + { + "epoch": 0.17794, + "grad_norm": 0.6298681750645757, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 17794 + }, + { + "epoch": 0.17795, + "grad_norm": 0.7422992519851486, + "learning_rate": 0.003, + "loss": 3.995, + "step": 17795 + }, + { + "epoch": 0.17796, + "grad_norm": 0.7154803979463898, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 17796 + }, + { + "epoch": 0.17797, + "grad_norm": 0.6531660537086172, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 17797 + }, + { + "epoch": 0.17798, + "grad_norm": 0.6572246123965538, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 17798 + }, + { + "epoch": 0.17799, + "grad_norm": 0.7406645740257729, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 17799 + }, + { + "epoch": 0.178, + "grad_norm": 0.8318964941469305, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 17800 + }, + { + "epoch": 0.17801, + "grad_norm": 0.9744477316941973, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 17801 + }, + { + "epoch": 0.17802, + "grad_norm": 1.2252738792332527, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 17802 + }, + { + "epoch": 0.17803, + "grad_norm": 0.9136863303236652, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 17803 + }, + { + "epoch": 0.17804, + "grad_norm": 0.8695397980993236, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 17804 + }, + { + "epoch": 0.17805, + "grad_norm": 0.8125808960465508, + "learning_rate": 0.003, + "loss": 4.023, + "step": 17805 + }, + { + "epoch": 0.17806, + "grad_norm": 0.8110643863429272, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 17806 + }, + { + "epoch": 0.17807, + "grad_norm": 0.8255740336733938, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 17807 + }, + { + "epoch": 0.17808, + "grad_norm": 0.7939640183771647, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 17808 + }, + { + "epoch": 0.17809, + "grad_norm": 0.836285107827856, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 17809 + }, + { + "epoch": 0.1781, + "grad_norm": 0.8792818067561591, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 17810 + }, + { + "epoch": 0.17811, + "grad_norm": 0.8587566016833209, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 17811 + }, + { + "epoch": 0.17812, + "grad_norm": 0.7994664990332865, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 17812 + }, + { + "epoch": 0.17813, + "grad_norm": 0.8959378707443332, + "learning_rate": 0.003, + "loss": 4.0543, + "step": 17813 + }, + { + "epoch": 0.17814, + "grad_norm": 0.8751539561352704, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 17814 + }, + { + "epoch": 0.17815, + "grad_norm": 0.8804440208036911, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 17815 + }, + { + "epoch": 0.17816, + "grad_norm": 0.7692982920995547, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17816 + }, + { + "epoch": 0.17817, + "grad_norm": 0.6915316330564767, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 17817 + }, + { + "epoch": 0.17818, + "grad_norm": 0.7410667567218665, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 17818 + }, + { + "epoch": 0.17819, + "grad_norm": 0.6930005604100906, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 17819 + }, + { + "epoch": 0.1782, + "grad_norm": 0.723061299278545, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 17820 + }, + { + "epoch": 0.17821, + "grad_norm": 0.6685243124811979, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 17821 + }, + { + "epoch": 0.17822, + "grad_norm": 0.5993266519091653, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 17822 + }, + { + "epoch": 0.17823, + "grad_norm": 0.5604940339962072, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 17823 + }, + { + "epoch": 0.17824, + "grad_norm": 0.6179341640077198, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 17824 + }, + { + "epoch": 0.17825, + "grad_norm": 0.7846915524072594, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 17825 + }, + { + "epoch": 0.17826, + "grad_norm": 0.8859304685181375, + "learning_rate": 0.003, + "loss": 4.015, + "step": 17826 + }, + { + "epoch": 0.17827, + "grad_norm": 0.9542775257277278, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 17827 + }, + { + "epoch": 0.17828, + "grad_norm": 1.0951711743187404, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 17828 + }, + { + "epoch": 0.17829, + "grad_norm": 0.8002577749411851, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 17829 + }, + { + "epoch": 0.1783, + "grad_norm": 0.5958616163817734, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 17830 + }, + { + "epoch": 0.17831, + "grad_norm": 0.7544555153632293, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 17831 + }, + { + "epoch": 0.17832, + "grad_norm": 0.8396772654133542, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 17832 + }, + { + "epoch": 0.17833, + "grad_norm": 0.7907083081040103, + "learning_rate": 0.003, + "loss": 4.004, + "step": 17833 + }, + { + "epoch": 0.17834, + "grad_norm": 0.7784182127993904, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 17834 + }, + { + "epoch": 0.17835, + "grad_norm": 0.8178577691833826, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 17835 + }, + { + "epoch": 0.17836, + "grad_norm": 0.8880945900944366, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 17836 + }, + { + "epoch": 0.17837, + "grad_norm": 0.948756638511432, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17837 + }, + { + "epoch": 0.17838, + "grad_norm": 0.9482889360212773, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 17838 + }, + { + "epoch": 0.17839, + "grad_norm": 0.9401454322920183, + "learning_rate": 0.003, + "loss": 4.0427, + "step": 17839 + }, + { + "epoch": 0.1784, + "grad_norm": 0.8822377295954403, + "learning_rate": 0.003, + "loss": 4.03, + "step": 17840 + }, + { + "epoch": 0.17841, + "grad_norm": 0.7760721517013988, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 17841 + }, + { + "epoch": 0.17842, + "grad_norm": 0.7303146137899575, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 17842 + }, + { + "epoch": 0.17843, + "grad_norm": 0.7490709695720057, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 17843 + }, + { + "epoch": 0.17844, + "grad_norm": 0.7017713707337356, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 17844 + }, + { + "epoch": 0.17845, + "grad_norm": 0.7025370509603573, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 17845 + }, + { + "epoch": 0.17846, + "grad_norm": 0.7117653828816721, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 17846 + }, + { + "epoch": 0.17847, + "grad_norm": 0.6728256145377987, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17847 + }, + { + "epoch": 0.17848, + "grad_norm": 0.6177217642201924, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17848 + }, + { + "epoch": 0.17849, + "grad_norm": 0.6122008731939478, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 17849 + }, + { + "epoch": 0.1785, + "grad_norm": 0.5701801043218622, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 17850 + }, + { + "epoch": 0.17851, + "grad_norm": 0.5866674960397902, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 17851 + }, + { + "epoch": 0.17852, + "grad_norm": 0.6408894660728612, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 17852 + }, + { + "epoch": 0.17853, + "grad_norm": 0.7657966343940249, + "learning_rate": 0.003, + "loss": 4.014, + "step": 17853 + }, + { + "epoch": 0.17854, + "grad_norm": 0.9073957593738344, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 17854 + }, + { + "epoch": 0.17855, + "grad_norm": 0.9403741306532452, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 17855 + }, + { + "epoch": 0.17856, + "grad_norm": 0.7956952940995581, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 17856 + }, + { + "epoch": 0.17857, + "grad_norm": 0.770611257475015, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 17857 + }, + { + "epoch": 0.17858, + "grad_norm": 0.8239170107051795, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 17858 + }, + { + "epoch": 0.17859, + "grad_norm": 0.9412037561079991, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 17859 + }, + { + "epoch": 0.1786, + "grad_norm": 0.9528541257954539, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 17860 + }, + { + "epoch": 0.17861, + "grad_norm": 0.9221964838544371, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 17861 + }, + { + "epoch": 0.17862, + "grad_norm": 0.6932495526022696, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 17862 + }, + { + "epoch": 0.17863, + "grad_norm": 0.6582424339453403, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 17863 + }, + { + "epoch": 0.17864, + "grad_norm": 0.6939060516830703, + "learning_rate": 0.003, + "loss": 4.036, + "step": 17864 + }, + { + "epoch": 0.17865, + "grad_norm": 0.6184948967746993, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 17865 + }, + { + "epoch": 0.17866, + "grad_norm": 0.6668734018811345, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 17866 + }, + { + "epoch": 0.17867, + "grad_norm": 0.6660234588973196, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 17867 + }, + { + "epoch": 0.17868, + "grad_norm": 0.6415563521315906, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 17868 + }, + { + "epoch": 0.17869, + "grad_norm": 0.6543048654412082, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 17869 + }, + { + "epoch": 0.1787, + "grad_norm": 0.6669387085200268, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 17870 + }, + { + "epoch": 0.17871, + "grad_norm": 0.7569358646390526, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 17871 + }, + { + "epoch": 0.17872, + "grad_norm": 0.7860840167120514, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 17872 + }, + { + "epoch": 0.17873, + "grad_norm": 0.7315023374742248, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 17873 + }, + { + "epoch": 0.17874, + "grad_norm": 0.6519654679103634, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 17874 + }, + { + "epoch": 0.17875, + "grad_norm": 0.5916885289098235, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 17875 + }, + { + "epoch": 0.17876, + "grad_norm": 0.7163641646417989, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 17876 + }, + { + "epoch": 0.17877, + "grad_norm": 0.7381558112794314, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 17877 + }, + { + "epoch": 0.17878, + "grad_norm": 0.819999889263831, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 17878 + }, + { + "epoch": 0.17879, + "grad_norm": 0.9515888997318757, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 17879 + }, + { + "epoch": 0.1788, + "grad_norm": 1.0507406160628907, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 17880 + }, + { + "epoch": 0.17881, + "grad_norm": 0.9305118759938117, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 17881 + }, + { + "epoch": 0.17882, + "grad_norm": 0.835887326154001, + "learning_rate": 0.003, + "loss": 3.987, + "step": 17882 + }, + { + "epoch": 0.17883, + "grad_norm": 0.7214992872131512, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 17883 + }, + { + "epoch": 0.17884, + "grad_norm": 0.7622853022324707, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 17884 + }, + { + "epoch": 0.17885, + "grad_norm": 0.7522828597919473, + "learning_rate": 0.003, + "loss": 4.002, + "step": 17885 + }, + { + "epoch": 0.17886, + "grad_norm": 0.6794997967716764, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 17886 + }, + { + "epoch": 0.17887, + "grad_norm": 0.5470882854193422, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 17887 + }, + { + "epoch": 0.17888, + "grad_norm": 0.5445860884472873, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 17888 + }, + { + "epoch": 0.17889, + "grad_norm": 0.6115356174020877, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 17889 + }, + { + "epoch": 0.1789, + "grad_norm": 0.6585522882959239, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 17890 + }, + { + "epoch": 0.17891, + "grad_norm": 0.6024319290298151, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 17891 + }, + { + "epoch": 0.17892, + "grad_norm": 0.6171528206280029, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 17892 + }, + { + "epoch": 0.17893, + "grad_norm": 0.62012256019983, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 17893 + }, + { + "epoch": 0.17894, + "grad_norm": 0.6825560765756906, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 17894 + }, + { + "epoch": 0.17895, + "grad_norm": 0.7452076257544399, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 17895 + }, + { + "epoch": 0.17896, + "grad_norm": 0.6635126404968203, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 17896 + }, + { + "epoch": 0.17897, + "grad_norm": 0.6856950970715819, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 17897 + }, + { + "epoch": 0.17898, + "grad_norm": 0.7904243167469791, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 17898 + }, + { + "epoch": 0.17899, + "grad_norm": 0.8222744637257274, + "learning_rate": 0.003, + "loss": 4.01, + "step": 17899 + }, + { + "epoch": 0.179, + "grad_norm": 0.8009412693964045, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 17900 + }, + { + "epoch": 0.17901, + "grad_norm": 0.7858150610325385, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 17901 + }, + { + "epoch": 0.17902, + "grad_norm": 0.6309225536931735, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 17902 + }, + { + "epoch": 0.17903, + "grad_norm": 0.5765129511798047, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 17903 + }, + { + "epoch": 0.17904, + "grad_norm": 0.548354059984998, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 17904 + }, + { + "epoch": 0.17905, + "grad_norm": 0.6874441153384213, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 17905 + }, + { + "epoch": 0.17906, + "grad_norm": 0.6834476778325045, + "learning_rate": 0.003, + "loss": 3.985, + "step": 17906 + }, + { + "epoch": 0.17907, + "grad_norm": 0.7329293373798162, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 17907 + }, + { + "epoch": 0.17908, + "grad_norm": 0.7946550375876247, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 17908 + }, + { + "epoch": 0.17909, + "grad_norm": 0.7835997502837185, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 17909 + }, + { + "epoch": 0.1791, + "grad_norm": 0.7452726885563133, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 17910 + }, + { + "epoch": 0.17911, + "grad_norm": 0.760608375339361, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 17911 + }, + { + "epoch": 0.17912, + "grad_norm": 0.7163769657527088, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 17912 + }, + { + "epoch": 0.17913, + "grad_norm": 0.7753358204308125, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 17913 + }, + { + "epoch": 0.17914, + "grad_norm": 0.8829497297166935, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 17914 + }, + { + "epoch": 0.17915, + "grad_norm": 0.9659384898001814, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 17915 + }, + { + "epoch": 0.17916, + "grad_norm": 0.9953354082129513, + "learning_rate": 0.003, + "loss": 4.033, + "step": 17916 + }, + { + "epoch": 0.17917, + "grad_norm": 1.0424911683369191, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 17917 + }, + { + "epoch": 0.17918, + "grad_norm": 0.8487595581172849, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 17918 + }, + { + "epoch": 0.17919, + "grad_norm": 0.8321714648302491, + "learning_rate": 0.003, + "loss": 3.996, + "step": 17919 + }, + { + "epoch": 0.1792, + "grad_norm": 0.8573170107218162, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 17920 + }, + { + "epoch": 0.17921, + "grad_norm": 0.878637068569802, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 17921 + }, + { + "epoch": 0.17922, + "grad_norm": 0.8819191743978197, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 17922 + }, + { + "epoch": 0.17923, + "grad_norm": 0.8976762245224811, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 17923 + }, + { + "epoch": 0.17924, + "grad_norm": 0.9860477967626696, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 17924 + }, + { + "epoch": 0.17925, + "grad_norm": 1.0133516314791782, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 17925 + }, + { + "epoch": 0.17926, + "grad_norm": 0.9582684977238166, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 17926 + }, + { + "epoch": 0.17927, + "grad_norm": 1.0364191636202347, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 17927 + }, + { + "epoch": 0.17928, + "grad_norm": 0.8696072371353448, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 17928 + }, + { + "epoch": 0.17929, + "grad_norm": 0.7685774072863095, + "learning_rate": 0.003, + "loss": 4.016, + "step": 17929 + }, + { + "epoch": 0.1793, + "grad_norm": 0.7669677509290224, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 17930 + }, + { + "epoch": 0.17931, + "grad_norm": 0.6550556522777877, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 17931 + }, + { + "epoch": 0.17932, + "grad_norm": 0.70824473116266, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 17932 + }, + { + "epoch": 0.17933, + "grad_norm": 0.7479336354943023, + "learning_rate": 0.003, + "loss": 4.005, + "step": 17933 + }, + { + "epoch": 0.17934, + "grad_norm": 0.8155657378858974, + "learning_rate": 0.003, + "loss": 4.0503, + "step": 17934 + }, + { + "epoch": 0.17935, + "grad_norm": 0.8273890692295404, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 17935 + }, + { + "epoch": 0.17936, + "grad_norm": 0.7080906628585593, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 17936 + }, + { + "epoch": 0.17937, + "grad_norm": 0.6922362226019355, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 17937 + }, + { + "epoch": 0.17938, + "grad_norm": 0.7924704582097438, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 17938 + }, + { + "epoch": 0.17939, + "grad_norm": 0.8000427949630681, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 17939 + }, + { + "epoch": 0.1794, + "grad_norm": 0.7952877621577026, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 17940 + }, + { + "epoch": 0.17941, + "grad_norm": 0.7898246014166608, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 17941 + }, + { + "epoch": 0.17942, + "grad_norm": 0.7291686921576386, + "learning_rate": 0.003, + "loss": 3.986, + "step": 17942 + }, + { + "epoch": 0.17943, + "grad_norm": 0.657534171985804, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 17943 + }, + { + "epoch": 0.17944, + "grad_norm": 0.6263651614834489, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 17944 + }, + { + "epoch": 0.17945, + "grad_norm": 0.5242940394522302, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 17945 + }, + { + "epoch": 0.17946, + "grad_norm": 0.4804777681446332, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 17946 + }, + { + "epoch": 0.17947, + "grad_norm": 0.5176266995516532, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 17947 + }, + { + "epoch": 0.17948, + "grad_norm": 0.472439751203431, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 17948 + }, + { + "epoch": 0.17949, + "grad_norm": 0.5116755293879676, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 17949 + }, + { + "epoch": 0.1795, + "grad_norm": 0.7227948820167466, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 17950 + }, + { + "epoch": 0.17951, + "grad_norm": 0.9088047593307038, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 17951 + }, + { + "epoch": 0.17952, + "grad_norm": 0.915474067424238, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 17952 + }, + { + "epoch": 0.17953, + "grad_norm": 0.8256048391449686, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 17953 + }, + { + "epoch": 0.17954, + "grad_norm": 0.8166971668520518, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 17954 + }, + { + "epoch": 0.17955, + "grad_norm": 0.7273594732959626, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 17955 + }, + { + "epoch": 0.17956, + "grad_norm": 0.6383917159063068, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 17956 + }, + { + "epoch": 0.17957, + "grad_norm": 0.6698249365037459, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 17957 + }, + { + "epoch": 0.17958, + "grad_norm": 0.670332733853036, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 17958 + }, + { + "epoch": 0.17959, + "grad_norm": 0.66732536498545, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 17959 + }, + { + "epoch": 0.1796, + "grad_norm": 0.7473653277809637, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 17960 + }, + { + "epoch": 0.17961, + "grad_norm": 0.8099817391359522, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 17961 + }, + { + "epoch": 0.17962, + "grad_norm": 0.7826880292790256, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 17962 + }, + { + "epoch": 0.17963, + "grad_norm": 0.7756249762516467, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 17963 + }, + { + "epoch": 0.17964, + "grad_norm": 0.6867033932123762, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 17964 + }, + { + "epoch": 0.17965, + "grad_norm": 0.7158291158875233, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 17965 + }, + { + "epoch": 0.17966, + "grad_norm": 0.7007027354159386, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 17966 + }, + { + "epoch": 0.17967, + "grad_norm": 0.7104850478327499, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 17967 + }, + { + "epoch": 0.17968, + "grad_norm": 0.64001013449228, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 17968 + }, + { + "epoch": 0.17969, + "grad_norm": 0.5614024173821011, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 17969 + }, + { + "epoch": 0.1797, + "grad_norm": 0.5213429776740167, + "learning_rate": 0.003, + "loss": 3.996, + "step": 17970 + }, + { + "epoch": 0.17971, + "grad_norm": 0.5201533130705263, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 17971 + }, + { + "epoch": 0.17972, + "grad_norm": 0.5547551441312963, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 17972 + }, + { + "epoch": 0.17973, + "grad_norm": 0.5288745304433453, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 17973 + }, + { + "epoch": 0.17974, + "grad_norm": 0.5741060376519835, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 17974 + }, + { + "epoch": 0.17975, + "grad_norm": 0.6015731442112457, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17975 + }, + { + "epoch": 0.17976, + "grad_norm": 0.641593250800454, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 17976 + }, + { + "epoch": 0.17977, + "grad_norm": 0.6934378073345926, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 17977 + }, + { + "epoch": 0.17978, + "grad_norm": 0.873726800636869, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 17978 + }, + { + "epoch": 0.17979, + "grad_norm": 0.9083880798240577, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 17979 + }, + { + "epoch": 0.1798, + "grad_norm": 0.6810828283699585, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 17980 + }, + { + "epoch": 0.17981, + "grad_norm": 0.7224249904207316, + "learning_rate": 0.003, + "loss": 3.9321, + "step": 17981 + }, + { + "epoch": 0.17982, + "grad_norm": 0.7693578985140149, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 17982 + }, + { + "epoch": 0.17983, + "grad_norm": 0.8722200562415913, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 17983 + }, + { + "epoch": 0.17984, + "grad_norm": 1.0152502250754447, + "learning_rate": 0.003, + "loss": 4.016, + "step": 17984 + }, + { + "epoch": 0.17985, + "grad_norm": 1.1065755987717405, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 17985 + }, + { + "epoch": 0.17986, + "grad_norm": 0.7975320978470276, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 17986 + }, + { + "epoch": 0.17987, + "grad_norm": 0.8344879924008952, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 17987 + }, + { + "epoch": 0.17988, + "grad_norm": 0.7879405352182317, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 17988 + }, + { + "epoch": 0.17989, + "grad_norm": 0.7664435928897555, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 17989 + }, + { + "epoch": 0.1799, + "grad_norm": 0.7553230186631742, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 17990 + }, + { + "epoch": 0.17991, + "grad_norm": 0.8208833567938487, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 17991 + }, + { + "epoch": 0.17992, + "grad_norm": 0.9827467288759022, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 17992 + }, + { + "epoch": 0.17993, + "grad_norm": 1.089355252162156, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 17993 + }, + { + "epoch": 0.17994, + "grad_norm": 0.8635295671886025, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 17994 + }, + { + "epoch": 0.17995, + "grad_norm": 0.8898697083007765, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 17995 + }, + { + "epoch": 0.17996, + "grad_norm": 0.9771708661068259, + "learning_rate": 0.003, + "loss": 3.999, + "step": 17996 + }, + { + "epoch": 0.17997, + "grad_norm": 1.0142568044476727, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 17997 + }, + { + "epoch": 0.17998, + "grad_norm": 1.0721466099351542, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 17998 + }, + { + "epoch": 0.17999, + "grad_norm": 0.8036036345937688, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 17999 + }, + { + "epoch": 0.18, + "grad_norm": 0.7643796556897622, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 18000 + }, + { + "epoch": 0.18001, + "grad_norm": 0.8821528135632517, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 18001 + }, + { + "epoch": 0.18002, + "grad_norm": 0.854638941981803, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 18002 + }, + { + "epoch": 0.18003, + "grad_norm": 0.8743062960092784, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 18003 + }, + { + "epoch": 0.18004, + "grad_norm": 0.7916022347463382, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 18004 + }, + { + "epoch": 0.18005, + "grad_norm": 0.730008381192327, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 18005 + }, + { + "epoch": 0.18006, + "grad_norm": 0.7675106471500588, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 18006 + }, + { + "epoch": 0.18007, + "grad_norm": 0.7434804568626352, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 18007 + }, + { + "epoch": 0.18008, + "grad_norm": 0.6573970255262219, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 18008 + }, + { + "epoch": 0.18009, + "grad_norm": 0.5855239349945962, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 18009 + }, + { + "epoch": 0.1801, + "grad_norm": 0.5693741614379216, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 18010 + }, + { + "epoch": 0.18011, + "grad_norm": 0.5476570877410413, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 18011 + }, + { + "epoch": 0.18012, + "grad_norm": 0.572609468019908, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 18012 + }, + { + "epoch": 0.18013, + "grad_norm": 0.5719197049981548, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18013 + }, + { + "epoch": 0.18014, + "grad_norm": 0.575973488312855, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 18014 + }, + { + "epoch": 0.18015, + "grad_norm": 0.5134710209483807, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 18015 + }, + { + "epoch": 0.18016, + "grad_norm": 0.537151621773828, + "learning_rate": 0.003, + "loss": 4.012, + "step": 18016 + }, + { + "epoch": 0.18017, + "grad_norm": 0.519328280476381, + "learning_rate": 0.003, + "loss": 4.0, + "step": 18017 + }, + { + "epoch": 0.18018, + "grad_norm": 0.5317514503886648, + "learning_rate": 0.003, + "loss": 3.992, + "step": 18018 + }, + { + "epoch": 0.18019, + "grad_norm": 0.5529363340316562, + "learning_rate": 0.003, + "loss": 3.981, + "step": 18019 + }, + { + "epoch": 0.1802, + "grad_norm": 0.6310609284343959, + "learning_rate": 0.003, + "loss": 4.002, + "step": 18020 + }, + { + "epoch": 0.18021, + "grad_norm": 0.7256137521990095, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 18021 + }, + { + "epoch": 0.18022, + "grad_norm": 0.7876427348065413, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 18022 + }, + { + "epoch": 0.18023, + "grad_norm": 0.8092270668730133, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 18023 + }, + { + "epoch": 0.18024, + "grad_norm": 0.747094919006115, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 18024 + }, + { + "epoch": 0.18025, + "grad_norm": 0.6824696482635638, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 18025 + }, + { + "epoch": 0.18026, + "grad_norm": 0.7362805051296984, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 18026 + }, + { + "epoch": 0.18027, + "grad_norm": 0.8931410110178623, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 18027 + }, + { + "epoch": 0.18028, + "grad_norm": 1.0125375904668168, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 18028 + }, + { + "epoch": 0.18029, + "grad_norm": 1.0347460705696219, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 18029 + }, + { + "epoch": 0.1803, + "grad_norm": 0.8555532678479654, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 18030 + }, + { + "epoch": 0.18031, + "grad_norm": 0.8853875225165655, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18031 + }, + { + "epoch": 0.18032, + "grad_norm": 0.8414419313470975, + "learning_rate": 0.003, + "loss": 4.0781, + "step": 18032 + }, + { + "epoch": 0.18033, + "grad_norm": 0.8871802127009487, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18033 + }, + { + "epoch": 0.18034, + "grad_norm": 0.82909086381241, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 18034 + }, + { + "epoch": 0.18035, + "grad_norm": 0.7998777958875034, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 18035 + }, + { + "epoch": 0.18036, + "grad_norm": 0.7584463630647444, + "learning_rate": 0.003, + "loss": 4.03, + "step": 18036 + }, + { + "epoch": 0.18037, + "grad_norm": 0.7575325714418603, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18037 + }, + { + "epoch": 0.18038, + "grad_norm": 0.7338710526097817, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18038 + }, + { + "epoch": 0.18039, + "grad_norm": 0.8489629210839104, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18039 + }, + { + "epoch": 0.1804, + "grad_norm": 0.789063811345953, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 18040 + }, + { + "epoch": 0.18041, + "grad_norm": 0.707015266710935, + "learning_rate": 0.003, + "loss": 4.008, + "step": 18041 + }, + { + "epoch": 0.18042, + "grad_norm": 0.7621086157163013, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 18042 + }, + { + "epoch": 0.18043, + "grad_norm": 0.819513537306387, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 18043 + }, + { + "epoch": 0.18044, + "grad_norm": 0.9738203323169643, + "learning_rate": 0.003, + "loss": 4.0647, + "step": 18044 + }, + { + "epoch": 0.18045, + "grad_norm": 1.0535899057795135, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 18045 + }, + { + "epoch": 0.18046, + "grad_norm": 0.9606283155578001, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 18046 + }, + { + "epoch": 0.18047, + "grad_norm": 0.822718525854737, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 18047 + }, + { + "epoch": 0.18048, + "grad_norm": 0.7030131645033313, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 18048 + }, + { + "epoch": 0.18049, + "grad_norm": 0.6566420508376823, + "learning_rate": 0.003, + "loss": 4.04, + "step": 18049 + }, + { + "epoch": 0.1805, + "grad_norm": 0.6830561121934892, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 18050 + }, + { + "epoch": 0.18051, + "grad_norm": 0.6362344922765131, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 18051 + }, + { + "epoch": 0.18052, + "grad_norm": 0.6090146730597985, + "learning_rate": 0.003, + "loss": 4.019, + "step": 18052 + }, + { + "epoch": 0.18053, + "grad_norm": 0.5785991604171256, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 18053 + }, + { + "epoch": 0.18054, + "grad_norm": 0.499573435272377, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 18054 + }, + { + "epoch": 0.18055, + "grad_norm": 0.49699663381700787, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 18055 + }, + { + "epoch": 0.18056, + "grad_norm": 0.510431021195116, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 18056 + }, + { + "epoch": 0.18057, + "grad_norm": 0.4672495617167407, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 18057 + }, + { + "epoch": 0.18058, + "grad_norm": 0.5163692020669651, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 18058 + }, + { + "epoch": 0.18059, + "grad_norm": 0.6018937098704723, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 18059 + }, + { + "epoch": 0.1806, + "grad_norm": 0.712213767983374, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 18060 + }, + { + "epoch": 0.18061, + "grad_norm": 0.8268499323820055, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 18061 + }, + { + "epoch": 0.18062, + "grad_norm": 1.0022025736240099, + "learning_rate": 0.003, + "loss": 4.004, + "step": 18062 + }, + { + "epoch": 0.18063, + "grad_norm": 0.9656596202360868, + "learning_rate": 0.003, + "loss": 4.001, + "step": 18063 + }, + { + "epoch": 0.18064, + "grad_norm": 0.6612547228470589, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 18064 + }, + { + "epoch": 0.18065, + "grad_norm": 0.6486731273756434, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18065 + }, + { + "epoch": 0.18066, + "grad_norm": 0.9571993039266985, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 18066 + }, + { + "epoch": 0.18067, + "grad_norm": 1.0029683415715365, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 18067 + }, + { + "epoch": 0.18068, + "grad_norm": 0.7413392802404187, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 18068 + }, + { + "epoch": 0.18069, + "grad_norm": 0.7915947461595811, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 18069 + }, + { + "epoch": 0.1807, + "grad_norm": 0.8059236157060583, + "learning_rate": 0.003, + "loss": 3.987, + "step": 18070 + }, + { + "epoch": 0.18071, + "grad_norm": 0.7516196682616636, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 18071 + }, + { + "epoch": 0.18072, + "grad_norm": 0.7948335832194884, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 18072 + }, + { + "epoch": 0.18073, + "grad_norm": 0.8209788163024052, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 18073 + }, + { + "epoch": 0.18074, + "grad_norm": 0.7485029411178221, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 18074 + }, + { + "epoch": 0.18075, + "grad_norm": 0.7368551033898066, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 18075 + }, + { + "epoch": 0.18076, + "grad_norm": 0.8499715996601263, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 18076 + }, + { + "epoch": 0.18077, + "grad_norm": 0.9204013173178125, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 18077 + }, + { + "epoch": 0.18078, + "grad_norm": 0.7993237591046582, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 18078 + }, + { + "epoch": 0.18079, + "grad_norm": 0.8487882741076059, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18079 + }, + { + "epoch": 0.1808, + "grad_norm": 0.7697898559268782, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 18080 + }, + { + "epoch": 0.18081, + "grad_norm": 0.7113251617406767, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 18081 + }, + { + "epoch": 0.18082, + "grad_norm": 0.6735406601059762, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 18082 + }, + { + "epoch": 0.18083, + "grad_norm": 0.6303128245768557, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 18083 + }, + { + "epoch": 0.18084, + "grad_norm": 0.6520309789974743, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 18084 + }, + { + "epoch": 0.18085, + "grad_norm": 0.5641481035743103, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 18085 + }, + { + "epoch": 0.18086, + "grad_norm": 0.5459680811531777, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 18086 + }, + { + "epoch": 0.18087, + "grad_norm": 0.5247828047348102, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 18087 + }, + { + "epoch": 0.18088, + "grad_norm": 0.49845293627830345, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 18088 + }, + { + "epoch": 0.18089, + "grad_norm": 0.604818864268055, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 18089 + }, + { + "epoch": 0.1809, + "grad_norm": 0.7599812647307324, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 18090 + }, + { + "epoch": 0.18091, + "grad_norm": 0.9469370608294416, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18091 + }, + { + "epoch": 0.18092, + "grad_norm": 0.846058531736194, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 18092 + }, + { + "epoch": 0.18093, + "grad_norm": 0.6529722508656386, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 18093 + }, + { + "epoch": 0.18094, + "grad_norm": 0.6474363213608875, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 18094 + }, + { + "epoch": 0.18095, + "grad_norm": 0.8864681878562892, + "learning_rate": 0.003, + "loss": 3.977, + "step": 18095 + }, + { + "epoch": 0.18096, + "grad_norm": 0.9527359662561671, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 18096 + }, + { + "epoch": 0.18097, + "grad_norm": 0.7344307503947034, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 18097 + }, + { + "epoch": 0.18098, + "grad_norm": 0.720394904716429, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 18098 + }, + { + "epoch": 0.18099, + "grad_norm": 0.667568265501774, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 18099 + }, + { + "epoch": 0.181, + "grad_norm": 0.7238100650245455, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 18100 + }, + { + "epoch": 0.18101, + "grad_norm": 0.6984936837782607, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 18101 + }, + { + "epoch": 0.18102, + "grad_norm": 0.5965111545657741, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 18102 + }, + { + "epoch": 0.18103, + "grad_norm": 0.6763002079171697, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 18103 + }, + { + "epoch": 0.18104, + "grad_norm": 0.6936014857097839, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 18104 + }, + { + "epoch": 0.18105, + "grad_norm": 0.632640917980509, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 18105 + }, + { + "epoch": 0.18106, + "grad_norm": 0.578134275169601, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 18106 + }, + { + "epoch": 0.18107, + "grad_norm": 0.5911864059845159, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 18107 + }, + { + "epoch": 0.18108, + "grad_norm": 0.5693994106342609, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 18108 + }, + { + "epoch": 0.18109, + "grad_norm": 0.6374026597043138, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 18109 + }, + { + "epoch": 0.1811, + "grad_norm": 0.694998643407546, + "learning_rate": 0.003, + "loss": 4.036, + "step": 18110 + }, + { + "epoch": 0.18111, + "grad_norm": 0.7063512335442994, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 18111 + }, + { + "epoch": 0.18112, + "grad_norm": 0.6365234287774746, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 18112 + }, + { + "epoch": 0.18113, + "grad_norm": 0.6889936174450955, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 18113 + }, + { + "epoch": 0.18114, + "grad_norm": 0.6641654820395331, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 18114 + }, + { + "epoch": 0.18115, + "grad_norm": 0.6054636651903239, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 18115 + }, + { + "epoch": 0.18116, + "grad_norm": 0.7173395595429682, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 18116 + }, + { + "epoch": 0.18117, + "grad_norm": 0.7390465685136314, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 18117 + }, + { + "epoch": 0.18118, + "grad_norm": 0.8544039421169902, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 18118 + }, + { + "epoch": 0.18119, + "grad_norm": 1.016421636341858, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 18119 + }, + { + "epoch": 0.1812, + "grad_norm": 0.8305000483900141, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18120 + }, + { + "epoch": 0.18121, + "grad_norm": 0.6710607454970792, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 18121 + }, + { + "epoch": 0.18122, + "grad_norm": 0.7375772505953633, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 18122 + }, + { + "epoch": 0.18123, + "grad_norm": 0.6797009507066167, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 18123 + }, + { + "epoch": 0.18124, + "grad_norm": 0.7790003304266641, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 18124 + }, + { + "epoch": 0.18125, + "grad_norm": 0.8416637887226556, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 18125 + }, + { + "epoch": 0.18126, + "grad_norm": 0.9292290314075363, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 18126 + }, + { + "epoch": 0.18127, + "grad_norm": 0.9620486586417466, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 18127 + }, + { + "epoch": 0.18128, + "grad_norm": 1.0025751568353911, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 18128 + }, + { + "epoch": 0.18129, + "grad_norm": 0.9132743139512034, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 18129 + }, + { + "epoch": 0.1813, + "grad_norm": 0.8490981796515266, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 18130 + }, + { + "epoch": 0.18131, + "grad_norm": 0.8966997758743974, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18131 + }, + { + "epoch": 0.18132, + "grad_norm": 0.846722930529131, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 18132 + }, + { + "epoch": 0.18133, + "grad_norm": 0.7879753448416549, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 18133 + }, + { + "epoch": 0.18134, + "grad_norm": 0.7800512750177002, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 18134 + }, + { + "epoch": 0.18135, + "grad_norm": 0.7536081172843799, + "learning_rate": 0.003, + "loss": 4.019, + "step": 18135 + }, + { + "epoch": 0.18136, + "grad_norm": 0.768673617051232, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 18136 + }, + { + "epoch": 0.18137, + "grad_norm": 0.8017129341205481, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 18137 + }, + { + "epoch": 0.18138, + "grad_norm": 0.7515953006594525, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 18138 + }, + { + "epoch": 0.18139, + "grad_norm": 0.7091580127789191, + "learning_rate": 0.003, + "loss": 4.008, + "step": 18139 + }, + { + "epoch": 0.1814, + "grad_norm": 0.657264915840839, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 18140 + }, + { + "epoch": 0.18141, + "grad_norm": 0.6002315578071462, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 18141 + }, + { + "epoch": 0.18142, + "grad_norm": 0.5868339483251617, + "learning_rate": 0.003, + "loss": 3.987, + "step": 18142 + }, + { + "epoch": 0.18143, + "grad_norm": 0.5479197505796369, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 18143 + }, + { + "epoch": 0.18144, + "grad_norm": 0.5925250831888058, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 18144 + }, + { + "epoch": 0.18145, + "grad_norm": 0.7495400896971207, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18145 + }, + { + "epoch": 0.18146, + "grad_norm": 0.8689663195435204, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 18146 + }, + { + "epoch": 0.18147, + "grad_norm": 0.9026128953491468, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 18147 + }, + { + "epoch": 0.18148, + "grad_norm": 0.7866971004031295, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 18148 + }, + { + "epoch": 0.18149, + "grad_norm": 0.6849801986218305, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18149 + }, + { + "epoch": 0.1815, + "grad_norm": 0.6584097572279665, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 18150 + }, + { + "epoch": 0.18151, + "grad_norm": 0.545464058571609, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 18151 + }, + { + "epoch": 0.18152, + "grad_norm": 0.5192004890296824, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 18152 + }, + { + "epoch": 0.18153, + "grad_norm": 0.5156220233593475, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 18153 + }, + { + "epoch": 0.18154, + "grad_norm": 0.5486602669825582, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 18154 + }, + { + "epoch": 0.18155, + "grad_norm": 0.6436293368858059, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18155 + }, + { + "epoch": 0.18156, + "grad_norm": 0.6536231576472259, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 18156 + }, + { + "epoch": 0.18157, + "grad_norm": 0.633543327862493, + "learning_rate": 0.003, + "loss": 4.011, + "step": 18157 + }, + { + "epoch": 0.18158, + "grad_norm": 0.743713660781897, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 18158 + }, + { + "epoch": 0.18159, + "grad_norm": 0.8133702293863125, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 18159 + }, + { + "epoch": 0.1816, + "grad_norm": 0.8821233402511396, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 18160 + }, + { + "epoch": 0.18161, + "grad_norm": 1.0270869596776613, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 18161 + }, + { + "epoch": 0.18162, + "grad_norm": 0.8562985694113451, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 18162 + }, + { + "epoch": 0.18163, + "grad_norm": 0.6896384647427507, + "learning_rate": 0.003, + "loss": 4.003, + "step": 18163 + }, + { + "epoch": 0.18164, + "grad_norm": 0.7350609701051134, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 18164 + }, + { + "epoch": 0.18165, + "grad_norm": 0.6590528519718943, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 18165 + }, + { + "epoch": 0.18166, + "grad_norm": 0.6004316345378495, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 18166 + }, + { + "epoch": 0.18167, + "grad_norm": 0.6439573150088764, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 18167 + }, + { + "epoch": 0.18168, + "grad_norm": 0.6542516770962303, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18168 + }, + { + "epoch": 0.18169, + "grad_norm": 0.7339656967214558, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 18169 + }, + { + "epoch": 0.1817, + "grad_norm": 0.8449865847230582, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 18170 + }, + { + "epoch": 0.18171, + "grad_norm": 0.9713326270473798, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 18171 + }, + { + "epoch": 0.18172, + "grad_norm": 1.0436397357833822, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 18172 + }, + { + "epoch": 0.18173, + "grad_norm": 0.8874373491848969, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 18173 + }, + { + "epoch": 0.18174, + "grad_norm": 0.8718669201581295, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 18174 + }, + { + "epoch": 0.18175, + "grad_norm": 0.7988029007112151, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 18175 + }, + { + "epoch": 0.18176, + "grad_norm": 0.7156055088099931, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 18176 + }, + { + "epoch": 0.18177, + "grad_norm": 0.8261209195674091, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 18177 + }, + { + "epoch": 0.18178, + "grad_norm": 0.846987257940766, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 18178 + }, + { + "epoch": 0.18179, + "grad_norm": 0.9027083591041437, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 18179 + }, + { + "epoch": 0.1818, + "grad_norm": 0.9177574735315799, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 18180 + }, + { + "epoch": 0.18181, + "grad_norm": 0.9458178748439781, + "learning_rate": 0.003, + "loss": 4.018, + "step": 18181 + }, + { + "epoch": 0.18182, + "grad_norm": 0.8481760041313409, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 18182 + }, + { + "epoch": 0.18183, + "grad_norm": 0.7619602406907509, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 18183 + }, + { + "epoch": 0.18184, + "grad_norm": 0.799643344930933, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 18184 + }, + { + "epoch": 0.18185, + "grad_norm": 0.7620006451776655, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 18185 + }, + { + "epoch": 0.18186, + "grad_norm": 0.7448042203699542, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 18186 + }, + { + "epoch": 0.18187, + "grad_norm": 0.7670383504443742, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 18187 + }, + { + "epoch": 0.18188, + "grad_norm": 0.7852961293097434, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 18188 + }, + { + "epoch": 0.18189, + "grad_norm": 0.8557565088361174, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 18189 + }, + { + "epoch": 0.1819, + "grad_norm": 0.7820493759674344, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 18190 + }, + { + "epoch": 0.18191, + "grad_norm": 0.9277866843148828, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 18191 + }, + { + "epoch": 0.18192, + "grad_norm": 0.8734367126641327, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 18192 + }, + { + "epoch": 0.18193, + "grad_norm": 0.762310057178126, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 18193 + }, + { + "epoch": 0.18194, + "grad_norm": 0.7095465769725653, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 18194 + }, + { + "epoch": 0.18195, + "grad_norm": 0.5565042843846906, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 18195 + }, + { + "epoch": 0.18196, + "grad_norm": 0.660455685423448, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 18196 + }, + { + "epoch": 0.18197, + "grad_norm": 0.6418763504384927, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 18197 + }, + { + "epoch": 0.18198, + "grad_norm": 0.655411205098473, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 18198 + }, + { + "epoch": 0.18199, + "grad_norm": 0.5899442691018162, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 18199 + }, + { + "epoch": 0.182, + "grad_norm": 0.5729480210690719, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 18200 + }, + { + "epoch": 0.18201, + "grad_norm": 0.589025434189014, + "learning_rate": 0.003, + "loss": 4.016, + "step": 18201 + }, + { + "epoch": 0.18202, + "grad_norm": 0.6885630833352038, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 18202 + }, + { + "epoch": 0.18203, + "grad_norm": 0.8598941953333745, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 18203 + }, + { + "epoch": 0.18204, + "grad_norm": 1.0713951623185236, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18204 + }, + { + "epoch": 0.18205, + "grad_norm": 0.9705656332964194, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 18205 + }, + { + "epoch": 0.18206, + "grad_norm": 0.6355771540711015, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 18206 + }, + { + "epoch": 0.18207, + "grad_norm": 0.5937069979306931, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 18207 + }, + { + "epoch": 0.18208, + "grad_norm": 0.6746796479955997, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 18208 + }, + { + "epoch": 0.18209, + "grad_norm": 0.7702524414215337, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 18209 + }, + { + "epoch": 0.1821, + "grad_norm": 0.7855482974023985, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18210 + }, + { + "epoch": 0.18211, + "grad_norm": 0.7703531149231344, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 18211 + }, + { + "epoch": 0.18212, + "grad_norm": 0.5820281319665591, + "learning_rate": 0.003, + "loss": 3.985, + "step": 18212 + }, + { + "epoch": 0.18213, + "grad_norm": 0.5405584056548465, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 18213 + }, + { + "epoch": 0.18214, + "grad_norm": 0.587201877666789, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 18214 + }, + { + "epoch": 0.18215, + "grad_norm": 0.7109465564823296, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 18215 + }, + { + "epoch": 0.18216, + "grad_norm": 0.741248800979584, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 18216 + }, + { + "epoch": 0.18217, + "grad_norm": 0.7252097092438075, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 18217 + }, + { + "epoch": 0.18218, + "grad_norm": 0.649230256259267, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 18218 + }, + { + "epoch": 0.18219, + "grad_norm": 0.6555633647353231, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 18219 + }, + { + "epoch": 0.1822, + "grad_norm": 0.7047827384962382, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 18220 + }, + { + "epoch": 0.18221, + "grad_norm": 0.7040163065894106, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 18221 + }, + { + "epoch": 0.18222, + "grad_norm": 0.6990586096757019, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18222 + }, + { + "epoch": 0.18223, + "grad_norm": 0.7165806849366285, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 18223 + }, + { + "epoch": 0.18224, + "grad_norm": 0.6711634601820953, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 18224 + }, + { + "epoch": 0.18225, + "grad_norm": 0.7250258297778163, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 18225 + }, + { + "epoch": 0.18226, + "grad_norm": 0.8049221656904009, + "learning_rate": 0.003, + "loss": 4.024, + "step": 18226 + }, + { + "epoch": 0.18227, + "grad_norm": 0.87019297573573, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 18227 + }, + { + "epoch": 0.18228, + "grad_norm": 0.8315471913020742, + "learning_rate": 0.003, + "loss": 4.02, + "step": 18228 + }, + { + "epoch": 0.18229, + "grad_norm": 0.7912251587370118, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 18229 + }, + { + "epoch": 0.1823, + "grad_norm": 0.7632283330944678, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 18230 + }, + { + "epoch": 0.18231, + "grad_norm": 0.8013475236488635, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 18231 + }, + { + "epoch": 0.18232, + "grad_norm": 0.7964141097127816, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 18232 + }, + { + "epoch": 0.18233, + "grad_norm": 0.8575257112422066, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 18233 + }, + { + "epoch": 0.18234, + "grad_norm": 0.8139771131065391, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 18234 + }, + { + "epoch": 0.18235, + "grad_norm": 0.9397315459546036, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 18235 + }, + { + "epoch": 0.18236, + "grad_norm": 1.0951918714093924, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 18236 + }, + { + "epoch": 0.18237, + "grad_norm": 0.8428202429745868, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 18237 + }, + { + "epoch": 0.18238, + "grad_norm": 0.7104388732606719, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 18238 + }, + { + "epoch": 0.18239, + "grad_norm": 0.6691105754154593, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 18239 + }, + { + "epoch": 0.1824, + "grad_norm": 0.6672351500037259, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 18240 + }, + { + "epoch": 0.18241, + "grad_norm": 0.6233565804435698, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 18241 + }, + { + "epoch": 0.18242, + "grad_norm": 0.6949530050641614, + "learning_rate": 0.003, + "loss": 3.966, + "step": 18242 + }, + { + "epoch": 0.18243, + "grad_norm": 0.7497304905341359, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 18243 + }, + { + "epoch": 0.18244, + "grad_norm": 0.7174164012508055, + "learning_rate": 0.003, + "loss": 4.008, + "step": 18244 + }, + { + "epoch": 0.18245, + "grad_norm": 0.6098804459773091, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 18245 + }, + { + "epoch": 0.18246, + "grad_norm": 0.56684300952393, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 18246 + }, + { + "epoch": 0.18247, + "grad_norm": 0.6880126953722274, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 18247 + }, + { + "epoch": 0.18248, + "grad_norm": 0.7901568649738481, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 18248 + }, + { + "epoch": 0.18249, + "grad_norm": 0.6819559302255755, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 18249 + }, + { + "epoch": 0.1825, + "grad_norm": 0.6618308247016369, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 18250 + }, + { + "epoch": 0.18251, + "grad_norm": 0.7241483106839138, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 18251 + }, + { + "epoch": 0.18252, + "grad_norm": 0.8122920937510538, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 18252 + }, + { + "epoch": 0.18253, + "grad_norm": 0.7722701906594204, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 18253 + }, + { + "epoch": 0.18254, + "grad_norm": 0.7710249341145875, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 18254 + }, + { + "epoch": 0.18255, + "grad_norm": 0.775587382732904, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 18255 + }, + { + "epoch": 0.18256, + "grad_norm": 0.8482215618625716, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 18256 + }, + { + "epoch": 0.18257, + "grad_norm": 0.8987620587641557, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 18257 + }, + { + "epoch": 0.18258, + "grad_norm": 0.9199539226685443, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 18258 + }, + { + "epoch": 0.18259, + "grad_norm": 0.9143838411562056, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 18259 + }, + { + "epoch": 0.1826, + "grad_norm": 0.88376581050473, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 18260 + }, + { + "epoch": 0.18261, + "grad_norm": 0.7664750608810522, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 18261 + }, + { + "epoch": 0.18262, + "grad_norm": 0.738088594692022, + "learning_rate": 0.003, + "loss": 4.005, + "step": 18262 + }, + { + "epoch": 0.18263, + "grad_norm": 0.7458376585575259, + "learning_rate": 0.003, + "loss": 4.016, + "step": 18263 + }, + { + "epoch": 0.18264, + "grad_norm": 0.7932161387617875, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 18264 + }, + { + "epoch": 0.18265, + "grad_norm": 0.8672186388185318, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 18265 + }, + { + "epoch": 0.18266, + "grad_norm": 0.9129484812280112, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 18266 + }, + { + "epoch": 0.18267, + "grad_norm": 0.7769380780593883, + "learning_rate": 0.003, + "loss": 4.021, + "step": 18267 + }, + { + "epoch": 0.18268, + "grad_norm": 0.6736874706187709, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18268 + }, + { + "epoch": 0.18269, + "grad_norm": 0.7138610950086627, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 18269 + }, + { + "epoch": 0.1827, + "grad_norm": 0.714461755208027, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 18270 + }, + { + "epoch": 0.18271, + "grad_norm": 0.796275214953874, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 18271 + }, + { + "epoch": 0.18272, + "grad_norm": 0.857511615216595, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 18272 + }, + { + "epoch": 0.18273, + "grad_norm": 0.8604551735722271, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 18273 + }, + { + "epoch": 0.18274, + "grad_norm": 0.8756137581731602, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 18274 + }, + { + "epoch": 0.18275, + "grad_norm": 1.035622431762434, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 18275 + }, + { + "epoch": 0.18276, + "grad_norm": 1.042950069842768, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 18276 + }, + { + "epoch": 0.18277, + "grad_norm": 0.7731844358886756, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 18277 + }, + { + "epoch": 0.18278, + "grad_norm": 0.622892354570165, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18278 + }, + { + "epoch": 0.18279, + "grad_norm": 0.6519486327542536, + "learning_rate": 0.003, + "loss": 4.017, + "step": 18279 + }, + { + "epoch": 0.1828, + "grad_norm": 0.6041559486688692, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 18280 + }, + { + "epoch": 0.18281, + "grad_norm": 0.6304966834391075, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 18281 + }, + { + "epoch": 0.18282, + "grad_norm": 0.6529850284746352, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 18282 + }, + { + "epoch": 0.18283, + "grad_norm": 0.597321020846399, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18283 + }, + { + "epoch": 0.18284, + "grad_norm": 0.5957545840823446, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 18284 + }, + { + "epoch": 0.18285, + "grad_norm": 0.6051733204221227, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 18285 + }, + { + "epoch": 0.18286, + "grad_norm": 0.5335878292955782, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18286 + }, + { + "epoch": 0.18287, + "grad_norm": 0.517970979611633, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 18287 + }, + { + "epoch": 0.18288, + "grad_norm": 0.5320560630017307, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 18288 + }, + { + "epoch": 0.18289, + "grad_norm": 0.5755933034178425, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 18289 + }, + { + "epoch": 0.1829, + "grad_norm": 0.5680354008870699, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 18290 + }, + { + "epoch": 0.18291, + "grad_norm": 0.6072966661787136, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 18291 + }, + { + "epoch": 0.18292, + "grad_norm": 0.701646907490007, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 18292 + }, + { + "epoch": 0.18293, + "grad_norm": 0.8875417414264732, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 18293 + }, + { + "epoch": 0.18294, + "grad_norm": 1.0946425023497532, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 18294 + }, + { + "epoch": 0.18295, + "grad_norm": 0.7682767064062607, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 18295 + }, + { + "epoch": 0.18296, + "grad_norm": 0.5955996906721339, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 18296 + }, + { + "epoch": 0.18297, + "grad_norm": 0.6995352717173645, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 18297 + }, + { + "epoch": 0.18298, + "grad_norm": 0.9103466187269743, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 18298 + }, + { + "epoch": 0.18299, + "grad_norm": 0.960347644374411, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 18299 + }, + { + "epoch": 0.183, + "grad_norm": 0.8303174603222648, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 18300 + }, + { + "epoch": 0.18301, + "grad_norm": 0.813745327702921, + "learning_rate": 0.003, + "loss": 4.012, + "step": 18301 + }, + { + "epoch": 0.18302, + "grad_norm": 0.8205259039707933, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 18302 + }, + { + "epoch": 0.18303, + "grad_norm": 0.8459480804682701, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 18303 + }, + { + "epoch": 0.18304, + "grad_norm": 0.882707177255301, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 18304 + }, + { + "epoch": 0.18305, + "grad_norm": 0.7801212217703488, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 18305 + }, + { + "epoch": 0.18306, + "grad_norm": 0.786178389521359, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 18306 + }, + { + "epoch": 0.18307, + "grad_norm": 0.7504546960599197, + "learning_rate": 0.003, + "loss": 3.986, + "step": 18307 + }, + { + "epoch": 0.18308, + "grad_norm": 0.8117823424619641, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 18308 + }, + { + "epoch": 0.18309, + "grad_norm": 0.8714310877054025, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 18309 + }, + { + "epoch": 0.1831, + "grad_norm": 0.8441347869901277, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 18310 + }, + { + "epoch": 0.18311, + "grad_norm": 0.7963149511437565, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 18311 + }, + { + "epoch": 0.18312, + "grad_norm": 0.7273949011104853, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 18312 + }, + { + "epoch": 0.18313, + "grad_norm": 0.6447283246649246, + "learning_rate": 0.003, + "loss": 3.9369, + "step": 18313 + }, + { + "epoch": 0.18314, + "grad_norm": 0.634682776429222, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 18314 + }, + { + "epoch": 0.18315, + "grad_norm": 0.810361215204305, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 18315 + }, + { + "epoch": 0.18316, + "grad_norm": 0.9460568394653357, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 18316 + }, + { + "epoch": 0.18317, + "grad_norm": 0.8369159085762914, + "learning_rate": 0.003, + "loss": 4.024, + "step": 18317 + }, + { + "epoch": 0.18318, + "grad_norm": 0.7923672182537222, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 18318 + }, + { + "epoch": 0.18319, + "grad_norm": 0.902059620580266, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 18319 + }, + { + "epoch": 0.1832, + "grad_norm": 0.875966982406173, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 18320 + }, + { + "epoch": 0.18321, + "grad_norm": 0.7523953483035991, + "learning_rate": 0.003, + "loss": 4.0708, + "step": 18321 + }, + { + "epoch": 0.18322, + "grad_norm": 0.8076990835825204, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 18322 + }, + { + "epoch": 0.18323, + "grad_norm": 0.872523648894515, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 18323 + }, + { + "epoch": 0.18324, + "grad_norm": 0.8602455724927075, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 18324 + }, + { + "epoch": 0.18325, + "grad_norm": 0.754993832594747, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 18325 + }, + { + "epoch": 0.18326, + "grad_norm": 0.6993453821554478, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 18326 + }, + { + "epoch": 0.18327, + "grad_norm": 0.6067207789035304, + "learning_rate": 0.003, + "loss": 4.0, + "step": 18327 + }, + { + "epoch": 0.18328, + "grad_norm": 0.647729789171814, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 18328 + }, + { + "epoch": 0.18329, + "grad_norm": 0.7815824074476979, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 18329 + }, + { + "epoch": 0.1833, + "grad_norm": 0.9983838373083846, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 18330 + }, + { + "epoch": 0.18331, + "grad_norm": 1.1338263142399814, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 18331 + }, + { + "epoch": 0.18332, + "grad_norm": 0.642041223601514, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18332 + }, + { + "epoch": 0.18333, + "grad_norm": 0.6636287669901222, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 18333 + }, + { + "epoch": 0.18334, + "grad_norm": 0.7526536244039632, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 18334 + }, + { + "epoch": 0.18335, + "grad_norm": 0.7649761255670432, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 18335 + }, + { + "epoch": 0.18336, + "grad_norm": 0.8503378821047559, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 18336 + }, + { + "epoch": 0.18337, + "grad_norm": 0.8964983399758536, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 18337 + }, + { + "epoch": 0.18338, + "grad_norm": 0.853145082186502, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 18338 + }, + { + "epoch": 0.18339, + "grad_norm": 0.7068367739706798, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 18339 + }, + { + "epoch": 0.1834, + "grad_norm": 0.7355218859365442, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 18340 + }, + { + "epoch": 0.18341, + "grad_norm": 0.6523743927762076, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 18341 + }, + { + "epoch": 0.18342, + "grad_norm": 0.6871708761900814, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 18342 + }, + { + "epoch": 0.18343, + "grad_norm": 0.6697001529628921, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 18343 + }, + { + "epoch": 0.18344, + "grad_norm": 0.7040177326191914, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 18344 + }, + { + "epoch": 0.18345, + "grad_norm": 0.6415528664839508, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 18345 + }, + { + "epoch": 0.18346, + "grad_norm": 0.5262562217388835, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 18346 + }, + { + "epoch": 0.18347, + "grad_norm": 0.4827513586142221, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 18347 + }, + { + "epoch": 0.18348, + "grad_norm": 0.5013778057307486, + "learning_rate": 0.003, + "loss": 4.017, + "step": 18348 + }, + { + "epoch": 0.18349, + "grad_norm": 0.5093917247694251, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 18349 + }, + { + "epoch": 0.1835, + "grad_norm": 0.4799901065071378, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 18350 + }, + { + "epoch": 0.18351, + "grad_norm": 0.6067893311662568, + "learning_rate": 0.003, + "loss": 4.002, + "step": 18351 + }, + { + "epoch": 0.18352, + "grad_norm": 0.6587169970214978, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 18352 + }, + { + "epoch": 0.18353, + "grad_norm": 0.7199780149238035, + "learning_rate": 0.003, + "loss": 4.004, + "step": 18353 + }, + { + "epoch": 0.18354, + "grad_norm": 0.8002879008860198, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 18354 + }, + { + "epoch": 0.18355, + "grad_norm": 0.8204552631058327, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18355 + }, + { + "epoch": 0.18356, + "grad_norm": 0.691196611178952, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 18356 + }, + { + "epoch": 0.18357, + "grad_norm": 0.5628600463727601, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 18357 + }, + { + "epoch": 0.18358, + "grad_norm": 0.4604152607591477, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 18358 + }, + { + "epoch": 0.18359, + "grad_norm": 0.49753437911702464, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 18359 + }, + { + "epoch": 0.1836, + "grad_norm": 0.5457434478971858, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 18360 + }, + { + "epoch": 0.18361, + "grad_norm": 0.6067623717948145, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 18361 + }, + { + "epoch": 0.18362, + "grad_norm": 0.6806053609022452, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 18362 + }, + { + "epoch": 0.18363, + "grad_norm": 0.8132563711681287, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 18363 + }, + { + "epoch": 0.18364, + "grad_norm": 0.8784422120694079, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 18364 + }, + { + "epoch": 0.18365, + "grad_norm": 1.0040190303790226, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 18365 + }, + { + "epoch": 0.18366, + "grad_norm": 1.1524128026100562, + "learning_rate": 0.003, + "loss": 4.017, + "step": 18366 + }, + { + "epoch": 0.18367, + "grad_norm": 0.9597016459887963, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 18367 + }, + { + "epoch": 0.18368, + "grad_norm": 0.9231604348894202, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 18368 + }, + { + "epoch": 0.18369, + "grad_norm": 0.8147942173977889, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 18369 + }, + { + "epoch": 0.1837, + "grad_norm": 0.7202284708182317, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 18370 + }, + { + "epoch": 0.18371, + "grad_norm": 0.6948919807685212, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 18371 + }, + { + "epoch": 0.18372, + "grad_norm": 0.6386059290318806, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 18372 + }, + { + "epoch": 0.18373, + "grad_norm": 0.6272118917221939, + "learning_rate": 0.003, + "loss": 4.026, + "step": 18373 + }, + { + "epoch": 0.18374, + "grad_norm": 0.6224776449110274, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18374 + }, + { + "epoch": 0.18375, + "grad_norm": 0.6397722722375582, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18375 + }, + { + "epoch": 0.18376, + "grad_norm": 0.5954944486315152, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 18376 + }, + { + "epoch": 0.18377, + "grad_norm": 0.6017773420968536, + "learning_rate": 0.003, + "loss": 3.979, + "step": 18377 + }, + { + "epoch": 0.18378, + "grad_norm": 0.6703279322882283, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 18378 + }, + { + "epoch": 0.18379, + "grad_norm": 0.7858446652014723, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 18379 + }, + { + "epoch": 0.1838, + "grad_norm": 0.7647620423457573, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 18380 + }, + { + "epoch": 0.18381, + "grad_norm": 0.6936290659976733, + "learning_rate": 0.003, + "loss": 4.028, + "step": 18381 + }, + { + "epoch": 0.18382, + "grad_norm": 0.5483218838472684, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 18382 + }, + { + "epoch": 0.18383, + "grad_norm": 0.5618094768503167, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 18383 + }, + { + "epoch": 0.18384, + "grad_norm": 0.7150666294960961, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 18384 + }, + { + "epoch": 0.18385, + "grad_norm": 0.8050393602385894, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 18385 + }, + { + "epoch": 0.18386, + "grad_norm": 0.7329053482709907, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 18386 + }, + { + "epoch": 0.18387, + "grad_norm": 0.6330199483104282, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 18387 + }, + { + "epoch": 0.18388, + "grad_norm": 0.6363890940717956, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 18388 + }, + { + "epoch": 0.18389, + "grad_norm": 0.5304315447039345, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 18389 + }, + { + "epoch": 0.1839, + "grad_norm": 0.476644532828935, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 18390 + }, + { + "epoch": 0.18391, + "grad_norm": 0.593306597783877, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 18391 + }, + { + "epoch": 0.18392, + "grad_norm": 0.8140899950905227, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 18392 + }, + { + "epoch": 0.18393, + "grad_norm": 1.035789845004156, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 18393 + }, + { + "epoch": 0.18394, + "grad_norm": 1.1058513492757815, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 18394 + }, + { + "epoch": 0.18395, + "grad_norm": 0.7358993919469842, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 18395 + }, + { + "epoch": 0.18396, + "grad_norm": 0.7042674953387211, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 18396 + }, + { + "epoch": 0.18397, + "grad_norm": 0.8705795829313966, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 18397 + }, + { + "epoch": 0.18398, + "grad_norm": 0.8886200046692391, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 18398 + }, + { + "epoch": 0.18399, + "grad_norm": 0.8782284294143409, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 18399 + }, + { + "epoch": 0.184, + "grad_norm": 0.8782005394071806, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 18400 + }, + { + "epoch": 0.18401, + "grad_norm": 0.846606521697059, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18401 + }, + { + "epoch": 0.18402, + "grad_norm": 0.898840196247324, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18402 + }, + { + "epoch": 0.18403, + "grad_norm": 0.9125067847525966, + "learning_rate": 0.003, + "loss": 3.983, + "step": 18403 + }, + { + "epoch": 0.18404, + "grad_norm": 0.8889619435574942, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18404 + }, + { + "epoch": 0.18405, + "grad_norm": 0.7654006923456778, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18405 + }, + { + "epoch": 0.18406, + "grad_norm": 0.7123665822194265, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 18406 + }, + { + "epoch": 0.18407, + "grad_norm": 0.6791859748751278, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18407 + }, + { + "epoch": 0.18408, + "grad_norm": 0.6678109008126418, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 18408 + }, + { + "epoch": 0.18409, + "grad_norm": 0.7310082342953522, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 18409 + }, + { + "epoch": 0.1841, + "grad_norm": 0.735597653547832, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 18410 + }, + { + "epoch": 0.18411, + "grad_norm": 0.7411674391946984, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 18411 + }, + { + "epoch": 0.18412, + "grad_norm": 0.8126371544106243, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 18412 + }, + { + "epoch": 0.18413, + "grad_norm": 0.7980326468939581, + "learning_rate": 0.003, + "loss": 3.999, + "step": 18413 + }, + { + "epoch": 0.18414, + "grad_norm": 0.7628681700908239, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 18414 + }, + { + "epoch": 0.18415, + "grad_norm": 0.7754119144366188, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 18415 + }, + { + "epoch": 0.18416, + "grad_norm": 0.8021495130390385, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 18416 + }, + { + "epoch": 0.18417, + "grad_norm": 0.7231655708498632, + "learning_rate": 0.003, + "loss": 4.04, + "step": 18417 + }, + { + "epoch": 0.18418, + "grad_norm": 0.6897591256874965, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 18418 + }, + { + "epoch": 0.18419, + "grad_norm": 0.694810223444298, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 18419 + }, + { + "epoch": 0.1842, + "grad_norm": 0.7103738751975557, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 18420 + }, + { + "epoch": 0.18421, + "grad_norm": 0.7446233373127674, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 18421 + }, + { + "epoch": 0.18422, + "grad_norm": 0.9236465053495261, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 18422 + }, + { + "epoch": 0.18423, + "grad_norm": 1.0950642753218505, + "learning_rate": 0.003, + "loss": 3.998, + "step": 18423 + }, + { + "epoch": 0.18424, + "grad_norm": 0.9481086146452716, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 18424 + }, + { + "epoch": 0.18425, + "grad_norm": 0.850608536782529, + "learning_rate": 0.003, + "loss": 3.977, + "step": 18425 + }, + { + "epoch": 0.18426, + "grad_norm": 0.765750018302042, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 18426 + }, + { + "epoch": 0.18427, + "grad_norm": 0.782849650684836, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 18427 + }, + { + "epoch": 0.18428, + "grad_norm": 0.7969848745676518, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 18428 + }, + { + "epoch": 0.18429, + "grad_norm": 0.7700288278292277, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18429 + }, + { + "epoch": 0.1843, + "grad_norm": 0.7525219468090872, + "learning_rate": 0.003, + "loss": 3.989, + "step": 18430 + }, + { + "epoch": 0.18431, + "grad_norm": 0.6812152401232157, + "learning_rate": 0.003, + "loss": 3.996, + "step": 18431 + }, + { + "epoch": 0.18432, + "grad_norm": 0.6141124461866013, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 18432 + }, + { + "epoch": 0.18433, + "grad_norm": 0.6581829879539546, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 18433 + }, + { + "epoch": 0.18434, + "grad_norm": 0.6814234833534859, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 18434 + }, + { + "epoch": 0.18435, + "grad_norm": 0.7330609457417562, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18435 + }, + { + "epoch": 0.18436, + "grad_norm": 0.7530397280385768, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18436 + }, + { + "epoch": 0.18437, + "grad_norm": 0.6779908787465703, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 18437 + }, + { + "epoch": 0.18438, + "grad_norm": 0.6926013119361407, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 18438 + }, + { + "epoch": 0.18439, + "grad_norm": 0.8042055342326166, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 18439 + }, + { + "epoch": 0.1844, + "grad_norm": 0.7855987262862552, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 18440 + }, + { + "epoch": 0.18441, + "grad_norm": 0.6620275714636101, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 18441 + }, + { + "epoch": 0.18442, + "grad_norm": 0.5714352400505971, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 18442 + }, + { + "epoch": 0.18443, + "grad_norm": 0.5379027505894953, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18443 + }, + { + "epoch": 0.18444, + "grad_norm": 0.5676755345019342, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 18444 + }, + { + "epoch": 0.18445, + "grad_norm": 0.543298485902165, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 18445 + }, + { + "epoch": 0.18446, + "grad_norm": 0.5840554835430268, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 18446 + }, + { + "epoch": 0.18447, + "grad_norm": 0.6965005796833791, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 18447 + }, + { + "epoch": 0.18448, + "grad_norm": 0.8196715672156324, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 18448 + }, + { + "epoch": 0.18449, + "grad_norm": 0.8843281904993552, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 18449 + }, + { + "epoch": 0.1845, + "grad_norm": 1.022502310277498, + "learning_rate": 0.003, + "loss": 3.994, + "step": 18450 + }, + { + "epoch": 0.18451, + "grad_norm": 1.0274960842747154, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 18451 + }, + { + "epoch": 0.18452, + "grad_norm": 0.8391874557153834, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 18452 + }, + { + "epoch": 0.18453, + "grad_norm": 0.738213628970356, + "learning_rate": 0.003, + "loss": 4.05, + "step": 18453 + }, + { + "epoch": 0.18454, + "grad_norm": 0.5713585124597532, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 18454 + }, + { + "epoch": 0.18455, + "grad_norm": 0.6188381198048861, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 18455 + }, + { + "epoch": 0.18456, + "grad_norm": 0.6404367568281909, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 18456 + }, + { + "epoch": 0.18457, + "grad_norm": 0.6412075089082425, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 18457 + }, + { + "epoch": 0.18458, + "grad_norm": 0.6540253536668592, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 18458 + }, + { + "epoch": 0.18459, + "grad_norm": 0.5768360437861895, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 18459 + }, + { + "epoch": 0.1846, + "grad_norm": 0.593805329522252, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 18460 + }, + { + "epoch": 0.18461, + "grad_norm": 0.71871156538892, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 18461 + }, + { + "epoch": 0.18462, + "grad_norm": 0.7836763027084894, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 18462 + }, + { + "epoch": 0.18463, + "grad_norm": 0.7452827617099861, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 18463 + }, + { + "epoch": 0.18464, + "grad_norm": 0.755349915078932, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 18464 + }, + { + "epoch": 0.18465, + "grad_norm": 0.7325062304078261, + "learning_rate": 0.003, + "loss": 4.022, + "step": 18465 + }, + { + "epoch": 0.18466, + "grad_norm": 0.6084019778830516, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 18466 + }, + { + "epoch": 0.18467, + "grad_norm": 0.5782311170871599, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 18467 + }, + { + "epoch": 0.18468, + "grad_norm": 0.5774875569189245, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 18468 + }, + { + "epoch": 0.18469, + "grad_norm": 0.6639847018850916, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 18469 + }, + { + "epoch": 0.1847, + "grad_norm": 0.7560196006041626, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 18470 + }, + { + "epoch": 0.18471, + "grad_norm": 0.8880091005750893, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 18471 + }, + { + "epoch": 0.18472, + "grad_norm": 0.9030983835331943, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 18472 + }, + { + "epoch": 0.18473, + "grad_norm": 1.0142013386252842, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 18473 + }, + { + "epoch": 0.18474, + "grad_norm": 1.1310655001230008, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 18474 + }, + { + "epoch": 0.18475, + "grad_norm": 1.1159271652849585, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 18475 + }, + { + "epoch": 0.18476, + "grad_norm": 0.8877080693120786, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 18476 + }, + { + "epoch": 0.18477, + "grad_norm": 0.9127606856811152, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 18477 + }, + { + "epoch": 0.18478, + "grad_norm": 0.8858856093149953, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 18478 + }, + { + "epoch": 0.18479, + "grad_norm": 0.7901040014399204, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 18479 + }, + { + "epoch": 0.1848, + "grad_norm": 0.7431866161689814, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 18480 + }, + { + "epoch": 0.18481, + "grad_norm": 0.8365388705936032, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 18481 + }, + { + "epoch": 0.18482, + "grad_norm": 0.8417759293600158, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 18482 + }, + { + "epoch": 0.18483, + "grad_norm": 0.7271887369924228, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 18483 + }, + { + "epoch": 0.18484, + "grad_norm": 0.8308610787580221, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 18484 + }, + { + "epoch": 0.18485, + "grad_norm": 0.843688828721087, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 18485 + }, + { + "epoch": 0.18486, + "grad_norm": 0.801788388354645, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 18486 + }, + { + "epoch": 0.18487, + "grad_norm": 0.8712009917703994, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 18487 + }, + { + "epoch": 0.18488, + "grad_norm": 0.930887907804385, + "learning_rate": 0.003, + "loss": 4.019, + "step": 18488 + }, + { + "epoch": 0.18489, + "grad_norm": 1.0832792548618644, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 18489 + }, + { + "epoch": 0.1849, + "grad_norm": 0.7263978054898605, + "learning_rate": 0.003, + "loss": 4.031, + "step": 18490 + }, + { + "epoch": 0.18491, + "grad_norm": 0.6887655954363225, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 18491 + }, + { + "epoch": 0.18492, + "grad_norm": 0.6835790586925067, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 18492 + }, + { + "epoch": 0.18493, + "grad_norm": 0.5921153845476231, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 18493 + }, + { + "epoch": 0.18494, + "grad_norm": 0.5133493356535744, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 18494 + }, + { + "epoch": 0.18495, + "grad_norm": 0.6135352051623705, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 18495 + }, + { + "epoch": 0.18496, + "grad_norm": 0.5978242022197211, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 18496 + }, + { + "epoch": 0.18497, + "grad_norm": 0.5784827496004279, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 18497 + }, + { + "epoch": 0.18498, + "grad_norm": 0.6771708556440991, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 18498 + }, + { + "epoch": 0.18499, + "grad_norm": 0.6823818993959406, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 18499 + }, + { + "epoch": 0.185, + "grad_norm": 0.5757173861732866, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 18500 + }, + { + "epoch": 0.18501, + "grad_norm": 0.5344078914871904, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 18501 + }, + { + "epoch": 0.18502, + "grad_norm": 0.5135388789872529, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18502 + }, + { + "epoch": 0.18503, + "grad_norm": 0.5914956200405949, + "learning_rate": 0.003, + "loss": 4.002, + "step": 18503 + }, + { + "epoch": 0.18504, + "grad_norm": 0.6694341261184009, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 18504 + }, + { + "epoch": 0.18505, + "grad_norm": 0.7674791998200291, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 18505 + }, + { + "epoch": 0.18506, + "grad_norm": 0.9341694841503271, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 18506 + }, + { + "epoch": 0.18507, + "grad_norm": 1.0176019368960014, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 18507 + }, + { + "epoch": 0.18508, + "grad_norm": 0.7331594257732688, + "learning_rate": 0.003, + "loss": 4.009, + "step": 18508 + }, + { + "epoch": 0.18509, + "grad_norm": 0.6647515160063896, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 18509 + }, + { + "epoch": 0.1851, + "grad_norm": 0.6957666252435012, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 18510 + }, + { + "epoch": 0.18511, + "grad_norm": 0.8372449832706363, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 18511 + }, + { + "epoch": 0.18512, + "grad_norm": 0.9374489258398074, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18512 + }, + { + "epoch": 0.18513, + "grad_norm": 0.8944850891056223, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 18513 + }, + { + "epoch": 0.18514, + "grad_norm": 0.7545056124471785, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 18514 + }, + { + "epoch": 0.18515, + "grad_norm": 0.6048034091155862, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 18515 + }, + { + "epoch": 0.18516, + "grad_norm": 0.649483178688814, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 18516 + }, + { + "epoch": 0.18517, + "grad_norm": 0.6244296543746121, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 18517 + }, + { + "epoch": 0.18518, + "grad_norm": 0.6464191056751218, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 18518 + }, + { + "epoch": 0.18519, + "grad_norm": 0.6102330510398791, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 18519 + }, + { + "epoch": 0.1852, + "grad_norm": 0.5954394613830718, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 18520 + }, + { + "epoch": 0.18521, + "grad_norm": 0.5110521259584845, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 18521 + }, + { + "epoch": 0.18522, + "grad_norm": 0.5046592573571453, + "learning_rate": 0.003, + "loss": 3.993, + "step": 18522 + }, + { + "epoch": 0.18523, + "grad_norm": 0.5582549918365517, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 18523 + }, + { + "epoch": 0.18524, + "grad_norm": 0.6697905181163559, + "learning_rate": 0.003, + "loss": 4.012, + "step": 18524 + }, + { + "epoch": 0.18525, + "grad_norm": 0.7598282602809256, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18525 + }, + { + "epoch": 0.18526, + "grad_norm": 0.8308661200178142, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 18526 + }, + { + "epoch": 0.18527, + "grad_norm": 0.872916939629396, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 18527 + }, + { + "epoch": 0.18528, + "grad_norm": 0.9305949224943858, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 18528 + }, + { + "epoch": 0.18529, + "grad_norm": 0.8530767763073737, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 18529 + }, + { + "epoch": 0.1853, + "grad_norm": 0.7614793028449756, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 18530 + }, + { + "epoch": 0.18531, + "grad_norm": 0.8983020519128377, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 18531 + }, + { + "epoch": 0.18532, + "grad_norm": 0.8618978080754732, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 18532 + }, + { + "epoch": 0.18533, + "grad_norm": 0.9096420155251231, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 18533 + }, + { + "epoch": 0.18534, + "grad_norm": 1.0949364399131667, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 18534 + }, + { + "epoch": 0.18535, + "grad_norm": 0.8531496328258401, + "learning_rate": 0.003, + "loss": 4.034, + "step": 18535 + }, + { + "epoch": 0.18536, + "grad_norm": 0.742860742706361, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 18536 + }, + { + "epoch": 0.18537, + "grad_norm": 0.735669193753493, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 18537 + }, + { + "epoch": 0.18538, + "grad_norm": 0.7517425852638916, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 18538 + }, + { + "epoch": 0.18539, + "grad_norm": 0.6282020403527632, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 18539 + }, + { + "epoch": 0.1854, + "grad_norm": 0.6571771559141921, + "learning_rate": 0.003, + "loss": 4.017, + "step": 18540 + }, + { + "epoch": 0.18541, + "grad_norm": 0.6783133572456582, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 18541 + }, + { + "epoch": 0.18542, + "grad_norm": 0.739192502223586, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 18542 + }, + { + "epoch": 0.18543, + "grad_norm": 0.8211260753341998, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 18543 + }, + { + "epoch": 0.18544, + "grad_norm": 0.805538097926538, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 18544 + }, + { + "epoch": 0.18545, + "grad_norm": 0.8226999789190816, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 18545 + }, + { + "epoch": 0.18546, + "grad_norm": 0.770722390599874, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 18546 + }, + { + "epoch": 0.18547, + "grad_norm": 0.6519043685159525, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 18547 + }, + { + "epoch": 0.18548, + "grad_norm": 0.6223726975603899, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 18548 + }, + { + "epoch": 0.18549, + "grad_norm": 0.650234592737104, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 18549 + }, + { + "epoch": 0.1855, + "grad_norm": 0.766925960499907, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 18550 + }, + { + "epoch": 0.18551, + "grad_norm": 0.7846840905333315, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 18551 + }, + { + "epoch": 0.18552, + "grad_norm": 0.9143391635179412, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 18552 + }, + { + "epoch": 0.18553, + "grad_norm": 1.2548611085642039, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18553 + }, + { + "epoch": 0.18554, + "grad_norm": 0.9094125713846363, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 18554 + }, + { + "epoch": 0.18555, + "grad_norm": 0.6878078145878197, + "learning_rate": 0.003, + "loss": 4.009, + "step": 18555 + }, + { + "epoch": 0.18556, + "grad_norm": 0.8412952250418608, + "learning_rate": 0.003, + "loss": 4.023, + "step": 18556 + }, + { + "epoch": 0.18557, + "grad_norm": 0.783133147573986, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 18557 + }, + { + "epoch": 0.18558, + "grad_norm": 0.7370385258606467, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 18558 + }, + { + "epoch": 0.18559, + "grad_norm": 0.6687552077094119, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 18559 + }, + { + "epoch": 0.1856, + "grad_norm": 0.6035252264419158, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 18560 + }, + { + "epoch": 0.18561, + "grad_norm": 0.5694021450872423, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 18561 + }, + { + "epoch": 0.18562, + "grad_norm": 0.49389995082316795, + "learning_rate": 0.003, + "loss": 3.995, + "step": 18562 + }, + { + "epoch": 0.18563, + "grad_norm": 0.47704123384265174, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 18563 + }, + { + "epoch": 0.18564, + "grad_norm": 0.4754263677332994, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 18564 + }, + { + "epoch": 0.18565, + "grad_norm": 0.50968808291251, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 18565 + }, + { + "epoch": 0.18566, + "grad_norm": 0.6616185650542236, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 18566 + }, + { + "epoch": 0.18567, + "grad_norm": 0.7824631550197145, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 18567 + }, + { + "epoch": 0.18568, + "grad_norm": 0.8856130385352554, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18568 + }, + { + "epoch": 0.18569, + "grad_norm": 0.9705927546724883, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 18569 + }, + { + "epoch": 0.1857, + "grad_norm": 0.92161711929861, + "learning_rate": 0.003, + "loss": 4.024, + "step": 18570 + }, + { + "epoch": 0.18571, + "grad_norm": 0.8083813834440592, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 18571 + }, + { + "epoch": 0.18572, + "grad_norm": 0.7254362246450182, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 18572 + }, + { + "epoch": 0.18573, + "grad_norm": 0.5159368823363767, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 18573 + }, + { + "epoch": 0.18574, + "grad_norm": 0.6426664853149341, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 18574 + }, + { + "epoch": 0.18575, + "grad_norm": 0.7684469958816703, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 18575 + }, + { + "epoch": 0.18576, + "grad_norm": 0.850707397578438, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 18576 + }, + { + "epoch": 0.18577, + "grad_norm": 0.8676702208551013, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 18577 + }, + { + "epoch": 0.18578, + "grad_norm": 0.8599209696683818, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 18578 + }, + { + "epoch": 0.18579, + "grad_norm": 0.8000075781982102, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 18579 + }, + { + "epoch": 0.1858, + "grad_norm": 0.7652043281352611, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 18580 + }, + { + "epoch": 0.18581, + "grad_norm": 0.6902965013648203, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 18581 + }, + { + "epoch": 0.18582, + "grad_norm": 0.6306394229884433, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 18582 + }, + { + "epoch": 0.18583, + "grad_norm": 0.717825615482299, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 18583 + }, + { + "epoch": 0.18584, + "grad_norm": 0.7415527004230156, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 18584 + }, + { + "epoch": 0.18585, + "grad_norm": 0.7158846821093546, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 18585 + }, + { + "epoch": 0.18586, + "grad_norm": 0.7779638423702474, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 18586 + }, + { + "epoch": 0.18587, + "grad_norm": 0.9046648020559163, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 18587 + }, + { + "epoch": 0.18588, + "grad_norm": 0.709366530897983, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 18588 + }, + { + "epoch": 0.18589, + "grad_norm": 0.6731591412608005, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 18589 + }, + { + "epoch": 0.1859, + "grad_norm": 0.7114221748594014, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 18590 + }, + { + "epoch": 0.18591, + "grad_norm": 0.6885460496125004, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 18591 + }, + { + "epoch": 0.18592, + "grad_norm": 0.7168045767739818, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 18592 + }, + { + "epoch": 0.18593, + "grad_norm": 0.7941082011051133, + "learning_rate": 0.003, + "loss": 4.031, + "step": 18593 + }, + { + "epoch": 0.18594, + "grad_norm": 0.9238911377252176, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 18594 + }, + { + "epoch": 0.18595, + "grad_norm": 0.9119467310977216, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 18595 + }, + { + "epoch": 0.18596, + "grad_norm": 0.7981284082438801, + "learning_rate": 0.003, + "loss": 4.0467, + "step": 18596 + }, + { + "epoch": 0.18597, + "grad_norm": 0.7317100407872701, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 18597 + }, + { + "epoch": 0.18598, + "grad_norm": 0.8384501861589906, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 18598 + }, + { + "epoch": 0.18599, + "grad_norm": 0.9060641595051969, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 18599 + }, + { + "epoch": 0.186, + "grad_norm": 0.8656662830845567, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 18600 + }, + { + "epoch": 0.18601, + "grad_norm": 0.9145019122353978, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 18601 + }, + { + "epoch": 0.18602, + "grad_norm": 1.0786605864509624, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 18602 + }, + { + "epoch": 0.18603, + "grad_norm": 1.1012021124754638, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 18603 + }, + { + "epoch": 0.18604, + "grad_norm": 0.7773685163382243, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 18604 + }, + { + "epoch": 0.18605, + "grad_norm": 0.7560407494924272, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 18605 + }, + { + "epoch": 0.18606, + "grad_norm": 0.7539816165424467, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 18606 + }, + { + "epoch": 0.18607, + "grad_norm": 0.7513012653609153, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 18607 + }, + { + "epoch": 0.18608, + "grad_norm": 0.6030558965203087, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 18608 + }, + { + "epoch": 0.18609, + "grad_norm": 0.613183621580848, + "learning_rate": 0.003, + "loss": 4.017, + "step": 18609 + }, + { + "epoch": 0.1861, + "grad_norm": 0.8607995891176324, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 18610 + }, + { + "epoch": 0.18611, + "grad_norm": 0.9910138661599887, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 18611 + }, + { + "epoch": 0.18612, + "grad_norm": 0.9820741530779314, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 18612 + }, + { + "epoch": 0.18613, + "grad_norm": 0.999536470860581, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 18613 + }, + { + "epoch": 0.18614, + "grad_norm": 0.9359896699666336, + "learning_rate": 0.003, + "loss": 3.975, + "step": 18614 + }, + { + "epoch": 0.18615, + "grad_norm": 0.8342582810107119, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 18615 + }, + { + "epoch": 0.18616, + "grad_norm": 0.7311063379460248, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 18616 + }, + { + "epoch": 0.18617, + "grad_norm": 0.682313261451543, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 18617 + }, + { + "epoch": 0.18618, + "grad_norm": 0.5822533447119164, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 18618 + }, + { + "epoch": 0.18619, + "grad_norm": 0.5361855541743135, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 18619 + }, + { + "epoch": 0.1862, + "grad_norm": 0.5242384554655722, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 18620 + }, + { + "epoch": 0.18621, + "grad_norm": 0.5343434069408941, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 18621 + }, + { + "epoch": 0.18622, + "grad_norm": 0.5327871203774499, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 18622 + }, + { + "epoch": 0.18623, + "grad_norm": 0.5035602936841013, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 18623 + }, + { + "epoch": 0.18624, + "grad_norm": 0.4995009940002232, + "learning_rate": 0.003, + "loss": 3.999, + "step": 18624 + }, + { + "epoch": 0.18625, + "grad_norm": 0.44323951517738663, + "learning_rate": 0.003, + "loss": 3.993, + "step": 18625 + }, + { + "epoch": 0.18626, + "grad_norm": 0.4703248215712079, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 18626 + }, + { + "epoch": 0.18627, + "grad_norm": 0.5294526626539999, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 18627 + }, + { + "epoch": 0.18628, + "grad_norm": 0.5811927225041419, + "learning_rate": 0.003, + "loss": 3.982, + "step": 18628 + }, + { + "epoch": 0.18629, + "grad_norm": 0.7214216316965267, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 18629 + }, + { + "epoch": 0.1863, + "grad_norm": 0.9265915150699503, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 18630 + }, + { + "epoch": 0.18631, + "grad_norm": 1.10504103494837, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 18631 + }, + { + "epoch": 0.18632, + "grad_norm": 0.7052630610286446, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 18632 + }, + { + "epoch": 0.18633, + "grad_norm": 0.5614751813312797, + "learning_rate": 0.003, + "loss": 3.993, + "step": 18633 + }, + { + "epoch": 0.18634, + "grad_norm": 0.6277367674972708, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 18634 + }, + { + "epoch": 0.18635, + "grad_norm": 0.7208287476361389, + "learning_rate": 0.003, + "loss": 4.029, + "step": 18635 + }, + { + "epoch": 0.18636, + "grad_norm": 0.8440415142639855, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 18636 + }, + { + "epoch": 0.18637, + "grad_norm": 0.8061196125247145, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 18637 + }, + { + "epoch": 0.18638, + "grad_norm": 0.9134726859463089, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 18638 + }, + { + "epoch": 0.18639, + "grad_norm": 0.9164198961098972, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 18639 + }, + { + "epoch": 0.1864, + "grad_norm": 0.9678471699751883, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 18640 + }, + { + "epoch": 0.18641, + "grad_norm": 1.0136752138179743, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 18641 + }, + { + "epoch": 0.18642, + "grad_norm": 1.0622914873501113, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 18642 + }, + { + "epoch": 0.18643, + "grad_norm": 0.917005478326722, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 18643 + }, + { + "epoch": 0.18644, + "grad_norm": 0.7979342086974802, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 18644 + }, + { + "epoch": 0.18645, + "grad_norm": 0.8157235407107448, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 18645 + }, + { + "epoch": 0.18646, + "grad_norm": 0.7308278274594475, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 18646 + }, + { + "epoch": 0.18647, + "grad_norm": 0.6523786056823121, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 18647 + }, + { + "epoch": 0.18648, + "grad_norm": 0.6952108073877713, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 18648 + }, + { + "epoch": 0.18649, + "grad_norm": 0.7974394934843421, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 18649 + }, + { + "epoch": 0.1865, + "grad_norm": 0.8606541999389266, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 18650 + }, + { + "epoch": 0.18651, + "grad_norm": 0.9844312255728548, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18651 + }, + { + "epoch": 0.18652, + "grad_norm": 1.001213534150566, + "learning_rate": 0.003, + "loss": 4.0429, + "step": 18652 + }, + { + "epoch": 0.18653, + "grad_norm": 1.023987750758092, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 18653 + }, + { + "epoch": 0.18654, + "grad_norm": 0.8599478688365298, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 18654 + }, + { + "epoch": 0.18655, + "grad_norm": 0.7705499934272618, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 18655 + }, + { + "epoch": 0.18656, + "grad_norm": 0.7809517847569285, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 18656 + }, + { + "epoch": 0.18657, + "grad_norm": 0.7597495705585862, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 18657 + }, + { + "epoch": 0.18658, + "grad_norm": 0.7302021884049115, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 18658 + }, + { + "epoch": 0.18659, + "grad_norm": 0.7382617014980813, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 18659 + }, + { + "epoch": 0.1866, + "grad_norm": 0.7600383060559108, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 18660 + }, + { + "epoch": 0.18661, + "grad_norm": 0.6429755731823346, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 18661 + }, + { + "epoch": 0.18662, + "grad_norm": 0.6675243459534613, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 18662 + }, + { + "epoch": 0.18663, + "grad_norm": 0.6606594972767804, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 18663 + }, + { + "epoch": 0.18664, + "grad_norm": 0.5281977964948973, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 18664 + }, + { + "epoch": 0.18665, + "grad_norm": 0.543538304875981, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 18665 + }, + { + "epoch": 0.18666, + "grad_norm": 0.526675182378391, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 18666 + }, + { + "epoch": 0.18667, + "grad_norm": 0.5353596999637199, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 18667 + }, + { + "epoch": 0.18668, + "grad_norm": 0.4838986593241969, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 18668 + }, + { + "epoch": 0.18669, + "grad_norm": 0.578115823769787, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 18669 + }, + { + "epoch": 0.1867, + "grad_norm": 0.6007779486923898, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 18670 + }, + { + "epoch": 0.18671, + "grad_norm": 0.6718398640305233, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18671 + }, + { + "epoch": 0.18672, + "grad_norm": 0.741671002982468, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 18672 + }, + { + "epoch": 0.18673, + "grad_norm": 0.7033783340784744, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 18673 + }, + { + "epoch": 0.18674, + "grad_norm": 0.6469827134164702, + "learning_rate": 0.003, + "loss": 3.971, + "step": 18674 + }, + { + "epoch": 0.18675, + "grad_norm": 0.60894209658333, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 18675 + }, + { + "epoch": 0.18676, + "grad_norm": 0.5691621396429606, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 18676 + }, + { + "epoch": 0.18677, + "grad_norm": 0.6419711369061717, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 18677 + }, + { + "epoch": 0.18678, + "grad_norm": 0.6750428506305743, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 18678 + }, + { + "epoch": 0.18679, + "grad_norm": 0.6232992691089504, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 18679 + }, + { + "epoch": 0.1868, + "grad_norm": 0.6830697926810245, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 18680 + }, + { + "epoch": 0.18681, + "grad_norm": 0.7633180692261908, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 18681 + }, + { + "epoch": 0.18682, + "grad_norm": 0.7242999099189277, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 18682 + }, + { + "epoch": 0.18683, + "grad_norm": 0.7838521948411215, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 18683 + }, + { + "epoch": 0.18684, + "grad_norm": 0.8900932957045854, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 18684 + }, + { + "epoch": 0.18685, + "grad_norm": 0.9707228240987424, + "learning_rate": 0.003, + "loss": 4.002, + "step": 18685 + }, + { + "epoch": 0.18686, + "grad_norm": 0.952970583073336, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 18686 + }, + { + "epoch": 0.18687, + "grad_norm": 0.9720133430323731, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 18687 + }, + { + "epoch": 0.18688, + "grad_norm": 0.9120607097159831, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 18688 + }, + { + "epoch": 0.18689, + "grad_norm": 0.8508618926344759, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 18689 + }, + { + "epoch": 0.1869, + "grad_norm": 0.7862141258135605, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 18690 + }, + { + "epoch": 0.18691, + "grad_norm": 0.811350493265396, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 18691 + }, + { + "epoch": 0.18692, + "grad_norm": 0.860367253233588, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18692 + }, + { + "epoch": 0.18693, + "grad_norm": 0.7943647740046552, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 18693 + }, + { + "epoch": 0.18694, + "grad_norm": 0.767389142016884, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 18694 + }, + { + "epoch": 0.18695, + "grad_norm": 0.8251822188310688, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 18695 + }, + { + "epoch": 0.18696, + "grad_norm": 0.8568262135387982, + "learning_rate": 0.003, + "loss": 4.031, + "step": 18696 + }, + { + "epoch": 0.18697, + "grad_norm": 0.8978646905927147, + "learning_rate": 0.003, + "loss": 3.997, + "step": 18697 + }, + { + "epoch": 0.18698, + "grad_norm": 0.9498707173821958, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 18698 + }, + { + "epoch": 0.18699, + "grad_norm": 0.8741769513312904, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 18699 + }, + { + "epoch": 0.187, + "grad_norm": 0.7568742524085, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 18700 + }, + { + "epoch": 0.18701, + "grad_norm": 0.6855264318437687, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 18701 + }, + { + "epoch": 0.18702, + "grad_norm": 0.6001067541896252, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 18702 + }, + { + "epoch": 0.18703, + "grad_norm": 0.6450632059677652, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 18703 + }, + { + "epoch": 0.18704, + "grad_norm": 0.6932013719291363, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 18704 + }, + { + "epoch": 0.18705, + "grad_norm": 0.6937331228686318, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 18705 + }, + { + "epoch": 0.18706, + "grad_norm": 0.7322271783213578, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 18706 + }, + { + "epoch": 0.18707, + "grad_norm": 0.7427646891736644, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 18707 + }, + { + "epoch": 0.18708, + "grad_norm": 0.69474568576222, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 18708 + }, + { + "epoch": 0.18709, + "grad_norm": 0.7569286121122587, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 18709 + }, + { + "epoch": 0.1871, + "grad_norm": 0.766104395124034, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 18710 + }, + { + "epoch": 0.18711, + "grad_norm": 0.7979795216784576, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 18711 + }, + { + "epoch": 0.18712, + "grad_norm": 0.7955363090794372, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 18712 + }, + { + "epoch": 0.18713, + "grad_norm": 0.751980308473347, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 18713 + }, + { + "epoch": 0.18714, + "grad_norm": 0.5581587295235377, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 18714 + }, + { + "epoch": 0.18715, + "grad_norm": 0.6207258212671234, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 18715 + }, + { + "epoch": 0.18716, + "grad_norm": 0.5828294867092607, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 18716 + }, + { + "epoch": 0.18717, + "grad_norm": 0.594888339579703, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 18717 + }, + { + "epoch": 0.18718, + "grad_norm": 0.5649271774322104, + "learning_rate": 0.003, + "loss": 3.979, + "step": 18718 + }, + { + "epoch": 0.18719, + "grad_norm": 0.6090506001786035, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 18719 + }, + { + "epoch": 0.1872, + "grad_norm": 0.5950005487963871, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 18720 + }, + { + "epoch": 0.18721, + "grad_norm": 0.7544624393658526, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 18721 + }, + { + "epoch": 0.18722, + "grad_norm": 0.9383546353076503, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 18722 + }, + { + "epoch": 0.18723, + "grad_norm": 1.1026622699542992, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 18723 + }, + { + "epoch": 0.18724, + "grad_norm": 0.7602121736287344, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18724 + }, + { + "epoch": 0.18725, + "grad_norm": 0.7153198723584484, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 18725 + }, + { + "epoch": 0.18726, + "grad_norm": 0.8352698145202425, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 18726 + }, + { + "epoch": 0.18727, + "grad_norm": 0.9018790312132974, + "learning_rate": 0.003, + "loss": 3.984, + "step": 18727 + }, + { + "epoch": 0.18728, + "grad_norm": 0.8752720998364965, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 18728 + }, + { + "epoch": 0.18729, + "grad_norm": 0.9633298710345001, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 18729 + }, + { + "epoch": 0.1873, + "grad_norm": 0.9583601042810744, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 18730 + }, + { + "epoch": 0.18731, + "grad_norm": 0.9311019448918602, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18731 + }, + { + "epoch": 0.18732, + "grad_norm": 1.060423711287027, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 18732 + }, + { + "epoch": 0.18733, + "grad_norm": 0.9164368117619744, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 18733 + }, + { + "epoch": 0.18734, + "grad_norm": 0.8007078392889425, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 18734 + }, + { + "epoch": 0.18735, + "grad_norm": 0.7888385363723291, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 18735 + }, + { + "epoch": 0.18736, + "grad_norm": 0.8654199031063029, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 18736 + }, + { + "epoch": 0.18737, + "grad_norm": 0.7948801256207583, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 18737 + }, + { + "epoch": 0.18738, + "grad_norm": 0.7299892583127783, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 18738 + }, + { + "epoch": 0.18739, + "grad_norm": 0.7083074259663926, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 18739 + }, + { + "epoch": 0.1874, + "grad_norm": 0.7736073098901901, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 18740 + }, + { + "epoch": 0.18741, + "grad_norm": 0.7127533121299838, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 18741 + }, + { + "epoch": 0.18742, + "grad_norm": 0.748645053120168, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 18742 + }, + { + "epoch": 0.18743, + "grad_norm": 0.7245931419640033, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 18743 + }, + { + "epoch": 0.18744, + "grad_norm": 0.6772917038374636, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 18744 + }, + { + "epoch": 0.18745, + "grad_norm": 0.6795011044503596, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 18745 + }, + { + "epoch": 0.18746, + "grad_norm": 0.6852595652595164, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 18746 + }, + { + "epoch": 0.18747, + "grad_norm": 0.6653395136148078, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 18747 + }, + { + "epoch": 0.18748, + "grad_norm": 0.7064432923841182, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 18748 + }, + { + "epoch": 0.18749, + "grad_norm": 0.6814162394462233, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 18749 + }, + { + "epoch": 0.1875, + "grad_norm": 0.6263604751219431, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 18750 + }, + { + "epoch": 0.18751, + "grad_norm": 0.5730228973465636, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 18751 + }, + { + "epoch": 0.18752, + "grad_norm": 0.6192965394747949, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 18752 + }, + { + "epoch": 0.18753, + "grad_norm": 0.6244967166050803, + "learning_rate": 0.003, + "loss": 4.024, + "step": 18753 + }, + { + "epoch": 0.18754, + "grad_norm": 0.6463928917102257, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 18754 + }, + { + "epoch": 0.18755, + "grad_norm": 0.678141281382781, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 18755 + }, + { + "epoch": 0.18756, + "grad_norm": 0.7435192779309362, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18756 + }, + { + "epoch": 0.18757, + "grad_norm": 0.8039277720054611, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 18757 + }, + { + "epoch": 0.18758, + "grad_norm": 0.7991072321650146, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18758 + }, + { + "epoch": 0.18759, + "grad_norm": 0.6424914814002511, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 18759 + }, + { + "epoch": 0.1876, + "grad_norm": 0.6500344843367954, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 18760 + }, + { + "epoch": 0.18761, + "grad_norm": 0.6929362193400054, + "learning_rate": 0.003, + "loss": 4.0, + "step": 18761 + }, + { + "epoch": 0.18762, + "grad_norm": 0.6798994858070492, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 18762 + }, + { + "epoch": 0.18763, + "grad_norm": 0.6857343112149886, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 18763 + }, + { + "epoch": 0.18764, + "grad_norm": 0.6214129895360323, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 18764 + }, + { + "epoch": 0.18765, + "grad_norm": 0.6209585240344359, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 18765 + }, + { + "epoch": 0.18766, + "grad_norm": 0.7243741782782389, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 18766 + }, + { + "epoch": 0.18767, + "grad_norm": 0.7637170500308076, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 18767 + }, + { + "epoch": 0.18768, + "grad_norm": 0.6906062869568229, + "learning_rate": 0.003, + "loss": 3.993, + "step": 18768 + }, + { + "epoch": 0.18769, + "grad_norm": 0.6801382414176258, + "learning_rate": 0.003, + "loss": 4.007, + "step": 18769 + }, + { + "epoch": 0.1877, + "grad_norm": 0.7542165663071692, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 18770 + }, + { + "epoch": 0.18771, + "grad_norm": 0.8370237081422546, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 18771 + }, + { + "epoch": 0.18772, + "grad_norm": 0.8255608344314023, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 18772 + }, + { + "epoch": 0.18773, + "grad_norm": 0.7498963341905988, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 18773 + }, + { + "epoch": 0.18774, + "grad_norm": 0.7378801952629225, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 18774 + }, + { + "epoch": 0.18775, + "grad_norm": 0.801159864820087, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 18775 + }, + { + "epoch": 0.18776, + "grad_norm": 1.0420504479948498, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 18776 + }, + { + "epoch": 0.18777, + "grad_norm": 0.9732570561609607, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 18777 + }, + { + "epoch": 0.18778, + "grad_norm": 1.20124770460446, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 18778 + }, + { + "epoch": 0.18779, + "grad_norm": 0.7718438178106606, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 18779 + }, + { + "epoch": 0.1878, + "grad_norm": 0.7326643824975992, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 18780 + }, + { + "epoch": 0.18781, + "grad_norm": 0.6326365560749702, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 18781 + }, + { + "epoch": 0.18782, + "grad_norm": 0.6616769354732115, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 18782 + }, + { + "epoch": 0.18783, + "grad_norm": 0.7513480058532606, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 18783 + }, + { + "epoch": 0.18784, + "grad_norm": 0.7963382197505524, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18784 + }, + { + "epoch": 0.18785, + "grad_norm": 0.8439689394224811, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 18785 + }, + { + "epoch": 0.18786, + "grad_norm": 0.8284278965670364, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 18786 + }, + { + "epoch": 0.18787, + "grad_norm": 0.9109816432281513, + "learning_rate": 0.003, + "loss": 4.015, + "step": 18787 + }, + { + "epoch": 0.18788, + "grad_norm": 0.8551363828683061, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 18788 + }, + { + "epoch": 0.18789, + "grad_norm": 0.8773997297511154, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 18789 + }, + { + "epoch": 0.1879, + "grad_norm": 0.9442994278559491, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 18790 + }, + { + "epoch": 0.18791, + "grad_norm": 0.9651176899529363, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 18791 + }, + { + "epoch": 0.18792, + "grad_norm": 1.0452352403648815, + "learning_rate": 0.003, + "loss": 4.018, + "step": 18792 + }, + { + "epoch": 0.18793, + "grad_norm": 0.9667215872094005, + "learning_rate": 0.003, + "loss": 4.0486, + "step": 18793 + }, + { + "epoch": 0.18794, + "grad_norm": 0.9205841692011495, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18794 + }, + { + "epoch": 0.18795, + "grad_norm": 0.8016210427374546, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 18795 + }, + { + "epoch": 0.18796, + "grad_norm": 0.7790436541799667, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 18796 + }, + { + "epoch": 0.18797, + "grad_norm": 0.8195514758694753, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 18797 + }, + { + "epoch": 0.18798, + "grad_norm": 0.8217268161784704, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 18798 + }, + { + "epoch": 0.18799, + "grad_norm": 0.8744034401723285, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 18799 + }, + { + "epoch": 0.188, + "grad_norm": 0.783137435999375, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 18800 + }, + { + "epoch": 0.18801, + "grad_norm": 0.6766782820866318, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18801 + }, + { + "epoch": 0.18802, + "grad_norm": 0.5960807903854132, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 18802 + }, + { + "epoch": 0.18803, + "grad_norm": 0.5631390428606453, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 18803 + }, + { + "epoch": 0.18804, + "grad_norm": 0.5615870122988219, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 18804 + }, + { + "epoch": 0.18805, + "grad_norm": 0.5192477261318682, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 18805 + }, + { + "epoch": 0.18806, + "grad_norm": 0.5466855086298749, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 18806 + }, + { + "epoch": 0.18807, + "grad_norm": 0.5305656461253773, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 18807 + }, + { + "epoch": 0.18808, + "grad_norm": 0.44413743853591403, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 18808 + }, + { + "epoch": 0.18809, + "grad_norm": 0.43174504282953957, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 18809 + }, + { + "epoch": 0.1881, + "grad_norm": 0.38646518167452826, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 18810 + }, + { + "epoch": 0.18811, + "grad_norm": 0.3803689003080482, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 18811 + }, + { + "epoch": 0.18812, + "grad_norm": 0.37778245540135125, + "learning_rate": 0.003, + "loss": 3.965, + "step": 18812 + }, + { + "epoch": 0.18813, + "grad_norm": 0.4553363962357483, + "learning_rate": 0.003, + "loss": 3.976, + "step": 18813 + }, + { + "epoch": 0.18814, + "grad_norm": 0.5492558599950933, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 18814 + }, + { + "epoch": 0.18815, + "grad_norm": 0.6187594324943728, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 18815 + }, + { + "epoch": 0.18816, + "grad_norm": 0.6293232506406246, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 18816 + }, + { + "epoch": 0.18817, + "grad_norm": 0.6231870525401215, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 18817 + }, + { + "epoch": 0.18818, + "grad_norm": 0.5819825345558965, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 18818 + }, + { + "epoch": 0.18819, + "grad_norm": 0.6400966180575421, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 18819 + }, + { + "epoch": 0.1882, + "grad_norm": 0.6347339968772474, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 18820 + }, + { + "epoch": 0.18821, + "grad_norm": 0.6548874482910525, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 18821 + }, + { + "epoch": 0.18822, + "grad_norm": 0.8694141631752428, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 18822 + }, + { + "epoch": 0.18823, + "grad_norm": 0.9351102572265193, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 18823 + }, + { + "epoch": 0.18824, + "grad_norm": 0.8532391798907555, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 18824 + }, + { + "epoch": 0.18825, + "grad_norm": 0.8608590220344958, + "learning_rate": 0.003, + "loss": 3.994, + "step": 18825 + }, + { + "epoch": 0.18826, + "grad_norm": 0.8559711064201767, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 18826 + }, + { + "epoch": 0.18827, + "grad_norm": 0.8600542493048954, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 18827 + }, + { + "epoch": 0.18828, + "grad_norm": 0.9265619593453834, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 18828 + }, + { + "epoch": 0.18829, + "grad_norm": 0.9131499987386412, + "learning_rate": 0.003, + "loss": 4.018, + "step": 18829 + }, + { + "epoch": 0.1883, + "grad_norm": 0.8268273901443607, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 18830 + }, + { + "epoch": 0.18831, + "grad_norm": 0.7459868697420596, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 18831 + }, + { + "epoch": 0.18832, + "grad_norm": 0.649240039201116, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 18832 + }, + { + "epoch": 0.18833, + "grad_norm": 0.6855601202769457, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 18833 + }, + { + "epoch": 0.18834, + "grad_norm": 0.6552535148563494, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 18834 + }, + { + "epoch": 0.18835, + "grad_norm": 0.7285924690241348, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 18835 + }, + { + "epoch": 0.18836, + "grad_norm": 0.8520970861247162, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 18836 + }, + { + "epoch": 0.18837, + "grad_norm": 1.050639180188081, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 18837 + }, + { + "epoch": 0.18838, + "grad_norm": 0.9407731604470052, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 18838 + }, + { + "epoch": 0.18839, + "grad_norm": 0.8621276335980478, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 18839 + }, + { + "epoch": 0.1884, + "grad_norm": 0.7876649558816456, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 18840 + }, + { + "epoch": 0.18841, + "grad_norm": 0.7263070452599937, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 18841 + }, + { + "epoch": 0.18842, + "grad_norm": 0.7037008007896113, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 18842 + }, + { + "epoch": 0.18843, + "grad_norm": 0.7787207565008692, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 18843 + }, + { + "epoch": 0.18844, + "grad_norm": 0.90224474310234, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 18844 + }, + { + "epoch": 0.18845, + "grad_norm": 0.7422542852360645, + "learning_rate": 0.003, + "loss": 4.002, + "step": 18845 + }, + { + "epoch": 0.18846, + "grad_norm": 0.5555696186973657, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 18846 + }, + { + "epoch": 0.18847, + "grad_norm": 0.6390214093455266, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 18847 + }, + { + "epoch": 0.18848, + "grad_norm": 0.704291791242021, + "learning_rate": 0.003, + "loss": 3.986, + "step": 18848 + }, + { + "epoch": 0.18849, + "grad_norm": 0.7043329798929148, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 18849 + }, + { + "epoch": 0.1885, + "grad_norm": 0.6463961111471129, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 18850 + }, + { + "epoch": 0.18851, + "grad_norm": 0.7196168708078042, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 18851 + }, + { + "epoch": 0.18852, + "grad_norm": 0.7729449035219305, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 18852 + }, + { + "epoch": 0.18853, + "grad_norm": 0.8162130142791245, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 18853 + }, + { + "epoch": 0.18854, + "grad_norm": 0.7974057586965564, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 18854 + }, + { + "epoch": 0.18855, + "grad_norm": 0.8082904518489755, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 18855 + }, + { + "epoch": 0.18856, + "grad_norm": 0.7489574345166621, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 18856 + }, + { + "epoch": 0.18857, + "grad_norm": 0.7682054874806807, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 18857 + }, + { + "epoch": 0.18858, + "grad_norm": 0.7285243135744893, + "learning_rate": 0.003, + "loss": 3.999, + "step": 18858 + }, + { + "epoch": 0.18859, + "grad_norm": 0.7030444833718954, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 18859 + }, + { + "epoch": 0.1886, + "grad_norm": 0.7434396468950654, + "learning_rate": 0.003, + "loss": 4.026, + "step": 18860 + }, + { + "epoch": 0.18861, + "grad_norm": 0.6134523188688009, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 18861 + }, + { + "epoch": 0.18862, + "grad_norm": 0.5160181373670276, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 18862 + }, + { + "epoch": 0.18863, + "grad_norm": 0.6016601943479785, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 18863 + }, + { + "epoch": 0.18864, + "grad_norm": 0.6904000690728449, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 18864 + }, + { + "epoch": 0.18865, + "grad_norm": 0.8211441960993832, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 18865 + }, + { + "epoch": 0.18866, + "grad_norm": 0.9201208400455994, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18866 + }, + { + "epoch": 0.18867, + "grad_norm": 0.9591674609878678, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 18867 + }, + { + "epoch": 0.18868, + "grad_norm": 0.8449594446547487, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 18868 + }, + { + "epoch": 0.18869, + "grad_norm": 0.8543100516268627, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 18869 + }, + { + "epoch": 0.1887, + "grad_norm": 0.878960122909512, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 18870 + }, + { + "epoch": 0.18871, + "grad_norm": 0.8230820389949723, + "learning_rate": 0.003, + "loss": 4.029, + "step": 18871 + }, + { + "epoch": 0.18872, + "grad_norm": 0.7691948871028476, + "learning_rate": 0.003, + "loss": 4.008, + "step": 18872 + }, + { + "epoch": 0.18873, + "grad_norm": 0.7417910587705115, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 18873 + }, + { + "epoch": 0.18874, + "grad_norm": 0.7180041904886646, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 18874 + }, + { + "epoch": 0.18875, + "grad_norm": 0.8545454188914136, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 18875 + }, + { + "epoch": 0.18876, + "grad_norm": 0.823875427109604, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 18876 + }, + { + "epoch": 0.18877, + "grad_norm": 0.7982093964760835, + "learning_rate": 0.003, + "loss": 4.013, + "step": 18877 + }, + { + "epoch": 0.18878, + "grad_norm": 0.8734105832903972, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 18878 + }, + { + "epoch": 0.18879, + "grad_norm": 0.9114231248555678, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 18879 + }, + { + "epoch": 0.1888, + "grad_norm": 0.8437482569830609, + "learning_rate": 0.003, + "loss": 4.02, + "step": 18880 + }, + { + "epoch": 0.18881, + "grad_norm": 0.6845022622897622, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 18881 + }, + { + "epoch": 0.18882, + "grad_norm": 0.7160877877915566, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 18882 + }, + { + "epoch": 0.18883, + "grad_norm": 0.6842900155445499, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 18883 + }, + { + "epoch": 0.18884, + "grad_norm": 0.5718956780400477, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 18884 + }, + { + "epoch": 0.18885, + "grad_norm": 0.5239450503801911, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 18885 + }, + { + "epoch": 0.18886, + "grad_norm": 0.5863983562690075, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 18886 + }, + { + "epoch": 0.18887, + "grad_norm": 0.6736267308641001, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 18887 + }, + { + "epoch": 0.18888, + "grad_norm": 0.7546655275473205, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 18888 + }, + { + "epoch": 0.18889, + "grad_norm": 0.785315020685868, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 18889 + }, + { + "epoch": 0.1889, + "grad_norm": 0.8489843749443374, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 18890 + }, + { + "epoch": 0.18891, + "grad_norm": 0.8692123739621872, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 18891 + }, + { + "epoch": 0.18892, + "grad_norm": 0.676169843890652, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 18892 + }, + { + "epoch": 0.18893, + "grad_norm": 0.5589136110239937, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 18893 + }, + { + "epoch": 0.18894, + "grad_norm": 0.5626599710560869, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 18894 + }, + { + "epoch": 0.18895, + "grad_norm": 0.6279121668941107, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 18895 + }, + { + "epoch": 0.18896, + "grad_norm": 0.7580378728830642, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 18896 + }, + { + "epoch": 0.18897, + "grad_norm": 0.8038739993627712, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 18897 + }, + { + "epoch": 0.18898, + "grad_norm": 0.8068880602783306, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 18898 + }, + { + "epoch": 0.18899, + "grad_norm": 0.6743167074557405, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 18899 + }, + { + "epoch": 0.189, + "grad_norm": 0.6635469905428019, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 18900 + }, + { + "epoch": 0.18901, + "grad_norm": 0.5622031871161361, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 18901 + }, + { + "epoch": 0.18902, + "grad_norm": 0.5737890869120024, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 18902 + }, + { + "epoch": 0.18903, + "grad_norm": 0.5554385192828459, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 18903 + }, + { + "epoch": 0.18904, + "grad_norm": 0.6416688288712712, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 18904 + }, + { + "epoch": 0.18905, + "grad_norm": 0.7049357041031346, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 18905 + }, + { + "epoch": 0.18906, + "grad_norm": 0.762718688925408, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 18906 + }, + { + "epoch": 0.18907, + "grad_norm": 0.7765065792893818, + "learning_rate": 0.003, + "loss": 3.99, + "step": 18907 + }, + { + "epoch": 0.18908, + "grad_norm": 0.8730712681020256, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 18908 + }, + { + "epoch": 0.18909, + "grad_norm": 0.9080887118858543, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 18909 + }, + { + "epoch": 0.1891, + "grad_norm": 0.9587194302679122, + "learning_rate": 0.003, + "loss": 4.033, + "step": 18910 + }, + { + "epoch": 0.18911, + "grad_norm": 0.9980853680652783, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 18911 + }, + { + "epoch": 0.18912, + "grad_norm": 0.827802360051384, + "learning_rate": 0.003, + "loss": 4.024, + "step": 18912 + }, + { + "epoch": 0.18913, + "grad_norm": 0.9146581329476952, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 18913 + }, + { + "epoch": 0.18914, + "grad_norm": 1.043108606136416, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 18914 + }, + { + "epoch": 0.18915, + "grad_norm": 1.0952686102532017, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 18915 + }, + { + "epoch": 0.18916, + "grad_norm": 0.8477600181825168, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 18916 + }, + { + "epoch": 0.18917, + "grad_norm": 0.8884452500592119, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 18917 + }, + { + "epoch": 0.18918, + "grad_norm": 0.9043021678542245, + "learning_rate": 0.003, + "loss": 4.006, + "step": 18918 + }, + { + "epoch": 0.18919, + "grad_norm": 0.9476040192441807, + "learning_rate": 0.003, + "loss": 4.0455, + "step": 18919 + }, + { + "epoch": 0.1892, + "grad_norm": 0.9217421812527545, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 18920 + }, + { + "epoch": 0.18921, + "grad_norm": 0.9072632836993808, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 18921 + }, + { + "epoch": 0.18922, + "grad_norm": 0.8506910893225509, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 18922 + }, + { + "epoch": 0.18923, + "grad_norm": 0.9034697521807814, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 18923 + }, + { + "epoch": 0.18924, + "grad_norm": 0.9378802803704416, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 18924 + }, + { + "epoch": 0.18925, + "grad_norm": 0.869775125263093, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 18925 + }, + { + "epoch": 0.18926, + "grad_norm": 0.8019855447663577, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 18926 + }, + { + "epoch": 0.18927, + "grad_norm": 0.8075259557804633, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 18927 + }, + { + "epoch": 0.18928, + "grad_norm": 0.8292596342762584, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 18928 + }, + { + "epoch": 0.18929, + "grad_norm": 0.675124036271808, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 18929 + }, + { + "epoch": 0.1893, + "grad_norm": 0.7053632948067855, + "learning_rate": 0.003, + "loss": 3.995, + "step": 18930 + }, + { + "epoch": 0.18931, + "grad_norm": 0.7609043595871048, + "learning_rate": 0.003, + "loss": 4.017, + "step": 18931 + }, + { + "epoch": 0.18932, + "grad_norm": 0.7925552819973807, + "learning_rate": 0.003, + "loss": 4.0548, + "step": 18932 + }, + { + "epoch": 0.18933, + "grad_norm": 0.6668565014513064, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 18933 + }, + { + "epoch": 0.18934, + "grad_norm": 0.5334825317077287, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 18934 + }, + { + "epoch": 0.18935, + "grad_norm": 0.4387249626967101, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 18935 + }, + { + "epoch": 0.18936, + "grad_norm": 0.47976955415013, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 18936 + }, + { + "epoch": 0.18937, + "grad_norm": 0.5396823158363736, + "learning_rate": 0.003, + "loss": 3.978, + "step": 18937 + }, + { + "epoch": 0.18938, + "grad_norm": 0.5334603624736158, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 18938 + }, + { + "epoch": 0.18939, + "grad_norm": 0.5706900031587143, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 18939 + }, + { + "epoch": 0.1894, + "grad_norm": 0.6519158774993224, + "learning_rate": 0.003, + "loss": 4.025, + "step": 18940 + }, + { + "epoch": 0.18941, + "grad_norm": 0.6833600290027028, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 18941 + }, + { + "epoch": 0.18942, + "grad_norm": 0.6495451187591241, + "learning_rate": 0.003, + "loss": 4.014, + "step": 18942 + }, + { + "epoch": 0.18943, + "grad_norm": 0.7111220106181038, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 18943 + }, + { + "epoch": 0.18944, + "grad_norm": 0.7360750819904539, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 18944 + }, + { + "epoch": 0.18945, + "grad_norm": 0.7666487822277542, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 18945 + }, + { + "epoch": 0.18946, + "grad_norm": 0.8154968172522751, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 18946 + }, + { + "epoch": 0.18947, + "grad_norm": 0.8594018673346592, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 18947 + }, + { + "epoch": 0.18948, + "grad_norm": 0.7954782269380947, + "learning_rate": 0.003, + "loss": 4.016, + "step": 18948 + }, + { + "epoch": 0.18949, + "grad_norm": 0.5977919821099205, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 18949 + }, + { + "epoch": 0.1895, + "grad_norm": 0.6340510718946033, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 18950 + }, + { + "epoch": 0.18951, + "grad_norm": 0.6878577585642967, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 18951 + }, + { + "epoch": 0.18952, + "grad_norm": 0.6992456817537874, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 18952 + }, + { + "epoch": 0.18953, + "grad_norm": 0.7345175011567718, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 18953 + }, + { + "epoch": 0.18954, + "grad_norm": 0.6326454894202036, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 18954 + }, + { + "epoch": 0.18955, + "grad_norm": 0.5842429618604259, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 18955 + }, + { + "epoch": 0.18956, + "grad_norm": 0.656578044402032, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 18956 + }, + { + "epoch": 0.18957, + "grad_norm": 0.8046573883132105, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 18957 + }, + { + "epoch": 0.18958, + "grad_norm": 0.9551723781468637, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 18958 + }, + { + "epoch": 0.18959, + "grad_norm": 1.1004209312219526, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 18959 + }, + { + "epoch": 0.1896, + "grad_norm": 0.851558819775649, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 18960 + }, + { + "epoch": 0.18961, + "grad_norm": 0.6822066018078725, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 18961 + }, + { + "epoch": 0.18962, + "grad_norm": 0.6126694406525554, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 18962 + }, + { + "epoch": 0.18963, + "grad_norm": 0.6921769400771112, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 18963 + }, + { + "epoch": 0.18964, + "grad_norm": 0.7229650650986484, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 18964 + }, + { + "epoch": 0.18965, + "grad_norm": 0.7825425375205395, + "learning_rate": 0.003, + "loss": 4.015, + "step": 18965 + }, + { + "epoch": 0.18966, + "grad_norm": 0.7495952747806705, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 18966 + }, + { + "epoch": 0.18967, + "grad_norm": 0.8255888529793396, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18967 + }, + { + "epoch": 0.18968, + "grad_norm": 0.8193059900373121, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 18968 + }, + { + "epoch": 0.18969, + "grad_norm": 0.8352566625222692, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 18969 + }, + { + "epoch": 0.1897, + "grad_norm": 0.9786284199441596, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 18970 + }, + { + "epoch": 0.18971, + "grad_norm": 1.0428914310543071, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 18971 + }, + { + "epoch": 0.18972, + "grad_norm": 1.093001006111784, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 18972 + }, + { + "epoch": 0.18973, + "grad_norm": 0.927594684346053, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 18973 + }, + { + "epoch": 0.18974, + "grad_norm": 0.9379063709355582, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 18974 + }, + { + "epoch": 0.18975, + "grad_norm": 0.8771259138661749, + "learning_rate": 0.003, + "loss": 4.016, + "step": 18975 + }, + { + "epoch": 0.18976, + "grad_norm": 0.8293653372910518, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 18976 + }, + { + "epoch": 0.18977, + "grad_norm": 0.8232919207967928, + "learning_rate": 0.003, + "loss": 4.0458, + "step": 18977 + }, + { + "epoch": 0.18978, + "grad_norm": 0.8371980888017003, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 18978 + }, + { + "epoch": 0.18979, + "grad_norm": 0.7525932525741005, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 18979 + }, + { + "epoch": 0.1898, + "grad_norm": 0.734188672884096, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 18980 + }, + { + "epoch": 0.18981, + "grad_norm": 0.7958642121391422, + "learning_rate": 0.003, + "loss": 4.017, + "step": 18981 + }, + { + "epoch": 0.18982, + "grad_norm": 0.8235747759247536, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 18982 + }, + { + "epoch": 0.18983, + "grad_norm": 0.745913397929208, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 18983 + }, + { + "epoch": 0.18984, + "grad_norm": 0.7272692057573827, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 18984 + }, + { + "epoch": 0.18985, + "grad_norm": 0.6440940872217832, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 18985 + }, + { + "epoch": 0.18986, + "grad_norm": 0.5908082700920034, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 18986 + }, + { + "epoch": 0.18987, + "grad_norm": 0.567038760563411, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 18987 + }, + { + "epoch": 0.18988, + "grad_norm": 0.6382276559128748, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 18988 + }, + { + "epoch": 0.18989, + "grad_norm": 0.7533248838179428, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 18989 + }, + { + "epoch": 0.1899, + "grad_norm": 0.8632728995494264, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 18990 + }, + { + "epoch": 0.18991, + "grad_norm": 0.795928248126452, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 18991 + }, + { + "epoch": 0.18992, + "grad_norm": 0.6542022409025758, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 18992 + }, + { + "epoch": 0.18993, + "grad_norm": 0.5937384390386119, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 18993 + }, + { + "epoch": 0.18994, + "grad_norm": 0.6735092125055353, + "learning_rate": 0.003, + "loss": 4.006, + "step": 18994 + }, + { + "epoch": 0.18995, + "grad_norm": 0.7346736895759813, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 18995 + }, + { + "epoch": 0.18996, + "grad_norm": 0.7252355986010839, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 18996 + }, + { + "epoch": 0.18997, + "grad_norm": 0.7717092087573135, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 18997 + }, + { + "epoch": 0.18998, + "grad_norm": 0.7710263275541609, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 18998 + }, + { + "epoch": 0.18999, + "grad_norm": 0.7465296095644909, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 18999 + }, + { + "epoch": 0.19, + "grad_norm": 0.7780264856079583, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 19000 + }, + { + "epoch": 0.19001, + "grad_norm": 0.689148707651334, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 19001 + }, + { + "epoch": 0.19002, + "grad_norm": 0.5735310618460556, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 19002 + }, + { + "epoch": 0.19003, + "grad_norm": 0.5608647567231988, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 19003 + }, + { + "epoch": 0.19004, + "grad_norm": 0.5908867648974973, + "learning_rate": 0.003, + "loss": 3.998, + "step": 19004 + }, + { + "epoch": 0.19005, + "grad_norm": 0.6809882361521168, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 19005 + }, + { + "epoch": 0.19006, + "grad_norm": 0.6570623857426389, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 19006 + }, + { + "epoch": 0.19007, + "grad_norm": 0.708840797082912, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 19007 + }, + { + "epoch": 0.19008, + "grad_norm": 0.7323487239875279, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 19008 + }, + { + "epoch": 0.19009, + "grad_norm": 0.6789547207461096, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 19009 + }, + { + "epoch": 0.1901, + "grad_norm": 0.6736947485833079, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 19010 + }, + { + "epoch": 0.19011, + "grad_norm": 0.7174189033531232, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 19011 + }, + { + "epoch": 0.19012, + "grad_norm": 0.6862752713469927, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 19012 + }, + { + "epoch": 0.19013, + "grad_norm": 0.6402372153580539, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 19013 + }, + { + "epoch": 0.19014, + "grad_norm": 0.533642298636183, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 19014 + }, + { + "epoch": 0.19015, + "grad_norm": 0.45713332374048027, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 19015 + }, + { + "epoch": 0.19016, + "grad_norm": 0.506291473002686, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 19016 + }, + { + "epoch": 0.19017, + "grad_norm": 0.5352936334682412, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 19017 + }, + { + "epoch": 0.19018, + "grad_norm": 0.6623944033748248, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 19018 + }, + { + "epoch": 0.19019, + "grad_norm": 0.92301418492652, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 19019 + }, + { + "epoch": 0.1902, + "grad_norm": 1.0888925265801683, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 19020 + }, + { + "epoch": 0.19021, + "grad_norm": 0.8630300199005617, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 19021 + }, + { + "epoch": 0.19022, + "grad_norm": 0.7225690750697273, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 19022 + }, + { + "epoch": 0.19023, + "grad_norm": 0.6312671001326444, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 19023 + }, + { + "epoch": 0.19024, + "grad_norm": 0.6297177989261042, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 19024 + }, + { + "epoch": 0.19025, + "grad_norm": 0.6248867543898163, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 19025 + }, + { + "epoch": 0.19026, + "grad_norm": 0.6592650593052806, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 19026 + }, + { + "epoch": 0.19027, + "grad_norm": 0.7897927803765525, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 19027 + }, + { + "epoch": 0.19028, + "grad_norm": 0.8844968854956691, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 19028 + }, + { + "epoch": 0.19029, + "grad_norm": 0.9233489857022502, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 19029 + }, + { + "epoch": 0.1903, + "grad_norm": 0.8566835005903751, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 19030 + }, + { + "epoch": 0.19031, + "grad_norm": 0.9351103570600363, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 19031 + }, + { + "epoch": 0.19032, + "grad_norm": 1.1632759938637085, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 19032 + }, + { + "epoch": 0.19033, + "grad_norm": 0.8310790355267728, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 19033 + }, + { + "epoch": 0.19034, + "grad_norm": 0.7656571490368123, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 19034 + }, + { + "epoch": 0.19035, + "grad_norm": 0.8895430619566485, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 19035 + }, + { + "epoch": 0.19036, + "grad_norm": 1.0715846451847735, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 19036 + }, + { + "epoch": 0.19037, + "grad_norm": 0.9263849884844452, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 19037 + }, + { + "epoch": 0.19038, + "grad_norm": 0.7808820992197361, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 19038 + }, + { + "epoch": 0.19039, + "grad_norm": 0.7732631125999377, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 19039 + }, + { + "epoch": 0.1904, + "grad_norm": 0.7804307626305859, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 19040 + }, + { + "epoch": 0.19041, + "grad_norm": 0.6429431132462742, + "learning_rate": 0.003, + "loss": 4.009, + "step": 19041 + }, + { + "epoch": 0.19042, + "grad_norm": 0.7166398541841839, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 19042 + }, + { + "epoch": 0.19043, + "grad_norm": 0.8215246864882297, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 19043 + }, + { + "epoch": 0.19044, + "grad_norm": 0.9525750927056484, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 19044 + }, + { + "epoch": 0.19045, + "grad_norm": 1.030543319969092, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 19045 + }, + { + "epoch": 0.19046, + "grad_norm": 1.0598404174640383, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 19046 + }, + { + "epoch": 0.19047, + "grad_norm": 0.7318804759940901, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 19047 + }, + { + "epoch": 0.19048, + "grad_norm": 0.6773359176325405, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 19048 + }, + { + "epoch": 0.19049, + "grad_norm": 0.7126441922542207, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 19049 + }, + { + "epoch": 0.1905, + "grad_norm": 0.7184255271827857, + "learning_rate": 0.003, + "loss": 4.0413, + "step": 19050 + }, + { + "epoch": 0.19051, + "grad_norm": 0.7058606522825595, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 19051 + }, + { + "epoch": 0.19052, + "grad_norm": 0.7013810194935293, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 19052 + }, + { + "epoch": 0.19053, + "grad_norm": 0.7273313736595908, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 19053 + }, + { + "epoch": 0.19054, + "grad_norm": 0.6561357277401056, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 19054 + }, + { + "epoch": 0.19055, + "grad_norm": 0.5855905453842972, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 19055 + }, + { + "epoch": 0.19056, + "grad_norm": 0.6162345048315252, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 19056 + }, + { + "epoch": 0.19057, + "grad_norm": 0.6769463767983404, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 19057 + }, + { + "epoch": 0.19058, + "grad_norm": 0.7628304765991879, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 19058 + }, + { + "epoch": 0.19059, + "grad_norm": 0.8729722430074068, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 19059 + }, + { + "epoch": 0.1906, + "grad_norm": 0.8829172087396173, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 19060 + }, + { + "epoch": 0.19061, + "grad_norm": 0.6597899858333224, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 19061 + }, + { + "epoch": 0.19062, + "grad_norm": 0.6805135867598191, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 19062 + }, + { + "epoch": 0.19063, + "grad_norm": 0.7753230008008181, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 19063 + }, + { + "epoch": 0.19064, + "grad_norm": 0.8094876399824424, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 19064 + }, + { + "epoch": 0.19065, + "grad_norm": 0.8059137334135541, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 19065 + }, + { + "epoch": 0.19066, + "grad_norm": 0.7298050486936826, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 19066 + }, + { + "epoch": 0.19067, + "grad_norm": 0.5935747416580236, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 19067 + }, + { + "epoch": 0.19068, + "grad_norm": 0.6150505490085798, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 19068 + }, + { + "epoch": 0.19069, + "grad_norm": 0.649537456068245, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 19069 + }, + { + "epoch": 0.1907, + "grad_norm": 0.6379861805926257, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 19070 + }, + { + "epoch": 0.19071, + "grad_norm": 0.6404285818205513, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 19071 + }, + { + "epoch": 0.19072, + "grad_norm": 0.7144576997635619, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 19072 + }, + { + "epoch": 0.19073, + "grad_norm": 0.8758886253449099, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 19073 + }, + { + "epoch": 0.19074, + "grad_norm": 0.9685331268011812, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 19074 + }, + { + "epoch": 0.19075, + "grad_norm": 0.996776709781083, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 19075 + }, + { + "epoch": 0.19076, + "grad_norm": 0.9326967855807528, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 19076 + }, + { + "epoch": 0.19077, + "grad_norm": 0.7991032771581802, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 19077 + }, + { + "epoch": 0.19078, + "grad_norm": 0.6952938450646481, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 19078 + }, + { + "epoch": 0.19079, + "grad_norm": 0.564406002290839, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 19079 + }, + { + "epoch": 0.1908, + "grad_norm": 0.5219759667532666, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 19080 + }, + { + "epoch": 0.19081, + "grad_norm": 0.5436308110151069, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 19081 + }, + { + "epoch": 0.19082, + "grad_norm": 0.5248041249344221, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 19082 + }, + { + "epoch": 0.19083, + "grad_norm": 0.5553987168127129, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 19083 + }, + { + "epoch": 0.19084, + "grad_norm": 0.6593341860540991, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 19084 + }, + { + "epoch": 0.19085, + "grad_norm": 0.6549286560642488, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 19085 + }, + { + "epoch": 0.19086, + "grad_norm": 0.6358004403503749, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 19086 + }, + { + "epoch": 0.19087, + "grad_norm": 0.7176314113126413, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 19087 + }, + { + "epoch": 0.19088, + "grad_norm": 0.8195563315157258, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 19088 + }, + { + "epoch": 0.19089, + "grad_norm": 0.8306386512849965, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 19089 + }, + { + "epoch": 0.1909, + "grad_norm": 0.7789949990781693, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 19090 + }, + { + "epoch": 0.19091, + "grad_norm": 0.7691142576620014, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 19091 + }, + { + "epoch": 0.19092, + "grad_norm": 0.8013215702936134, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 19092 + }, + { + "epoch": 0.19093, + "grad_norm": 0.8381509986540164, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 19093 + }, + { + "epoch": 0.19094, + "grad_norm": 0.7820641560333031, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 19094 + }, + { + "epoch": 0.19095, + "grad_norm": 0.8062916440374298, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 19095 + }, + { + "epoch": 0.19096, + "grad_norm": 0.9669249262440933, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 19096 + }, + { + "epoch": 0.19097, + "grad_norm": 1.0515012074826697, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 19097 + }, + { + "epoch": 0.19098, + "grad_norm": 1.0086636396739284, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 19098 + }, + { + "epoch": 0.19099, + "grad_norm": 1.0680511391095744, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 19099 + }, + { + "epoch": 0.191, + "grad_norm": 0.7839594655050686, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 19100 + }, + { + "epoch": 0.19101, + "grad_norm": 0.6674212088508157, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 19101 + }, + { + "epoch": 0.19102, + "grad_norm": 0.6555821911264834, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 19102 + }, + { + "epoch": 0.19103, + "grad_norm": 0.7086252876976309, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 19103 + }, + { + "epoch": 0.19104, + "grad_norm": 0.7430477201630983, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 19104 + }, + { + "epoch": 0.19105, + "grad_norm": 0.8220273414635798, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 19105 + }, + { + "epoch": 0.19106, + "grad_norm": 0.7842896267880082, + "learning_rate": 0.003, + "loss": 4.004, + "step": 19106 + }, + { + "epoch": 0.19107, + "grad_norm": 0.6914084454977361, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 19107 + }, + { + "epoch": 0.19108, + "grad_norm": 0.679721293874475, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 19108 + }, + { + "epoch": 0.19109, + "grad_norm": 0.6422493550709218, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 19109 + }, + { + "epoch": 0.1911, + "grad_norm": 0.6265900455636845, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 19110 + }, + { + "epoch": 0.19111, + "grad_norm": 0.6394818627872147, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 19111 + }, + { + "epoch": 0.19112, + "grad_norm": 0.6111607218821922, + "learning_rate": 0.003, + "loss": 4.018, + "step": 19112 + }, + { + "epoch": 0.19113, + "grad_norm": 0.6489103084049878, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 19113 + }, + { + "epoch": 0.19114, + "grad_norm": 0.7080266670424573, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 19114 + }, + { + "epoch": 0.19115, + "grad_norm": 0.7046346429403711, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 19115 + }, + { + "epoch": 0.19116, + "grad_norm": 0.7484935383989108, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 19116 + }, + { + "epoch": 0.19117, + "grad_norm": 0.7476519606635612, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 19117 + }, + { + "epoch": 0.19118, + "grad_norm": 0.6824650627889394, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 19118 + }, + { + "epoch": 0.19119, + "grad_norm": 0.6823925278973962, + "learning_rate": 0.003, + "loss": 4.0435, + "step": 19119 + }, + { + "epoch": 0.1912, + "grad_norm": 0.5944705146407733, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 19120 + }, + { + "epoch": 0.19121, + "grad_norm": 0.624004472186464, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 19121 + }, + { + "epoch": 0.19122, + "grad_norm": 0.677614646384963, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 19122 + }, + { + "epoch": 0.19123, + "grad_norm": 0.7526296796548863, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 19123 + }, + { + "epoch": 0.19124, + "grad_norm": 0.7036167789519913, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 19124 + }, + { + "epoch": 0.19125, + "grad_norm": 0.591817299495074, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 19125 + }, + { + "epoch": 0.19126, + "grad_norm": 0.5866342656132048, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 19126 + }, + { + "epoch": 0.19127, + "grad_norm": 0.5468364457412994, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 19127 + }, + { + "epoch": 0.19128, + "grad_norm": 0.6127122445221858, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 19128 + }, + { + "epoch": 0.19129, + "grad_norm": 0.7549136503088882, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 19129 + }, + { + "epoch": 0.1913, + "grad_norm": 0.945028439911932, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 19130 + }, + { + "epoch": 0.19131, + "grad_norm": 1.0930525116212304, + "learning_rate": 0.003, + "loss": 4.003, + "step": 19131 + }, + { + "epoch": 0.19132, + "grad_norm": 0.881188163383802, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 19132 + }, + { + "epoch": 0.19133, + "grad_norm": 0.7916304993618916, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 19133 + }, + { + "epoch": 0.19134, + "grad_norm": 0.8646505494437072, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 19134 + }, + { + "epoch": 0.19135, + "grad_norm": 0.906428961445756, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 19135 + }, + { + "epoch": 0.19136, + "grad_norm": 0.8096010204448394, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 19136 + }, + { + "epoch": 0.19137, + "grad_norm": 0.7463341473645309, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 19137 + }, + { + "epoch": 0.19138, + "grad_norm": 0.8491790179063561, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 19138 + }, + { + "epoch": 0.19139, + "grad_norm": 0.9145507057810345, + "learning_rate": 0.003, + "loss": 4.039, + "step": 19139 + }, + { + "epoch": 0.1914, + "grad_norm": 0.774758144448813, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 19140 + }, + { + "epoch": 0.19141, + "grad_norm": 0.7899383285649403, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 19141 + }, + { + "epoch": 0.19142, + "grad_norm": 0.8369306505578721, + "learning_rate": 0.003, + "loss": 4.007, + "step": 19142 + }, + { + "epoch": 0.19143, + "grad_norm": 0.8545529638941368, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 19143 + }, + { + "epoch": 0.19144, + "grad_norm": 0.7810496838699009, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 19144 + }, + { + "epoch": 0.19145, + "grad_norm": 0.6908087363016453, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 19145 + }, + { + "epoch": 0.19146, + "grad_norm": 0.6434688068587873, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 19146 + }, + { + "epoch": 0.19147, + "grad_norm": 0.6371625753853151, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 19147 + }, + { + "epoch": 0.19148, + "grad_norm": 0.7962263066194022, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 19148 + }, + { + "epoch": 0.19149, + "grad_norm": 0.8075239479790189, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 19149 + }, + { + "epoch": 0.1915, + "grad_norm": 0.8738143462141499, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 19150 + }, + { + "epoch": 0.19151, + "grad_norm": 0.9316718919010668, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 19151 + }, + { + "epoch": 0.19152, + "grad_norm": 0.840650522668778, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 19152 + }, + { + "epoch": 0.19153, + "grad_norm": 0.790506752683176, + "learning_rate": 0.003, + "loss": 4.018, + "step": 19153 + }, + { + "epoch": 0.19154, + "grad_norm": 0.85834906853454, + "learning_rate": 0.003, + "loss": 4.033, + "step": 19154 + }, + { + "epoch": 0.19155, + "grad_norm": 0.8775201731111041, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 19155 + }, + { + "epoch": 0.19156, + "grad_norm": 0.8823279701773312, + "learning_rate": 0.003, + "loss": 4.032, + "step": 19156 + }, + { + "epoch": 0.19157, + "grad_norm": 0.8693984266536867, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 19157 + }, + { + "epoch": 0.19158, + "grad_norm": 0.9491978246756176, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 19158 + }, + { + "epoch": 0.19159, + "grad_norm": 0.9124371601700461, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 19159 + }, + { + "epoch": 0.1916, + "grad_norm": 0.9184385238230427, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 19160 + }, + { + "epoch": 0.19161, + "grad_norm": 0.8955999627218225, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 19161 + }, + { + "epoch": 0.19162, + "grad_norm": 0.8066000677972759, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 19162 + }, + { + "epoch": 0.19163, + "grad_norm": 0.8043570146528062, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 19163 + }, + { + "epoch": 0.19164, + "grad_norm": 0.7750640750848351, + "learning_rate": 0.003, + "loss": 4.0664, + "step": 19164 + }, + { + "epoch": 0.19165, + "grad_norm": 0.7043552441982787, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 19165 + }, + { + "epoch": 0.19166, + "grad_norm": 0.6831014524592453, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 19166 + }, + { + "epoch": 0.19167, + "grad_norm": 0.6254851620308229, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 19167 + }, + { + "epoch": 0.19168, + "grad_norm": 0.6262484892920389, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 19168 + }, + { + "epoch": 0.19169, + "grad_norm": 0.7596178790658732, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 19169 + }, + { + "epoch": 0.1917, + "grad_norm": 0.7715233932419178, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 19170 + }, + { + "epoch": 0.19171, + "grad_norm": 0.7256530966819816, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 19171 + }, + { + "epoch": 0.19172, + "grad_norm": 0.8234856996083915, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 19172 + }, + { + "epoch": 0.19173, + "grad_norm": 0.7950041263353835, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 19173 + }, + { + "epoch": 0.19174, + "grad_norm": 0.7346597865278666, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 19174 + }, + { + "epoch": 0.19175, + "grad_norm": 0.7495225132448612, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 19175 + }, + { + "epoch": 0.19176, + "grad_norm": 0.7355787583658155, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 19176 + }, + { + "epoch": 0.19177, + "grad_norm": 0.7532067656162682, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 19177 + }, + { + "epoch": 0.19178, + "grad_norm": 0.7538301570647516, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 19178 + }, + { + "epoch": 0.19179, + "grad_norm": 0.6570452737635055, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 19179 + }, + { + "epoch": 0.1918, + "grad_norm": 0.6310053171919779, + "learning_rate": 0.003, + "loss": 3.981, + "step": 19180 + }, + { + "epoch": 0.19181, + "grad_norm": 0.5840715201431911, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 19181 + }, + { + "epoch": 0.19182, + "grad_norm": 0.6059164230366256, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 19182 + }, + { + "epoch": 0.19183, + "grad_norm": 0.554366346912067, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 19183 + }, + { + "epoch": 0.19184, + "grad_norm": 0.5366071344499657, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 19184 + }, + { + "epoch": 0.19185, + "grad_norm": 0.595639482562865, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 19185 + }, + { + "epoch": 0.19186, + "grad_norm": 0.6324710363053769, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 19186 + }, + { + "epoch": 0.19187, + "grad_norm": 0.7294643535261784, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 19187 + }, + { + "epoch": 0.19188, + "grad_norm": 0.7818079173404648, + "learning_rate": 0.003, + "loss": 3.996, + "step": 19188 + }, + { + "epoch": 0.19189, + "grad_norm": 0.7789674379625573, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 19189 + }, + { + "epoch": 0.1919, + "grad_norm": 0.8123895431946101, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 19190 + }, + { + "epoch": 0.19191, + "grad_norm": 0.7693468812500356, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 19191 + }, + { + "epoch": 0.19192, + "grad_norm": 0.761005093444604, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 19192 + }, + { + "epoch": 0.19193, + "grad_norm": 0.8108307432558706, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 19193 + }, + { + "epoch": 0.19194, + "grad_norm": 0.8748929909161637, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 19194 + }, + { + "epoch": 0.19195, + "grad_norm": 0.9894020766593868, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 19195 + }, + { + "epoch": 0.19196, + "grad_norm": 0.9580985485725522, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 19196 + }, + { + "epoch": 0.19197, + "grad_norm": 0.8955564973171809, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 19197 + }, + { + "epoch": 0.19198, + "grad_norm": 0.9504877859967438, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 19198 + }, + { + "epoch": 0.19199, + "grad_norm": 0.8630493415082274, + "learning_rate": 0.003, + "loss": 4.039, + "step": 19199 + }, + { + "epoch": 0.192, + "grad_norm": 0.7471652403431296, + "learning_rate": 0.003, + "loss": 4.0516, + "step": 19200 + }, + { + "epoch": 0.19201, + "grad_norm": 0.6647933754044645, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 19201 + }, + { + "epoch": 0.19202, + "grad_norm": 0.6521588062991923, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 19202 + }, + { + "epoch": 0.19203, + "grad_norm": 0.5332166820544656, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 19203 + }, + { + "epoch": 0.19204, + "grad_norm": 0.520458220499647, + "learning_rate": 0.003, + "loss": 3.995, + "step": 19204 + }, + { + "epoch": 0.19205, + "grad_norm": 0.4954172605627414, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 19205 + }, + { + "epoch": 0.19206, + "grad_norm": 0.43093954532031725, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 19206 + }, + { + "epoch": 0.19207, + "grad_norm": 0.44575042081767924, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 19207 + }, + { + "epoch": 0.19208, + "grad_norm": 0.47040473049379794, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 19208 + }, + { + "epoch": 0.19209, + "grad_norm": 0.48042801876840335, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 19209 + }, + { + "epoch": 0.1921, + "grad_norm": 0.5618422403189889, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 19210 + }, + { + "epoch": 0.19211, + "grad_norm": 0.6450891374021275, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 19211 + }, + { + "epoch": 0.19212, + "grad_norm": 0.6798447806274542, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 19212 + }, + { + "epoch": 0.19213, + "grad_norm": 0.7741963032281262, + "learning_rate": 0.003, + "loss": 4.008, + "step": 19213 + }, + { + "epoch": 0.19214, + "grad_norm": 0.7865885664417167, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 19214 + }, + { + "epoch": 0.19215, + "grad_norm": 0.9609672529045892, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 19215 + }, + { + "epoch": 0.19216, + "grad_norm": 1.2450385614563504, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 19216 + }, + { + "epoch": 0.19217, + "grad_norm": 0.8490064828501218, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 19217 + }, + { + "epoch": 0.19218, + "grad_norm": 0.7417635750223123, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 19218 + }, + { + "epoch": 0.19219, + "grad_norm": 0.7226136351664606, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 19219 + }, + { + "epoch": 0.1922, + "grad_norm": 0.608230613368551, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 19220 + }, + { + "epoch": 0.19221, + "grad_norm": 0.6054654236034096, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 19221 + }, + { + "epoch": 0.19222, + "grad_norm": 0.6336917906003374, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 19222 + }, + { + "epoch": 0.19223, + "grad_norm": 0.7383060001522851, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 19223 + }, + { + "epoch": 0.19224, + "grad_norm": 0.9135351280546246, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 19224 + }, + { + "epoch": 0.19225, + "grad_norm": 1.0327216764576783, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 19225 + }, + { + "epoch": 0.19226, + "grad_norm": 0.9729972041365594, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 19226 + }, + { + "epoch": 0.19227, + "grad_norm": 0.715599900504076, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 19227 + }, + { + "epoch": 0.19228, + "grad_norm": 0.6711213877409073, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 19228 + }, + { + "epoch": 0.19229, + "grad_norm": 0.6551300014047873, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 19229 + }, + { + "epoch": 0.1923, + "grad_norm": 0.6233433067755414, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 19230 + }, + { + "epoch": 0.19231, + "grad_norm": 0.6232040650254479, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 19231 + }, + { + "epoch": 0.19232, + "grad_norm": 0.6283612380147721, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 19232 + }, + { + "epoch": 0.19233, + "grad_norm": 0.6587409368335579, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 19233 + }, + { + "epoch": 0.19234, + "grad_norm": 0.6657284070260111, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 19234 + }, + { + "epoch": 0.19235, + "grad_norm": 0.6989331996460008, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 19235 + }, + { + "epoch": 0.19236, + "grad_norm": 0.7219493086073241, + "learning_rate": 0.003, + "loss": 4.021, + "step": 19236 + }, + { + "epoch": 0.19237, + "grad_norm": 0.6907173092024301, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 19237 + }, + { + "epoch": 0.19238, + "grad_norm": 0.6962082620096226, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 19238 + }, + { + "epoch": 0.19239, + "grad_norm": 0.7173513436937835, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 19239 + }, + { + "epoch": 0.1924, + "grad_norm": 0.6988324831921603, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 19240 + }, + { + "epoch": 0.19241, + "grad_norm": 0.6676884740839915, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 19241 + }, + { + "epoch": 0.19242, + "grad_norm": 0.6787226894333233, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 19242 + }, + { + "epoch": 0.19243, + "grad_norm": 0.6758227667619502, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 19243 + }, + { + "epoch": 0.19244, + "grad_norm": 0.6871703114138528, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 19244 + }, + { + "epoch": 0.19245, + "grad_norm": 0.7260468271456799, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 19245 + }, + { + "epoch": 0.19246, + "grad_norm": 0.7962445127109903, + "learning_rate": 0.003, + "loss": 4.009, + "step": 19246 + }, + { + "epoch": 0.19247, + "grad_norm": 0.8372244159346, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 19247 + }, + { + "epoch": 0.19248, + "grad_norm": 0.9486492243638456, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 19248 + }, + { + "epoch": 0.19249, + "grad_norm": 0.8933575926984627, + "learning_rate": 0.003, + "loss": 3.984, + "step": 19249 + }, + { + "epoch": 0.1925, + "grad_norm": 0.8902603519467372, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 19250 + }, + { + "epoch": 0.19251, + "grad_norm": 0.858730161030556, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 19251 + }, + { + "epoch": 0.19252, + "grad_norm": 0.7773358640547423, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 19252 + }, + { + "epoch": 0.19253, + "grad_norm": 0.7413616804845438, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 19253 + }, + { + "epoch": 0.19254, + "grad_norm": 0.734638648942607, + "learning_rate": 0.003, + "loss": 4.0494, + "step": 19254 + }, + { + "epoch": 0.19255, + "grad_norm": 0.7709146949096248, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 19255 + }, + { + "epoch": 0.19256, + "grad_norm": 0.7838821165569148, + "learning_rate": 0.003, + "loss": 4.001, + "step": 19256 + }, + { + "epoch": 0.19257, + "grad_norm": 0.7340581396954181, + "learning_rate": 0.003, + "loss": 3.978, + "step": 19257 + }, + { + "epoch": 0.19258, + "grad_norm": 0.6958746937725968, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 19258 + }, + { + "epoch": 0.19259, + "grad_norm": 0.7079030636348089, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 19259 + }, + { + "epoch": 0.1926, + "grad_norm": 0.7565740077259201, + "learning_rate": 0.003, + "loss": 4.012, + "step": 19260 + }, + { + "epoch": 0.19261, + "grad_norm": 0.7344689770184313, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 19261 + }, + { + "epoch": 0.19262, + "grad_norm": 0.6660731819978915, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 19262 + }, + { + "epoch": 0.19263, + "grad_norm": 0.7627387240174507, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 19263 + }, + { + "epoch": 0.19264, + "grad_norm": 0.8816068371208893, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 19264 + }, + { + "epoch": 0.19265, + "grad_norm": 1.0038810754645258, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 19265 + }, + { + "epoch": 0.19266, + "grad_norm": 0.9713467242292174, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 19266 + }, + { + "epoch": 0.19267, + "grad_norm": 1.0045962696893327, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 19267 + }, + { + "epoch": 0.19268, + "grad_norm": 0.9764019986561282, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 19268 + }, + { + "epoch": 0.19269, + "grad_norm": 0.8792427013378326, + "learning_rate": 0.003, + "loss": 4.034, + "step": 19269 + }, + { + "epoch": 0.1927, + "grad_norm": 0.7845576452956575, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 19270 + }, + { + "epoch": 0.19271, + "grad_norm": 0.7490018633519255, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 19271 + }, + { + "epoch": 0.19272, + "grad_norm": 0.6735606294895752, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 19272 + }, + { + "epoch": 0.19273, + "grad_norm": 0.6289690580105658, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 19273 + }, + { + "epoch": 0.19274, + "grad_norm": 0.6084241532658018, + "learning_rate": 0.003, + "loss": 4.037, + "step": 19274 + }, + { + "epoch": 0.19275, + "grad_norm": 0.5998630833655659, + "learning_rate": 0.003, + "loss": 3.995, + "step": 19275 + }, + { + "epoch": 0.19276, + "grad_norm": 0.5818895371958196, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 19276 + }, + { + "epoch": 0.19277, + "grad_norm": 0.6126717624451588, + "learning_rate": 0.003, + "loss": 4.042, + "step": 19277 + }, + { + "epoch": 0.19278, + "grad_norm": 0.5885840800823705, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 19278 + }, + { + "epoch": 0.19279, + "grad_norm": 0.6260003030503422, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 19279 + }, + { + "epoch": 0.1928, + "grad_norm": 0.676258105638988, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 19280 + }, + { + "epoch": 0.19281, + "grad_norm": 0.6526456728103947, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 19281 + }, + { + "epoch": 0.19282, + "grad_norm": 0.7094252651764399, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 19282 + }, + { + "epoch": 0.19283, + "grad_norm": 0.8214062629299486, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 19283 + }, + { + "epoch": 0.19284, + "grad_norm": 0.8142067331391974, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 19284 + }, + { + "epoch": 0.19285, + "grad_norm": 0.8101899409565294, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 19285 + }, + { + "epoch": 0.19286, + "grad_norm": 0.8225320903775878, + "learning_rate": 0.003, + "loss": 3.977, + "step": 19286 + }, + { + "epoch": 0.19287, + "grad_norm": 0.8576065600999572, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 19287 + }, + { + "epoch": 0.19288, + "grad_norm": 0.9000559678379374, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 19288 + }, + { + "epoch": 0.19289, + "grad_norm": 0.9569356021389782, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 19289 + }, + { + "epoch": 0.1929, + "grad_norm": 0.9285980033573457, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 19290 + }, + { + "epoch": 0.19291, + "grad_norm": 0.8983185152984823, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 19291 + }, + { + "epoch": 0.19292, + "grad_norm": 0.8609024890337953, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 19292 + }, + { + "epoch": 0.19293, + "grad_norm": 0.8595308478355322, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 19293 + }, + { + "epoch": 0.19294, + "grad_norm": 0.7938205960423625, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 19294 + }, + { + "epoch": 0.19295, + "grad_norm": 0.6916162409787557, + "learning_rate": 0.003, + "loss": 3.997, + "step": 19295 + }, + { + "epoch": 0.19296, + "grad_norm": 0.6028211083047526, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 19296 + }, + { + "epoch": 0.19297, + "grad_norm": 0.6353048454335649, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 19297 + }, + { + "epoch": 0.19298, + "grad_norm": 0.5507060966004189, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 19298 + }, + { + "epoch": 0.19299, + "grad_norm": 0.5687060400424155, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 19299 + }, + { + "epoch": 0.193, + "grad_norm": 0.6565957665901332, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 19300 + }, + { + "epoch": 0.19301, + "grad_norm": 0.6819202774734004, + "learning_rate": 0.003, + "loss": 4.031, + "step": 19301 + }, + { + "epoch": 0.19302, + "grad_norm": 0.6796122890012234, + "learning_rate": 0.003, + "loss": 4.002, + "step": 19302 + }, + { + "epoch": 0.19303, + "grad_norm": 0.6630644685601884, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 19303 + }, + { + "epoch": 0.19304, + "grad_norm": 0.6989257402538083, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19304 + }, + { + "epoch": 0.19305, + "grad_norm": 0.6412718751069395, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 19305 + }, + { + "epoch": 0.19306, + "grad_norm": 0.6010283524764974, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 19306 + }, + { + "epoch": 0.19307, + "grad_norm": 0.544677335185735, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 19307 + }, + { + "epoch": 0.19308, + "grad_norm": 0.536986124698817, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 19308 + }, + { + "epoch": 0.19309, + "grad_norm": 0.4707724227893927, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 19309 + }, + { + "epoch": 0.1931, + "grad_norm": 0.522657115248014, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19310 + }, + { + "epoch": 0.19311, + "grad_norm": 0.5257576276042324, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 19311 + }, + { + "epoch": 0.19312, + "grad_norm": 0.5944469874372587, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 19312 + }, + { + "epoch": 0.19313, + "grad_norm": 0.7825764486001341, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 19313 + }, + { + "epoch": 0.19314, + "grad_norm": 1.064353210695835, + "learning_rate": 0.003, + "loss": 4.012, + "step": 19314 + }, + { + "epoch": 0.19315, + "grad_norm": 1.0605526875149942, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 19315 + }, + { + "epoch": 0.19316, + "grad_norm": 0.8540565997270588, + "learning_rate": 0.003, + "loss": 3.988, + "step": 19316 + }, + { + "epoch": 0.19317, + "grad_norm": 0.8600913386605066, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 19317 + }, + { + "epoch": 0.19318, + "grad_norm": 0.9413801698976074, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 19318 + }, + { + "epoch": 0.19319, + "grad_norm": 0.9275206750699673, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 19319 + }, + { + "epoch": 0.1932, + "grad_norm": 0.8068898743128018, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 19320 + }, + { + "epoch": 0.19321, + "grad_norm": 0.7465241906027335, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 19321 + }, + { + "epoch": 0.19322, + "grad_norm": 0.818368636265983, + "learning_rate": 0.003, + "loss": 4.028, + "step": 19322 + }, + { + "epoch": 0.19323, + "grad_norm": 0.9278579994904966, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 19323 + }, + { + "epoch": 0.19324, + "grad_norm": 0.8549968792733643, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 19324 + }, + { + "epoch": 0.19325, + "grad_norm": 0.9273334123936273, + "learning_rate": 0.003, + "loss": 4.012, + "step": 19325 + }, + { + "epoch": 0.19326, + "grad_norm": 0.9837284304727744, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 19326 + }, + { + "epoch": 0.19327, + "grad_norm": 1.041032532563577, + "learning_rate": 0.003, + "loss": 4.04, + "step": 19327 + }, + { + "epoch": 0.19328, + "grad_norm": 0.9089328600487329, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 19328 + }, + { + "epoch": 0.19329, + "grad_norm": 0.7490036538258049, + "learning_rate": 0.003, + "loss": 4.031, + "step": 19329 + }, + { + "epoch": 0.1933, + "grad_norm": 0.82055656617776, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 19330 + }, + { + "epoch": 0.19331, + "grad_norm": 0.8471419011858651, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 19331 + }, + { + "epoch": 0.19332, + "grad_norm": 0.8370372255612256, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 19332 + }, + { + "epoch": 0.19333, + "grad_norm": 0.9619788846896309, + "learning_rate": 0.003, + "loss": 4.03, + "step": 19333 + }, + { + "epoch": 0.19334, + "grad_norm": 0.9174398624012856, + "learning_rate": 0.003, + "loss": 3.989, + "step": 19334 + }, + { + "epoch": 0.19335, + "grad_norm": 0.8883633083754231, + "learning_rate": 0.003, + "loss": 3.998, + "step": 19335 + }, + { + "epoch": 0.19336, + "grad_norm": 0.7974706804087697, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 19336 + }, + { + "epoch": 0.19337, + "grad_norm": 0.768388768586143, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 19337 + }, + { + "epoch": 0.19338, + "grad_norm": 0.7730040662555178, + "learning_rate": 0.003, + "loss": 4.053, + "step": 19338 + }, + { + "epoch": 0.19339, + "grad_norm": 0.9063322473993962, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 19339 + }, + { + "epoch": 0.1934, + "grad_norm": 0.9654727543925578, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 19340 + }, + { + "epoch": 0.19341, + "grad_norm": 0.9930343556801654, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 19341 + }, + { + "epoch": 0.19342, + "grad_norm": 1.0137814603990158, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 19342 + }, + { + "epoch": 0.19343, + "grad_norm": 1.011112911690136, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 19343 + }, + { + "epoch": 0.19344, + "grad_norm": 0.87185465443592, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 19344 + }, + { + "epoch": 0.19345, + "grad_norm": 0.8837027753983738, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 19345 + }, + { + "epoch": 0.19346, + "grad_norm": 0.6991543647715207, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 19346 + }, + { + "epoch": 0.19347, + "grad_norm": 0.5643476814716483, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 19347 + }, + { + "epoch": 0.19348, + "grad_norm": 0.6377799913004762, + "learning_rate": 0.003, + "loss": 4.0586, + "step": 19348 + }, + { + "epoch": 0.19349, + "grad_norm": 0.6940608925794935, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 19349 + }, + { + "epoch": 0.1935, + "grad_norm": 0.6500432136200792, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 19350 + }, + { + "epoch": 0.19351, + "grad_norm": 0.6406946190725612, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 19351 + }, + { + "epoch": 0.19352, + "grad_norm": 0.5773320371891991, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 19352 + }, + { + "epoch": 0.19353, + "grad_norm": 0.5200512865095946, + "learning_rate": 0.003, + "loss": 4.005, + "step": 19353 + }, + { + "epoch": 0.19354, + "grad_norm": 0.5017026172859754, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 19354 + }, + { + "epoch": 0.19355, + "grad_norm": 0.5491529130995463, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 19355 + }, + { + "epoch": 0.19356, + "grad_norm": 0.5345195182646756, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 19356 + }, + { + "epoch": 0.19357, + "grad_norm": 0.6559244945349181, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 19357 + }, + { + "epoch": 0.19358, + "grad_norm": 0.7993029481560875, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 19358 + }, + { + "epoch": 0.19359, + "grad_norm": 0.9610740919238538, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 19359 + }, + { + "epoch": 0.1936, + "grad_norm": 0.9022751421648237, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 19360 + }, + { + "epoch": 0.19361, + "grad_norm": 0.7496969442210469, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 19361 + }, + { + "epoch": 0.19362, + "grad_norm": 0.711633271428676, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 19362 + }, + { + "epoch": 0.19363, + "grad_norm": 0.729112206141669, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 19363 + }, + { + "epoch": 0.19364, + "grad_norm": 0.7113311870684272, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 19364 + }, + { + "epoch": 0.19365, + "grad_norm": 0.7322062287006504, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 19365 + }, + { + "epoch": 0.19366, + "grad_norm": 0.7109702856399656, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 19366 + }, + { + "epoch": 0.19367, + "grad_norm": 0.7107236313123559, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 19367 + }, + { + "epoch": 0.19368, + "grad_norm": 0.6830494876923621, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 19368 + }, + { + "epoch": 0.19369, + "grad_norm": 0.6910265410707636, + "learning_rate": 0.003, + "loss": 3.993, + "step": 19369 + }, + { + "epoch": 0.1937, + "grad_norm": 0.6144471382333515, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 19370 + }, + { + "epoch": 0.19371, + "grad_norm": 0.5828134463693273, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 19371 + }, + { + "epoch": 0.19372, + "grad_norm": 0.5400263228105101, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 19372 + }, + { + "epoch": 0.19373, + "grad_norm": 0.47055768808222637, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 19373 + }, + { + "epoch": 0.19374, + "grad_norm": 0.48820204886480645, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 19374 + }, + { + "epoch": 0.19375, + "grad_norm": 0.45674607113312693, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 19375 + }, + { + "epoch": 0.19376, + "grad_norm": 0.4903032204494939, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 19376 + }, + { + "epoch": 0.19377, + "grad_norm": 0.5780612525553998, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 19377 + }, + { + "epoch": 0.19378, + "grad_norm": 0.6502562301725849, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 19378 + }, + { + "epoch": 0.19379, + "grad_norm": 0.6733386897719978, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 19379 + }, + { + "epoch": 0.1938, + "grad_norm": 0.7288755994528715, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 19380 + }, + { + "epoch": 0.19381, + "grad_norm": 0.9080265659664325, + "learning_rate": 0.003, + "loss": 3.995, + "step": 19381 + }, + { + "epoch": 0.19382, + "grad_norm": 0.96942921076844, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 19382 + }, + { + "epoch": 0.19383, + "grad_norm": 0.9821196894394981, + "learning_rate": 0.003, + "loss": 3.983, + "step": 19383 + }, + { + "epoch": 0.19384, + "grad_norm": 0.8954981065516148, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 19384 + }, + { + "epoch": 0.19385, + "grad_norm": 0.7013790911757647, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 19385 + }, + { + "epoch": 0.19386, + "grad_norm": 0.5946885739718285, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 19386 + }, + { + "epoch": 0.19387, + "grad_norm": 0.6570493378462273, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 19387 + }, + { + "epoch": 0.19388, + "grad_norm": 0.6568868214212662, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 19388 + }, + { + "epoch": 0.19389, + "grad_norm": 0.6949548884210348, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 19389 + }, + { + "epoch": 0.1939, + "grad_norm": 0.819194274699085, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 19390 + }, + { + "epoch": 0.19391, + "grad_norm": 0.997420054421183, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 19391 + }, + { + "epoch": 0.19392, + "grad_norm": 0.9080821920324523, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 19392 + }, + { + "epoch": 0.19393, + "grad_norm": 0.7603860257065277, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 19393 + }, + { + "epoch": 0.19394, + "grad_norm": 0.8251973021408404, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 19394 + }, + { + "epoch": 0.19395, + "grad_norm": 0.803978817201257, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 19395 + }, + { + "epoch": 0.19396, + "grad_norm": 0.8285094319689371, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 19396 + }, + { + "epoch": 0.19397, + "grad_norm": 0.7610106792906679, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 19397 + }, + { + "epoch": 0.19398, + "grad_norm": 0.7794722063910924, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 19398 + }, + { + "epoch": 0.19399, + "grad_norm": 0.8171576067930062, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 19399 + }, + { + "epoch": 0.194, + "grad_norm": 0.7306443507401335, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 19400 + }, + { + "epoch": 0.19401, + "grad_norm": 0.7886724355121876, + "learning_rate": 0.003, + "loss": 3.998, + "step": 19401 + }, + { + "epoch": 0.19402, + "grad_norm": 0.9015779610285097, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 19402 + }, + { + "epoch": 0.19403, + "grad_norm": 0.8461352737902409, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 19403 + }, + { + "epoch": 0.19404, + "grad_norm": 0.7068419416155095, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 19404 + }, + { + "epoch": 0.19405, + "grad_norm": 0.6687565438388842, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 19405 + }, + { + "epoch": 0.19406, + "grad_norm": 0.6184730328316479, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 19406 + }, + { + "epoch": 0.19407, + "grad_norm": 0.5113618962658322, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 19407 + }, + { + "epoch": 0.19408, + "grad_norm": 0.4749068387420251, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 19408 + }, + { + "epoch": 0.19409, + "grad_norm": 0.5598106696749793, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 19409 + }, + { + "epoch": 0.1941, + "grad_norm": 0.7270237827645364, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 19410 + }, + { + "epoch": 0.19411, + "grad_norm": 0.8708983677339341, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 19411 + }, + { + "epoch": 0.19412, + "grad_norm": 1.1568439604898655, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 19412 + }, + { + "epoch": 0.19413, + "grad_norm": 0.8802186667948575, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 19413 + }, + { + "epoch": 0.19414, + "grad_norm": 0.9598093766191728, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 19414 + }, + { + "epoch": 0.19415, + "grad_norm": 0.8752948710472198, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 19415 + }, + { + "epoch": 0.19416, + "grad_norm": 0.7811037122520624, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 19416 + }, + { + "epoch": 0.19417, + "grad_norm": 0.7384010687154396, + "learning_rate": 0.003, + "loss": 4.024, + "step": 19417 + }, + { + "epoch": 0.19418, + "grad_norm": 0.6525193339349907, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 19418 + }, + { + "epoch": 0.19419, + "grad_norm": 0.6072929798402118, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 19419 + }, + { + "epoch": 0.1942, + "grad_norm": 0.6499870847074347, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 19420 + }, + { + "epoch": 0.19421, + "grad_norm": 0.6464200653530645, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 19421 + }, + { + "epoch": 0.19422, + "grad_norm": 0.7085487540435073, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 19422 + }, + { + "epoch": 0.19423, + "grad_norm": 0.7336022201078685, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 19423 + }, + { + "epoch": 0.19424, + "grad_norm": 0.756056359848618, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 19424 + }, + { + "epoch": 0.19425, + "grad_norm": 0.8059401505918246, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 19425 + }, + { + "epoch": 0.19426, + "grad_norm": 0.7870805237114605, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 19426 + }, + { + "epoch": 0.19427, + "grad_norm": 0.7519215493275969, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 19427 + }, + { + "epoch": 0.19428, + "grad_norm": 0.7254073946076948, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 19428 + }, + { + "epoch": 0.19429, + "grad_norm": 0.7169850645887481, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 19429 + }, + { + "epoch": 0.1943, + "grad_norm": 0.8022478668791584, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 19430 + }, + { + "epoch": 0.19431, + "grad_norm": 0.8874546323433145, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 19431 + }, + { + "epoch": 0.19432, + "grad_norm": 0.9495346053017534, + "learning_rate": 0.003, + "loss": 4.016, + "step": 19432 + }, + { + "epoch": 0.19433, + "grad_norm": 0.8624257650902457, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 19433 + }, + { + "epoch": 0.19434, + "grad_norm": 0.6235100443479586, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 19434 + }, + { + "epoch": 0.19435, + "grad_norm": 0.5951534871654856, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 19435 + }, + { + "epoch": 0.19436, + "grad_norm": 0.7063611790766269, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 19436 + }, + { + "epoch": 0.19437, + "grad_norm": 0.613867867401437, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 19437 + }, + { + "epoch": 0.19438, + "grad_norm": 0.5825405614858788, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 19438 + }, + { + "epoch": 0.19439, + "grad_norm": 0.6137641155638572, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 19439 + }, + { + "epoch": 0.1944, + "grad_norm": 0.6206944744584669, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 19440 + }, + { + "epoch": 0.19441, + "grad_norm": 0.6910660947989866, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 19441 + }, + { + "epoch": 0.19442, + "grad_norm": 0.7154620888726876, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 19442 + }, + { + "epoch": 0.19443, + "grad_norm": 0.6197349155506331, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 19443 + }, + { + "epoch": 0.19444, + "grad_norm": 0.6471881439531998, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 19444 + }, + { + "epoch": 0.19445, + "grad_norm": 0.7539909337597296, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 19445 + }, + { + "epoch": 0.19446, + "grad_norm": 0.7911338677996304, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 19446 + }, + { + "epoch": 0.19447, + "grad_norm": 0.8075153900752847, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 19447 + }, + { + "epoch": 0.19448, + "grad_norm": 0.7491139570463061, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 19448 + }, + { + "epoch": 0.19449, + "grad_norm": 0.7205306193538538, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 19449 + }, + { + "epoch": 0.1945, + "grad_norm": 0.7093827561230829, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 19450 + }, + { + "epoch": 0.19451, + "grad_norm": 0.7792836404271147, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 19451 + }, + { + "epoch": 0.19452, + "grad_norm": 0.8011449416918651, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 19452 + }, + { + "epoch": 0.19453, + "grad_norm": 0.6707397685081288, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 19453 + }, + { + "epoch": 0.19454, + "grad_norm": 0.7431640345845412, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 19454 + }, + { + "epoch": 0.19455, + "grad_norm": 0.7223506941359111, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 19455 + }, + { + "epoch": 0.19456, + "grad_norm": 0.8227157768776646, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 19456 + }, + { + "epoch": 0.19457, + "grad_norm": 1.0713600010075646, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 19457 + }, + { + "epoch": 0.19458, + "grad_norm": 1.0966129418133135, + "learning_rate": 0.003, + "loss": 4.023, + "step": 19458 + }, + { + "epoch": 0.19459, + "grad_norm": 0.8306355538459066, + "learning_rate": 0.003, + "loss": 4.014, + "step": 19459 + }, + { + "epoch": 0.1946, + "grad_norm": 0.9645321755978409, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 19460 + }, + { + "epoch": 0.19461, + "grad_norm": 0.9447890486054726, + "learning_rate": 0.003, + "loss": 4.0629, + "step": 19461 + }, + { + "epoch": 0.19462, + "grad_norm": 0.8715824607052706, + "learning_rate": 0.003, + "loss": 4.023, + "step": 19462 + }, + { + "epoch": 0.19463, + "grad_norm": 0.7787762676467825, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 19463 + }, + { + "epoch": 0.19464, + "grad_norm": 0.7575341339771651, + "learning_rate": 0.003, + "loss": 3.979, + "step": 19464 + }, + { + "epoch": 0.19465, + "grad_norm": 0.7567028225707095, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 19465 + }, + { + "epoch": 0.19466, + "grad_norm": 0.7389691233714966, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 19466 + }, + { + "epoch": 0.19467, + "grad_norm": 0.712311550497252, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 19467 + }, + { + "epoch": 0.19468, + "grad_norm": 0.6782402388123921, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 19468 + }, + { + "epoch": 0.19469, + "grad_norm": 0.6043977729082749, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 19469 + }, + { + "epoch": 0.1947, + "grad_norm": 0.5297012049125829, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 19470 + }, + { + "epoch": 0.19471, + "grad_norm": 0.5712163524964026, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 19471 + }, + { + "epoch": 0.19472, + "grad_norm": 0.6015756374170848, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 19472 + }, + { + "epoch": 0.19473, + "grad_norm": 0.6879300277879287, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 19473 + }, + { + "epoch": 0.19474, + "grad_norm": 0.7903140595832849, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 19474 + }, + { + "epoch": 0.19475, + "grad_norm": 0.7764454588238762, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 19475 + }, + { + "epoch": 0.19476, + "grad_norm": 0.8497432340248974, + "learning_rate": 0.003, + "loss": 3.999, + "step": 19476 + }, + { + "epoch": 0.19477, + "grad_norm": 0.8336805074402966, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 19477 + }, + { + "epoch": 0.19478, + "grad_norm": 0.764772585112821, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 19478 + }, + { + "epoch": 0.19479, + "grad_norm": 0.8020763093589947, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 19479 + }, + { + "epoch": 0.1948, + "grad_norm": 0.8197989013549999, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 19480 + }, + { + "epoch": 0.19481, + "grad_norm": 0.788482976051111, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 19481 + }, + { + "epoch": 0.19482, + "grad_norm": 0.8261707417514862, + "learning_rate": 0.003, + "loss": 4.021, + "step": 19482 + }, + { + "epoch": 0.19483, + "grad_norm": 0.8459170681595094, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 19483 + }, + { + "epoch": 0.19484, + "grad_norm": 0.8759670445604393, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 19484 + }, + { + "epoch": 0.19485, + "grad_norm": 1.117577077726077, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 19485 + }, + { + "epoch": 0.19486, + "grad_norm": 1.1177384088203026, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 19486 + }, + { + "epoch": 0.19487, + "grad_norm": 0.8151430115142477, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 19487 + }, + { + "epoch": 0.19488, + "grad_norm": 0.7293161294219742, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 19488 + }, + { + "epoch": 0.19489, + "grad_norm": 0.7022309077945561, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 19489 + }, + { + "epoch": 0.1949, + "grad_norm": 0.7328916357166834, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 19490 + }, + { + "epoch": 0.19491, + "grad_norm": 0.7125985852389112, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 19491 + }, + { + "epoch": 0.19492, + "grad_norm": 0.5836421274088484, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 19492 + }, + { + "epoch": 0.19493, + "grad_norm": 0.557004259923469, + "learning_rate": 0.003, + "loss": 4.004, + "step": 19493 + }, + { + "epoch": 0.19494, + "grad_norm": 0.4946845196351458, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 19494 + }, + { + "epoch": 0.19495, + "grad_norm": 0.5263536940564171, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 19495 + }, + { + "epoch": 0.19496, + "grad_norm": 0.48249923550337526, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 19496 + }, + { + "epoch": 0.19497, + "grad_norm": 0.5248017586728824, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 19497 + }, + { + "epoch": 0.19498, + "grad_norm": 0.542622044686344, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 19498 + }, + { + "epoch": 0.19499, + "grad_norm": 0.5111794873407947, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 19499 + }, + { + "epoch": 0.195, + "grad_norm": 0.48107453642099407, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 19500 + }, + { + "epoch": 0.19501, + "grad_norm": 0.5178762871910413, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 19501 + }, + { + "epoch": 0.19502, + "grad_norm": 0.5542840556923857, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 19502 + }, + { + "epoch": 0.19503, + "grad_norm": 0.6546974448348233, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 19503 + }, + { + "epoch": 0.19504, + "grad_norm": 0.8637298797783544, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 19504 + }, + { + "epoch": 0.19505, + "grad_norm": 1.0662051327463302, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 19505 + }, + { + "epoch": 0.19506, + "grad_norm": 0.836580515719222, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 19506 + }, + { + "epoch": 0.19507, + "grad_norm": 0.6644030265823605, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 19507 + }, + { + "epoch": 0.19508, + "grad_norm": 0.6617254793190253, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 19508 + }, + { + "epoch": 0.19509, + "grad_norm": 0.740098948778372, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 19509 + }, + { + "epoch": 0.1951, + "grad_norm": 0.8970356415317019, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 19510 + }, + { + "epoch": 0.19511, + "grad_norm": 1.085470345118336, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 19511 + }, + { + "epoch": 0.19512, + "grad_norm": 0.9990327192987164, + "learning_rate": 0.003, + "loss": 4.027, + "step": 19512 + }, + { + "epoch": 0.19513, + "grad_norm": 0.9086614182635668, + "learning_rate": 0.003, + "loss": 4.029, + "step": 19513 + }, + { + "epoch": 0.19514, + "grad_norm": 0.8188159707049019, + "learning_rate": 0.003, + "loss": 4.0463, + "step": 19514 + }, + { + "epoch": 0.19515, + "grad_norm": 0.8536968455489518, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 19515 + }, + { + "epoch": 0.19516, + "grad_norm": 0.9834237060293421, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 19516 + }, + { + "epoch": 0.19517, + "grad_norm": 0.9718277871736687, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 19517 + }, + { + "epoch": 0.19518, + "grad_norm": 0.85923467934552, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 19518 + }, + { + "epoch": 0.19519, + "grad_norm": 0.9263113132650002, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 19519 + }, + { + "epoch": 0.1952, + "grad_norm": 1.1133637866198407, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 19520 + }, + { + "epoch": 0.19521, + "grad_norm": 0.8572805307052898, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 19521 + }, + { + "epoch": 0.19522, + "grad_norm": 0.7373161513865372, + "learning_rate": 0.003, + "loss": 4.045, + "step": 19522 + }, + { + "epoch": 0.19523, + "grad_norm": 0.8162147741295604, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 19523 + }, + { + "epoch": 0.19524, + "grad_norm": 0.7926086926998741, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 19524 + }, + { + "epoch": 0.19525, + "grad_norm": 0.7218683506057267, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 19525 + }, + { + "epoch": 0.19526, + "grad_norm": 0.6957803591036568, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 19526 + }, + { + "epoch": 0.19527, + "grad_norm": 0.6170796577297714, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 19527 + }, + { + "epoch": 0.19528, + "grad_norm": 0.5932115393833945, + "learning_rate": 0.003, + "loss": 4.011, + "step": 19528 + }, + { + "epoch": 0.19529, + "grad_norm": 0.5643377566551923, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 19529 + }, + { + "epoch": 0.1953, + "grad_norm": 0.5974879051369113, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 19530 + }, + { + "epoch": 0.19531, + "grad_norm": 0.572620680913876, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 19531 + }, + { + "epoch": 0.19532, + "grad_norm": 0.5317884680982963, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 19532 + }, + { + "epoch": 0.19533, + "grad_norm": 0.5229109363684576, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 19533 + }, + { + "epoch": 0.19534, + "grad_norm": 0.5199754912977684, + "learning_rate": 0.003, + "loss": 4.028, + "step": 19534 + }, + { + "epoch": 0.19535, + "grad_norm": 0.5688043229042344, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 19535 + }, + { + "epoch": 0.19536, + "grad_norm": 0.6740795804614993, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 19536 + }, + { + "epoch": 0.19537, + "grad_norm": 0.8501341342352702, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 19537 + }, + { + "epoch": 0.19538, + "grad_norm": 1.0198271679409052, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 19538 + }, + { + "epoch": 0.19539, + "grad_norm": 0.9532589229953153, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 19539 + }, + { + "epoch": 0.1954, + "grad_norm": 0.7704630564170838, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 19540 + }, + { + "epoch": 0.19541, + "grad_norm": 0.7447512854441931, + "learning_rate": 0.003, + "loss": 4.012, + "step": 19541 + }, + { + "epoch": 0.19542, + "grad_norm": 0.8002510650550025, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 19542 + }, + { + "epoch": 0.19543, + "grad_norm": 0.7513616366262406, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 19543 + }, + { + "epoch": 0.19544, + "grad_norm": 0.545114267066449, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 19544 + }, + { + "epoch": 0.19545, + "grad_norm": 0.6097812460441925, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 19545 + }, + { + "epoch": 0.19546, + "grad_norm": 0.6290037524835842, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 19546 + }, + { + "epoch": 0.19547, + "grad_norm": 0.6960619251284846, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 19547 + }, + { + "epoch": 0.19548, + "grad_norm": 0.7689758092901491, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 19548 + }, + { + "epoch": 0.19549, + "grad_norm": 0.7950452222861367, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 19549 + }, + { + "epoch": 0.1955, + "grad_norm": 0.6757144139036952, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 19550 + }, + { + "epoch": 0.19551, + "grad_norm": 0.665493690900344, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 19551 + }, + { + "epoch": 0.19552, + "grad_norm": 0.6317229424894896, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 19552 + }, + { + "epoch": 0.19553, + "grad_norm": 0.6334480946697623, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 19553 + }, + { + "epoch": 0.19554, + "grad_norm": 0.681968442774545, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 19554 + }, + { + "epoch": 0.19555, + "grad_norm": 0.6454465548951903, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 19555 + }, + { + "epoch": 0.19556, + "grad_norm": 0.814010451241069, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 19556 + }, + { + "epoch": 0.19557, + "grad_norm": 0.9072355598516718, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 19557 + }, + { + "epoch": 0.19558, + "grad_norm": 0.8880047546501386, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 19558 + }, + { + "epoch": 0.19559, + "grad_norm": 0.7307465878411658, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 19559 + }, + { + "epoch": 0.1956, + "grad_norm": 0.6828736463105738, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 19560 + }, + { + "epoch": 0.19561, + "grad_norm": 0.6876082809205213, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 19561 + }, + { + "epoch": 0.19562, + "grad_norm": 0.6550568957454802, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 19562 + }, + { + "epoch": 0.19563, + "grad_norm": 0.642244014772672, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 19563 + }, + { + "epoch": 0.19564, + "grad_norm": 0.6475238621458269, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 19564 + }, + { + "epoch": 0.19565, + "grad_norm": 0.8002760945505607, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 19565 + }, + { + "epoch": 0.19566, + "grad_norm": 1.0550484119185148, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 19566 + }, + { + "epoch": 0.19567, + "grad_norm": 1.1176507491189869, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 19567 + }, + { + "epoch": 0.19568, + "grad_norm": 0.8296648991054744, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 19568 + }, + { + "epoch": 0.19569, + "grad_norm": 0.7837984727456854, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 19569 + }, + { + "epoch": 0.1957, + "grad_norm": 0.7319647629198586, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 19570 + }, + { + "epoch": 0.19571, + "grad_norm": 0.7134434064218816, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 19571 + }, + { + "epoch": 0.19572, + "grad_norm": 0.7572028155565036, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 19572 + }, + { + "epoch": 0.19573, + "grad_norm": 0.7250633407169258, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 19573 + }, + { + "epoch": 0.19574, + "grad_norm": 0.7291847427256833, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 19574 + }, + { + "epoch": 0.19575, + "grad_norm": 0.7931721098734371, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 19575 + }, + { + "epoch": 0.19576, + "grad_norm": 0.8237024234538658, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 19576 + }, + { + "epoch": 0.19577, + "grad_norm": 0.8576049240665503, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 19577 + }, + { + "epoch": 0.19578, + "grad_norm": 0.9475103775189337, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 19578 + }, + { + "epoch": 0.19579, + "grad_norm": 0.9495601178033163, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 19579 + }, + { + "epoch": 0.1958, + "grad_norm": 0.9908686686978567, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 19580 + }, + { + "epoch": 0.19581, + "grad_norm": 1.1470337936634203, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 19581 + }, + { + "epoch": 0.19582, + "grad_norm": 0.9994042192286776, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 19582 + }, + { + "epoch": 0.19583, + "grad_norm": 1.1147634596786307, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 19583 + }, + { + "epoch": 0.19584, + "grad_norm": 0.9944853908876652, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 19584 + }, + { + "epoch": 0.19585, + "grad_norm": 0.8813221577577892, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 19585 + }, + { + "epoch": 0.19586, + "grad_norm": 0.7591352502821492, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 19586 + }, + { + "epoch": 0.19587, + "grad_norm": 0.8079386405777979, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 19587 + }, + { + "epoch": 0.19588, + "grad_norm": 0.7901077405324101, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 19588 + }, + { + "epoch": 0.19589, + "grad_norm": 0.8474467587837506, + "learning_rate": 0.003, + "loss": 4.03, + "step": 19589 + }, + { + "epoch": 0.1959, + "grad_norm": 0.90980747894296, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 19590 + }, + { + "epoch": 0.19591, + "grad_norm": 0.9202839006156817, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 19591 + }, + { + "epoch": 0.19592, + "grad_norm": 0.7606349958252887, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 19592 + }, + { + "epoch": 0.19593, + "grad_norm": 0.5759630915875942, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 19593 + }, + { + "epoch": 0.19594, + "grad_norm": 0.5696072362270092, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 19594 + }, + { + "epoch": 0.19595, + "grad_norm": 0.5719111159806295, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 19595 + }, + { + "epoch": 0.19596, + "grad_norm": 0.6531677803727686, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 19596 + }, + { + "epoch": 0.19597, + "grad_norm": 0.6355093259221939, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 19597 + }, + { + "epoch": 0.19598, + "grad_norm": 0.6962268954187959, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 19598 + }, + { + "epoch": 0.19599, + "grad_norm": 0.6139869859400612, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 19599 + }, + { + "epoch": 0.196, + "grad_norm": 0.5139068631670963, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 19600 + }, + { + "epoch": 0.19601, + "grad_norm": 0.4780631529483705, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 19601 + }, + { + "epoch": 0.19602, + "grad_norm": 0.5164746699560882, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 19602 + }, + { + "epoch": 0.19603, + "grad_norm": 0.5360945244568178, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 19603 + }, + { + "epoch": 0.19604, + "grad_norm": 0.5893170425926788, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 19604 + }, + { + "epoch": 0.19605, + "grad_norm": 0.5957789813527742, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 19605 + }, + { + "epoch": 0.19606, + "grad_norm": 0.5224308751347464, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 19606 + }, + { + "epoch": 0.19607, + "grad_norm": 0.5639985052453903, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 19607 + }, + { + "epoch": 0.19608, + "grad_norm": 0.6369267331296804, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 19608 + }, + { + "epoch": 0.19609, + "grad_norm": 0.7662954983599883, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 19609 + }, + { + "epoch": 0.1961, + "grad_norm": 0.8123888828614815, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 19610 + }, + { + "epoch": 0.19611, + "grad_norm": 0.8618233998151218, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 19611 + }, + { + "epoch": 0.19612, + "grad_norm": 0.9195941693810764, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 19612 + }, + { + "epoch": 0.19613, + "grad_norm": 0.9816825956896699, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 19613 + }, + { + "epoch": 0.19614, + "grad_norm": 0.9367246794569034, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 19614 + }, + { + "epoch": 0.19615, + "grad_norm": 0.8401314083901141, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 19615 + }, + { + "epoch": 0.19616, + "grad_norm": 0.739994404924749, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 19616 + }, + { + "epoch": 0.19617, + "grad_norm": 0.8169306103578905, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 19617 + }, + { + "epoch": 0.19618, + "grad_norm": 0.7591297676499528, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 19618 + }, + { + "epoch": 0.19619, + "grad_norm": 0.7318400255824185, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 19619 + }, + { + "epoch": 0.1962, + "grad_norm": 0.6830815531626494, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 19620 + }, + { + "epoch": 0.19621, + "grad_norm": 0.6328930914494715, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 19621 + }, + { + "epoch": 0.19622, + "grad_norm": 0.6915997330226559, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 19622 + }, + { + "epoch": 0.19623, + "grad_norm": 0.7654811485661136, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 19623 + }, + { + "epoch": 0.19624, + "grad_norm": 0.7443839289820983, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 19624 + }, + { + "epoch": 0.19625, + "grad_norm": 0.8130344214408592, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 19625 + }, + { + "epoch": 0.19626, + "grad_norm": 0.898096508903261, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 19626 + }, + { + "epoch": 0.19627, + "grad_norm": 0.980054420946815, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 19627 + }, + { + "epoch": 0.19628, + "grad_norm": 0.8373671654211381, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 19628 + }, + { + "epoch": 0.19629, + "grad_norm": 0.7620330406729565, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 19629 + }, + { + "epoch": 0.1963, + "grad_norm": 0.6998543608385374, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 19630 + }, + { + "epoch": 0.19631, + "grad_norm": 0.6940023623319768, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 19631 + }, + { + "epoch": 0.19632, + "grad_norm": 0.5754313165742077, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 19632 + }, + { + "epoch": 0.19633, + "grad_norm": 0.5829982661135854, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 19633 + }, + { + "epoch": 0.19634, + "grad_norm": 0.511249623517591, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 19634 + }, + { + "epoch": 0.19635, + "grad_norm": 0.557874468448235, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 19635 + }, + { + "epoch": 0.19636, + "grad_norm": 0.6478573675949048, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 19636 + }, + { + "epoch": 0.19637, + "grad_norm": 0.7787583415867051, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 19637 + }, + { + "epoch": 0.19638, + "grad_norm": 0.9334623435233643, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 19638 + }, + { + "epoch": 0.19639, + "grad_norm": 0.9874084719569178, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 19639 + }, + { + "epoch": 0.1964, + "grad_norm": 0.9267814600062754, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 19640 + }, + { + "epoch": 0.19641, + "grad_norm": 0.7398559916844136, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 19641 + }, + { + "epoch": 0.19642, + "grad_norm": 0.6750979135420361, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 19642 + }, + { + "epoch": 0.19643, + "grad_norm": 0.7926194375590001, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 19643 + }, + { + "epoch": 0.19644, + "grad_norm": 0.8191956813928805, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 19644 + }, + { + "epoch": 0.19645, + "grad_norm": 0.8047179371895393, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 19645 + }, + { + "epoch": 0.19646, + "grad_norm": 0.8458786356901118, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 19646 + }, + { + "epoch": 0.19647, + "grad_norm": 0.856139156141248, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 19647 + }, + { + "epoch": 0.19648, + "grad_norm": 0.9256637869533163, + "learning_rate": 0.003, + "loss": 4.028, + "step": 19648 + }, + { + "epoch": 0.19649, + "grad_norm": 0.9212274675836353, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 19649 + }, + { + "epoch": 0.1965, + "grad_norm": 0.8696548651803382, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 19650 + }, + { + "epoch": 0.19651, + "grad_norm": 0.8319607085871725, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 19651 + }, + { + "epoch": 0.19652, + "grad_norm": 0.805980429348326, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 19652 + }, + { + "epoch": 0.19653, + "grad_norm": 0.7621264472428813, + "learning_rate": 0.003, + "loss": 3.994, + "step": 19653 + }, + { + "epoch": 0.19654, + "grad_norm": 0.6808521780170396, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 19654 + }, + { + "epoch": 0.19655, + "grad_norm": 0.7050879601478148, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 19655 + }, + { + "epoch": 0.19656, + "grad_norm": 0.7576038043923169, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 19656 + }, + { + "epoch": 0.19657, + "grad_norm": 0.6746994443631628, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 19657 + }, + { + "epoch": 0.19658, + "grad_norm": 0.679885948085455, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 19658 + }, + { + "epoch": 0.19659, + "grad_norm": 0.7159230181035187, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 19659 + }, + { + "epoch": 0.1966, + "grad_norm": 0.6711563341989192, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 19660 + }, + { + "epoch": 0.19661, + "grad_norm": 0.6092167725041495, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 19661 + }, + { + "epoch": 0.19662, + "grad_norm": 0.5026129142459835, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 19662 + }, + { + "epoch": 0.19663, + "grad_norm": 0.5274523699067979, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 19663 + }, + { + "epoch": 0.19664, + "grad_norm": 0.5236838083426235, + "learning_rate": 0.003, + "loss": 3.993, + "step": 19664 + }, + { + "epoch": 0.19665, + "grad_norm": 0.48040986120076484, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 19665 + }, + { + "epoch": 0.19666, + "grad_norm": 0.4907839374700705, + "learning_rate": 0.003, + "loss": 3.999, + "step": 19666 + }, + { + "epoch": 0.19667, + "grad_norm": 0.49537076396380486, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 19667 + }, + { + "epoch": 0.19668, + "grad_norm": 0.5838151050712009, + "learning_rate": 0.003, + "loss": 3.984, + "step": 19668 + }, + { + "epoch": 0.19669, + "grad_norm": 0.6580070278472224, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 19669 + }, + { + "epoch": 0.1967, + "grad_norm": 0.6242081101098885, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 19670 + }, + { + "epoch": 0.19671, + "grad_norm": 0.6156666540174427, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 19671 + }, + { + "epoch": 0.19672, + "grad_norm": 0.7115147432762647, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 19672 + }, + { + "epoch": 0.19673, + "grad_norm": 0.7586846309686108, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 19673 + }, + { + "epoch": 0.19674, + "grad_norm": 0.8019613447290489, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 19674 + }, + { + "epoch": 0.19675, + "grad_norm": 0.8359303959120182, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 19675 + }, + { + "epoch": 0.19676, + "grad_norm": 0.9080395916490704, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 19676 + }, + { + "epoch": 0.19677, + "grad_norm": 0.8828723181681345, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 19677 + }, + { + "epoch": 0.19678, + "grad_norm": 0.7480882555116566, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 19678 + }, + { + "epoch": 0.19679, + "grad_norm": 0.6760329520219313, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 19679 + }, + { + "epoch": 0.1968, + "grad_norm": 0.7462878545309836, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 19680 + }, + { + "epoch": 0.19681, + "grad_norm": 0.6699157862972626, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 19681 + }, + { + "epoch": 0.19682, + "grad_norm": 0.6720786800076232, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 19682 + }, + { + "epoch": 0.19683, + "grad_norm": 0.6636766987329581, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 19683 + }, + { + "epoch": 0.19684, + "grad_norm": 0.668770968847305, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 19684 + }, + { + "epoch": 0.19685, + "grad_norm": 0.8385152304383162, + "learning_rate": 0.003, + "loss": 3.962, + "step": 19685 + }, + { + "epoch": 0.19686, + "grad_norm": 0.9714316853108154, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 19686 + }, + { + "epoch": 0.19687, + "grad_norm": 0.9409040987772078, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 19687 + }, + { + "epoch": 0.19688, + "grad_norm": 0.8134943111025021, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 19688 + }, + { + "epoch": 0.19689, + "grad_norm": 0.7598891224384864, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 19689 + }, + { + "epoch": 0.1969, + "grad_norm": 0.7471827053636374, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 19690 + }, + { + "epoch": 0.19691, + "grad_norm": 0.8458477049377191, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 19691 + }, + { + "epoch": 0.19692, + "grad_norm": 1.0208509638952972, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 19692 + }, + { + "epoch": 0.19693, + "grad_norm": 1.0031346322013814, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 19693 + }, + { + "epoch": 0.19694, + "grad_norm": 0.9298939260243351, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 19694 + }, + { + "epoch": 0.19695, + "grad_norm": 0.9682882581726052, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 19695 + }, + { + "epoch": 0.19696, + "grad_norm": 1.0821825044211233, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 19696 + }, + { + "epoch": 0.19697, + "grad_norm": 1.0824636026269714, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 19697 + }, + { + "epoch": 0.19698, + "grad_norm": 0.8452256457153889, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 19698 + }, + { + "epoch": 0.19699, + "grad_norm": 0.677002330439613, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 19699 + }, + { + "epoch": 0.197, + "grad_norm": 0.7287167051343875, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 19700 + }, + { + "epoch": 0.19701, + "grad_norm": 0.6828067267802234, + "learning_rate": 0.003, + "loss": 4.0396, + "step": 19701 + }, + { + "epoch": 0.19702, + "grad_norm": 0.7222947638635108, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 19702 + }, + { + "epoch": 0.19703, + "grad_norm": 0.7350233969448137, + "learning_rate": 0.003, + "loss": 4.0388, + "step": 19703 + }, + { + "epoch": 0.19704, + "grad_norm": 0.710666266603408, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 19704 + }, + { + "epoch": 0.19705, + "grad_norm": 0.7261980477118221, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 19705 + }, + { + "epoch": 0.19706, + "grad_norm": 0.7428361487571875, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 19706 + }, + { + "epoch": 0.19707, + "grad_norm": 0.7343029291897075, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 19707 + }, + { + "epoch": 0.19708, + "grad_norm": 0.64951504028847, + "learning_rate": 0.003, + "loss": 4.017, + "step": 19708 + }, + { + "epoch": 0.19709, + "grad_norm": 0.6596948721994004, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 19709 + }, + { + "epoch": 0.1971, + "grad_norm": 0.5905760292681833, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 19710 + }, + { + "epoch": 0.19711, + "grad_norm": 0.5086583591682448, + "learning_rate": 0.003, + "loss": 4.015, + "step": 19711 + }, + { + "epoch": 0.19712, + "grad_norm": 0.5267651309044322, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 19712 + }, + { + "epoch": 0.19713, + "grad_norm": 0.5460880245026265, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 19713 + }, + { + "epoch": 0.19714, + "grad_norm": 0.597962672865829, + "learning_rate": 0.003, + "loss": 3.968, + "step": 19714 + }, + { + "epoch": 0.19715, + "grad_norm": 0.6369255756552854, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 19715 + }, + { + "epoch": 0.19716, + "grad_norm": 0.6747051004532915, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 19716 + }, + { + "epoch": 0.19717, + "grad_norm": 0.6662123729288968, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 19717 + }, + { + "epoch": 0.19718, + "grad_norm": 0.7750885335625288, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 19718 + }, + { + "epoch": 0.19719, + "grad_norm": 0.7309460913456585, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 19719 + }, + { + "epoch": 0.1972, + "grad_norm": 0.7091965464099179, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 19720 + }, + { + "epoch": 0.19721, + "grad_norm": 0.8163497486175038, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 19721 + }, + { + "epoch": 0.19722, + "grad_norm": 1.0570637534612466, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 19722 + }, + { + "epoch": 0.19723, + "grad_norm": 1.1690705258844447, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 19723 + }, + { + "epoch": 0.19724, + "grad_norm": 0.6869987325677623, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 19724 + }, + { + "epoch": 0.19725, + "grad_norm": 0.5438796012046976, + "learning_rate": 0.003, + "loss": 3.989, + "step": 19725 + }, + { + "epoch": 0.19726, + "grad_norm": 0.682657643061729, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 19726 + }, + { + "epoch": 0.19727, + "grad_norm": 0.7220320413540162, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 19727 + }, + { + "epoch": 0.19728, + "grad_norm": 0.7188620771167151, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 19728 + }, + { + "epoch": 0.19729, + "grad_norm": 0.6787787901288609, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 19729 + }, + { + "epoch": 0.1973, + "grad_norm": 0.6520952820473302, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 19730 + }, + { + "epoch": 0.19731, + "grad_norm": 0.6062322282248126, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 19731 + }, + { + "epoch": 0.19732, + "grad_norm": 0.5188834180246629, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 19732 + }, + { + "epoch": 0.19733, + "grad_norm": 0.57933166380261, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 19733 + }, + { + "epoch": 0.19734, + "grad_norm": 0.5740257150900832, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 19734 + }, + { + "epoch": 0.19735, + "grad_norm": 0.6216932968323405, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 19735 + }, + { + "epoch": 0.19736, + "grad_norm": 0.6775979743817007, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 19736 + }, + { + "epoch": 0.19737, + "grad_norm": 0.8415184379894424, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 19737 + }, + { + "epoch": 0.19738, + "grad_norm": 1.0131153103077666, + "learning_rate": 0.003, + "loss": 3.996, + "step": 19738 + }, + { + "epoch": 0.19739, + "grad_norm": 0.973981788840608, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 19739 + }, + { + "epoch": 0.1974, + "grad_norm": 0.7111476979302211, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 19740 + }, + { + "epoch": 0.19741, + "grad_norm": 0.7573855023568276, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 19741 + }, + { + "epoch": 0.19742, + "grad_norm": 0.8995237971419915, + "learning_rate": 0.003, + "loss": 4.001, + "step": 19742 + }, + { + "epoch": 0.19743, + "grad_norm": 0.9393003582991586, + "learning_rate": 0.003, + "loss": 4.005, + "step": 19743 + }, + { + "epoch": 0.19744, + "grad_norm": 0.8410413433685606, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 19744 + }, + { + "epoch": 0.19745, + "grad_norm": 0.7254934289470732, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 19745 + }, + { + "epoch": 0.19746, + "grad_norm": 0.8703829649424863, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 19746 + }, + { + "epoch": 0.19747, + "grad_norm": 0.9604791175359089, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 19747 + }, + { + "epoch": 0.19748, + "grad_norm": 0.9808459395323208, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 19748 + }, + { + "epoch": 0.19749, + "grad_norm": 0.9152690849647185, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 19749 + }, + { + "epoch": 0.1975, + "grad_norm": 0.7219007527833088, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 19750 + }, + { + "epoch": 0.19751, + "grad_norm": 0.5893545384690463, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 19751 + }, + { + "epoch": 0.19752, + "grad_norm": 0.5814401232375386, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 19752 + }, + { + "epoch": 0.19753, + "grad_norm": 0.5811802596212895, + "learning_rate": 0.003, + "loss": 3.995, + "step": 19753 + }, + { + "epoch": 0.19754, + "grad_norm": 0.6455862170694832, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 19754 + }, + { + "epoch": 0.19755, + "grad_norm": 0.6524911830853564, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 19755 + }, + { + "epoch": 0.19756, + "grad_norm": 0.6475578342912681, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 19756 + }, + { + "epoch": 0.19757, + "grad_norm": 0.7475198188986673, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 19757 + }, + { + "epoch": 0.19758, + "grad_norm": 0.7294192880948349, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 19758 + }, + { + "epoch": 0.19759, + "grad_norm": 0.6632747722909648, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 19759 + }, + { + "epoch": 0.1976, + "grad_norm": 0.692494831752627, + "learning_rate": 0.003, + "loss": 4.0492, + "step": 19760 + }, + { + "epoch": 0.19761, + "grad_norm": 0.7284045999247868, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 19761 + }, + { + "epoch": 0.19762, + "grad_norm": 0.8198741209607893, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 19762 + }, + { + "epoch": 0.19763, + "grad_norm": 0.769869886881409, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 19763 + }, + { + "epoch": 0.19764, + "grad_norm": 0.7691946281863159, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 19764 + }, + { + "epoch": 0.19765, + "grad_norm": 0.7825647074022373, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 19765 + }, + { + "epoch": 0.19766, + "grad_norm": 0.9117423367780271, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 19766 + }, + { + "epoch": 0.19767, + "grad_norm": 0.8331537190785516, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 19767 + }, + { + "epoch": 0.19768, + "grad_norm": 0.7649406504955011, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 19768 + }, + { + "epoch": 0.19769, + "grad_norm": 0.7479501258985118, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 19769 + }, + { + "epoch": 0.1977, + "grad_norm": 0.8272718240000017, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 19770 + }, + { + "epoch": 0.19771, + "grad_norm": 0.8786083654798017, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 19771 + }, + { + "epoch": 0.19772, + "grad_norm": 0.9587818820274385, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 19772 + }, + { + "epoch": 0.19773, + "grad_norm": 0.9235418806973662, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 19773 + }, + { + "epoch": 0.19774, + "grad_norm": 0.7863212803470927, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 19774 + }, + { + "epoch": 0.19775, + "grad_norm": 0.8321572638679195, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 19775 + }, + { + "epoch": 0.19776, + "grad_norm": 0.9424505691494677, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 19776 + }, + { + "epoch": 0.19777, + "grad_norm": 1.010053394975937, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 19777 + }, + { + "epoch": 0.19778, + "grad_norm": 0.9108095776572783, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 19778 + }, + { + "epoch": 0.19779, + "grad_norm": 0.8427261334735189, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 19779 + }, + { + "epoch": 0.1978, + "grad_norm": 0.8421745486181434, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 19780 + }, + { + "epoch": 0.19781, + "grad_norm": 0.6681607961186082, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 19781 + }, + { + "epoch": 0.19782, + "grad_norm": 0.6786642856831999, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 19782 + }, + { + "epoch": 0.19783, + "grad_norm": 0.733868328730909, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 19783 + }, + { + "epoch": 0.19784, + "grad_norm": 0.7503369607738795, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 19784 + }, + { + "epoch": 0.19785, + "grad_norm": 0.7290840191387361, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 19785 + }, + { + "epoch": 0.19786, + "grad_norm": 0.5932710915739767, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 19786 + }, + { + "epoch": 0.19787, + "grad_norm": 0.6157472426315096, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 19787 + }, + { + "epoch": 0.19788, + "grad_norm": 0.6964617830849921, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 19788 + }, + { + "epoch": 0.19789, + "grad_norm": 0.7332021409201672, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 19789 + }, + { + "epoch": 0.1979, + "grad_norm": 0.6932019734432724, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 19790 + }, + { + "epoch": 0.19791, + "grad_norm": 0.684864427560063, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 19791 + }, + { + "epoch": 0.19792, + "grad_norm": 0.659881147240048, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 19792 + }, + { + "epoch": 0.19793, + "grad_norm": 0.7315581440254829, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 19793 + }, + { + "epoch": 0.19794, + "grad_norm": 0.6922048616521578, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 19794 + }, + { + "epoch": 0.19795, + "grad_norm": 0.6368260140614496, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 19795 + }, + { + "epoch": 0.19796, + "grad_norm": 0.6803691213012633, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 19796 + }, + { + "epoch": 0.19797, + "grad_norm": 0.8170641554686527, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 19797 + }, + { + "epoch": 0.19798, + "grad_norm": 0.8977105845753985, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 19798 + }, + { + "epoch": 0.19799, + "grad_norm": 0.9316641711256429, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 19799 + }, + { + "epoch": 0.198, + "grad_norm": 0.8235269907284808, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 19800 + }, + { + "epoch": 0.19801, + "grad_norm": 0.7972790989487337, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 19801 + }, + { + "epoch": 0.19802, + "grad_norm": 0.8491913475120846, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 19802 + }, + { + "epoch": 0.19803, + "grad_norm": 0.8423554031741456, + "learning_rate": 0.003, + "loss": 4.009, + "step": 19803 + }, + { + "epoch": 0.19804, + "grad_norm": 0.8050282642895005, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 19804 + }, + { + "epoch": 0.19805, + "grad_norm": 0.6852654461268213, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 19805 + }, + { + "epoch": 0.19806, + "grad_norm": 0.6841481930251384, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 19806 + }, + { + "epoch": 0.19807, + "grad_norm": 0.6161530969020022, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 19807 + }, + { + "epoch": 0.19808, + "grad_norm": 0.6184644582638749, + "learning_rate": 0.003, + "loss": 4.0502, + "step": 19808 + }, + { + "epoch": 0.19809, + "grad_norm": 0.572158952244445, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 19809 + }, + { + "epoch": 0.1981, + "grad_norm": 0.5529319438981039, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 19810 + }, + { + "epoch": 0.19811, + "grad_norm": 0.48489854910738245, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 19811 + }, + { + "epoch": 0.19812, + "grad_norm": 0.47624449177972994, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 19812 + }, + { + "epoch": 0.19813, + "grad_norm": 0.48787125401522385, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 19813 + }, + { + "epoch": 0.19814, + "grad_norm": 0.4873245695794914, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 19814 + }, + { + "epoch": 0.19815, + "grad_norm": 0.49155294268322997, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 19815 + }, + { + "epoch": 0.19816, + "grad_norm": 0.5613388897392712, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 19816 + }, + { + "epoch": 0.19817, + "grad_norm": 0.6000486280566922, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 19817 + }, + { + "epoch": 0.19818, + "grad_norm": 0.7043348883858006, + "learning_rate": 0.003, + "loss": 3.977, + "step": 19818 + }, + { + "epoch": 0.19819, + "grad_norm": 0.8702518859724835, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 19819 + }, + { + "epoch": 0.1982, + "grad_norm": 1.1632325388959237, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 19820 + }, + { + "epoch": 0.19821, + "grad_norm": 0.7490476171273479, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 19821 + }, + { + "epoch": 0.19822, + "grad_norm": 0.5896345620138947, + "learning_rate": 0.003, + "loss": 3.978, + "step": 19822 + }, + { + "epoch": 0.19823, + "grad_norm": 0.5740877866840668, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 19823 + }, + { + "epoch": 0.19824, + "grad_norm": 0.7282683001711344, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 19824 + }, + { + "epoch": 0.19825, + "grad_norm": 0.8311376150978471, + "learning_rate": 0.003, + "loss": 3.99, + "step": 19825 + }, + { + "epoch": 0.19826, + "grad_norm": 0.9059356572482312, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 19826 + }, + { + "epoch": 0.19827, + "grad_norm": 1.0192191946462021, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 19827 + }, + { + "epoch": 0.19828, + "grad_norm": 1.149746657351863, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 19828 + }, + { + "epoch": 0.19829, + "grad_norm": 0.9894069139165625, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 19829 + }, + { + "epoch": 0.1983, + "grad_norm": 0.9971648732131199, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 19830 + }, + { + "epoch": 0.19831, + "grad_norm": 0.8544620048687979, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 19831 + }, + { + "epoch": 0.19832, + "grad_norm": 0.926678811739357, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 19832 + }, + { + "epoch": 0.19833, + "grad_norm": 1.0580519272199826, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 19833 + }, + { + "epoch": 0.19834, + "grad_norm": 0.9617440068901573, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 19834 + }, + { + "epoch": 0.19835, + "grad_norm": 1.0246268547414932, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 19835 + }, + { + "epoch": 0.19836, + "grad_norm": 0.9082697082725325, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 19836 + }, + { + "epoch": 0.19837, + "grad_norm": 0.8138065833612894, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 19837 + }, + { + "epoch": 0.19838, + "grad_norm": 0.713007301800989, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 19838 + }, + { + "epoch": 0.19839, + "grad_norm": 0.6545860585700856, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 19839 + }, + { + "epoch": 0.1984, + "grad_norm": 0.6246025645122217, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 19840 + }, + { + "epoch": 0.19841, + "grad_norm": 0.6664329009736881, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 19841 + }, + { + "epoch": 0.19842, + "grad_norm": 0.7543561539361447, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 19842 + }, + { + "epoch": 0.19843, + "grad_norm": 0.7447818851003185, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 19843 + }, + { + "epoch": 0.19844, + "grad_norm": 0.7092760576967008, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 19844 + }, + { + "epoch": 0.19845, + "grad_norm": 0.8080240898604911, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 19845 + }, + { + "epoch": 0.19846, + "grad_norm": 0.9023796852695622, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 19846 + }, + { + "epoch": 0.19847, + "grad_norm": 1.0113692858499315, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 19847 + }, + { + "epoch": 0.19848, + "grad_norm": 0.8805554503895532, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 19848 + }, + { + "epoch": 0.19849, + "grad_norm": 0.8111007376794092, + "learning_rate": 0.003, + "loss": 4.013, + "step": 19849 + }, + { + "epoch": 0.1985, + "grad_norm": 0.761376553000959, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 19850 + }, + { + "epoch": 0.19851, + "grad_norm": 0.6388753855118183, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 19851 + }, + { + "epoch": 0.19852, + "grad_norm": 0.5815784636185805, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 19852 + }, + { + "epoch": 0.19853, + "grad_norm": 0.5096560450299075, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 19853 + }, + { + "epoch": 0.19854, + "grad_norm": 0.5446317116068978, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 19854 + }, + { + "epoch": 0.19855, + "grad_norm": 0.5854939892619184, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 19855 + }, + { + "epoch": 0.19856, + "grad_norm": 0.595674122605613, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 19856 + }, + { + "epoch": 0.19857, + "grad_norm": 0.5925639999410025, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 19857 + }, + { + "epoch": 0.19858, + "grad_norm": 0.6638216978434285, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 19858 + }, + { + "epoch": 0.19859, + "grad_norm": 0.7702793918899653, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 19859 + }, + { + "epoch": 0.1986, + "grad_norm": 0.9019085063149986, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 19860 + }, + { + "epoch": 0.19861, + "grad_norm": 1.0119675769644763, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 19861 + }, + { + "epoch": 0.19862, + "grad_norm": 0.9302779510338499, + "learning_rate": 0.003, + "loss": 4.003, + "step": 19862 + }, + { + "epoch": 0.19863, + "grad_norm": 0.63907652663846, + "learning_rate": 0.003, + "loss": 4.011, + "step": 19863 + }, + { + "epoch": 0.19864, + "grad_norm": 0.6378627680943163, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 19864 + }, + { + "epoch": 0.19865, + "grad_norm": 0.6757647047393013, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 19865 + }, + { + "epoch": 0.19866, + "grad_norm": 0.758325570450026, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 19866 + }, + { + "epoch": 0.19867, + "grad_norm": 0.7309990091656502, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 19867 + }, + { + "epoch": 0.19868, + "grad_norm": 0.6474761101656349, + "learning_rate": 0.003, + "loss": 4.014, + "step": 19868 + }, + { + "epoch": 0.19869, + "grad_norm": 0.5383984749325873, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 19869 + }, + { + "epoch": 0.1987, + "grad_norm": 0.5297368276117025, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 19870 + }, + { + "epoch": 0.19871, + "grad_norm": 0.5251499083902544, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 19871 + }, + { + "epoch": 0.19872, + "grad_norm": 0.5011631131998576, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 19872 + }, + { + "epoch": 0.19873, + "grad_norm": 0.48694286842259726, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 19873 + }, + { + "epoch": 0.19874, + "grad_norm": 0.5493281367968355, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 19874 + }, + { + "epoch": 0.19875, + "grad_norm": 0.5740054674757689, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 19875 + }, + { + "epoch": 0.19876, + "grad_norm": 0.5717010879598212, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 19876 + }, + { + "epoch": 0.19877, + "grad_norm": 0.683502707297754, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 19877 + }, + { + "epoch": 0.19878, + "grad_norm": 0.7940561582540057, + "learning_rate": 0.003, + "loss": 4.003, + "step": 19878 + }, + { + "epoch": 0.19879, + "grad_norm": 0.8301547967862078, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 19879 + }, + { + "epoch": 0.1988, + "grad_norm": 0.7814422157989738, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 19880 + }, + { + "epoch": 0.19881, + "grad_norm": 0.7855708071906573, + "learning_rate": 0.003, + "loss": 3.997, + "step": 19881 + }, + { + "epoch": 0.19882, + "grad_norm": 0.7751173798430153, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 19882 + }, + { + "epoch": 0.19883, + "grad_norm": 0.8190678946912858, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 19883 + }, + { + "epoch": 0.19884, + "grad_norm": 0.7686684490927204, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 19884 + }, + { + "epoch": 0.19885, + "grad_norm": 0.7517814933776005, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 19885 + }, + { + "epoch": 0.19886, + "grad_norm": 0.7136204120252189, + "learning_rate": 0.003, + "loss": 3.984, + "step": 19886 + }, + { + "epoch": 0.19887, + "grad_norm": 0.7148214073671192, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 19887 + }, + { + "epoch": 0.19888, + "grad_norm": 0.7000467528548802, + "learning_rate": 0.003, + "loss": 4.001, + "step": 19888 + }, + { + "epoch": 0.19889, + "grad_norm": 0.8788656541816834, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 19889 + }, + { + "epoch": 0.1989, + "grad_norm": 1.0633470701587533, + "learning_rate": 0.003, + "loss": 4.027, + "step": 19890 + }, + { + "epoch": 0.19891, + "grad_norm": 1.108201223273882, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 19891 + }, + { + "epoch": 0.19892, + "grad_norm": 0.8053373560233912, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 19892 + }, + { + "epoch": 0.19893, + "grad_norm": 0.6563063070875835, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 19893 + }, + { + "epoch": 0.19894, + "grad_norm": 0.5827339180914727, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 19894 + }, + { + "epoch": 0.19895, + "grad_norm": 0.6393990551527465, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 19895 + }, + { + "epoch": 0.19896, + "grad_norm": 0.766585402608744, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 19896 + }, + { + "epoch": 0.19897, + "grad_norm": 0.6878321525334704, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 19897 + }, + { + "epoch": 0.19898, + "grad_norm": 0.6228209000862689, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 19898 + }, + { + "epoch": 0.19899, + "grad_norm": 0.6969345026091316, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 19899 + }, + { + "epoch": 0.199, + "grad_norm": 0.7583625337042028, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 19900 + }, + { + "epoch": 0.19901, + "grad_norm": 0.7673132034118683, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 19901 + }, + { + "epoch": 0.19902, + "grad_norm": 0.8145627209951705, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 19902 + }, + { + "epoch": 0.19903, + "grad_norm": 0.8592291073133245, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 19903 + }, + { + "epoch": 0.19904, + "grad_norm": 0.8361776656605461, + "learning_rate": 0.003, + "loss": 4.009, + "step": 19904 + }, + { + "epoch": 0.19905, + "grad_norm": 0.8544774537329541, + "learning_rate": 0.003, + "loss": 4.014, + "step": 19905 + }, + { + "epoch": 0.19906, + "grad_norm": 0.7914839817866952, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 19906 + }, + { + "epoch": 0.19907, + "grad_norm": 0.7858842693210603, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 19907 + }, + { + "epoch": 0.19908, + "grad_norm": 0.8037086906253803, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 19908 + }, + { + "epoch": 0.19909, + "grad_norm": 0.903505134237489, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 19909 + }, + { + "epoch": 0.1991, + "grad_norm": 0.8687745457902438, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 19910 + }, + { + "epoch": 0.19911, + "grad_norm": 0.8861239120348208, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 19911 + }, + { + "epoch": 0.19912, + "grad_norm": 0.9274569648521516, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 19912 + }, + { + "epoch": 0.19913, + "grad_norm": 0.8404570271380242, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 19913 + }, + { + "epoch": 0.19914, + "grad_norm": 0.9519862101108483, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 19914 + }, + { + "epoch": 0.19915, + "grad_norm": 0.904213459979311, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 19915 + }, + { + "epoch": 0.19916, + "grad_norm": 0.8254954542702581, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 19916 + }, + { + "epoch": 0.19917, + "grad_norm": 0.771162316416315, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 19917 + }, + { + "epoch": 0.19918, + "grad_norm": 0.8254404281716408, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 19918 + }, + { + "epoch": 0.19919, + "grad_norm": 0.7996562287055933, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 19919 + }, + { + "epoch": 0.1992, + "grad_norm": 0.6482030714475148, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 19920 + }, + { + "epoch": 0.19921, + "grad_norm": 0.6498691652364097, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 19921 + }, + { + "epoch": 0.19922, + "grad_norm": 0.6851581468151091, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 19922 + }, + { + "epoch": 0.19923, + "grad_norm": 0.7362043607867199, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 19923 + }, + { + "epoch": 0.19924, + "grad_norm": 0.8553305497575365, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 19924 + }, + { + "epoch": 0.19925, + "grad_norm": 0.8454884067018811, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 19925 + }, + { + "epoch": 0.19926, + "grad_norm": 0.7305046011886166, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 19926 + }, + { + "epoch": 0.19927, + "grad_norm": 0.6873782275143836, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 19927 + }, + { + "epoch": 0.19928, + "grad_norm": 0.8077085942110626, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 19928 + }, + { + "epoch": 0.19929, + "grad_norm": 0.8464017670284668, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 19929 + }, + { + "epoch": 0.1993, + "grad_norm": 0.7925541967837743, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 19930 + }, + { + "epoch": 0.19931, + "grad_norm": 0.7144483445110732, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 19931 + }, + { + "epoch": 0.19932, + "grad_norm": 0.6601158019737527, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 19932 + }, + { + "epoch": 0.19933, + "grad_norm": 0.5902980360026695, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 19933 + }, + { + "epoch": 0.19934, + "grad_norm": 0.6384218436272451, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 19934 + }, + { + "epoch": 0.19935, + "grad_norm": 0.732556411716997, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 19935 + }, + { + "epoch": 0.19936, + "grad_norm": 0.7784948453451356, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 19936 + }, + { + "epoch": 0.19937, + "grad_norm": 0.8691767927002182, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 19937 + }, + { + "epoch": 0.19938, + "grad_norm": 0.9724713254787617, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 19938 + }, + { + "epoch": 0.19939, + "grad_norm": 0.9372366284293433, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 19939 + }, + { + "epoch": 0.1994, + "grad_norm": 0.6938975916280076, + "learning_rate": 0.003, + "loss": 3.979, + "step": 19940 + }, + { + "epoch": 0.19941, + "grad_norm": 0.6142369837650004, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 19941 + }, + { + "epoch": 0.19942, + "grad_norm": 0.6708109367320131, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 19942 + }, + { + "epoch": 0.19943, + "grad_norm": 0.6966632458204243, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 19943 + }, + { + "epoch": 0.19944, + "grad_norm": 0.6916021228474729, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 19944 + }, + { + "epoch": 0.19945, + "grad_norm": 0.7394714437465869, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 19945 + }, + { + "epoch": 0.19946, + "grad_norm": 0.7128913153597678, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 19946 + }, + { + "epoch": 0.19947, + "grad_norm": 0.7613083330083891, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 19947 + }, + { + "epoch": 0.19948, + "grad_norm": 0.8216502057830504, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 19948 + }, + { + "epoch": 0.19949, + "grad_norm": 0.7655665069194446, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 19949 + }, + { + "epoch": 0.1995, + "grad_norm": 0.7494011170745235, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 19950 + }, + { + "epoch": 0.19951, + "grad_norm": 0.7601547489203127, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 19951 + }, + { + "epoch": 0.19952, + "grad_norm": 0.843273219731475, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 19952 + }, + { + "epoch": 0.19953, + "grad_norm": 0.7623360458912716, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 19953 + }, + { + "epoch": 0.19954, + "grad_norm": 0.8316577246439916, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 19954 + }, + { + "epoch": 0.19955, + "grad_norm": 0.869699258802325, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 19955 + }, + { + "epoch": 0.19956, + "grad_norm": 0.9408412668126317, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 19956 + }, + { + "epoch": 0.19957, + "grad_norm": 0.9753153029253502, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 19957 + }, + { + "epoch": 0.19958, + "grad_norm": 0.8887977235405068, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 19958 + }, + { + "epoch": 0.19959, + "grad_norm": 0.8062355254500571, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 19959 + }, + { + "epoch": 0.1996, + "grad_norm": 0.7233455415624709, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 19960 + }, + { + "epoch": 0.19961, + "grad_norm": 0.6985242147908187, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 19961 + }, + { + "epoch": 0.19962, + "grad_norm": 0.6020989542569066, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 19962 + }, + { + "epoch": 0.19963, + "grad_norm": 0.6714296877458833, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 19963 + }, + { + "epoch": 0.19964, + "grad_norm": 0.7377786972259592, + "learning_rate": 0.003, + "loss": 4.002, + "step": 19964 + }, + { + "epoch": 0.19965, + "grad_norm": 0.9706605720199071, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 19965 + }, + { + "epoch": 0.19966, + "grad_norm": 1.0338751751160564, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 19966 + }, + { + "epoch": 0.19967, + "grad_norm": 0.8884688516061275, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 19967 + }, + { + "epoch": 0.19968, + "grad_norm": 0.9687401220205296, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 19968 + }, + { + "epoch": 0.19969, + "grad_norm": 0.8719697073267451, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 19969 + }, + { + "epoch": 0.1997, + "grad_norm": 0.8047380792139012, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 19970 + }, + { + "epoch": 0.19971, + "grad_norm": 0.7699167047736817, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 19971 + }, + { + "epoch": 0.19972, + "grad_norm": 0.8941167433376193, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 19972 + }, + { + "epoch": 0.19973, + "grad_norm": 0.7872129084728906, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 19973 + }, + { + "epoch": 0.19974, + "grad_norm": 0.7521433562414533, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 19974 + }, + { + "epoch": 0.19975, + "grad_norm": 0.646596715514071, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 19975 + }, + { + "epoch": 0.19976, + "grad_norm": 0.6691092394142149, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 19976 + }, + { + "epoch": 0.19977, + "grad_norm": 0.5904185586368084, + "learning_rate": 0.003, + "loss": 3.968, + "step": 19977 + }, + { + "epoch": 0.19978, + "grad_norm": 0.6149430730618042, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 19978 + }, + { + "epoch": 0.19979, + "grad_norm": 0.6530732890731843, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 19979 + }, + { + "epoch": 0.1998, + "grad_norm": 0.6439677318459098, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 19980 + }, + { + "epoch": 0.19981, + "grad_norm": 0.6585599899834014, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 19981 + }, + { + "epoch": 0.19982, + "grad_norm": 0.7630972176352662, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 19982 + }, + { + "epoch": 0.19983, + "grad_norm": 0.6327143533014286, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 19983 + }, + { + "epoch": 0.19984, + "grad_norm": 0.5297058947278986, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 19984 + }, + { + "epoch": 0.19985, + "grad_norm": 0.5812132737441251, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 19985 + }, + { + "epoch": 0.19986, + "grad_norm": 0.6109912514005639, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 19986 + }, + { + "epoch": 0.19987, + "grad_norm": 0.6476558293884387, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 19987 + }, + { + "epoch": 0.19988, + "grad_norm": 0.6824356240421943, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 19988 + }, + { + "epoch": 0.19989, + "grad_norm": 0.6883539459172531, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 19989 + }, + { + "epoch": 0.1999, + "grad_norm": 0.6571255302905733, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 19990 + }, + { + "epoch": 0.19991, + "grad_norm": 0.47837143065561377, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 19991 + }, + { + "epoch": 0.19992, + "grad_norm": 0.413802507735596, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 19992 + }, + { + "epoch": 0.19993, + "grad_norm": 0.42381801842362654, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 19993 + }, + { + "epoch": 0.19994, + "grad_norm": 0.5327548486369519, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 19994 + }, + { + "epoch": 0.19995, + "grad_norm": 0.6576506583518112, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 19995 + }, + { + "epoch": 0.19996, + "grad_norm": 0.7341199833369395, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 19996 + }, + { + "epoch": 0.19997, + "grad_norm": 0.7209393709620573, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 19997 + }, + { + "epoch": 0.19998, + "grad_norm": 0.7551959657524542, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 19998 + }, + { + "epoch": 0.19999, + "grad_norm": 0.7219313730716942, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 19999 + }, + { + "epoch": 0.2, + "grad_norm": 0.6586625837495784, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 20000 + }, + { + "epoch": 0.20001, + "grad_norm": 0.7304197773587682, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 20001 + }, + { + "epoch": 0.20002, + "grad_norm": 0.8307558944814868, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 20002 + }, + { + "epoch": 0.20003, + "grad_norm": 0.825641908484607, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 20003 + }, + { + "epoch": 0.20004, + "grad_norm": 0.8216617200292919, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 20004 + }, + { + "epoch": 0.20005, + "grad_norm": 0.8893097902630962, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 20005 + }, + { + "epoch": 0.20006, + "grad_norm": 0.8927636543781022, + "learning_rate": 0.003, + "loss": 3.972, + "step": 20006 + }, + { + "epoch": 0.20007, + "grad_norm": 0.8793889877763003, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 20007 + }, + { + "epoch": 0.20008, + "grad_norm": 0.8210468411654966, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 20008 + }, + { + "epoch": 0.20009, + "grad_norm": 0.751767285149273, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 20009 + }, + { + "epoch": 0.2001, + "grad_norm": 0.9348290953512702, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 20010 + }, + { + "epoch": 0.20011, + "grad_norm": 0.9847308282343222, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 20011 + }, + { + "epoch": 0.20012, + "grad_norm": 0.871742882207395, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 20012 + }, + { + "epoch": 0.20013, + "grad_norm": 0.7226583155028748, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 20013 + }, + { + "epoch": 0.20014, + "grad_norm": 0.6866910362420582, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 20014 + }, + { + "epoch": 0.20015, + "grad_norm": 0.8195246731061213, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 20015 + }, + { + "epoch": 0.20016, + "grad_norm": 0.8921729125287859, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 20016 + }, + { + "epoch": 0.20017, + "grad_norm": 0.9406714211166461, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 20017 + }, + { + "epoch": 0.20018, + "grad_norm": 0.9671947538712713, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 20018 + }, + { + "epoch": 0.20019, + "grad_norm": 0.9255529448514197, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 20019 + }, + { + "epoch": 0.2002, + "grad_norm": 1.0427526976942925, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 20020 + }, + { + "epoch": 0.20021, + "grad_norm": 0.970033070477658, + "learning_rate": 0.003, + "loss": 4.0553, + "step": 20021 + }, + { + "epoch": 0.20022, + "grad_norm": 1.045652396619192, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 20022 + }, + { + "epoch": 0.20023, + "grad_norm": 0.9785695734842935, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 20023 + }, + { + "epoch": 0.20024, + "grad_norm": 0.8457030453393247, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 20024 + }, + { + "epoch": 0.20025, + "grad_norm": 0.7760865280287129, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 20025 + }, + { + "epoch": 0.20026, + "grad_norm": 0.6808764127654939, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 20026 + }, + { + "epoch": 0.20027, + "grad_norm": 0.5961077292276731, + "learning_rate": 0.003, + "loss": 4.015, + "step": 20027 + }, + { + "epoch": 0.20028, + "grad_norm": 0.5524507843947023, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 20028 + }, + { + "epoch": 0.20029, + "grad_norm": 0.526434360636963, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 20029 + }, + { + "epoch": 0.2003, + "grad_norm": 0.6842198010119879, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 20030 + }, + { + "epoch": 0.20031, + "grad_norm": 0.8475115339108801, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 20031 + }, + { + "epoch": 0.20032, + "grad_norm": 0.8978058743823586, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 20032 + }, + { + "epoch": 0.20033, + "grad_norm": 0.8583216802462685, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 20033 + }, + { + "epoch": 0.20034, + "grad_norm": 0.7901396247168336, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 20034 + }, + { + "epoch": 0.20035, + "grad_norm": 0.6831322803519015, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 20035 + }, + { + "epoch": 0.20036, + "grad_norm": 0.555135588560528, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 20036 + }, + { + "epoch": 0.20037, + "grad_norm": 0.5151061644543719, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 20037 + }, + { + "epoch": 0.20038, + "grad_norm": 0.5630542684275427, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 20038 + }, + { + "epoch": 0.20039, + "grad_norm": 0.6020503723129139, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 20039 + }, + { + "epoch": 0.2004, + "grad_norm": 0.5767365320737842, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 20040 + }, + { + "epoch": 0.20041, + "grad_norm": 0.6592390707858953, + "learning_rate": 0.003, + "loss": 4.005, + "step": 20041 + }, + { + "epoch": 0.20042, + "grad_norm": 0.7797019502189348, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 20042 + }, + { + "epoch": 0.20043, + "grad_norm": 0.8218492413884088, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 20043 + }, + { + "epoch": 0.20044, + "grad_norm": 0.843722215784383, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 20044 + }, + { + "epoch": 0.20045, + "grad_norm": 0.9123657671878324, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 20045 + }, + { + "epoch": 0.20046, + "grad_norm": 0.9118940522571418, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 20046 + }, + { + "epoch": 0.20047, + "grad_norm": 0.780462426324903, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 20047 + }, + { + "epoch": 0.20048, + "grad_norm": 0.6247022960537182, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 20048 + }, + { + "epoch": 0.20049, + "grad_norm": 0.6451925602347914, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 20049 + }, + { + "epoch": 0.2005, + "grad_norm": 0.6463765779393614, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 20050 + }, + { + "epoch": 0.20051, + "grad_norm": 0.6430203702996614, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 20051 + }, + { + "epoch": 0.20052, + "grad_norm": 0.6174423324153124, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 20052 + }, + { + "epoch": 0.20053, + "grad_norm": 0.7082317809388395, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 20053 + }, + { + "epoch": 0.20054, + "grad_norm": 0.8175270096709649, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 20054 + }, + { + "epoch": 0.20055, + "grad_norm": 0.8773554392530601, + "learning_rate": 0.003, + "loss": 3.993, + "step": 20055 + }, + { + "epoch": 0.20056, + "grad_norm": 0.8284291496420298, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 20056 + }, + { + "epoch": 0.20057, + "grad_norm": 0.7597320749700079, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 20057 + }, + { + "epoch": 0.20058, + "grad_norm": 0.8209448500710851, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 20058 + }, + { + "epoch": 0.20059, + "grad_norm": 0.7678452991417477, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 20059 + }, + { + "epoch": 0.2006, + "grad_norm": 0.6452770561064738, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 20060 + }, + { + "epoch": 0.20061, + "grad_norm": 0.5763117423514129, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 20061 + }, + { + "epoch": 0.20062, + "grad_norm": 0.570438691497318, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 20062 + }, + { + "epoch": 0.20063, + "grad_norm": 0.5398039392759587, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 20063 + }, + { + "epoch": 0.20064, + "grad_norm": 0.5698746320543828, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 20064 + }, + { + "epoch": 0.20065, + "grad_norm": 0.6089764368626703, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 20065 + }, + { + "epoch": 0.20066, + "grad_norm": 0.5945633916894264, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 20066 + }, + { + "epoch": 0.20067, + "grad_norm": 0.5714765233237329, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 20067 + }, + { + "epoch": 0.20068, + "grad_norm": 0.5859113823732492, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 20068 + }, + { + "epoch": 0.20069, + "grad_norm": 0.6283094179016786, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 20069 + }, + { + "epoch": 0.2007, + "grad_norm": 0.7339560627366473, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 20070 + }, + { + "epoch": 0.20071, + "grad_norm": 0.8868955753372766, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 20071 + }, + { + "epoch": 0.20072, + "grad_norm": 1.0960800738545007, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 20072 + }, + { + "epoch": 0.20073, + "grad_norm": 0.7922776791265033, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 20073 + }, + { + "epoch": 0.20074, + "grad_norm": 0.6768260895802481, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 20074 + }, + { + "epoch": 0.20075, + "grad_norm": 0.6852721679296879, + "learning_rate": 0.003, + "loss": 3.991, + "step": 20075 + }, + { + "epoch": 0.20076, + "grad_norm": 0.6885101071112152, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 20076 + }, + { + "epoch": 0.20077, + "grad_norm": 0.7576387914382314, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 20077 + }, + { + "epoch": 0.20078, + "grad_norm": 0.7775132320594396, + "learning_rate": 0.003, + "loss": 4.007, + "step": 20078 + }, + { + "epoch": 0.20079, + "grad_norm": 0.6198892661582283, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 20079 + }, + { + "epoch": 0.2008, + "grad_norm": 0.6328471081573763, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 20080 + }, + { + "epoch": 0.20081, + "grad_norm": 0.7417751974000814, + "learning_rate": 0.003, + "loss": 3.986, + "step": 20081 + }, + { + "epoch": 0.20082, + "grad_norm": 0.845646129432644, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 20082 + }, + { + "epoch": 0.20083, + "grad_norm": 0.9222967171484822, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 20083 + }, + { + "epoch": 0.20084, + "grad_norm": 0.873249826658085, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 20084 + }, + { + "epoch": 0.20085, + "grad_norm": 0.7916328732882948, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 20085 + }, + { + "epoch": 0.20086, + "grad_norm": 0.7061060110097825, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 20086 + }, + { + "epoch": 0.20087, + "grad_norm": 0.8056062413423868, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 20087 + }, + { + "epoch": 0.20088, + "grad_norm": 0.8380686730602036, + "learning_rate": 0.003, + "loss": 3.988, + "step": 20088 + }, + { + "epoch": 0.20089, + "grad_norm": 0.8805013150746424, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 20089 + }, + { + "epoch": 0.2009, + "grad_norm": 0.8907748329281961, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 20090 + }, + { + "epoch": 0.20091, + "grad_norm": 0.8097461673854649, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 20091 + }, + { + "epoch": 0.20092, + "grad_norm": 0.7478370700673146, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 20092 + }, + { + "epoch": 0.20093, + "grad_norm": 0.8347361510535135, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 20093 + }, + { + "epoch": 0.20094, + "grad_norm": 0.8284062674350586, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 20094 + }, + { + "epoch": 0.20095, + "grad_norm": 0.6864655372978774, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 20095 + }, + { + "epoch": 0.20096, + "grad_norm": 0.7102465460939545, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 20096 + }, + { + "epoch": 0.20097, + "grad_norm": 0.7665201152722105, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 20097 + }, + { + "epoch": 0.20098, + "grad_norm": 0.8750082431411185, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 20098 + }, + { + "epoch": 0.20099, + "grad_norm": 0.8164725550729609, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 20099 + }, + { + "epoch": 0.201, + "grad_norm": 0.6884856980649281, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 20100 + }, + { + "epoch": 0.20101, + "grad_norm": 0.6837243544645405, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 20101 + }, + { + "epoch": 0.20102, + "grad_norm": 0.6309418943923986, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 20102 + }, + { + "epoch": 0.20103, + "grad_norm": 0.6781410050580906, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 20103 + }, + { + "epoch": 0.20104, + "grad_norm": 0.7583636434709431, + "learning_rate": 0.003, + "loss": 4.013, + "step": 20104 + }, + { + "epoch": 0.20105, + "grad_norm": 0.866186468669112, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 20105 + }, + { + "epoch": 0.20106, + "grad_norm": 0.8088000493267465, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 20106 + }, + { + "epoch": 0.20107, + "grad_norm": 0.7766855680807473, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 20107 + }, + { + "epoch": 0.20108, + "grad_norm": 0.8576677766370858, + "learning_rate": 0.003, + "loss": 4.03, + "step": 20108 + }, + { + "epoch": 0.20109, + "grad_norm": 0.9423720025770883, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 20109 + }, + { + "epoch": 0.2011, + "grad_norm": 0.6921778414527199, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 20110 + }, + { + "epoch": 0.20111, + "grad_norm": 0.7178717583468016, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 20111 + }, + { + "epoch": 0.20112, + "grad_norm": 0.6390615653287623, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 20112 + }, + { + "epoch": 0.20113, + "grad_norm": 0.7308221461873914, + "learning_rate": 0.003, + "loss": 4.019, + "step": 20113 + }, + { + "epoch": 0.20114, + "grad_norm": 0.7280964767051464, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 20114 + }, + { + "epoch": 0.20115, + "grad_norm": 0.700090421666984, + "learning_rate": 0.003, + "loss": 4.038, + "step": 20115 + }, + { + "epoch": 0.20116, + "grad_norm": 0.7875888336751482, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 20116 + }, + { + "epoch": 0.20117, + "grad_norm": 0.9976553734289991, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 20117 + }, + { + "epoch": 0.20118, + "grad_norm": 0.9549864359075987, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 20118 + }, + { + "epoch": 0.20119, + "grad_norm": 0.8072813722584153, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 20119 + }, + { + "epoch": 0.2012, + "grad_norm": 0.7794095719344727, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 20120 + }, + { + "epoch": 0.20121, + "grad_norm": 0.7721471510242187, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 20121 + }, + { + "epoch": 0.20122, + "grad_norm": 0.7884934462322606, + "learning_rate": 0.003, + "loss": 4.043, + "step": 20122 + }, + { + "epoch": 0.20123, + "grad_norm": 0.7909826707015193, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 20123 + }, + { + "epoch": 0.20124, + "grad_norm": 0.9485647809805782, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 20124 + }, + { + "epoch": 0.20125, + "grad_norm": 1.0170056274161436, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 20125 + }, + { + "epoch": 0.20126, + "grad_norm": 0.9981031759736696, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 20126 + }, + { + "epoch": 0.20127, + "grad_norm": 0.8797733101845219, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 20127 + }, + { + "epoch": 0.20128, + "grad_norm": 0.7548072147786502, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 20128 + }, + { + "epoch": 0.20129, + "grad_norm": 0.7427617065271436, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 20129 + }, + { + "epoch": 0.2013, + "grad_norm": 0.6780946476426718, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 20130 + }, + { + "epoch": 0.20131, + "grad_norm": 0.7493604500612826, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 20131 + }, + { + "epoch": 0.20132, + "grad_norm": 0.969622173571818, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 20132 + }, + { + "epoch": 0.20133, + "grad_norm": 1.0412724344834599, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 20133 + }, + { + "epoch": 0.20134, + "grad_norm": 0.978954528312447, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 20134 + }, + { + "epoch": 0.20135, + "grad_norm": 0.8503560563237506, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 20135 + }, + { + "epoch": 0.20136, + "grad_norm": 0.7515655864524747, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 20136 + }, + { + "epoch": 0.20137, + "grad_norm": 0.7653315244052966, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 20137 + }, + { + "epoch": 0.20138, + "grad_norm": 0.6782411773437332, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 20138 + }, + { + "epoch": 0.20139, + "grad_norm": 0.5744792920540219, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 20139 + }, + { + "epoch": 0.2014, + "grad_norm": 0.5172282881286189, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 20140 + }, + { + "epoch": 0.20141, + "grad_norm": 0.5190242289701001, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 20141 + }, + { + "epoch": 0.20142, + "grad_norm": 0.5325460638213865, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 20142 + }, + { + "epoch": 0.20143, + "grad_norm": 0.48333102092878705, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 20143 + }, + { + "epoch": 0.20144, + "grad_norm": 0.47389572249758843, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 20144 + }, + { + "epoch": 0.20145, + "grad_norm": 0.4815428109341166, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 20145 + }, + { + "epoch": 0.20146, + "grad_norm": 0.5018453613759905, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 20146 + }, + { + "epoch": 0.20147, + "grad_norm": 0.5600156738661245, + "learning_rate": 0.003, + "loss": 4.002, + "step": 20147 + }, + { + "epoch": 0.20148, + "grad_norm": 0.5853296256920438, + "learning_rate": 0.003, + "loss": 4.002, + "step": 20148 + }, + { + "epoch": 0.20149, + "grad_norm": 0.6263049361225668, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 20149 + }, + { + "epoch": 0.2015, + "grad_norm": 0.6219402386583994, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 20150 + }, + { + "epoch": 0.20151, + "grad_norm": 0.5835684478405019, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 20151 + }, + { + "epoch": 0.20152, + "grad_norm": 0.6865346406840501, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 20152 + }, + { + "epoch": 0.20153, + "grad_norm": 0.9320635500432851, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 20153 + }, + { + "epoch": 0.20154, + "grad_norm": 1.1181078853193314, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 20154 + }, + { + "epoch": 0.20155, + "grad_norm": 0.80487255320592, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 20155 + }, + { + "epoch": 0.20156, + "grad_norm": 0.6881227303915238, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 20156 + }, + { + "epoch": 0.20157, + "grad_norm": 0.5630576684435927, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 20157 + }, + { + "epoch": 0.20158, + "grad_norm": 0.539260651796501, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 20158 + }, + { + "epoch": 0.20159, + "grad_norm": 0.5647999959011761, + "learning_rate": 0.003, + "loss": 3.961, + "step": 20159 + }, + { + "epoch": 0.2016, + "grad_norm": 0.6167997968841262, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 20160 + }, + { + "epoch": 0.20161, + "grad_norm": 0.6461228327159546, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 20161 + }, + { + "epoch": 0.20162, + "grad_norm": 0.738360174180246, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 20162 + }, + { + "epoch": 0.20163, + "grad_norm": 0.787408523429592, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 20163 + }, + { + "epoch": 0.20164, + "grad_norm": 0.8089709227175854, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 20164 + }, + { + "epoch": 0.20165, + "grad_norm": 0.8669904468810457, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 20165 + }, + { + "epoch": 0.20166, + "grad_norm": 0.8562840990287662, + "learning_rate": 0.003, + "loss": 4.016, + "step": 20166 + }, + { + "epoch": 0.20167, + "grad_norm": 0.8183034693083585, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 20167 + }, + { + "epoch": 0.20168, + "grad_norm": 0.7921256829673585, + "learning_rate": 0.003, + "loss": 4.003, + "step": 20168 + }, + { + "epoch": 0.20169, + "grad_norm": 0.7579615874567596, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 20169 + }, + { + "epoch": 0.2017, + "grad_norm": 0.7099445719281334, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 20170 + }, + { + "epoch": 0.20171, + "grad_norm": 0.7443026658609964, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 20171 + }, + { + "epoch": 0.20172, + "grad_norm": 0.7751995389744157, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 20172 + }, + { + "epoch": 0.20173, + "grad_norm": 0.9305057143302977, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 20173 + }, + { + "epoch": 0.20174, + "grad_norm": 0.9849733066971307, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 20174 + }, + { + "epoch": 0.20175, + "grad_norm": 0.9711894518860762, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 20175 + }, + { + "epoch": 0.20176, + "grad_norm": 0.8280435796890636, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 20176 + }, + { + "epoch": 0.20177, + "grad_norm": 0.6142373806449386, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 20177 + }, + { + "epoch": 0.20178, + "grad_norm": 0.6833776761592387, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 20178 + }, + { + "epoch": 0.20179, + "grad_norm": 0.7107750944453681, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 20179 + }, + { + "epoch": 0.2018, + "grad_norm": 0.7402967474957347, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 20180 + }, + { + "epoch": 0.20181, + "grad_norm": 0.7428227333741181, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 20181 + }, + { + "epoch": 0.20182, + "grad_norm": 0.7274291562848075, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 20182 + }, + { + "epoch": 0.20183, + "grad_norm": 0.6561704742408772, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 20183 + }, + { + "epoch": 0.20184, + "grad_norm": 0.6272162788640231, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 20184 + }, + { + "epoch": 0.20185, + "grad_norm": 0.6477878074100514, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 20185 + }, + { + "epoch": 0.20186, + "grad_norm": 0.655237021410275, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 20186 + }, + { + "epoch": 0.20187, + "grad_norm": 0.7166340374252154, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 20187 + }, + { + "epoch": 0.20188, + "grad_norm": 0.7460493156630065, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 20188 + }, + { + "epoch": 0.20189, + "grad_norm": 0.8666646498177367, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 20189 + }, + { + "epoch": 0.2019, + "grad_norm": 0.8732628618060658, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 20190 + }, + { + "epoch": 0.20191, + "grad_norm": 0.7972146873596704, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 20191 + }, + { + "epoch": 0.20192, + "grad_norm": 0.6814140699645804, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 20192 + }, + { + "epoch": 0.20193, + "grad_norm": 0.7951553517542045, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 20193 + }, + { + "epoch": 0.20194, + "grad_norm": 0.8963851288134455, + "learning_rate": 0.003, + "loss": 4.017, + "step": 20194 + }, + { + "epoch": 0.20195, + "grad_norm": 1.00022128278293, + "learning_rate": 0.003, + "loss": 4.0676, + "step": 20195 + }, + { + "epoch": 0.20196, + "grad_norm": 0.9660305295827913, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 20196 + }, + { + "epoch": 0.20197, + "grad_norm": 0.7754576454178924, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 20197 + }, + { + "epoch": 0.20198, + "grad_norm": 0.725059680627483, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 20198 + }, + { + "epoch": 0.20199, + "grad_norm": 0.6454771347249456, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 20199 + }, + { + "epoch": 0.202, + "grad_norm": 0.5856416321569354, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 20200 + }, + { + "epoch": 0.20201, + "grad_norm": 0.5853499483202828, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 20201 + }, + { + "epoch": 0.20202, + "grad_norm": 0.6002634304614478, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 20202 + }, + { + "epoch": 0.20203, + "grad_norm": 0.6256790280554895, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 20203 + }, + { + "epoch": 0.20204, + "grad_norm": 0.6679408568434012, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 20204 + }, + { + "epoch": 0.20205, + "grad_norm": 0.9028198031214248, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 20205 + }, + { + "epoch": 0.20206, + "grad_norm": 1.196781635718897, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 20206 + }, + { + "epoch": 0.20207, + "grad_norm": 0.843819277967857, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 20207 + }, + { + "epoch": 0.20208, + "grad_norm": 0.8726897345056958, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 20208 + }, + { + "epoch": 0.20209, + "grad_norm": 0.7603066053569538, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 20209 + }, + { + "epoch": 0.2021, + "grad_norm": 0.6904486136303193, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 20210 + }, + { + "epoch": 0.20211, + "grad_norm": 0.6978539579520997, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 20211 + }, + { + "epoch": 0.20212, + "grad_norm": 0.6297319691409771, + "learning_rate": 0.003, + "loss": 4.002, + "step": 20212 + }, + { + "epoch": 0.20213, + "grad_norm": 0.6018520982108679, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 20213 + }, + { + "epoch": 0.20214, + "grad_norm": 0.5672683232009093, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 20214 + }, + { + "epoch": 0.20215, + "grad_norm": 0.5677681169784736, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 20215 + }, + { + "epoch": 0.20216, + "grad_norm": 0.4394269591815414, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 20216 + }, + { + "epoch": 0.20217, + "grad_norm": 0.44309297236117656, + "learning_rate": 0.003, + "loss": 3.978, + "step": 20217 + }, + { + "epoch": 0.20218, + "grad_norm": 0.503594451246959, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 20218 + }, + { + "epoch": 0.20219, + "grad_norm": 0.6411198525893247, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 20219 + }, + { + "epoch": 0.2022, + "grad_norm": 0.856242361850307, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 20220 + }, + { + "epoch": 0.20221, + "grad_norm": 1.0731956739782746, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 20221 + }, + { + "epoch": 0.20222, + "grad_norm": 0.8621294576609458, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 20222 + }, + { + "epoch": 0.20223, + "grad_norm": 0.749271239759895, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 20223 + }, + { + "epoch": 0.20224, + "grad_norm": 0.7634863786419059, + "learning_rate": 0.003, + "loss": 4.038, + "step": 20224 + }, + { + "epoch": 0.20225, + "grad_norm": 0.819858244449352, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 20225 + }, + { + "epoch": 0.20226, + "grad_norm": 0.926861861839321, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 20226 + }, + { + "epoch": 0.20227, + "grad_norm": 0.901426283109545, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 20227 + }, + { + "epoch": 0.20228, + "grad_norm": 0.6455763739465206, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 20228 + }, + { + "epoch": 0.20229, + "grad_norm": 0.6164087988976089, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 20229 + }, + { + "epoch": 0.2023, + "grad_norm": 0.6545788584341283, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 20230 + }, + { + "epoch": 0.20231, + "grad_norm": 0.7251441949119408, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 20231 + }, + { + "epoch": 0.20232, + "grad_norm": 0.7207367134968705, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 20232 + }, + { + "epoch": 0.20233, + "grad_norm": 0.7886956460849524, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 20233 + }, + { + "epoch": 0.20234, + "grad_norm": 0.9726795046994932, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 20234 + }, + { + "epoch": 0.20235, + "grad_norm": 0.9777439682842188, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 20235 + }, + { + "epoch": 0.20236, + "grad_norm": 0.8743080722656865, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 20236 + }, + { + "epoch": 0.20237, + "grad_norm": 0.9279717042528518, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 20237 + }, + { + "epoch": 0.20238, + "grad_norm": 0.9289849842074324, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 20238 + }, + { + "epoch": 0.20239, + "grad_norm": 1.0411092708577394, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 20239 + }, + { + "epoch": 0.2024, + "grad_norm": 0.8549415774629084, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 20240 + }, + { + "epoch": 0.20241, + "grad_norm": 0.8611922670327885, + "learning_rate": 0.003, + "loss": 4.021, + "step": 20241 + }, + { + "epoch": 0.20242, + "grad_norm": 0.9234471171883933, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 20242 + }, + { + "epoch": 0.20243, + "grad_norm": 1.0868344511608774, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 20243 + }, + { + "epoch": 0.20244, + "grad_norm": 0.8334510735022014, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 20244 + }, + { + "epoch": 0.20245, + "grad_norm": 0.7640094513898174, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 20245 + }, + { + "epoch": 0.20246, + "grad_norm": 0.734008424005918, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 20246 + }, + { + "epoch": 0.20247, + "grad_norm": 0.6531974491896422, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 20247 + }, + { + "epoch": 0.20248, + "grad_norm": 0.688559116661125, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 20248 + }, + { + "epoch": 0.20249, + "grad_norm": 0.587256705722287, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 20249 + }, + { + "epoch": 0.2025, + "grad_norm": 0.6555690427446274, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 20250 + }, + { + "epoch": 0.20251, + "grad_norm": 0.6595126686035362, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 20251 + }, + { + "epoch": 0.20252, + "grad_norm": 0.6446257362961476, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 20252 + }, + { + "epoch": 0.20253, + "grad_norm": 0.6192335695958366, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 20253 + }, + { + "epoch": 0.20254, + "grad_norm": 0.7259076214179866, + "learning_rate": 0.003, + "loss": 4.0465, + "step": 20254 + }, + { + "epoch": 0.20255, + "grad_norm": 0.773020602773177, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 20255 + }, + { + "epoch": 0.20256, + "grad_norm": 0.6968384799988623, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 20256 + }, + { + "epoch": 0.20257, + "grad_norm": 0.7276249645655477, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 20257 + }, + { + "epoch": 0.20258, + "grad_norm": 0.9511382651976585, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 20258 + }, + { + "epoch": 0.20259, + "grad_norm": 1.0511208785389274, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 20259 + }, + { + "epoch": 0.2026, + "grad_norm": 0.9608367616420654, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 20260 + }, + { + "epoch": 0.20261, + "grad_norm": 0.801484809340342, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 20261 + }, + { + "epoch": 0.20262, + "grad_norm": 0.6825884065625404, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 20262 + }, + { + "epoch": 0.20263, + "grad_norm": 0.6532302663677254, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 20263 + }, + { + "epoch": 0.20264, + "grad_norm": 0.6594540258904026, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 20264 + }, + { + "epoch": 0.20265, + "grad_norm": 0.6976101148358909, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 20265 + }, + { + "epoch": 0.20266, + "grad_norm": 0.6969818960267535, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 20266 + }, + { + "epoch": 0.20267, + "grad_norm": 0.6612631864637771, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 20267 + }, + { + "epoch": 0.20268, + "grad_norm": 0.61870804648751, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 20268 + }, + { + "epoch": 0.20269, + "grad_norm": 0.5668922190541575, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 20269 + }, + { + "epoch": 0.2027, + "grad_norm": 0.6080554471975597, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 20270 + }, + { + "epoch": 0.20271, + "grad_norm": 0.6511168541975132, + "learning_rate": 0.003, + "loss": 4.02, + "step": 20271 + }, + { + "epoch": 0.20272, + "grad_norm": 0.656017449583945, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 20272 + }, + { + "epoch": 0.20273, + "grad_norm": 0.6815463245666242, + "learning_rate": 0.003, + "loss": 3.982, + "step": 20273 + }, + { + "epoch": 0.20274, + "grad_norm": 0.6916571187589557, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 20274 + }, + { + "epoch": 0.20275, + "grad_norm": 0.6789690344941298, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 20275 + }, + { + "epoch": 0.20276, + "grad_norm": 0.866286508716408, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 20276 + }, + { + "epoch": 0.20277, + "grad_norm": 0.9473753399209757, + "learning_rate": 0.003, + "loss": 4.027, + "step": 20277 + }, + { + "epoch": 0.20278, + "grad_norm": 0.812512817362784, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 20278 + }, + { + "epoch": 0.20279, + "grad_norm": 0.6288941215470314, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 20279 + }, + { + "epoch": 0.2028, + "grad_norm": 0.7600707926586033, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 20280 + }, + { + "epoch": 0.20281, + "grad_norm": 0.7206688071877103, + "learning_rate": 0.003, + "loss": 3.98, + "step": 20281 + }, + { + "epoch": 0.20282, + "grad_norm": 0.678867087887113, + "learning_rate": 0.003, + "loss": 4.002, + "step": 20282 + }, + { + "epoch": 0.20283, + "grad_norm": 0.7565984368990963, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 20283 + }, + { + "epoch": 0.20284, + "grad_norm": 0.8448421882510097, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 20284 + }, + { + "epoch": 0.20285, + "grad_norm": 0.7878607927945817, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 20285 + }, + { + "epoch": 0.20286, + "grad_norm": 0.8928041359337932, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 20286 + }, + { + "epoch": 0.20287, + "grad_norm": 0.9397411480717106, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 20287 + }, + { + "epoch": 0.20288, + "grad_norm": 0.9456173592375957, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 20288 + }, + { + "epoch": 0.20289, + "grad_norm": 1.0126275495998667, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 20289 + }, + { + "epoch": 0.2029, + "grad_norm": 0.9619024767592103, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 20290 + }, + { + "epoch": 0.20291, + "grad_norm": 0.8775144888515453, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 20291 + }, + { + "epoch": 0.20292, + "grad_norm": 0.8823928791318055, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 20292 + }, + { + "epoch": 0.20293, + "grad_norm": 0.738557381378013, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 20293 + }, + { + "epoch": 0.20294, + "grad_norm": 0.6883062006832396, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 20294 + }, + { + "epoch": 0.20295, + "grad_norm": 0.7431810168207105, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 20295 + }, + { + "epoch": 0.20296, + "grad_norm": 0.6908545109207357, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 20296 + }, + { + "epoch": 0.20297, + "grad_norm": 0.5751894351648036, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 20297 + }, + { + "epoch": 0.20298, + "grad_norm": 0.6062923632020374, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 20298 + }, + { + "epoch": 0.20299, + "grad_norm": 0.5979260319756273, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 20299 + }, + { + "epoch": 0.203, + "grad_norm": 0.5381429471826685, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 20300 + }, + { + "epoch": 0.20301, + "grad_norm": 0.5829705154680007, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 20301 + }, + { + "epoch": 0.20302, + "grad_norm": 0.6385651995891346, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 20302 + }, + { + "epoch": 0.20303, + "grad_norm": 0.6266784801487898, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 20303 + }, + { + "epoch": 0.20304, + "grad_norm": 0.6098241191700651, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 20304 + }, + { + "epoch": 0.20305, + "grad_norm": 0.6424390938683116, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 20305 + }, + { + "epoch": 0.20306, + "grad_norm": 0.6544827710974398, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 20306 + }, + { + "epoch": 0.20307, + "grad_norm": 0.6626599947348886, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 20307 + }, + { + "epoch": 0.20308, + "grad_norm": 0.7927526919219389, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 20308 + }, + { + "epoch": 0.20309, + "grad_norm": 0.9310423528474954, + "learning_rate": 0.003, + "loss": 3.995, + "step": 20309 + }, + { + "epoch": 0.2031, + "grad_norm": 0.9453198058694835, + "learning_rate": 0.003, + "loss": 3.991, + "step": 20310 + }, + { + "epoch": 0.20311, + "grad_norm": 0.9390606537245991, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 20311 + }, + { + "epoch": 0.20312, + "grad_norm": 0.8870620661170724, + "learning_rate": 0.003, + "loss": 3.985, + "step": 20312 + }, + { + "epoch": 0.20313, + "grad_norm": 0.9179158899728895, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 20313 + }, + { + "epoch": 0.20314, + "grad_norm": 0.8947620853547339, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 20314 + }, + { + "epoch": 0.20315, + "grad_norm": 0.7819293993420867, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 20315 + }, + { + "epoch": 0.20316, + "grad_norm": 0.694590746420825, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 20316 + }, + { + "epoch": 0.20317, + "grad_norm": 0.8157327133078143, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 20317 + }, + { + "epoch": 0.20318, + "grad_norm": 0.8135429882495062, + "learning_rate": 0.003, + "loss": 4.005, + "step": 20318 + }, + { + "epoch": 0.20319, + "grad_norm": 0.8889204865945199, + "learning_rate": 0.003, + "loss": 4.017, + "step": 20319 + }, + { + "epoch": 0.2032, + "grad_norm": 0.8742967319677405, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 20320 + }, + { + "epoch": 0.20321, + "grad_norm": 0.8725738753345575, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 20321 + }, + { + "epoch": 0.20322, + "grad_norm": 0.8148063447634799, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 20322 + }, + { + "epoch": 0.20323, + "grad_norm": 0.7805150305368848, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 20323 + }, + { + "epoch": 0.20324, + "grad_norm": 0.8289579420612636, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 20324 + }, + { + "epoch": 0.20325, + "grad_norm": 0.9197215681237512, + "learning_rate": 0.003, + "loss": 4.005, + "step": 20325 + }, + { + "epoch": 0.20326, + "grad_norm": 0.8831428943179759, + "learning_rate": 0.003, + "loss": 3.972, + "step": 20326 + }, + { + "epoch": 0.20327, + "grad_norm": 0.7721837969718944, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 20327 + }, + { + "epoch": 0.20328, + "grad_norm": 0.7119503778639102, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 20328 + }, + { + "epoch": 0.20329, + "grad_norm": 0.6000065291081916, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 20329 + }, + { + "epoch": 0.2033, + "grad_norm": 0.5519106039194688, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 20330 + }, + { + "epoch": 0.20331, + "grad_norm": 0.6091893490062195, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 20331 + }, + { + "epoch": 0.20332, + "grad_norm": 0.6402900808734804, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 20332 + }, + { + "epoch": 0.20333, + "grad_norm": 0.7400231224944204, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 20333 + }, + { + "epoch": 0.20334, + "grad_norm": 0.8887798969590981, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 20334 + }, + { + "epoch": 0.20335, + "grad_norm": 0.8935271694779896, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 20335 + }, + { + "epoch": 0.20336, + "grad_norm": 0.7553210914001396, + "learning_rate": 0.003, + "loss": 4.001, + "step": 20336 + }, + { + "epoch": 0.20337, + "grad_norm": 0.6862898281056273, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 20337 + }, + { + "epoch": 0.20338, + "grad_norm": 0.5751289339926011, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 20338 + }, + { + "epoch": 0.20339, + "grad_norm": 0.5533284942327815, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 20339 + }, + { + "epoch": 0.2034, + "grad_norm": 0.6105238514680617, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 20340 + }, + { + "epoch": 0.20341, + "grad_norm": 0.6425035816359367, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 20341 + }, + { + "epoch": 0.20342, + "grad_norm": 0.6098920716252937, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 20342 + }, + { + "epoch": 0.20343, + "grad_norm": 0.7028212068665379, + "learning_rate": 0.003, + "loss": 4.014, + "step": 20343 + }, + { + "epoch": 0.20344, + "grad_norm": 0.7204996136589821, + "learning_rate": 0.003, + "loss": 3.995, + "step": 20344 + }, + { + "epoch": 0.20345, + "grad_norm": 0.7395634361151247, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 20345 + }, + { + "epoch": 0.20346, + "grad_norm": 0.7360268187793625, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 20346 + }, + { + "epoch": 0.20347, + "grad_norm": 0.6446374635057549, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 20347 + }, + { + "epoch": 0.20348, + "grad_norm": 0.5578094818800717, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 20348 + }, + { + "epoch": 0.20349, + "grad_norm": 0.5187984671069017, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 20349 + }, + { + "epoch": 0.2035, + "grad_norm": 0.586453356251306, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 20350 + }, + { + "epoch": 0.20351, + "grad_norm": 0.5580077745583748, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 20351 + }, + { + "epoch": 0.20352, + "grad_norm": 0.545114792089224, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 20352 + }, + { + "epoch": 0.20353, + "grad_norm": 0.6101582038514494, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 20353 + }, + { + "epoch": 0.20354, + "grad_norm": 0.7287874476305538, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 20354 + }, + { + "epoch": 0.20355, + "grad_norm": 0.8778749510217199, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 20355 + }, + { + "epoch": 0.20356, + "grad_norm": 0.9898253351434215, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 20356 + }, + { + "epoch": 0.20357, + "grad_norm": 1.0217171530242957, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 20357 + }, + { + "epoch": 0.20358, + "grad_norm": 0.9933515515469676, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 20358 + }, + { + "epoch": 0.20359, + "grad_norm": 0.7520410398659747, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 20359 + }, + { + "epoch": 0.2036, + "grad_norm": 0.7003947049210878, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 20360 + }, + { + "epoch": 0.20361, + "grad_norm": 0.7090980706807122, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 20361 + }, + { + "epoch": 0.20362, + "grad_norm": 0.7512072724237825, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 20362 + }, + { + "epoch": 0.20363, + "grad_norm": 0.7912585149715438, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 20363 + }, + { + "epoch": 0.20364, + "grad_norm": 0.7687954695449859, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 20364 + }, + { + "epoch": 0.20365, + "grad_norm": 0.7693002359173295, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 20365 + }, + { + "epoch": 0.20366, + "grad_norm": 0.7571331435604816, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 20366 + }, + { + "epoch": 0.20367, + "grad_norm": 0.7109870292843715, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 20367 + }, + { + "epoch": 0.20368, + "grad_norm": 0.751971187972105, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 20368 + }, + { + "epoch": 0.20369, + "grad_norm": 0.727036305922877, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 20369 + }, + { + "epoch": 0.2037, + "grad_norm": 0.7542182034833308, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 20370 + }, + { + "epoch": 0.20371, + "grad_norm": 0.713101905916998, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 20371 + }, + { + "epoch": 0.20372, + "grad_norm": 0.5838672971291737, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 20372 + }, + { + "epoch": 0.20373, + "grad_norm": 0.6855692999426602, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 20373 + }, + { + "epoch": 0.20374, + "grad_norm": 0.8184892329934007, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 20374 + }, + { + "epoch": 0.20375, + "grad_norm": 0.8765237247737941, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 20375 + }, + { + "epoch": 0.20376, + "grad_norm": 0.8814937435751463, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 20376 + }, + { + "epoch": 0.20377, + "grad_norm": 0.9651391581475478, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 20377 + }, + { + "epoch": 0.20378, + "grad_norm": 1.0306833040104637, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 20378 + }, + { + "epoch": 0.20379, + "grad_norm": 0.925083820285513, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 20379 + }, + { + "epoch": 0.2038, + "grad_norm": 0.8425866056204823, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 20380 + }, + { + "epoch": 0.20381, + "grad_norm": 0.9092679912468006, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 20381 + }, + { + "epoch": 0.20382, + "grad_norm": 0.7246314668140275, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 20382 + }, + { + "epoch": 0.20383, + "grad_norm": 0.7188588235229608, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 20383 + }, + { + "epoch": 0.20384, + "grad_norm": 0.7886577286859852, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 20384 + }, + { + "epoch": 0.20385, + "grad_norm": 0.9090837683089773, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 20385 + }, + { + "epoch": 0.20386, + "grad_norm": 0.8839897198648974, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 20386 + }, + { + "epoch": 0.20387, + "grad_norm": 0.7876138915422394, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 20387 + }, + { + "epoch": 0.20388, + "grad_norm": 0.6850195619917252, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 20388 + }, + { + "epoch": 0.20389, + "grad_norm": 0.6664634547446611, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 20389 + }, + { + "epoch": 0.2039, + "grad_norm": 0.6095732109548043, + "learning_rate": 0.003, + "loss": 4.03, + "step": 20390 + }, + { + "epoch": 0.20391, + "grad_norm": 0.5820759615749149, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 20391 + }, + { + "epoch": 0.20392, + "grad_norm": 0.6815972090904134, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 20392 + }, + { + "epoch": 0.20393, + "grad_norm": 0.8984131565469776, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 20393 + }, + { + "epoch": 0.20394, + "grad_norm": 1.1268299884815454, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 20394 + }, + { + "epoch": 0.20395, + "grad_norm": 0.9969254216839267, + "learning_rate": 0.003, + "loss": 4.037, + "step": 20395 + }, + { + "epoch": 0.20396, + "grad_norm": 0.8545996424056714, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 20396 + }, + { + "epoch": 0.20397, + "grad_norm": 0.6523792068770117, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 20397 + }, + { + "epoch": 0.20398, + "grad_norm": 0.5910169637623587, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 20398 + }, + { + "epoch": 0.20399, + "grad_norm": 0.7300731503637806, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 20399 + }, + { + "epoch": 0.204, + "grad_norm": 0.8120566709690162, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 20400 + }, + { + "epoch": 0.20401, + "grad_norm": 0.7353265621949143, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 20401 + }, + { + "epoch": 0.20402, + "grad_norm": 0.6130155375485205, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 20402 + }, + { + "epoch": 0.20403, + "grad_norm": 0.5837958097323526, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 20403 + }, + { + "epoch": 0.20404, + "grad_norm": 0.509384605368924, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 20404 + }, + { + "epoch": 0.20405, + "grad_norm": 0.5110561361772817, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 20405 + }, + { + "epoch": 0.20406, + "grad_norm": 0.5164027094962281, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 20406 + }, + { + "epoch": 0.20407, + "grad_norm": 0.5298283554680007, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 20407 + }, + { + "epoch": 0.20408, + "grad_norm": 0.5480031613310662, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 20408 + }, + { + "epoch": 0.20409, + "grad_norm": 0.5286459470288499, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 20409 + }, + { + "epoch": 0.2041, + "grad_norm": 0.5567068121813995, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 20410 + }, + { + "epoch": 0.20411, + "grad_norm": 0.5751754319167662, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 20411 + }, + { + "epoch": 0.20412, + "grad_norm": 0.5417231648808106, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 20412 + }, + { + "epoch": 0.20413, + "grad_norm": 0.49596129958568863, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 20413 + }, + { + "epoch": 0.20414, + "grad_norm": 0.5631342190274198, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 20414 + }, + { + "epoch": 0.20415, + "grad_norm": 0.6100843744366684, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 20415 + }, + { + "epoch": 0.20416, + "grad_norm": 0.6961299584172244, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 20416 + }, + { + "epoch": 0.20417, + "grad_norm": 0.8574457381703728, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 20417 + }, + { + "epoch": 0.20418, + "grad_norm": 0.9321442769416256, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 20418 + }, + { + "epoch": 0.20419, + "grad_norm": 0.8811798560448874, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 20419 + }, + { + "epoch": 0.2042, + "grad_norm": 0.658941647436994, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 20420 + }, + { + "epoch": 0.20421, + "grad_norm": 0.553401476203413, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 20421 + }, + { + "epoch": 0.20422, + "grad_norm": 0.5955142659196577, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 20422 + }, + { + "epoch": 0.20423, + "grad_norm": 0.7279972002078986, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 20423 + }, + { + "epoch": 0.20424, + "grad_norm": 0.7999960788725436, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 20424 + }, + { + "epoch": 0.20425, + "grad_norm": 0.6977307779724798, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 20425 + }, + { + "epoch": 0.20426, + "grad_norm": 0.7717644387953171, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 20426 + }, + { + "epoch": 0.20427, + "grad_norm": 0.9504823569030115, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 20427 + }, + { + "epoch": 0.20428, + "grad_norm": 0.986535152463494, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 20428 + }, + { + "epoch": 0.20429, + "grad_norm": 0.9829941388351892, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 20429 + }, + { + "epoch": 0.2043, + "grad_norm": 1.1231092229966069, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 20430 + }, + { + "epoch": 0.20431, + "grad_norm": 1.0212425615432377, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 20431 + }, + { + "epoch": 0.20432, + "grad_norm": 0.9524288596413643, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 20432 + }, + { + "epoch": 0.20433, + "grad_norm": 0.9605735016626462, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 20433 + }, + { + "epoch": 0.20434, + "grad_norm": 0.9351698151341415, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 20434 + }, + { + "epoch": 0.20435, + "grad_norm": 0.8662390575454001, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 20435 + }, + { + "epoch": 0.20436, + "grad_norm": 0.8138979243701925, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 20436 + }, + { + "epoch": 0.20437, + "grad_norm": 0.7644694442478366, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 20437 + }, + { + "epoch": 0.20438, + "grad_norm": 0.7533190648373956, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 20438 + }, + { + "epoch": 0.20439, + "grad_norm": 0.7032411042201437, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 20439 + }, + { + "epoch": 0.2044, + "grad_norm": 0.7404836223108764, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 20440 + }, + { + "epoch": 0.20441, + "grad_norm": 0.7581499041620451, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 20441 + }, + { + "epoch": 0.20442, + "grad_norm": 0.6491531387031728, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 20442 + }, + { + "epoch": 0.20443, + "grad_norm": 0.6741806778044386, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 20443 + }, + { + "epoch": 0.20444, + "grad_norm": 0.7253782767675879, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 20444 + }, + { + "epoch": 0.20445, + "grad_norm": 0.6103344716020384, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 20445 + }, + { + "epoch": 0.20446, + "grad_norm": 0.6090013578016026, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 20446 + }, + { + "epoch": 0.20447, + "grad_norm": 0.5480336710508134, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 20447 + }, + { + "epoch": 0.20448, + "grad_norm": 0.5710278265563182, + "learning_rate": 0.003, + "loss": 4.008, + "step": 20448 + }, + { + "epoch": 0.20449, + "grad_norm": 0.6163404080598192, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 20449 + }, + { + "epoch": 0.2045, + "grad_norm": 0.6433106190631811, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 20450 + }, + { + "epoch": 0.20451, + "grad_norm": 0.7311648581115475, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 20451 + }, + { + "epoch": 0.20452, + "grad_norm": 0.8079256934591055, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 20452 + }, + { + "epoch": 0.20453, + "grad_norm": 0.9179182611000268, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 20453 + }, + { + "epoch": 0.20454, + "grad_norm": 1.173126902586321, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 20454 + }, + { + "epoch": 0.20455, + "grad_norm": 1.0203440721957817, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 20455 + }, + { + "epoch": 0.20456, + "grad_norm": 0.9133337991665741, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 20456 + }, + { + "epoch": 0.20457, + "grad_norm": 0.7204705803900616, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 20457 + }, + { + "epoch": 0.20458, + "grad_norm": 0.7044649524168134, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 20458 + }, + { + "epoch": 0.20459, + "grad_norm": 0.6278882344923268, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 20459 + }, + { + "epoch": 0.2046, + "grad_norm": 0.7218822197212736, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 20460 + }, + { + "epoch": 0.20461, + "grad_norm": 0.7344154914165129, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 20461 + }, + { + "epoch": 0.20462, + "grad_norm": 0.6953856825283439, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 20462 + }, + { + "epoch": 0.20463, + "grad_norm": 0.6629066623849867, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 20463 + }, + { + "epoch": 0.20464, + "grad_norm": 0.6442801438003857, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 20464 + }, + { + "epoch": 0.20465, + "grad_norm": 0.580668782455788, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 20465 + }, + { + "epoch": 0.20466, + "grad_norm": 0.5904979719966044, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 20466 + }, + { + "epoch": 0.20467, + "grad_norm": 0.5141760545297963, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 20467 + }, + { + "epoch": 0.20468, + "grad_norm": 0.46187258485482413, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 20468 + }, + { + "epoch": 0.20469, + "grad_norm": 0.5800795407273315, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 20469 + }, + { + "epoch": 0.2047, + "grad_norm": 0.602691420793041, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 20470 + }, + { + "epoch": 0.20471, + "grad_norm": 0.6643105853145905, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 20471 + }, + { + "epoch": 0.20472, + "grad_norm": 0.7163219718559465, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 20472 + }, + { + "epoch": 0.20473, + "grad_norm": 0.7988088464722334, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 20473 + }, + { + "epoch": 0.20474, + "grad_norm": 0.9372191748613335, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 20474 + }, + { + "epoch": 0.20475, + "grad_norm": 0.9195149238928961, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 20475 + }, + { + "epoch": 0.20476, + "grad_norm": 0.8244997729789432, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 20476 + }, + { + "epoch": 0.20477, + "grad_norm": 0.8052469218807514, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 20477 + }, + { + "epoch": 0.20478, + "grad_norm": 0.6939877636529218, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 20478 + }, + { + "epoch": 0.20479, + "grad_norm": 0.6094433532230155, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 20479 + }, + { + "epoch": 0.2048, + "grad_norm": 0.6525551667119596, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 20480 + }, + { + "epoch": 0.20481, + "grad_norm": 0.607103447193405, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 20481 + }, + { + "epoch": 0.20482, + "grad_norm": 0.6536757122757433, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 20482 + }, + { + "epoch": 0.20483, + "grad_norm": 0.7330008687286129, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 20483 + }, + { + "epoch": 0.20484, + "grad_norm": 0.7257036344477961, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 20484 + }, + { + "epoch": 0.20485, + "grad_norm": 0.7329867250702582, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 20485 + }, + { + "epoch": 0.20486, + "grad_norm": 0.7949056073745722, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 20486 + }, + { + "epoch": 0.20487, + "grad_norm": 0.8377639269630608, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 20487 + }, + { + "epoch": 0.20488, + "grad_norm": 0.8395058456663194, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 20488 + }, + { + "epoch": 0.20489, + "grad_norm": 0.7624403306692193, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 20489 + }, + { + "epoch": 0.2049, + "grad_norm": 0.7249762100814089, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 20490 + }, + { + "epoch": 0.20491, + "grad_norm": 0.7234008944273093, + "learning_rate": 0.003, + "loss": 4.007, + "step": 20491 + }, + { + "epoch": 0.20492, + "grad_norm": 0.6852052784191084, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 20492 + }, + { + "epoch": 0.20493, + "grad_norm": 0.7385090800602114, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 20493 + }, + { + "epoch": 0.20494, + "grad_norm": 0.8662522661046329, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 20494 + }, + { + "epoch": 0.20495, + "grad_norm": 1.0181849204688709, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 20495 + }, + { + "epoch": 0.20496, + "grad_norm": 0.8831114522995218, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 20496 + }, + { + "epoch": 0.20497, + "grad_norm": 0.9124123107685226, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 20497 + }, + { + "epoch": 0.20498, + "grad_norm": 0.8725329175579076, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 20498 + }, + { + "epoch": 0.20499, + "grad_norm": 0.8677771399687253, + "learning_rate": 0.003, + "loss": 4.035, + "step": 20499 + }, + { + "epoch": 0.205, + "grad_norm": 0.7358761456572323, + "learning_rate": 0.003, + "loss": 4.016, + "step": 20500 + }, + { + "epoch": 0.20501, + "grad_norm": 0.7286963078635518, + "learning_rate": 0.003, + "loss": 4.012, + "step": 20501 + }, + { + "epoch": 0.20502, + "grad_norm": 0.6912240397940084, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 20502 + }, + { + "epoch": 0.20503, + "grad_norm": 0.6158490502211658, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 20503 + }, + { + "epoch": 0.20504, + "grad_norm": 0.7147264570183101, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 20504 + }, + { + "epoch": 0.20505, + "grad_norm": 0.6847930542968079, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 20505 + }, + { + "epoch": 0.20506, + "grad_norm": 0.682740272359546, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 20506 + }, + { + "epoch": 0.20507, + "grad_norm": 0.644022532054537, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 20507 + }, + { + "epoch": 0.20508, + "grad_norm": 0.6368802027145423, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 20508 + }, + { + "epoch": 0.20509, + "grad_norm": 0.6613705625656321, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 20509 + }, + { + "epoch": 0.2051, + "grad_norm": 0.7225109952623768, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 20510 + }, + { + "epoch": 0.20511, + "grad_norm": 0.7818333874654488, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 20511 + }, + { + "epoch": 0.20512, + "grad_norm": 0.7861138471821506, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 20512 + }, + { + "epoch": 0.20513, + "grad_norm": 0.9153734997077961, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 20513 + }, + { + "epoch": 0.20514, + "grad_norm": 0.8202747167477028, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 20514 + }, + { + "epoch": 0.20515, + "grad_norm": 0.6711701657042827, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 20515 + }, + { + "epoch": 0.20516, + "grad_norm": 0.6577618334373466, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 20516 + }, + { + "epoch": 0.20517, + "grad_norm": 0.7645136538607437, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 20517 + }, + { + "epoch": 0.20518, + "grad_norm": 0.791261025325342, + "learning_rate": 0.003, + "loss": 3.988, + "step": 20518 + }, + { + "epoch": 0.20519, + "grad_norm": 0.7461819557194748, + "learning_rate": 0.003, + "loss": 4.011, + "step": 20519 + }, + { + "epoch": 0.2052, + "grad_norm": 0.8481653238025152, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 20520 + }, + { + "epoch": 0.20521, + "grad_norm": 0.9354377842080615, + "learning_rate": 0.003, + "loss": 3.979, + "step": 20521 + }, + { + "epoch": 0.20522, + "grad_norm": 1.0032477082084421, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 20522 + }, + { + "epoch": 0.20523, + "grad_norm": 1.0049389338759382, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 20523 + }, + { + "epoch": 0.20524, + "grad_norm": 1.0444593999768454, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 20524 + }, + { + "epoch": 0.20525, + "grad_norm": 0.8663864999218167, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 20525 + }, + { + "epoch": 0.20526, + "grad_norm": 0.8485993681936183, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 20526 + }, + { + "epoch": 0.20527, + "grad_norm": 0.9485661159647358, + "learning_rate": 0.003, + "loss": 4.018, + "step": 20527 + }, + { + "epoch": 0.20528, + "grad_norm": 0.8885750895135974, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 20528 + }, + { + "epoch": 0.20529, + "grad_norm": 0.7308743516394882, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 20529 + }, + { + "epoch": 0.2053, + "grad_norm": 0.7117854329486243, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 20530 + }, + { + "epoch": 0.20531, + "grad_norm": 0.8279636430676434, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 20531 + }, + { + "epoch": 0.20532, + "grad_norm": 0.9213049687717919, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 20532 + }, + { + "epoch": 0.20533, + "grad_norm": 0.9592903875018273, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 20533 + }, + { + "epoch": 0.20534, + "grad_norm": 0.8750477516222729, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 20534 + }, + { + "epoch": 0.20535, + "grad_norm": 0.8680324748263442, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 20535 + }, + { + "epoch": 0.20536, + "grad_norm": 0.6664550163729996, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 20536 + }, + { + "epoch": 0.20537, + "grad_norm": 0.7200060324917812, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 20537 + }, + { + "epoch": 0.20538, + "grad_norm": 0.6692300647571199, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 20538 + }, + { + "epoch": 0.20539, + "grad_norm": 0.7183338137576324, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 20539 + }, + { + "epoch": 0.2054, + "grad_norm": 0.6188270124855052, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 20540 + }, + { + "epoch": 0.20541, + "grad_norm": 0.574051364944443, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 20541 + }, + { + "epoch": 0.20542, + "grad_norm": 0.6725297595627925, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 20542 + }, + { + "epoch": 0.20543, + "grad_norm": 0.7246739782406196, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 20543 + }, + { + "epoch": 0.20544, + "grad_norm": 0.8049500446074254, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 20544 + }, + { + "epoch": 0.20545, + "grad_norm": 0.9042431224321713, + "learning_rate": 0.003, + "loss": 3.995, + "step": 20545 + }, + { + "epoch": 0.20546, + "grad_norm": 1.066270288781755, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 20546 + }, + { + "epoch": 0.20547, + "grad_norm": 0.8415106537910824, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 20547 + }, + { + "epoch": 0.20548, + "grad_norm": 0.7169447785852349, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 20548 + }, + { + "epoch": 0.20549, + "grad_norm": 0.7712008610748476, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 20549 + }, + { + "epoch": 0.2055, + "grad_norm": 0.8493832013984113, + "learning_rate": 0.003, + "loss": 4.02, + "step": 20550 + }, + { + "epoch": 0.20551, + "grad_norm": 0.8902215757851335, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 20551 + }, + { + "epoch": 0.20552, + "grad_norm": 0.8077765899633021, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 20552 + }, + { + "epoch": 0.20553, + "grad_norm": 0.7426853906423934, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 20553 + }, + { + "epoch": 0.20554, + "grad_norm": 0.6252814519524289, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 20554 + }, + { + "epoch": 0.20555, + "grad_norm": 0.5618230663009974, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 20555 + }, + { + "epoch": 0.20556, + "grad_norm": 0.5441334286174121, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 20556 + }, + { + "epoch": 0.20557, + "grad_norm": 0.5470288495182117, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 20557 + }, + { + "epoch": 0.20558, + "grad_norm": 0.5586359258088992, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 20558 + }, + { + "epoch": 0.20559, + "grad_norm": 0.556371525494898, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 20559 + }, + { + "epoch": 0.2056, + "grad_norm": 0.6058702549566778, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 20560 + }, + { + "epoch": 0.20561, + "grad_norm": 0.6750283098532999, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 20561 + }, + { + "epoch": 0.20562, + "grad_norm": 0.752526381278843, + "learning_rate": 0.003, + "loss": 4.003, + "step": 20562 + }, + { + "epoch": 0.20563, + "grad_norm": 0.7887792593389894, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 20563 + }, + { + "epoch": 0.20564, + "grad_norm": 0.8376640737698638, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 20564 + }, + { + "epoch": 0.20565, + "grad_norm": 0.7841170408309618, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 20565 + }, + { + "epoch": 0.20566, + "grad_norm": 0.7477099075484258, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 20566 + }, + { + "epoch": 0.20567, + "grad_norm": 0.7158364794085459, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 20567 + }, + { + "epoch": 0.20568, + "grad_norm": 0.6890945300482874, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 20568 + }, + { + "epoch": 0.20569, + "grad_norm": 0.6424436848434893, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 20569 + }, + { + "epoch": 0.2057, + "grad_norm": 0.6678028746568943, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 20570 + }, + { + "epoch": 0.20571, + "grad_norm": 0.7523854610332635, + "learning_rate": 0.003, + "loss": 4.0, + "step": 20571 + }, + { + "epoch": 0.20572, + "grad_norm": 0.782796090574714, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 20572 + }, + { + "epoch": 0.20573, + "grad_norm": 0.8005506058812727, + "learning_rate": 0.003, + "loss": 4.01, + "step": 20573 + }, + { + "epoch": 0.20574, + "grad_norm": 0.8339978776880876, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 20574 + }, + { + "epoch": 0.20575, + "grad_norm": 0.7642051471741542, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 20575 + }, + { + "epoch": 0.20576, + "grad_norm": 0.6927350008744555, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 20576 + }, + { + "epoch": 0.20577, + "grad_norm": 0.7604282703434849, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 20577 + }, + { + "epoch": 0.20578, + "grad_norm": 0.7615113347041298, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 20578 + }, + { + "epoch": 0.20579, + "grad_norm": 0.8371007509303824, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 20579 + }, + { + "epoch": 0.2058, + "grad_norm": 0.9380947154224465, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 20580 + }, + { + "epoch": 0.20581, + "grad_norm": 0.9710989246121678, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 20581 + }, + { + "epoch": 0.20582, + "grad_norm": 0.9122111542355176, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 20582 + }, + { + "epoch": 0.20583, + "grad_norm": 0.929456267770815, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 20583 + }, + { + "epoch": 0.20584, + "grad_norm": 0.8036131884151604, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 20584 + }, + { + "epoch": 0.20585, + "grad_norm": 0.7408147684254146, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 20585 + }, + { + "epoch": 0.20586, + "grad_norm": 0.8512561696166622, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 20586 + }, + { + "epoch": 0.20587, + "grad_norm": 1.0079076502526911, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 20587 + }, + { + "epoch": 0.20588, + "grad_norm": 1.145978145245854, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 20588 + }, + { + "epoch": 0.20589, + "grad_norm": 0.7757848976620231, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 20589 + }, + { + "epoch": 0.2059, + "grad_norm": 0.7110634215282196, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 20590 + }, + { + "epoch": 0.20591, + "grad_norm": 0.6024779520886702, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 20591 + }, + { + "epoch": 0.20592, + "grad_norm": 0.6613527318417329, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 20592 + }, + { + "epoch": 0.20593, + "grad_norm": 0.567682890513594, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 20593 + }, + { + "epoch": 0.20594, + "grad_norm": 0.5378953055496669, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 20594 + }, + { + "epoch": 0.20595, + "grad_norm": 0.4394778198563002, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 20595 + }, + { + "epoch": 0.20596, + "grad_norm": 0.4486368326600456, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 20596 + }, + { + "epoch": 0.20597, + "grad_norm": 0.47361903005242784, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 20597 + }, + { + "epoch": 0.20598, + "grad_norm": 0.460873890956977, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 20598 + }, + { + "epoch": 0.20599, + "grad_norm": 0.4435837087019383, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 20599 + }, + { + "epoch": 0.206, + "grad_norm": 0.4222590453083363, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 20600 + }, + { + "epoch": 0.20601, + "grad_norm": 0.4020653967244327, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 20601 + }, + { + "epoch": 0.20602, + "grad_norm": 0.4065777224459865, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 20602 + }, + { + "epoch": 0.20603, + "grad_norm": 0.4699360372159785, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 20603 + }, + { + "epoch": 0.20604, + "grad_norm": 0.5393547033351903, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 20604 + }, + { + "epoch": 0.20605, + "grad_norm": 0.6736220750479196, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 20605 + }, + { + "epoch": 0.20606, + "grad_norm": 0.8718696228449045, + "learning_rate": 0.003, + "loss": 3.968, + "step": 20606 + }, + { + "epoch": 0.20607, + "grad_norm": 1.0603175237373976, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 20607 + }, + { + "epoch": 0.20608, + "grad_norm": 0.8805134579835843, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 20608 + }, + { + "epoch": 0.20609, + "grad_norm": 0.7728995906006011, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 20609 + }, + { + "epoch": 0.2061, + "grad_norm": 0.8080467508398075, + "learning_rate": 0.003, + "loss": 3.993, + "step": 20610 + }, + { + "epoch": 0.20611, + "grad_norm": 0.7781215733171803, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 20611 + }, + { + "epoch": 0.20612, + "grad_norm": 0.8583825081545716, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 20612 + }, + { + "epoch": 0.20613, + "grad_norm": 0.8643013917630427, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 20613 + }, + { + "epoch": 0.20614, + "grad_norm": 0.8452417940528376, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 20614 + }, + { + "epoch": 0.20615, + "grad_norm": 0.8513712451332828, + "learning_rate": 0.003, + "loss": 3.972, + "step": 20615 + }, + { + "epoch": 0.20616, + "grad_norm": 0.9555452051224066, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 20616 + }, + { + "epoch": 0.20617, + "grad_norm": 1.1208260454570567, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 20617 + }, + { + "epoch": 0.20618, + "grad_norm": 1.0216823201347096, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 20618 + }, + { + "epoch": 0.20619, + "grad_norm": 1.0454696516959052, + "learning_rate": 0.003, + "loss": 4.0609, + "step": 20619 + }, + { + "epoch": 0.2062, + "grad_norm": 1.1244179866238118, + "learning_rate": 0.003, + "loss": 4.027, + "step": 20620 + }, + { + "epoch": 0.20621, + "grad_norm": 1.0725544852503082, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 20621 + }, + { + "epoch": 0.20622, + "grad_norm": 1.1044241069760543, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 20622 + }, + { + "epoch": 0.20623, + "grad_norm": 0.9103764695278943, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 20623 + }, + { + "epoch": 0.20624, + "grad_norm": 0.857044847277381, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 20624 + }, + { + "epoch": 0.20625, + "grad_norm": 0.8383868206826671, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 20625 + }, + { + "epoch": 0.20626, + "grad_norm": 0.845910400382633, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 20626 + }, + { + "epoch": 0.20627, + "grad_norm": 0.8348454751958028, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 20627 + }, + { + "epoch": 0.20628, + "grad_norm": 0.9758122185505153, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 20628 + }, + { + "epoch": 0.20629, + "grad_norm": 0.9623190009815188, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 20629 + }, + { + "epoch": 0.2063, + "grad_norm": 0.9321822341407345, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 20630 + }, + { + "epoch": 0.20631, + "grad_norm": 0.944132693375354, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 20631 + }, + { + "epoch": 0.20632, + "grad_norm": 0.8170170622330424, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 20632 + }, + { + "epoch": 0.20633, + "grad_norm": 0.7344039555617329, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 20633 + }, + { + "epoch": 0.20634, + "grad_norm": 0.6907555186652871, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 20634 + }, + { + "epoch": 0.20635, + "grad_norm": 0.6613258197730745, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 20635 + }, + { + "epoch": 0.20636, + "grad_norm": 0.5814958736278305, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 20636 + }, + { + "epoch": 0.20637, + "grad_norm": 0.5385898535208863, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 20637 + }, + { + "epoch": 0.20638, + "grad_norm": 0.49689092127458556, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 20638 + }, + { + "epoch": 0.20639, + "grad_norm": 0.48124571846732095, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 20639 + }, + { + "epoch": 0.2064, + "grad_norm": 0.4286037955554338, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 20640 + }, + { + "epoch": 0.20641, + "grad_norm": 0.4856196195027733, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 20641 + }, + { + "epoch": 0.20642, + "grad_norm": 0.4841291755095502, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 20642 + }, + { + "epoch": 0.20643, + "grad_norm": 0.5432901830942246, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 20643 + }, + { + "epoch": 0.20644, + "grad_norm": 0.5574860416534666, + "learning_rate": 0.003, + "loss": 4.014, + "step": 20644 + }, + { + "epoch": 0.20645, + "grad_norm": 0.5452845249496047, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 20645 + }, + { + "epoch": 0.20646, + "grad_norm": 0.594559321949997, + "learning_rate": 0.003, + "loss": 3.999, + "step": 20646 + }, + { + "epoch": 0.20647, + "grad_norm": 0.6119803735066547, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 20647 + }, + { + "epoch": 0.20648, + "grad_norm": 0.6601017084407726, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 20648 + }, + { + "epoch": 0.20649, + "grad_norm": 0.677958413468039, + "learning_rate": 0.003, + "loss": 3.995, + "step": 20649 + }, + { + "epoch": 0.2065, + "grad_norm": 0.6637899885605804, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 20650 + }, + { + "epoch": 0.20651, + "grad_norm": 0.7656751057365477, + "learning_rate": 0.003, + "loss": 4.002, + "step": 20651 + }, + { + "epoch": 0.20652, + "grad_norm": 0.8893562991418551, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 20652 + }, + { + "epoch": 0.20653, + "grad_norm": 1.0267811016569919, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 20653 + }, + { + "epoch": 0.20654, + "grad_norm": 1.0973076864835005, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 20654 + }, + { + "epoch": 0.20655, + "grad_norm": 0.7391888164640223, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 20655 + }, + { + "epoch": 0.20656, + "grad_norm": 0.717947365925848, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 20656 + }, + { + "epoch": 0.20657, + "grad_norm": 0.7795638661510661, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 20657 + }, + { + "epoch": 0.20658, + "grad_norm": 0.7528038302731817, + "learning_rate": 0.003, + "loss": 4.004, + "step": 20658 + }, + { + "epoch": 0.20659, + "grad_norm": 0.6055299561217218, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 20659 + }, + { + "epoch": 0.2066, + "grad_norm": 0.6519273153646985, + "learning_rate": 0.003, + "loss": 3.974, + "step": 20660 + }, + { + "epoch": 0.20661, + "grad_norm": 0.6637749075573153, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 20661 + }, + { + "epoch": 0.20662, + "grad_norm": 0.6823064529064204, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 20662 + }, + { + "epoch": 0.20663, + "grad_norm": 0.6512057059555628, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 20663 + }, + { + "epoch": 0.20664, + "grad_norm": 0.7656572384119096, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 20664 + }, + { + "epoch": 0.20665, + "grad_norm": 0.7614877118108421, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 20665 + }, + { + "epoch": 0.20666, + "grad_norm": 0.621284504946149, + "learning_rate": 0.003, + "loss": 3.998, + "step": 20666 + }, + { + "epoch": 0.20667, + "grad_norm": 0.609910947815715, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 20667 + }, + { + "epoch": 0.20668, + "grad_norm": 0.626007639166271, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 20668 + }, + { + "epoch": 0.20669, + "grad_norm": 0.7230361785364402, + "learning_rate": 0.003, + "loss": 3.961, + "step": 20669 + }, + { + "epoch": 0.2067, + "grad_norm": 0.8219486807002696, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 20670 + }, + { + "epoch": 0.20671, + "grad_norm": 0.9422692746330305, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 20671 + }, + { + "epoch": 0.20672, + "grad_norm": 1.0204395865033555, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 20672 + }, + { + "epoch": 0.20673, + "grad_norm": 0.975596156059832, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 20673 + }, + { + "epoch": 0.20674, + "grad_norm": 0.719097042352778, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 20674 + }, + { + "epoch": 0.20675, + "grad_norm": 0.6399435306086456, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 20675 + }, + { + "epoch": 0.20676, + "grad_norm": 0.6822550350368889, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 20676 + }, + { + "epoch": 0.20677, + "grad_norm": 0.688990981238369, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 20677 + }, + { + "epoch": 0.20678, + "grad_norm": 0.7431417279472007, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 20678 + }, + { + "epoch": 0.20679, + "grad_norm": 0.6875094207711759, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 20679 + }, + { + "epoch": 0.2068, + "grad_norm": 0.7473141437450511, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 20680 + }, + { + "epoch": 0.20681, + "grad_norm": 0.7686180918780854, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 20681 + }, + { + "epoch": 0.20682, + "grad_norm": 0.7285046658998481, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 20682 + }, + { + "epoch": 0.20683, + "grad_norm": 0.7306031858035114, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 20683 + }, + { + "epoch": 0.20684, + "grad_norm": 0.787399589875625, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 20684 + }, + { + "epoch": 0.20685, + "grad_norm": 0.858172464650602, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 20685 + }, + { + "epoch": 0.20686, + "grad_norm": 0.8692845176128227, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 20686 + }, + { + "epoch": 0.20687, + "grad_norm": 0.8040044471495489, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 20687 + }, + { + "epoch": 0.20688, + "grad_norm": 0.7864112645287474, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 20688 + }, + { + "epoch": 0.20689, + "grad_norm": 0.7385337740069797, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 20689 + }, + { + "epoch": 0.2069, + "grad_norm": 0.7757817478425395, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 20690 + }, + { + "epoch": 0.20691, + "grad_norm": 0.7841018125886556, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 20691 + }, + { + "epoch": 0.20692, + "grad_norm": 0.8477317816447412, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 20692 + }, + { + "epoch": 0.20693, + "grad_norm": 0.9284316678936951, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 20693 + }, + { + "epoch": 0.20694, + "grad_norm": 0.9690867212273199, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 20694 + }, + { + "epoch": 0.20695, + "grad_norm": 1.0462741059470724, + "learning_rate": 0.003, + "loss": 4.005, + "step": 20695 + }, + { + "epoch": 0.20696, + "grad_norm": 0.8749881074119724, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 20696 + }, + { + "epoch": 0.20697, + "grad_norm": 0.8854758175592693, + "learning_rate": 0.003, + "loss": 4.006, + "step": 20697 + }, + { + "epoch": 0.20698, + "grad_norm": 0.9261395760401454, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 20698 + }, + { + "epoch": 0.20699, + "grad_norm": 0.7668614252079934, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 20699 + }, + { + "epoch": 0.207, + "grad_norm": 0.7565294003353837, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 20700 + }, + { + "epoch": 0.20701, + "grad_norm": 0.7349767848631408, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 20701 + }, + { + "epoch": 0.20702, + "grad_norm": 0.8562641105591589, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 20702 + }, + { + "epoch": 0.20703, + "grad_norm": 0.9630032464744976, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 20703 + }, + { + "epoch": 0.20704, + "grad_norm": 0.9099063650319696, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 20704 + }, + { + "epoch": 0.20705, + "grad_norm": 0.7431771205879535, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 20705 + }, + { + "epoch": 0.20706, + "grad_norm": 0.5966469529942441, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 20706 + }, + { + "epoch": 0.20707, + "grad_norm": 0.6639334729718969, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 20707 + }, + { + "epoch": 0.20708, + "grad_norm": 0.7026126828301318, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 20708 + }, + { + "epoch": 0.20709, + "grad_norm": 0.7005647854405611, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 20709 + }, + { + "epoch": 0.2071, + "grad_norm": 0.7342141739314672, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 20710 + }, + { + "epoch": 0.20711, + "grad_norm": 0.7781786407910052, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 20711 + }, + { + "epoch": 0.20712, + "grad_norm": 0.7442282476392599, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 20712 + }, + { + "epoch": 0.20713, + "grad_norm": 0.6691703058852645, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 20713 + }, + { + "epoch": 0.20714, + "grad_norm": 0.6525298833483166, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 20714 + }, + { + "epoch": 0.20715, + "grad_norm": 0.634849336180056, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 20715 + }, + { + "epoch": 0.20716, + "grad_norm": 0.5671628800371192, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 20716 + }, + { + "epoch": 0.20717, + "grad_norm": 0.5704220959948728, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 20717 + }, + { + "epoch": 0.20718, + "grad_norm": 0.5924068982209562, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 20718 + }, + { + "epoch": 0.20719, + "grad_norm": 0.5994605690429917, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 20719 + }, + { + "epoch": 0.2072, + "grad_norm": 0.7187104193869976, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 20720 + }, + { + "epoch": 0.20721, + "grad_norm": 0.7032485319375108, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 20721 + }, + { + "epoch": 0.20722, + "grad_norm": 0.7142106226959795, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 20722 + }, + { + "epoch": 0.20723, + "grad_norm": 0.7949184519245363, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 20723 + }, + { + "epoch": 0.20724, + "grad_norm": 0.8792178246327845, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 20724 + }, + { + "epoch": 0.20725, + "grad_norm": 0.9077165363604695, + "learning_rate": 0.003, + "loss": 4.033, + "step": 20725 + }, + { + "epoch": 0.20726, + "grad_norm": 0.7841620511946787, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 20726 + }, + { + "epoch": 0.20727, + "grad_norm": 0.7174503240198384, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 20727 + }, + { + "epoch": 0.20728, + "grad_norm": 0.6856871682339953, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 20728 + }, + { + "epoch": 0.20729, + "grad_norm": 0.6678508204985093, + "learning_rate": 0.003, + "loss": 3.979, + "step": 20729 + }, + { + "epoch": 0.2073, + "grad_norm": 0.6558720244883144, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 20730 + }, + { + "epoch": 0.20731, + "grad_norm": 0.6538346573646416, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 20731 + }, + { + "epoch": 0.20732, + "grad_norm": 0.6102319756470612, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 20732 + }, + { + "epoch": 0.20733, + "grad_norm": 0.604937182697797, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 20733 + }, + { + "epoch": 0.20734, + "grad_norm": 0.622776260233972, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 20734 + }, + { + "epoch": 0.20735, + "grad_norm": 0.6411852210655435, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 20735 + }, + { + "epoch": 0.20736, + "grad_norm": 0.7224124324684432, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 20736 + }, + { + "epoch": 0.20737, + "grad_norm": 0.7172675506143695, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 20737 + }, + { + "epoch": 0.20738, + "grad_norm": 0.7149441796347479, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 20738 + }, + { + "epoch": 0.20739, + "grad_norm": 0.7655807466644746, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 20739 + }, + { + "epoch": 0.2074, + "grad_norm": 0.663571386455626, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 20740 + }, + { + "epoch": 0.20741, + "grad_norm": 0.6542026406223461, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 20741 + }, + { + "epoch": 0.20742, + "grad_norm": 0.6393238135275731, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 20742 + }, + { + "epoch": 0.20743, + "grad_norm": 0.7738588706639761, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 20743 + }, + { + "epoch": 0.20744, + "grad_norm": 0.9686225589791919, + "learning_rate": 0.003, + "loss": 4.0365, + "step": 20744 + }, + { + "epoch": 0.20745, + "grad_norm": 1.1530790936812518, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 20745 + }, + { + "epoch": 0.20746, + "grad_norm": 0.9592192822803449, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 20746 + }, + { + "epoch": 0.20747, + "grad_norm": 0.8739380940702879, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 20747 + }, + { + "epoch": 0.20748, + "grad_norm": 0.8694296557967807, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 20748 + }, + { + "epoch": 0.20749, + "grad_norm": 0.9776685491677483, + "learning_rate": 0.003, + "loss": 4.008, + "step": 20749 + }, + { + "epoch": 0.2075, + "grad_norm": 1.0966623581662875, + "learning_rate": 0.003, + "loss": 4.046, + "step": 20750 + }, + { + "epoch": 0.20751, + "grad_norm": 0.9109177694374928, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 20751 + }, + { + "epoch": 0.20752, + "grad_norm": 0.8620522777237102, + "learning_rate": 0.003, + "loss": 4.0441, + "step": 20752 + }, + { + "epoch": 0.20753, + "grad_norm": 0.8694374110488925, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 20753 + }, + { + "epoch": 0.20754, + "grad_norm": 0.8229719059276351, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 20754 + }, + { + "epoch": 0.20755, + "grad_norm": 0.7817776420303066, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 20755 + }, + { + "epoch": 0.20756, + "grad_norm": 0.6601560290784039, + "learning_rate": 0.003, + "loss": 3.997, + "step": 20756 + }, + { + "epoch": 0.20757, + "grad_norm": 0.6442357775951242, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 20757 + }, + { + "epoch": 0.20758, + "grad_norm": 0.5801675262253561, + "learning_rate": 0.003, + "loss": 3.999, + "step": 20758 + }, + { + "epoch": 0.20759, + "grad_norm": 0.5937422886764925, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 20759 + }, + { + "epoch": 0.2076, + "grad_norm": 0.5279203452184759, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 20760 + }, + { + "epoch": 0.20761, + "grad_norm": 0.49998016537479284, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 20761 + }, + { + "epoch": 0.20762, + "grad_norm": 0.5000814973215891, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 20762 + }, + { + "epoch": 0.20763, + "grad_norm": 0.6403081651130709, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 20763 + }, + { + "epoch": 0.20764, + "grad_norm": 0.694307552084044, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 20764 + }, + { + "epoch": 0.20765, + "grad_norm": 0.6927699252177045, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 20765 + }, + { + "epoch": 0.20766, + "grad_norm": 0.6162235194903171, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 20766 + }, + { + "epoch": 0.20767, + "grad_norm": 0.6920521359271882, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 20767 + }, + { + "epoch": 0.20768, + "grad_norm": 0.7270622795116948, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 20768 + }, + { + "epoch": 0.20769, + "grad_norm": 0.714724722857881, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 20769 + }, + { + "epoch": 0.2077, + "grad_norm": 0.7736061691504463, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 20770 + }, + { + "epoch": 0.20771, + "grad_norm": 0.8667321241294689, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 20771 + }, + { + "epoch": 0.20772, + "grad_norm": 1.009979308843053, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 20772 + }, + { + "epoch": 0.20773, + "grad_norm": 1.0674529173073988, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 20773 + }, + { + "epoch": 0.20774, + "grad_norm": 0.8466858346137931, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 20774 + }, + { + "epoch": 0.20775, + "grad_norm": 0.7529368180768445, + "learning_rate": 0.003, + "loss": 4.005, + "step": 20775 + }, + { + "epoch": 0.20776, + "grad_norm": 0.6699194663285156, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 20776 + }, + { + "epoch": 0.20777, + "grad_norm": 0.7305925481978442, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 20777 + }, + { + "epoch": 0.20778, + "grad_norm": 0.8069316974385079, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 20778 + }, + { + "epoch": 0.20779, + "grad_norm": 0.7200295924789037, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 20779 + }, + { + "epoch": 0.2078, + "grad_norm": 0.8516644247389931, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 20780 + }, + { + "epoch": 0.20781, + "grad_norm": 1.0787446812489878, + "learning_rate": 0.003, + "loss": 4.021, + "step": 20781 + }, + { + "epoch": 0.20782, + "grad_norm": 0.9306772016444536, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 20782 + }, + { + "epoch": 0.20783, + "grad_norm": 0.7647179330138082, + "learning_rate": 0.003, + "loss": 4.001, + "step": 20783 + }, + { + "epoch": 0.20784, + "grad_norm": 0.8947454887001847, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 20784 + }, + { + "epoch": 0.20785, + "grad_norm": 0.8380117963487906, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 20785 + }, + { + "epoch": 0.20786, + "grad_norm": 0.702558094570999, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 20786 + }, + { + "epoch": 0.20787, + "grad_norm": 0.6640818724942916, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 20787 + }, + { + "epoch": 0.20788, + "grad_norm": 0.658913154129988, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 20788 + }, + { + "epoch": 0.20789, + "grad_norm": 0.7369228785622575, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 20789 + }, + { + "epoch": 0.2079, + "grad_norm": 0.9224064613824003, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 20790 + }, + { + "epoch": 0.20791, + "grad_norm": 1.0418397128907633, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 20791 + }, + { + "epoch": 0.20792, + "grad_norm": 1.0453788851393115, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 20792 + }, + { + "epoch": 0.20793, + "grad_norm": 0.9623154867486063, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 20793 + }, + { + "epoch": 0.20794, + "grad_norm": 0.8748943303003389, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 20794 + }, + { + "epoch": 0.20795, + "grad_norm": 0.7330923046356688, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 20795 + }, + { + "epoch": 0.20796, + "grad_norm": 0.7501069116021372, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 20796 + }, + { + "epoch": 0.20797, + "grad_norm": 0.6120998815310423, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 20797 + }, + { + "epoch": 0.20798, + "grad_norm": 0.554536946110641, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 20798 + }, + { + "epoch": 0.20799, + "grad_norm": 0.512258848836708, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 20799 + }, + { + "epoch": 0.208, + "grad_norm": 0.4934076429911454, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 20800 + }, + { + "epoch": 0.20801, + "grad_norm": 0.5376365703674734, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 20801 + }, + { + "epoch": 0.20802, + "grad_norm": 0.5608451844488169, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 20802 + }, + { + "epoch": 0.20803, + "grad_norm": 0.6389847315901573, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 20803 + }, + { + "epoch": 0.20804, + "grad_norm": 0.6535064564370512, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 20804 + }, + { + "epoch": 0.20805, + "grad_norm": 0.7206601327263827, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 20805 + }, + { + "epoch": 0.20806, + "grad_norm": 0.681179207788442, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 20806 + }, + { + "epoch": 0.20807, + "grad_norm": 0.6470553682393775, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 20807 + }, + { + "epoch": 0.20808, + "grad_norm": 0.6499074871949573, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 20808 + }, + { + "epoch": 0.20809, + "grad_norm": 0.7109497913101789, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 20809 + }, + { + "epoch": 0.2081, + "grad_norm": 0.8583231868942125, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 20810 + }, + { + "epoch": 0.20811, + "grad_norm": 0.960035866320112, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 20811 + }, + { + "epoch": 0.20812, + "grad_norm": 0.911824734408362, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 20812 + }, + { + "epoch": 0.20813, + "grad_norm": 0.7953851062569295, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 20813 + }, + { + "epoch": 0.20814, + "grad_norm": 0.7341274946268906, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 20814 + }, + { + "epoch": 0.20815, + "grad_norm": 0.6764719984646284, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 20815 + }, + { + "epoch": 0.20816, + "grad_norm": 0.6893124353549757, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 20816 + }, + { + "epoch": 0.20817, + "grad_norm": 0.7065602096101486, + "learning_rate": 0.003, + "loss": 4.003, + "step": 20817 + }, + { + "epoch": 0.20818, + "grad_norm": 0.707616363385007, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 20818 + }, + { + "epoch": 0.20819, + "grad_norm": 0.6675253612810456, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 20819 + }, + { + "epoch": 0.2082, + "grad_norm": 0.6580623011411619, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 20820 + }, + { + "epoch": 0.20821, + "grad_norm": 0.5497252559559639, + "learning_rate": 0.003, + "loss": 3.975, + "step": 20821 + }, + { + "epoch": 0.20822, + "grad_norm": 0.5381889642884761, + "learning_rate": 0.003, + "loss": 4.021, + "step": 20822 + }, + { + "epoch": 0.20823, + "grad_norm": 0.5567014585668276, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 20823 + }, + { + "epoch": 0.20824, + "grad_norm": 0.5393640782289372, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 20824 + }, + { + "epoch": 0.20825, + "grad_norm": 0.6464292805056859, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 20825 + }, + { + "epoch": 0.20826, + "grad_norm": 0.7379978854836742, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 20826 + }, + { + "epoch": 0.20827, + "grad_norm": 0.7984545931594577, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 20827 + }, + { + "epoch": 0.20828, + "grad_norm": 0.7987445856403812, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 20828 + }, + { + "epoch": 0.20829, + "grad_norm": 0.716903960605346, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 20829 + }, + { + "epoch": 0.2083, + "grad_norm": 0.6939702239727649, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 20830 + }, + { + "epoch": 0.20831, + "grad_norm": 0.6732284984705267, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 20831 + }, + { + "epoch": 0.20832, + "grad_norm": 0.7580760729951688, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 20832 + }, + { + "epoch": 0.20833, + "grad_norm": 0.8213647195618985, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 20833 + }, + { + "epoch": 0.20834, + "grad_norm": 0.7271855219084686, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 20834 + }, + { + "epoch": 0.20835, + "grad_norm": 0.7971350696814967, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 20835 + }, + { + "epoch": 0.20836, + "grad_norm": 0.9259716392793169, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 20836 + }, + { + "epoch": 0.20837, + "grad_norm": 1.0256845877124325, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 20837 + }, + { + "epoch": 0.20838, + "grad_norm": 1.0234660127296373, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 20838 + }, + { + "epoch": 0.20839, + "grad_norm": 0.8710783345470126, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 20839 + }, + { + "epoch": 0.2084, + "grad_norm": 0.8048520173806204, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 20840 + }, + { + "epoch": 0.20841, + "grad_norm": 0.7529244125129501, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 20841 + }, + { + "epoch": 0.20842, + "grad_norm": 0.8318147410449092, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 20842 + }, + { + "epoch": 0.20843, + "grad_norm": 0.8646142624345001, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 20843 + }, + { + "epoch": 0.20844, + "grad_norm": 0.9323879204425197, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 20844 + }, + { + "epoch": 0.20845, + "grad_norm": 0.9868189048881486, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 20845 + }, + { + "epoch": 0.20846, + "grad_norm": 1.0092201781482701, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 20846 + }, + { + "epoch": 0.20847, + "grad_norm": 0.8446969752348443, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 20847 + }, + { + "epoch": 0.20848, + "grad_norm": 0.6654642299001395, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 20848 + }, + { + "epoch": 0.20849, + "grad_norm": 0.6824988694484712, + "learning_rate": 0.003, + "loss": 3.999, + "step": 20849 + }, + { + "epoch": 0.2085, + "grad_norm": 0.7234788268077534, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 20850 + }, + { + "epoch": 0.20851, + "grad_norm": 0.7377697229993281, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 20851 + }, + { + "epoch": 0.20852, + "grad_norm": 0.7493727391560796, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 20852 + }, + { + "epoch": 0.20853, + "grad_norm": 0.7216511775753078, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 20853 + }, + { + "epoch": 0.20854, + "grad_norm": 0.7333043808799856, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 20854 + }, + { + "epoch": 0.20855, + "grad_norm": 0.6922624348353381, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 20855 + }, + { + "epoch": 0.20856, + "grad_norm": 0.6888223104394309, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 20856 + }, + { + "epoch": 0.20857, + "grad_norm": 0.6884816140409558, + "learning_rate": 0.003, + "loss": 4.02, + "step": 20857 + }, + { + "epoch": 0.20858, + "grad_norm": 0.6555538160199587, + "learning_rate": 0.003, + "loss": 4.02, + "step": 20858 + }, + { + "epoch": 0.20859, + "grad_norm": 0.7331503174721685, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 20859 + }, + { + "epoch": 0.2086, + "grad_norm": 0.7076887996468585, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 20860 + }, + { + "epoch": 0.20861, + "grad_norm": 0.6786693112356252, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 20861 + }, + { + "epoch": 0.20862, + "grad_norm": 0.8477344912182458, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 20862 + }, + { + "epoch": 0.20863, + "grad_norm": 0.9914240366817453, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 20863 + }, + { + "epoch": 0.20864, + "grad_norm": 0.9368427260256922, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 20864 + }, + { + "epoch": 0.20865, + "grad_norm": 0.9261496489963956, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 20865 + }, + { + "epoch": 0.20866, + "grad_norm": 0.825011446515112, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 20866 + }, + { + "epoch": 0.20867, + "grad_norm": 0.7844423976893438, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 20867 + }, + { + "epoch": 0.20868, + "grad_norm": 0.619757197377466, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 20868 + }, + { + "epoch": 0.20869, + "grad_norm": 0.595437145844322, + "learning_rate": 0.003, + "loss": 4.017, + "step": 20869 + }, + { + "epoch": 0.2087, + "grad_norm": 0.673433715230681, + "learning_rate": 0.003, + "loss": 4.004, + "step": 20870 + }, + { + "epoch": 0.20871, + "grad_norm": 0.7121729536671715, + "learning_rate": 0.003, + "loss": 4.025, + "step": 20871 + }, + { + "epoch": 0.20872, + "grad_norm": 0.5877647704898378, + "learning_rate": 0.003, + "loss": 4.034, + "step": 20872 + }, + { + "epoch": 0.20873, + "grad_norm": 0.500347287001371, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 20873 + }, + { + "epoch": 0.20874, + "grad_norm": 0.44242193093502685, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 20874 + }, + { + "epoch": 0.20875, + "grad_norm": 0.46071605890476824, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 20875 + }, + { + "epoch": 0.20876, + "grad_norm": 0.4618762650573843, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 20876 + }, + { + "epoch": 0.20877, + "grad_norm": 0.49727610101287845, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 20877 + }, + { + "epoch": 0.20878, + "grad_norm": 0.5509630917448674, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 20878 + }, + { + "epoch": 0.20879, + "grad_norm": 0.6438151595326801, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 20879 + }, + { + "epoch": 0.2088, + "grad_norm": 0.7232815979576777, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 20880 + }, + { + "epoch": 0.20881, + "grad_norm": 0.7257934842398682, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 20881 + }, + { + "epoch": 0.20882, + "grad_norm": 0.7530184308057969, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 20882 + }, + { + "epoch": 0.20883, + "grad_norm": 0.7615198932728324, + "learning_rate": 0.003, + "loss": 4.001, + "step": 20883 + }, + { + "epoch": 0.20884, + "grad_norm": 0.6410949335143684, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 20884 + }, + { + "epoch": 0.20885, + "grad_norm": 0.6562122166232585, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 20885 + }, + { + "epoch": 0.20886, + "grad_norm": 0.8460310490761246, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 20886 + }, + { + "epoch": 0.20887, + "grad_norm": 0.9260686628807436, + "learning_rate": 0.003, + "loss": 4.0417, + "step": 20887 + }, + { + "epoch": 0.20888, + "grad_norm": 1.2027185513990395, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 20888 + }, + { + "epoch": 0.20889, + "grad_norm": 0.954593814573625, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 20889 + }, + { + "epoch": 0.2089, + "grad_norm": 0.9658129409239519, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 20890 + }, + { + "epoch": 0.20891, + "grad_norm": 0.8581879374152503, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 20891 + }, + { + "epoch": 0.20892, + "grad_norm": 0.8906650483649206, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 20892 + }, + { + "epoch": 0.20893, + "grad_norm": 0.9639703554396021, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 20893 + }, + { + "epoch": 0.20894, + "grad_norm": 0.9583189766916249, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 20894 + }, + { + "epoch": 0.20895, + "grad_norm": 1.0480591215758979, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 20895 + }, + { + "epoch": 0.20896, + "grad_norm": 0.9067714227245176, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 20896 + }, + { + "epoch": 0.20897, + "grad_norm": 1.0760745574055761, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 20897 + }, + { + "epoch": 0.20898, + "grad_norm": 1.0422063069087546, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 20898 + }, + { + "epoch": 0.20899, + "grad_norm": 0.8719391103558431, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 20899 + }, + { + "epoch": 0.209, + "grad_norm": 0.7438897179831354, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 20900 + }, + { + "epoch": 0.20901, + "grad_norm": 0.6515924706019836, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 20901 + }, + { + "epoch": 0.20902, + "grad_norm": 0.7054066035249528, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 20902 + }, + { + "epoch": 0.20903, + "grad_norm": 0.7490869128361373, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 20903 + }, + { + "epoch": 0.20904, + "grad_norm": 0.6891030072751542, + "learning_rate": 0.003, + "loss": 4.006, + "step": 20904 + }, + { + "epoch": 0.20905, + "grad_norm": 0.6804789789636855, + "learning_rate": 0.003, + "loss": 4.051, + "step": 20905 + }, + { + "epoch": 0.20906, + "grad_norm": 0.6980282802742136, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 20906 + }, + { + "epoch": 0.20907, + "grad_norm": 0.6664715542327716, + "learning_rate": 0.003, + "loss": 4.018, + "step": 20907 + }, + { + "epoch": 0.20908, + "grad_norm": 0.5983422763935576, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 20908 + }, + { + "epoch": 0.20909, + "grad_norm": 0.5018529228083116, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 20909 + }, + { + "epoch": 0.2091, + "grad_norm": 0.5365175766509066, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 20910 + }, + { + "epoch": 0.20911, + "grad_norm": 0.5060367550320816, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 20911 + }, + { + "epoch": 0.20912, + "grad_norm": 0.5227299910255248, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 20912 + }, + { + "epoch": 0.20913, + "grad_norm": 0.5504569387437639, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 20913 + }, + { + "epoch": 0.20914, + "grad_norm": 0.5809067583361618, + "learning_rate": 0.003, + "loss": 3.996, + "step": 20914 + }, + { + "epoch": 0.20915, + "grad_norm": 0.5692107526172272, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 20915 + }, + { + "epoch": 0.20916, + "grad_norm": 0.5828394268675717, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 20916 + }, + { + "epoch": 0.20917, + "grad_norm": 0.5154598069556916, + "learning_rate": 0.003, + "loss": 3.97, + "step": 20917 + }, + { + "epoch": 0.20918, + "grad_norm": 0.5230724815179841, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 20918 + }, + { + "epoch": 0.20919, + "grad_norm": 0.6222439251060663, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 20919 + }, + { + "epoch": 0.2092, + "grad_norm": 0.7240808750880454, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 20920 + }, + { + "epoch": 0.20921, + "grad_norm": 0.7517466607473207, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 20921 + }, + { + "epoch": 0.20922, + "grad_norm": 0.796432726803999, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 20922 + }, + { + "epoch": 0.20923, + "grad_norm": 0.9258107529194182, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 20923 + }, + { + "epoch": 0.20924, + "grad_norm": 0.9946715378657085, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 20924 + }, + { + "epoch": 0.20925, + "grad_norm": 1.2300600615793102, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 20925 + }, + { + "epoch": 0.20926, + "grad_norm": 0.7116402387659458, + "learning_rate": 0.003, + "loss": 4.004, + "step": 20926 + }, + { + "epoch": 0.20927, + "grad_norm": 0.6376117875414932, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 20927 + }, + { + "epoch": 0.20928, + "grad_norm": 0.6552456558733806, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 20928 + }, + { + "epoch": 0.20929, + "grad_norm": 0.6087701482357425, + "learning_rate": 0.003, + "loss": 4.008, + "step": 20929 + }, + { + "epoch": 0.2093, + "grad_norm": 0.6717119391415858, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 20930 + }, + { + "epoch": 0.20931, + "grad_norm": 0.6729377209162777, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 20931 + }, + { + "epoch": 0.20932, + "grad_norm": 0.6331628866016462, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 20932 + }, + { + "epoch": 0.20933, + "grad_norm": 0.7371232822899574, + "learning_rate": 0.003, + "loss": 3.976, + "step": 20933 + }, + { + "epoch": 0.20934, + "grad_norm": 0.7623414032177308, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 20934 + }, + { + "epoch": 0.20935, + "grad_norm": 0.7785422882961544, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 20935 + }, + { + "epoch": 0.20936, + "grad_norm": 0.776800502600593, + "learning_rate": 0.003, + "loss": 3.965, + "step": 20936 + }, + { + "epoch": 0.20937, + "grad_norm": 0.8171623514152386, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 20937 + }, + { + "epoch": 0.20938, + "grad_norm": 0.8244043092682203, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 20938 + }, + { + "epoch": 0.20939, + "grad_norm": 0.7558005160452077, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 20939 + }, + { + "epoch": 0.2094, + "grad_norm": 0.713295910361685, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 20940 + }, + { + "epoch": 0.20941, + "grad_norm": 0.7605474412727585, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 20941 + }, + { + "epoch": 0.20942, + "grad_norm": 0.9213725271564825, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 20942 + }, + { + "epoch": 0.20943, + "grad_norm": 1.1318035521325334, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 20943 + }, + { + "epoch": 0.20944, + "grad_norm": 0.9239153313597757, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 20944 + }, + { + "epoch": 0.20945, + "grad_norm": 0.8301149665789103, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 20945 + }, + { + "epoch": 0.20946, + "grad_norm": 0.8032272566360151, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 20946 + }, + { + "epoch": 0.20947, + "grad_norm": 0.7820181809684242, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 20947 + }, + { + "epoch": 0.20948, + "grad_norm": 0.7447748652230458, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 20948 + }, + { + "epoch": 0.20949, + "grad_norm": 0.8391258137939391, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 20949 + }, + { + "epoch": 0.2095, + "grad_norm": 0.9728839714381725, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 20950 + }, + { + "epoch": 0.20951, + "grad_norm": 1.1793699772257986, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 20951 + }, + { + "epoch": 0.20952, + "grad_norm": 0.7847511133915612, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 20952 + }, + { + "epoch": 0.20953, + "grad_norm": 0.7162107259247007, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 20953 + }, + { + "epoch": 0.20954, + "grad_norm": 0.7276276977144239, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 20954 + }, + { + "epoch": 0.20955, + "grad_norm": 0.862203993792162, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 20955 + }, + { + "epoch": 0.20956, + "grad_norm": 0.9108022944128619, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 20956 + }, + { + "epoch": 0.20957, + "grad_norm": 0.7047243874844601, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 20957 + }, + { + "epoch": 0.20958, + "grad_norm": 0.7630518820949418, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 20958 + }, + { + "epoch": 0.20959, + "grad_norm": 0.8380374046361619, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 20959 + }, + { + "epoch": 0.2096, + "grad_norm": 0.8507141254709075, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 20960 + }, + { + "epoch": 0.20961, + "grad_norm": 0.7230685571727478, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 20961 + }, + { + "epoch": 0.20962, + "grad_norm": 0.6350001060145711, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 20962 + }, + { + "epoch": 0.20963, + "grad_norm": 0.6384334971959221, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 20963 + }, + { + "epoch": 0.20964, + "grad_norm": 0.6394858584644065, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 20964 + }, + { + "epoch": 0.20965, + "grad_norm": 0.6378033828978453, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 20965 + }, + { + "epoch": 0.20966, + "grad_norm": 0.6016634184217268, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 20966 + }, + { + "epoch": 0.20967, + "grad_norm": 0.4939321633359444, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 20967 + }, + { + "epoch": 0.20968, + "grad_norm": 0.38467927345477904, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 20968 + }, + { + "epoch": 0.20969, + "grad_norm": 0.47388274288254595, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 20969 + }, + { + "epoch": 0.2097, + "grad_norm": 0.5458749713361009, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 20970 + }, + { + "epoch": 0.20971, + "grad_norm": 0.6157475945413651, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 20971 + }, + { + "epoch": 0.20972, + "grad_norm": 0.6292924079642748, + "learning_rate": 0.003, + "loss": 4.011, + "step": 20972 + }, + { + "epoch": 0.20973, + "grad_norm": 0.7379312729385584, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 20973 + }, + { + "epoch": 0.20974, + "grad_norm": 0.8671396086287233, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 20974 + }, + { + "epoch": 0.20975, + "grad_norm": 0.863105276232074, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 20975 + }, + { + "epoch": 0.20976, + "grad_norm": 0.7420551843648684, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 20976 + }, + { + "epoch": 0.20977, + "grad_norm": 0.654240171468961, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 20977 + }, + { + "epoch": 0.20978, + "grad_norm": 0.6619159401057324, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 20978 + }, + { + "epoch": 0.20979, + "grad_norm": 0.7246628070698304, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 20979 + }, + { + "epoch": 0.2098, + "grad_norm": 0.7710388236370247, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 20980 + }, + { + "epoch": 0.20981, + "grad_norm": 0.707440274114697, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 20981 + }, + { + "epoch": 0.20982, + "grad_norm": 0.6464117629727263, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 20982 + }, + { + "epoch": 0.20983, + "grad_norm": 0.6981419334384464, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 20983 + }, + { + "epoch": 0.20984, + "grad_norm": 0.6770556332156483, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 20984 + }, + { + "epoch": 0.20985, + "grad_norm": 0.709957336862496, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 20985 + }, + { + "epoch": 0.20986, + "grad_norm": 0.7848031327564378, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 20986 + }, + { + "epoch": 0.20987, + "grad_norm": 0.9176299293982443, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 20987 + }, + { + "epoch": 0.20988, + "grad_norm": 1.0533364212904954, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 20988 + }, + { + "epoch": 0.20989, + "grad_norm": 0.966847047235216, + "learning_rate": 0.003, + "loss": 3.997, + "step": 20989 + }, + { + "epoch": 0.2099, + "grad_norm": 0.8754017125528573, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 20990 + }, + { + "epoch": 0.20991, + "grad_norm": 0.7599007221742755, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 20991 + }, + { + "epoch": 0.20992, + "grad_norm": 0.6315470528008028, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 20992 + }, + { + "epoch": 0.20993, + "grad_norm": 0.7481283220868374, + "learning_rate": 0.003, + "loss": 4.06, + "step": 20993 + }, + { + "epoch": 0.20994, + "grad_norm": 0.7640819945687202, + "learning_rate": 0.003, + "loss": 4.021, + "step": 20994 + }, + { + "epoch": 0.20995, + "grad_norm": 0.9770979550059589, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 20995 + }, + { + "epoch": 0.20996, + "grad_norm": 1.0904975476992498, + "learning_rate": 0.003, + "loss": 4.02, + "step": 20996 + }, + { + "epoch": 0.20997, + "grad_norm": 1.020535661571175, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 20997 + }, + { + "epoch": 0.20998, + "grad_norm": 0.8410735399173201, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 20998 + }, + { + "epoch": 0.20999, + "grad_norm": 0.8595310118112839, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 20999 + }, + { + "epoch": 0.21, + "grad_norm": 0.8291125326846455, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 21000 + }, + { + "epoch": 0.21001, + "grad_norm": 0.8696029205707217, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 21001 + }, + { + "epoch": 0.21002, + "grad_norm": 0.8762243722540096, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 21002 + }, + { + "epoch": 0.21003, + "grad_norm": 1.0675300681846718, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 21003 + }, + { + "epoch": 0.21004, + "grad_norm": 0.7997679466742341, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 21004 + }, + { + "epoch": 0.21005, + "grad_norm": 0.6611639355079719, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 21005 + }, + { + "epoch": 0.21006, + "grad_norm": 0.6477168985358737, + "learning_rate": 0.003, + "loss": 4.003, + "step": 21006 + }, + { + "epoch": 0.21007, + "grad_norm": 0.6241214277359238, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 21007 + }, + { + "epoch": 0.21008, + "grad_norm": 0.6352831881387736, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 21008 + }, + { + "epoch": 0.21009, + "grad_norm": 0.6971166642989427, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 21009 + }, + { + "epoch": 0.2101, + "grad_norm": 0.7150821161856583, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 21010 + }, + { + "epoch": 0.21011, + "grad_norm": 0.7874828521116144, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 21011 + }, + { + "epoch": 0.21012, + "grad_norm": 0.9560962868211619, + "learning_rate": 0.003, + "loss": 4.031, + "step": 21012 + }, + { + "epoch": 0.21013, + "grad_norm": 0.878572385159276, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 21013 + }, + { + "epoch": 0.21014, + "grad_norm": 0.7163482877324255, + "learning_rate": 0.003, + "loss": 4.019, + "step": 21014 + }, + { + "epoch": 0.21015, + "grad_norm": 0.7218151272705788, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 21015 + }, + { + "epoch": 0.21016, + "grad_norm": 0.6947011656395556, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 21016 + }, + { + "epoch": 0.21017, + "grad_norm": 0.6720907264804272, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 21017 + }, + { + "epoch": 0.21018, + "grad_norm": 0.6976487697939356, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 21018 + }, + { + "epoch": 0.21019, + "grad_norm": 0.6134489686991222, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 21019 + }, + { + "epoch": 0.2102, + "grad_norm": 0.6436229425698038, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 21020 + }, + { + "epoch": 0.21021, + "grad_norm": 0.6686845539208095, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 21021 + }, + { + "epoch": 0.21022, + "grad_norm": 0.7129789872557434, + "learning_rate": 0.003, + "loss": 3.995, + "step": 21022 + }, + { + "epoch": 0.21023, + "grad_norm": 0.6727096261891775, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 21023 + }, + { + "epoch": 0.21024, + "grad_norm": 0.6454384203757174, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 21024 + }, + { + "epoch": 0.21025, + "grad_norm": 0.6645864029319164, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 21025 + }, + { + "epoch": 0.21026, + "grad_norm": 0.7472687962993371, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 21026 + }, + { + "epoch": 0.21027, + "grad_norm": 0.8647758454882478, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 21027 + }, + { + "epoch": 0.21028, + "grad_norm": 0.9185057870955434, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 21028 + }, + { + "epoch": 0.21029, + "grad_norm": 0.8378790984680934, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 21029 + }, + { + "epoch": 0.2103, + "grad_norm": 0.6442546990759301, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 21030 + }, + { + "epoch": 0.21031, + "grad_norm": 0.8595776890808869, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 21031 + }, + { + "epoch": 0.21032, + "grad_norm": 0.9554408365075745, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 21032 + }, + { + "epoch": 0.21033, + "grad_norm": 1.0074606274224762, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 21033 + }, + { + "epoch": 0.21034, + "grad_norm": 0.9744096417572639, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 21034 + }, + { + "epoch": 0.21035, + "grad_norm": 0.7964769066807219, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 21035 + }, + { + "epoch": 0.21036, + "grad_norm": 0.7308420902640176, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 21036 + }, + { + "epoch": 0.21037, + "grad_norm": 0.7552750706451398, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 21037 + }, + { + "epoch": 0.21038, + "grad_norm": 0.7816553654407847, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 21038 + }, + { + "epoch": 0.21039, + "grad_norm": 0.7879726657779169, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 21039 + }, + { + "epoch": 0.2104, + "grad_norm": 0.7810519161213411, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 21040 + }, + { + "epoch": 0.21041, + "grad_norm": 0.7710968800698415, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 21041 + }, + { + "epoch": 0.21042, + "grad_norm": 0.6653261456756071, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 21042 + }, + { + "epoch": 0.21043, + "grad_norm": 0.6906164880605293, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 21043 + }, + { + "epoch": 0.21044, + "grad_norm": 0.6316685169677996, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 21044 + }, + { + "epoch": 0.21045, + "grad_norm": 0.5607584788415311, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 21045 + }, + { + "epoch": 0.21046, + "grad_norm": 0.5548492594430922, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 21046 + }, + { + "epoch": 0.21047, + "grad_norm": 0.5261171398357389, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 21047 + }, + { + "epoch": 0.21048, + "grad_norm": 0.4776011111136378, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 21048 + }, + { + "epoch": 0.21049, + "grad_norm": 0.5250397190473844, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 21049 + }, + { + "epoch": 0.2105, + "grad_norm": 0.5415230561397675, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 21050 + }, + { + "epoch": 0.21051, + "grad_norm": 0.6401617626734338, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 21051 + }, + { + "epoch": 0.21052, + "grad_norm": 0.7821905926424524, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 21052 + }, + { + "epoch": 0.21053, + "grad_norm": 0.867840886613997, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 21053 + }, + { + "epoch": 0.21054, + "grad_norm": 0.8942115816253537, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 21054 + }, + { + "epoch": 0.21055, + "grad_norm": 0.7435929940669416, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 21055 + }, + { + "epoch": 0.21056, + "grad_norm": 0.7881546297177118, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 21056 + }, + { + "epoch": 0.21057, + "grad_norm": 0.9491891379197605, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 21057 + }, + { + "epoch": 0.21058, + "grad_norm": 0.9298271695610014, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 21058 + }, + { + "epoch": 0.21059, + "grad_norm": 0.8028064964676196, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 21059 + }, + { + "epoch": 0.2106, + "grad_norm": 0.6910215756735963, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 21060 + }, + { + "epoch": 0.21061, + "grad_norm": 0.8011741630065974, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 21061 + }, + { + "epoch": 0.21062, + "grad_norm": 0.8160736083955444, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 21062 + }, + { + "epoch": 0.21063, + "grad_norm": 0.9167134672719457, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 21063 + }, + { + "epoch": 0.21064, + "grad_norm": 0.8703916603601536, + "learning_rate": 0.003, + "loss": 4.026, + "step": 21064 + }, + { + "epoch": 0.21065, + "grad_norm": 0.8539583741322381, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 21065 + }, + { + "epoch": 0.21066, + "grad_norm": 0.7661854409428724, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 21066 + }, + { + "epoch": 0.21067, + "grad_norm": 0.7821181561304025, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 21067 + }, + { + "epoch": 0.21068, + "grad_norm": 0.838179778431643, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 21068 + }, + { + "epoch": 0.21069, + "grad_norm": 0.8568516156320897, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 21069 + }, + { + "epoch": 0.2107, + "grad_norm": 0.7793089377237146, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 21070 + }, + { + "epoch": 0.21071, + "grad_norm": 0.7361615077664392, + "learning_rate": 0.003, + "loss": 3.975, + "step": 21071 + }, + { + "epoch": 0.21072, + "grad_norm": 0.7402931210701514, + "learning_rate": 0.003, + "loss": 4.005, + "step": 21072 + }, + { + "epoch": 0.21073, + "grad_norm": 0.7240033830114518, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 21073 + }, + { + "epoch": 0.21074, + "grad_norm": 0.7039514273779341, + "learning_rate": 0.003, + "loss": 4.008, + "step": 21074 + }, + { + "epoch": 0.21075, + "grad_norm": 0.682034032812649, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 21075 + }, + { + "epoch": 0.21076, + "grad_norm": 0.6526334549023143, + "learning_rate": 0.003, + "loss": 4.005, + "step": 21076 + }, + { + "epoch": 0.21077, + "grad_norm": 0.6159295382239347, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 21077 + }, + { + "epoch": 0.21078, + "grad_norm": 0.6106263741006914, + "learning_rate": 0.003, + "loss": 3.992, + "step": 21078 + }, + { + "epoch": 0.21079, + "grad_norm": 0.63903778695696, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 21079 + }, + { + "epoch": 0.2108, + "grad_norm": 0.7095978453825822, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 21080 + }, + { + "epoch": 0.21081, + "grad_norm": 0.7880126304595153, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 21081 + }, + { + "epoch": 0.21082, + "grad_norm": 0.7740135083861598, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 21082 + }, + { + "epoch": 0.21083, + "grad_norm": 0.759435427703446, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 21083 + }, + { + "epoch": 0.21084, + "grad_norm": 0.8326361202731105, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 21084 + }, + { + "epoch": 0.21085, + "grad_norm": 0.9084544073860805, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 21085 + }, + { + "epoch": 0.21086, + "grad_norm": 0.9491343989857676, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 21086 + }, + { + "epoch": 0.21087, + "grad_norm": 0.9257445976430423, + "learning_rate": 0.003, + "loss": 4.02, + "step": 21087 + }, + { + "epoch": 0.21088, + "grad_norm": 0.8520281488011412, + "learning_rate": 0.003, + "loss": 3.988, + "step": 21088 + }, + { + "epoch": 0.21089, + "grad_norm": 0.8302724539746508, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 21089 + }, + { + "epoch": 0.2109, + "grad_norm": 0.7611763114376592, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 21090 + }, + { + "epoch": 0.21091, + "grad_norm": 0.6773709541888677, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 21091 + }, + { + "epoch": 0.21092, + "grad_norm": 0.6489695857949672, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 21092 + }, + { + "epoch": 0.21093, + "grad_norm": 0.5985841464830493, + "learning_rate": 0.003, + "loss": 4.017, + "step": 21093 + }, + { + "epoch": 0.21094, + "grad_norm": 0.6686921638728043, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 21094 + }, + { + "epoch": 0.21095, + "grad_norm": 0.8242628144442832, + "learning_rate": 0.003, + "loss": 4.015, + "step": 21095 + }, + { + "epoch": 0.21096, + "grad_norm": 0.8364418051651534, + "learning_rate": 0.003, + "loss": 3.99, + "step": 21096 + }, + { + "epoch": 0.21097, + "grad_norm": 0.7776573604260649, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 21097 + }, + { + "epoch": 0.21098, + "grad_norm": 0.7722065446393532, + "learning_rate": 0.003, + "loss": 4.002, + "step": 21098 + }, + { + "epoch": 0.21099, + "grad_norm": 0.8051359753566685, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 21099 + }, + { + "epoch": 0.211, + "grad_norm": 0.8404930333316829, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 21100 + }, + { + "epoch": 0.21101, + "grad_norm": 0.7594342674451281, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 21101 + }, + { + "epoch": 0.21102, + "grad_norm": 0.8294586505312483, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 21102 + }, + { + "epoch": 0.21103, + "grad_norm": 0.732838027546723, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 21103 + }, + { + "epoch": 0.21104, + "grad_norm": 0.7315336583973028, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 21104 + }, + { + "epoch": 0.21105, + "grad_norm": 0.7829920905328737, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 21105 + }, + { + "epoch": 0.21106, + "grad_norm": 0.7598384682976986, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 21106 + }, + { + "epoch": 0.21107, + "grad_norm": 0.6864155703910533, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 21107 + }, + { + "epoch": 0.21108, + "grad_norm": 0.7444217656357819, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 21108 + }, + { + "epoch": 0.21109, + "grad_norm": 0.7159119612898976, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 21109 + }, + { + "epoch": 0.2111, + "grad_norm": 0.7727545522910183, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 21110 + }, + { + "epoch": 0.21111, + "grad_norm": 0.8008561295805986, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 21111 + }, + { + "epoch": 0.21112, + "grad_norm": 0.6751968863043512, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 21112 + }, + { + "epoch": 0.21113, + "grad_norm": 0.7198696288994434, + "learning_rate": 0.003, + "loss": 4.034, + "step": 21113 + }, + { + "epoch": 0.21114, + "grad_norm": 0.7297958189316107, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 21114 + }, + { + "epoch": 0.21115, + "grad_norm": 0.7753618760056464, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 21115 + }, + { + "epoch": 0.21116, + "grad_norm": 0.92384625273085, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 21116 + }, + { + "epoch": 0.21117, + "grad_norm": 1.0313426298183233, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 21117 + }, + { + "epoch": 0.21118, + "grad_norm": 1.0369835456511376, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 21118 + }, + { + "epoch": 0.21119, + "grad_norm": 1.001179464409606, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 21119 + }, + { + "epoch": 0.2112, + "grad_norm": 0.83253474245824, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 21120 + }, + { + "epoch": 0.21121, + "grad_norm": 0.7706054284116322, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 21121 + }, + { + "epoch": 0.21122, + "grad_norm": 0.7718502105865878, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 21122 + }, + { + "epoch": 0.21123, + "grad_norm": 0.7426717667588533, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 21123 + }, + { + "epoch": 0.21124, + "grad_norm": 0.8051493317967281, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 21124 + }, + { + "epoch": 0.21125, + "grad_norm": 0.8597476988530678, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 21125 + }, + { + "epoch": 0.21126, + "grad_norm": 0.8217628105722796, + "learning_rate": 0.003, + "loss": 4.025, + "step": 21126 + }, + { + "epoch": 0.21127, + "grad_norm": 0.7724536365387886, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 21127 + }, + { + "epoch": 0.21128, + "grad_norm": 0.7209562567331936, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 21128 + }, + { + "epoch": 0.21129, + "grad_norm": 0.6495773354874671, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 21129 + }, + { + "epoch": 0.2113, + "grad_norm": 0.6116795467970129, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 21130 + }, + { + "epoch": 0.21131, + "grad_norm": 0.622447561004982, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 21131 + }, + { + "epoch": 0.21132, + "grad_norm": 0.6457765963593725, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 21132 + }, + { + "epoch": 0.21133, + "grad_norm": 0.6677615113183154, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 21133 + }, + { + "epoch": 0.21134, + "grad_norm": 0.673464426754585, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 21134 + }, + { + "epoch": 0.21135, + "grad_norm": 0.7779037803317472, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 21135 + }, + { + "epoch": 0.21136, + "grad_norm": 0.9182473863337539, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 21136 + }, + { + "epoch": 0.21137, + "grad_norm": 1.0480974379194172, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 21137 + }, + { + "epoch": 0.21138, + "grad_norm": 0.8176705094335274, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 21138 + }, + { + "epoch": 0.21139, + "grad_norm": 0.6848970054152724, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 21139 + }, + { + "epoch": 0.2114, + "grad_norm": 0.7253305589793063, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 21140 + }, + { + "epoch": 0.21141, + "grad_norm": 0.7969919451386863, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 21141 + }, + { + "epoch": 0.21142, + "grad_norm": 0.7518065099864636, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 21142 + }, + { + "epoch": 0.21143, + "grad_norm": 0.5889582968927505, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 21143 + }, + { + "epoch": 0.21144, + "grad_norm": 0.6077187689414575, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 21144 + }, + { + "epoch": 0.21145, + "grad_norm": 0.6148319735149551, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 21145 + }, + { + "epoch": 0.21146, + "grad_norm": 0.5728990759131777, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 21146 + }, + { + "epoch": 0.21147, + "grad_norm": 0.5449866042402753, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 21147 + }, + { + "epoch": 0.21148, + "grad_norm": 0.4945052134217132, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 21148 + }, + { + "epoch": 0.21149, + "grad_norm": 0.5278165951196229, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 21149 + }, + { + "epoch": 0.2115, + "grad_norm": 0.5589772222661037, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 21150 + }, + { + "epoch": 0.21151, + "grad_norm": 0.6188913017608378, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 21151 + }, + { + "epoch": 0.21152, + "grad_norm": 0.6415013873471072, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 21152 + }, + { + "epoch": 0.21153, + "grad_norm": 0.6304551493049014, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 21153 + }, + { + "epoch": 0.21154, + "grad_norm": 0.5774281135653314, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 21154 + }, + { + "epoch": 0.21155, + "grad_norm": 0.6022620739552237, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 21155 + }, + { + "epoch": 0.21156, + "grad_norm": 0.6588024547196103, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 21156 + }, + { + "epoch": 0.21157, + "grad_norm": 0.7136002868026131, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 21157 + }, + { + "epoch": 0.21158, + "grad_norm": 0.6834206083782766, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 21158 + }, + { + "epoch": 0.21159, + "grad_norm": 0.7479545999698405, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 21159 + }, + { + "epoch": 0.2116, + "grad_norm": 0.7658191158831663, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 21160 + }, + { + "epoch": 0.21161, + "grad_norm": 0.8838425132311322, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 21161 + }, + { + "epoch": 0.21162, + "grad_norm": 0.8901956818039152, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 21162 + }, + { + "epoch": 0.21163, + "grad_norm": 0.7870792994753886, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 21163 + }, + { + "epoch": 0.21164, + "grad_norm": 0.8252518608523257, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 21164 + }, + { + "epoch": 0.21165, + "grad_norm": 1.0741049537988587, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 21165 + }, + { + "epoch": 0.21166, + "grad_norm": 1.1758359493288368, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 21166 + }, + { + "epoch": 0.21167, + "grad_norm": 0.921412137035485, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 21167 + }, + { + "epoch": 0.21168, + "grad_norm": 0.8575468519586663, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 21168 + }, + { + "epoch": 0.21169, + "grad_norm": 0.8300438516399384, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 21169 + }, + { + "epoch": 0.2117, + "grad_norm": 0.8035770876361386, + "learning_rate": 0.003, + "loss": 4.011, + "step": 21170 + }, + { + "epoch": 0.21171, + "grad_norm": 0.8027261747695508, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 21171 + }, + { + "epoch": 0.21172, + "grad_norm": 0.7967771109217966, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 21172 + }, + { + "epoch": 0.21173, + "grad_norm": 0.7665931682759033, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 21173 + }, + { + "epoch": 0.21174, + "grad_norm": 0.8094896005333072, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 21174 + }, + { + "epoch": 0.21175, + "grad_norm": 0.7608420438097473, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 21175 + }, + { + "epoch": 0.21176, + "grad_norm": 0.8040394045595575, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 21176 + }, + { + "epoch": 0.21177, + "grad_norm": 0.8183810912023611, + "learning_rate": 0.003, + "loss": 3.998, + "step": 21177 + }, + { + "epoch": 0.21178, + "grad_norm": 0.6995795283142019, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 21178 + }, + { + "epoch": 0.21179, + "grad_norm": 0.6314441045442178, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 21179 + }, + { + "epoch": 0.2118, + "grad_norm": 0.6597094863097918, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 21180 + }, + { + "epoch": 0.21181, + "grad_norm": 0.5971314831674602, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 21181 + }, + { + "epoch": 0.21182, + "grad_norm": 0.6551926725474634, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 21182 + }, + { + "epoch": 0.21183, + "grad_norm": 0.6974563154181969, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 21183 + }, + { + "epoch": 0.21184, + "grad_norm": 0.7673179053050981, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 21184 + }, + { + "epoch": 0.21185, + "grad_norm": 0.7301663633592467, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 21185 + }, + { + "epoch": 0.21186, + "grad_norm": 0.6739073978174709, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 21186 + }, + { + "epoch": 0.21187, + "grad_norm": 0.7051849381324293, + "learning_rate": 0.003, + "loss": 3.986, + "step": 21187 + }, + { + "epoch": 0.21188, + "grad_norm": 0.7701419964582039, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 21188 + }, + { + "epoch": 0.21189, + "grad_norm": 0.8136978512396181, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 21189 + }, + { + "epoch": 0.2119, + "grad_norm": 0.7958517828207269, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 21190 + }, + { + "epoch": 0.21191, + "grad_norm": 0.6695954592559675, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 21191 + }, + { + "epoch": 0.21192, + "grad_norm": 0.7416239936348183, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 21192 + }, + { + "epoch": 0.21193, + "grad_norm": 0.8682143379952193, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 21193 + }, + { + "epoch": 0.21194, + "grad_norm": 1.1107736257183565, + "learning_rate": 0.003, + "loss": 4.044, + "step": 21194 + }, + { + "epoch": 0.21195, + "grad_norm": 0.8392272301195406, + "learning_rate": 0.003, + "loss": 4.008, + "step": 21195 + }, + { + "epoch": 0.21196, + "grad_norm": 0.7702457970286006, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 21196 + }, + { + "epoch": 0.21197, + "grad_norm": 0.7366012192620823, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 21197 + }, + { + "epoch": 0.21198, + "grad_norm": 0.7028979183044922, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 21198 + }, + { + "epoch": 0.21199, + "grad_norm": 0.6890285495383338, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 21199 + }, + { + "epoch": 0.212, + "grad_norm": 0.7280641489395326, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 21200 + }, + { + "epoch": 0.21201, + "grad_norm": 0.753327330205989, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 21201 + }, + { + "epoch": 0.21202, + "grad_norm": 0.7261936639416036, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 21202 + }, + { + "epoch": 0.21203, + "grad_norm": 0.7181292540619717, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 21203 + }, + { + "epoch": 0.21204, + "grad_norm": 0.7678435577348738, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 21204 + }, + { + "epoch": 0.21205, + "grad_norm": 0.92346759621294, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 21205 + }, + { + "epoch": 0.21206, + "grad_norm": 0.9149440844265848, + "learning_rate": 0.003, + "loss": 4.015, + "step": 21206 + }, + { + "epoch": 0.21207, + "grad_norm": 0.9021919757065507, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 21207 + }, + { + "epoch": 0.21208, + "grad_norm": 0.8389966178044302, + "learning_rate": 0.003, + "loss": 4.0422, + "step": 21208 + }, + { + "epoch": 0.21209, + "grad_norm": 0.8810383928104675, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 21209 + }, + { + "epoch": 0.2121, + "grad_norm": 1.0306345131231522, + "learning_rate": 0.003, + "loss": 4.022, + "step": 21210 + }, + { + "epoch": 0.21211, + "grad_norm": 1.188385668985935, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 21211 + }, + { + "epoch": 0.21212, + "grad_norm": 0.6844403303107398, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 21212 + }, + { + "epoch": 0.21213, + "grad_norm": 0.5930132234102086, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 21213 + }, + { + "epoch": 0.21214, + "grad_norm": 0.6660756734616493, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 21214 + }, + { + "epoch": 0.21215, + "grad_norm": 0.7132172898201817, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 21215 + }, + { + "epoch": 0.21216, + "grad_norm": 0.7908408429573668, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 21216 + }, + { + "epoch": 0.21217, + "grad_norm": 0.9229704209443196, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 21217 + }, + { + "epoch": 0.21218, + "grad_norm": 0.9442148070478474, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 21218 + }, + { + "epoch": 0.21219, + "grad_norm": 0.7330120492691109, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 21219 + }, + { + "epoch": 0.2122, + "grad_norm": 0.6779904721735232, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 21220 + }, + { + "epoch": 0.21221, + "grad_norm": 0.681343208973799, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 21221 + }, + { + "epoch": 0.21222, + "grad_norm": 0.5416817524148476, + "learning_rate": 0.003, + "loss": 3.992, + "step": 21222 + }, + { + "epoch": 0.21223, + "grad_norm": 0.5087732468087656, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 21223 + }, + { + "epoch": 0.21224, + "grad_norm": 0.5308559567084177, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 21224 + }, + { + "epoch": 0.21225, + "grad_norm": 0.4881842512213147, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 21225 + }, + { + "epoch": 0.21226, + "grad_norm": 0.4595455560616375, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 21226 + }, + { + "epoch": 0.21227, + "grad_norm": 0.5157037996309854, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 21227 + }, + { + "epoch": 0.21228, + "grad_norm": 0.5889660290694342, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 21228 + }, + { + "epoch": 0.21229, + "grad_norm": 0.7223971170163143, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 21229 + }, + { + "epoch": 0.2123, + "grad_norm": 0.8235193104255707, + "learning_rate": 0.003, + "loss": 3.992, + "step": 21230 + }, + { + "epoch": 0.21231, + "grad_norm": 0.9253319303087225, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 21231 + }, + { + "epoch": 0.21232, + "grad_norm": 1.0070462943773488, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 21232 + }, + { + "epoch": 0.21233, + "grad_norm": 0.8594581897411523, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 21233 + }, + { + "epoch": 0.21234, + "grad_norm": 0.7303941774164513, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 21234 + }, + { + "epoch": 0.21235, + "grad_norm": 0.7399487350013746, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 21235 + }, + { + "epoch": 0.21236, + "grad_norm": 0.7207233076403369, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 21236 + }, + { + "epoch": 0.21237, + "grad_norm": 0.7289514334158066, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 21237 + }, + { + "epoch": 0.21238, + "grad_norm": 0.7845517267158967, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 21238 + }, + { + "epoch": 0.21239, + "grad_norm": 0.7396560129288562, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 21239 + }, + { + "epoch": 0.2124, + "grad_norm": 0.5672098794355709, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 21240 + }, + { + "epoch": 0.21241, + "grad_norm": 0.4708033906679925, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 21241 + }, + { + "epoch": 0.21242, + "grad_norm": 0.5159348470737366, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 21242 + }, + { + "epoch": 0.21243, + "grad_norm": 0.43998815027478144, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 21243 + }, + { + "epoch": 0.21244, + "grad_norm": 0.5483463582858307, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 21244 + }, + { + "epoch": 0.21245, + "grad_norm": 0.6177465666198227, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 21245 + }, + { + "epoch": 0.21246, + "grad_norm": 0.6604348810706319, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 21246 + }, + { + "epoch": 0.21247, + "grad_norm": 0.6143382887641559, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 21247 + }, + { + "epoch": 0.21248, + "grad_norm": 0.6960378071696819, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 21248 + }, + { + "epoch": 0.21249, + "grad_norm": 0.8277922913448131, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 21249 + }, + { + "epoch": 0.2125, + "grad_norm": 0.9430393660372661, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 21250 + }, + { + "epoch": 0.21251, + "grad_norm": 0.8646429025868304, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 21251 + }, + { + "epoch": 0.21252, + "grad_norm": 0.7657880657127946, + "learning_rate": 0.003, + "loss": 4.015, + "step": 21252 + }, + { + "epoch": 0.21253, + "grad_norm": 0.7468715324444968, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 21253 + }, + { + "epoch": 0.21254, + "grad_norm": 0.7652571177423412, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 21254 + }, + { + "epoch": 0.21255, + "grad_norm": 0.7449713086197498, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 21255 + }, + { + "epoch": 0.21256, + "grad_norm": 0.6714456573434074, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 21256 + }, + { + "epoch": 0.21257, + "grad_norm": 0.7066777230515512, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 21257 + }, + { + "epoch": 0.21258, + "grad_norm": 0.7339322437877827, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 21258 + }, + { + "epoch": 0.21259, + "grad_norm": 0.6778441379775698, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 21259 + }, + { + "epoch": 0.2126, + "grad_norm": 0.6978076278987281, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 21260 + }, + { + "epoch": 0.21261, + "grad_norm": 0.7353820572125973, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 21261 + }, + { + "epoch": 0.21262, + "grad_norm": 0.707248942039301, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 21262 + }, + { + "epoch": 0.21263, + "grad_norm": 0.7313157901051507, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 21263 + }, + { + "epoch": 0.21264, + "grad_norm": 0.7614133574429482, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 21264 + }, + { + "epoch": 0.21265, + "grad_norm": 0.7426950765039022, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 21265 + }, + { + "epoch": 0.21266, + "grad_norm": 0.8306139613285717, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 21266 + }, + { + "epoch": 0.21267, + "grad_norm": 1.104890507183012, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 21267 + }, + { + "epoch": 0.21268, + "grad_norm": 0.8911625120775236, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 21268 + }, + { + "epoch": 0.21269, + "grad_norm": 0.8952807581743756, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 21269 + }, + { + "epoch": 0.2127, + "grad_norm": 1.0110745346246293, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 21270 + }, + { + "epoch": 0.21271, + "grad_norm": 1.046103221816639, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 21271 + }, + { + "epoch": 0.21272, + "grad_norm": 0.9502275600990853, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 21272 + }, + { + "epoch": 0.21273, + "grad_norm": 1.0360628292107277, + "learning_rate": 0.003, + "loss": 4.016, + "step": 21273 + }, + { + "epoch": 0.21274, + "grad_norm": 1.0849019290850723, + "learning_rate": 0.003, + "loss": 4.0505, + "step": 21274 + }, + { + "epoch": 0.21275, + "grad_norm": 0.8595575598108344, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 21275 + }, + { + "epoch": 0.21276, + "grad_norm": 0.7206390238974861, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 21276 + }, + { + "epoch": 0.21277, + "grad_norm": 0.6605561572945347, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 21277 + }, + { + "epoch": 0.21278, + "grad_norm": 0.6776063385015252, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 21278 + }, + { + "epoch": 0.21279, + "grad_norm": 0.7054647520180929, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 21279 + }, + { + "epoch": 0.2128, + "grad_norm": 0.7695106991433768, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 21280 + }, + { + "epoch": 0.21281, + "grad_norm": 0.7486872545825607, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 21281 + }, + { + "epoch": 0.21282, + "grad_norm": 0.7880622837154316, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 21282 + }, + { + "epoch": 0.21283, + "grad_norm": 0.8712610763473334, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 21283 + }, + { + "epoch": 0.21284, + "grad_norm": 0.9108278560161395, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 21284 + }, + { + "epoch": 0.21285, + "grad_norm": 1.0548547310123586, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 21285 + }, + { + "epoch": 0.21286, + "grad_norm": 0.7560550505619682, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 21286 + }, + { + "epoch": 0.21287, + "grad_norm": 0.6168607565363047, + "learning_rate": 0.003, + "loss": 3.996, + "step": 21287 + }, + { + "epoch": 0.21288, + "grad_norm": 0.6127883584233202, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 21288 + }, + { + "epoch": 0.21289, + "grad_norm": 0.5943593291619933, + "learning_rate": 0.003, + "loss": 3.997, + "step": 21289 + }, + { + "epoch": 0.2129, + "grad_norm": 0.6356461188032935, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 21290 + }, + { + "epoch": 0.21291, + "grad_norm": 0.6201877528975985, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 21291 + }, + { + "epoch": 0.21292, + "grad_norm": 0.6229672386652408, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 21292 + }, + { + "epoch": 0.21293, + "grad_norm": 0.6884133589788007, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 21293 + }, + { + "epoch": 0.21294, + "grad_norm": 0.7981800830751599, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 21294 + }, + { + "epoch": 0.21295, + "grad_norm": 0.9365333846987801, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 21295 + }, + { + "epoch": 0.21296, + "grad_norm": 0.9243900320284418, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 21296 + }, + { + "epoch": 0.21297, + "grad_norm": 0.7990115159482198, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 21297 + }, + { + "epoch": 0.21298, + "grad_norm": 0.7475042247080013, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 21298 + }, + { + "epoch": 0.21299, + "grad_norm": 0.7355664914901763, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 21299 + }, + { + "epoch": 0.213, + "grad_norm": 0.7017901012959129, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 21300 + }, + { + "epoch": 0.21301, + "grad_norm": 0.6526917179000888, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 21301 + }, + { + "epoch": 0.21302, + "grad_norm": 0.6409160353063567, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 21302 + }, + { + "epoch": 0.21303, + "grad_norm": 0.6289197762163127, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 21303 + }, + { + "epoch": 0.21304, + "grad_norm": 0.5789086873234663, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 21304 + }, + { + "epoch": 0.21305, + "grad_norm": 0.6794466325996396, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 21305 + }, + { + "epoch": 0.21306, + "grad_norm": 0.6974334360553806, + "learning_rate": 0.003, + "loss": 3.993, + "step": 21306 + }, + { + "epoch": 0.21307, + "grad_norm": 0.7636869952234036, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 21307 + }, + { + "epoch": 0.21308, + "grad_norm": 0.6567871225246517, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 21308 + }, + { + "epoch": 0.21309, + "grad_norm": 0.5979913239203345, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 21309 + }, + { + "epoch": 0.2131, + "grad_norm": 0.563397418911059, + "learning_rate": 0.003, + "loss": 3.97, + "step": 21310 + }, + { + "epoch": 0.21311, + "grad_norm": 0.6022926241614336, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 21311 + }, + { + "epoch": 0.21312, + "grad_norm": 0.623917733188052, + "learning_rate": 0.003, + "loss": 3.987, + "step": 21312 + }, + { + "epoch": 0.21313, + "grad_norm": 0.7139008540702251, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 21313 + }, + { + "epoch": 0.21314, + "grad_norm": 0.7699425164979291, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 21314 + }, + { + "epoch": 0.21315, + "grad_norm": 0.765211682072158, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 21315 + }, + { + "epoch": 0.21316, + "grad_norm": 0.6717149900049555, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 21316 + }, + { + "epoch": 0.21317, + "grad_norm": 0.6402899097111917, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 21317 + }, + { + "epoch": 0.21318, + "grad_norm": 0.7167056697915837, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 21318 + }, + { + "epoch": 0.21319, + "grad_norm": 0.7387065913389457, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 21319 + }, + { + "epoch": 0.2132, + "grad_norm": 0.7672531913488818, + "learning_rate": 0.003, + "loss": 4.013, + "step": 21320 + }, + { + "epoch": 0.21321, + "grad_norm": 0.9073724866630101, + "learning_rate": 0.003, + "loss": 3.995, + "step": 21321 + }, + { + "epoch": 0.21322, + "grad_norm": 1.0929211111684272, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 21322 + }, + { + "epoch": 0.21323, + "grad_norm": 0.9964889565795976, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 21323 + }, + { + "epoch": 0.21324, + "grad_norm": 0.9134807600981945, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 21324 + }, + { + "epoch": 0.21325, + "grad_norm": 0.9877599239637012, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 21325 + }, + { + "epoch": 0.21326, + "grad_norm": 0.9336790228108993, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 21326 + }, + { + "epoch": 0.21327, + "grad_norm": 1.0357847314970714, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 21327 + }, + { + "epoch": 0.21328, + "grad_norm": 0.9365042319558649, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 21328 + }, + { + "epoch": 0.21329, + "grad_norm": 0.9248958390780881, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 21329 + }, + { + "epoch": 0.2133, + "grad_norm": 0.8152841685563631, + "learning_rate": 0.003, + "loss": 4.022, + "step": 21330 + }, + { + "epoch": 0.21331, + "grad_norm": 0.8345633928670285, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 21331 + }, + { + "epoch": 0.21332, + "grad_norm": 0.7408349420253109, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 21332 + }, + { + "epoch": 0.21333, + "grad_norm": 0.7162493154080661, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 21333 + }, + { + "epoch": 0.21334, + "grad_norm": 0.7484051780255503, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 21334 + }, + { + "epoch": 0.21335, + "grad_norm": 0.6678109701358339, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 21335 + }, + { + "epoch": 0.21336, + "grad_norm": 0.6657879889490574, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 21336 + }, + { + "epoch": 0.21337, + "grad_norm": 0.6585014756022497, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 21337 + }, + { + "epoch": 0.21338, + "grad_norm": 0.7551116087775288, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 21338 + }, + { + "epoch": 0.21339, + "grad_norm": 0.9035058886997495, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 21339 + }, + { + "epoch": 0.2134, + "grad_norm": 0.9514087869255032, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 21340 + }, + { + "epoch": 0.21341, + "grad_norm": 0.8569587581761868, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 21341 + }, + { + "epoch": 0.21342, + "grad_norm": 0.8243027892806556, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 21342 + }, + { + "epoch": 0.21343, + "grad_norm": 0.8294926737309619, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 21343 + }, + { + "epoch": 0.21344, + "grad_norm": 0.9103329692940378, + "learning_rate": 0.003, + "loss": 4.014, + "step": 21344 + }, + { + "epoch": 0.21345, + "grad_norm": 0.9613973660857908, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 21345 + }, + { + "epoch": 0.21346, + "grad_norm": 0.909465423034283, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 21346 + }, + { + "epoch": 0.21347, + "grad_norm": 0.9498258744271787, + "learning_rate": 0.003, + "loss": 4.032, + "step": 21347 + }, + { + "epoch": 0.21348, + "grad_norm": 0.9349917759006421, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 21348 + }, + { + "epoch": 0.21349, + "grad_norm": 0.840428956054503, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 21349 + }, + { + "epoch": 0.2135, + "grad_norm": 0.9123203551728608, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 21350 + }, + { + "epoch": 0.21351, + "grad_norm": 0.8726195460055507, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 21351 + }, + { + "epoch": 0.21352, + "grad_norm": 0.755379381184055, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 21352 + }, + { + "epoch": 0.21353, + "grad_norm": 0.6265042003608046, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 21353 + }, + { + "epoch": 0.21354, + "grad_norm": 0.604298926911801, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 21354 + }, + { + "epoch": 0.21355, + "grad_norm": 0.5928892106041816, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 21355 + }, + { + "epoch": 0.21356, + "grad_norm": 0.5564567907633999, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 21356 + }, + { + "epoch": 0.21357, + "grad_norm": 0.6605337275044377, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 21357 + }, + { + "epoch": 0.21358, + "grad_norm": 0.6923796904848879, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 21358 + }, + { + "epoch": 0.21359, + "grad_norm": 0.5769380142195534, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 21359 + }, + { + "epoch": 0.2136, + "grad_norm": 0.556430933572112, + "learning_rate": 0.003, + "loss": 3.996, + "step": 21360 + }, + { + "epoch": 0.21361, + "grad_norm": 0.6755910450051472, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 21361 + }, + { + "epoch": 0.21362, + "grad_norm": 0.7352912153553687, + "learning_rate": 0.003, + "loss": 4.008, + "step": 21362 + }, + { + "epoch": 0.21363, + "grad_norm": 0.7711821416170355, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 21363 + }, + { + "epoch": 0.21364, + "grad_norm": 0.8217017965475246, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 21364 + }, + { + "epoch": 0.21365, + "grad_norm": 0.8966740752667176, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 21365 + }, + { + "epoch": 0.21366, + "grad_norm": 0.8682531569990433, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 21366 + }, + { + "epoch": 0.21367, + "grad_norm": 0.6564354653744766, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 21367 + }, + { + "epoch": 0.21368, + "grad_norm": 0.6561938158036852, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 21368 + }, + { + "epoch": 0.21369, + "grad_norm": 0.6637715224263456, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 21369 + }, + { + "epoch": 0.2137, + "grad_norm": 0.7395045260189796, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 21370 + }, + { + "epoch": 0.21371, + "grad_norm": 0.6988255847746381, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 21371 + }, + { + "epoch": 0.21372, + "grad_norm": 0.6583843178828505, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 21372 + }, + { + "epoch": 0.21373, + "grad_norm": 0.6014411977917853, + "learning_rate": 0.003, + "loss": 4.0, + "step": 21373 + }, + { + "epoch": 0.21374, + "grad_norm": 0.6028793729904728, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 21374 + }, + { + "epoch": 0.21375, + "grad_norm": 0.5445101389964084, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 21375 + }, + { + "epoch": 0.21376, + "grad_norm": 0.5458139704119405, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 21376 + }, + { + "epoch": 0.21377, + "grad_norm": 0.6449235950415018, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 21377 + }, + { + "epoch": 0.21378, + "grad_norm": 0.8256823987346648, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 21378 + }, + { + "epoch": 0.21379, + "grad_norm": 0.7864916136248876, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 21379 + }, + { + "epoch": 0.2138, + "grad_norm": 0.6957079308347857, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 21380 + }, + { + "epoch": 0.21381, + "grad_norm": 0.7312796549364278, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 21381 + }, + { + "epoch": 0.21382, + "grad_norm": 0.7006243697276449, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 21382 + }, + { + "epoch": 0.21383, + "grad_norm": 0.7024840129040919, + "learning_rate": 0.003, + "loss": 3.991, + "step": 21383 + }, + { + "epoch": 0.21384, + "grad_norm": 0.6803971607755047, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 21384 + }, + { + "epoch": 0.21385, + "grad_norm": 0.7463134881483733, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 21385 + }, + { + "epoch": 0.21386, + "grad_norm": 0.831724483411426, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 21386 + }, + { + "epoch": 0.21387, + "grad_norm": 0.9887849463465099, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 21387 + }, + { + "epoch": 0.21388, + "grad_norm": 1.1535473951915083, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 21388 + }, + { + "epoch": 0.21389, + "grad_norm": 0.740634468427631, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 21389 + }, + { + "epoch": 0.2139, + "grad_norm": 0.6303197348805434, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 21390 + }, + { + "epoch": 0.21391, + "grad_norm": 0.6177161454163773, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 21391 + }, + { + "epoch": 0.21392, + "grad_norm": 0.6450116028276679, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 21392 + }, + { + "epoch": 0.21393, + "grad_norm": 0.8214795955626432, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 21393 + }, + { + "epoch": 0.21394, + "grad_norm": 0.9419059576975803, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 21394 + }, + { + "epoch": 0.21395, + "grad_norm": 0.9781140548187814, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 21395 + }, + { + "epoch": 0.21396, + "grad_norm": 0.8781879056467814, + "learning_rate": 0.003, + "loss": 3.99, + "step": 21396 + }, + { + "epoch": 0.21397, + "grad_norm": 0.7974107153814656, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 21397 + }, + { + "epoch": 0.21398, + "grad_norm": 0.814494328157504, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 21398 + }, + { + "epoch": 0.21399, + "grad_norm": 0.8286534862923118, + "learning_rate": 0.003, + "loss": 4.001, + "step": 21399 + }, + { + "epoch": 0.214, + "grad_norm": 0.7245896117448004, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 21400 + }, + { + "epoch": 0.21401, + "grad_norm": 0.6836340752115273, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 21401 + }, + { + "epoch": 0.21402, + "grad_norm": 0.7714892993662004, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 21402 + }, + { + "epoch": 0.21403, + "grad_norm": 0.7013153308513087, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 21403 + }, + { + "epoch": 0.21404, + "grad_norm": 0.6482044291749826, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 21404 + }, + { + "epoch": 0.21405, + "grad_norm": 0.6200232723378064, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 21405 + }, + { + "epoch": 0.21406, + "grad_norm": 0.6544267268117995, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 21406 + }, + { + "epoch": 0.21407, + "grad_norm": 0.6086675177189118, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 21407 + }, + { + "epoch": 0.21408, + "grad_norm": 0.616282164789504, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 21408 + }, + { + "epoch": 0.21409, + "grad_norm": 0.5915732966497594, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 21409 + }, + { + "epoch": 0.2141, + "grad_norm": 0.6143732619863317, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 21410 + }, + { + "epoch": 0.21411, + "grad_norm": 0.622163469832088, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 21411 + }, + { + "epoch": 0.21412, + "grad_norm": 0.6594164491594627, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 21412 + }, + { + "epoch": 0.21413, + "grad_norm": 0.6871762287888006, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 21413 + }, + { + "epoch": 0.21414, + "grad_norm": 0.6827355420893858, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 21414 + }, + { + "epoch": 0.21415, + "grad_norm": 0.597955776877428, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 21415 + }, + { + "epoch": 0.21416, + "grad_norm": 0.5139806366603069, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 21416 + }, + { + "epoch": 0.21417, + "grad_norm": 0.4990872266373614, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 21417 + }, + { + "epoch": 0.21418, + "grad_norm": 0.5589842048780835, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 21418 + }, + { + "epoch": 0.21419, + "grad_norm": 0.7774524638458069, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 21419 + }, + { + "epoch": 0.2142, + "grad_norm": 1.1126492635202123, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 21420 + }, + { + "epoch": 0.21421, + "grad_norm": 0.9851203656323791, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 21421 + }, + { + "epoch": 0.21422, + "grad_norm": 0.8356968871646117, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 21422 + }, + { + "epoch": 0.21423, + "grad_norm": 0.8126412837435751, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 21423 + }, + { + "epoch": 0.21424, + "grad_norm": 0.7391517010285852, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 21424 + }, + { + "epoch": 0.21425, + "grad_norm": 0.7465757814569199, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 21425 + }, + { + "epoch": 0.21426, + "grad_norm": 0.8389453997575557, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 21426 + }, + { + "epoch": 0.21427, + "grad_norm": 0.8521607531690488, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 21427 + }, + { + "epoch": 0.21428, + "grad_norm": 0.7572174176272635, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 21428 + }, + { + "epoch": 0.21429, + "grad_norm": 0.8914651105945512, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 21429 + }, + { + "epoch": 0.2143, + "grad_norm": 0.8975814381725932, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 21430 + }, + { + "epoch": 0.21431, + "grad_norm": 0.888724792685067, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 21431 + }, + { + "epoch": 0.21432, + "grad_norm": 0.8916691608182202, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 21432 + }, + { + "epoch": 0.21433, + "grad_norm": 0.7976778438815707, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 21433 + }, + { + "epoch": 0.21434, + "grad_norm": 0.7617657913902505, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 21434 + }, + { + "epoch": 0.21435, + "grad_norm": 0.7836055472537208, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 21435 + }, + { + "epoch": 0.21436, + "grad_norm": 0.8157388642390792, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 21436 + }, + { + "epoch": 0.21437, + "grad_norm": 0.7803967393007577, + "learning_rate": 0.003, + "loss": 3.994, + "step": 21437 + }, + { + "epoch": 0.21438, + "grad_norm": 0.7699526222778034, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 21438 + }, + { + "epoch": 0.21439, + "grad_norm": 0.7181975176138564, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 21439 + }, + { + "epoch": 0.2144, + "grad_norm": 0.8392789915473722, + "learning_rate": 0.003, + "loss": 3.983, + "step": 21440 + }, + { + "epoch": 0.21441, + "grad_norm": 0.9380310553916709, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 21441 + }, + { + "epoch": 0.21442, + "grad_norm": 0.8855426382656504, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 21442 + }, + { + "epoch": 0.21443, + "grad_norm": 0.7564157596430318, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 21443 + }, + { + "epoch": 0.21444, + "grad_norm": 0.7498434507148967, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 21444 + }, + { + "epoch": 0.21445, + "grad_norm": 0.6877113145481485, + "learning_rate": 0.003, + "loss": 4.029, + "step": 21445 + }, + { + "epoch": 0.21446, + "grad_norm": 0.6245261703982543, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 21446 + }, + { + "epoch": 0.21447, + "grad_norm": 0.5951852863241447, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 21447 + }, + { + "epoch": 0.21448, + "grad_norm": 0.6041897138831139, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 21448 + }, + { + "epoch": 0.21449, + "grad_norm": 0.6518301786277225, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 21449 + }, + { + "epoch": 0.2145, + "grad_norm": 0.737419914236776, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 21450 + }, + { + "epoch": 0.21451, + "grad_norm": 0.8141748788369693, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 21451 + }, + { + "epoch": 0.21452, + "grad_norm": 1.0120322972158473, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 21452 + }, + { + "epoch": 0.21453, + "grad_norm": 1.1944406014205826, + "learning_rate": 0.003, + "loss": 4.002, + "step": 21453 + }, + { + "epoch": 0.21454, + "grad_norm": 0.7431284674026994, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 21454 + }, + { + "epoch": 0.21455, + "grad_norm": 0.5811597735977068, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 21455 + }, + { + "epoch": 0.21456, + "grad_norm": 0.5928604013991334, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 21456 + }, + { + "epoch": 0.21457, + "grad_norm": 0.5802218715171527, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 21457 + }, + { + "epoch": 0.21458, + "grad_norm": 0.6060864456990884, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 21458 + }, + { + "epoch": 0.21459, + "grad_norm": 0.6133597894292214, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 21459 + }, + { + "epoch": 0.2146, + "grad_norm": 0.659093054397003, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 21460 + }, + { + "epoch": 0.21461, + "grad_norm": 0.6933736186111963, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 21461 + }, + { + "epoch": 0.21462, + "grad_norm": 0.6683662037537489, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 21462 + }, + { + "epoch": 0.21463, + "grad_norm": 0.7037842057165956, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 21463 + }, + { + "epoch": 0.21464, + "grad_norm": 0.6854933567951138, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 21464 + }, + { + "epoch": 0.21465, + "grad_norm": 0.7432709148996588, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 21465 + }, + { + "epoch": 0.21466, + "grad_norm": 0.9080292335153212, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 21466 + }, + { + "epoch": 0.21467, + "grad_norm": 0.9286952096997215, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 21467 + }, + { + "epoch": 0.21468, + "grad_norm": 0.8331170705540931, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 21468 + }, + { + "epoch": 0.21469, + "grad_norm": 0.6926982460620369, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 21469 + }, + { + "epoch": 0.2147, + "grad_norm": 0.7008330601400681, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 21470 + }, + { + "epoch": 0.21471, + "grad_norm": 0.6994166608632137, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 21471 + }, + { + "epoch": 0.21472, + "grad_norm": 0.7061444028817064, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 21472 + }, + { + "epoch": 0.21473, + "grad_norm": 0.7260803110821386, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 21473 + }, + { + "epoch": 0.21474, + "grad_norm": 0.7475819834949654, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 21474 + }, + { + "epoch": 0.21475, + "grad_norm": 0.7474336071942116, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 21475 + }, + { + "epoch": 0.21476, + "grad_norm": 0.7561028275428602, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 21476 + }, + { + "epoch": 0.21477, + "grad_norm": 0.9542360848353326, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 21477 + }, + { + "epoch": 0.21478, + "grad_norm": 0.8978127833355327, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 21478 + }, + { + "epoch": 0.21479, + "grad_norm": 0.7752100567082332, + "learning_rate": 0.003, + "loss": 3.991, + "step": 21479 + }, + { + "epoch": 0.2148, + "grad_norm": 0.8740672148185527, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 21480 + }, + { + "epoch": 0.21481, + "grad_norm": 0.8092058158555184, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 21481 + }, + { + "epoch": 0.21482, + "grad_norm": 0.8884615415108519, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 21482 + }, + { + "epoch": 0.21483, + "grad_norm": 1.0470158145145092, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 21483 + }, + { + "epoch": 0.21484, + "grad_norm": 0.9887046487498045, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 21484 + }, + { + "epoch": 0.21485, + "grad_norm": 0.9912463933901394, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 21485 + }, + { + "epoch": 0.21486, + "grad_norm": 0.9997561562161101, + "learning_rate": 0.003, + "loss": 4.0468, + "step": 21486 + }, + { + "epoch": 0.21487, + "grad_norm": 1.0450423468484908, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 21487 + }, + { + "epoch": 0.21488, + "grad_norm": 0.8952384323638121, + "learning_rate": 0.003, + "loss": 4.0541, + "step": 21488 + }, + { + "epoch": 0.21489, + "grad_norm": 0.8527356365467927, + "learning_rate": 0.003, + "loss": 4.018, + "step": 21489 + }, + { + "epoch": 0.2149, + "grad_norm": 0.8651171639238671, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 21490 + }, + { + "epoch": 0.21491, + "grad_norm": 0.7597239645627192, + "learning_rate": 0.003, + "loss": 4.057, + "step": 21491 + }, + { + "epoch": 0.21492, + "grad_norm": 0.6679547600843685, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 21492 + }, + { + "epoch": 0.21493, + "grad_norm": 0.6472723065859465, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 21493 + }, + { + "epoch": 0.21494, + "grad_norm": 0.6879948748999399, + "learning_rate": 0.003, + "loss": 4.0508, + "step": 21494 + }, + { + "epoch": 0.21495, + "grad_norm": 0.6665932421842824, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 21495 + }, + { + "epoch": 0.21496, + "grad_norm": 0.7072292744277681, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 21496 + }, + { + "epoch": 0.21497, + "grad_norm": 0.6378668228407942, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 21497 + }, + { + "epoch": 0.21498, + "grad_norm": 0.5185995949278933, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 21498 + }, + { + "epoch": 0.21499, + "grad_norm": 0.5313767045811006, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 21499 + }, + { + "epoch": 0.215, + "grad_norm": 0.48849631656426146, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 21500 + }, + { + "epoch": 0.21501, + "grad_norm": 0.528037046900229, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 21501 + }, + { + "epoch": 0.21502, + "grad_norm": 0.5531711574923928, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 21502 + }, + { + "epoch": 0.21503, + "grad_norm": 0.5777281868566417, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 21503 + }, + { + "epoch": 0.21504, + "grad_norm": 0.5695710959962891, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 21504 + }, + { + "epoch": 0.21505, + "grad_norm": 0.5396226521493964, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 21505 + }, + { + "epoch": 0.21506, + "grad_norm": 0.5213650929510825, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 21506 + }, + { + "epoch": 0.21507, + "grad_norm": 0.49651917523850225, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 21507 + }, + { + "epoch": 0.21508, + "grad_norm": 0.47729759210664513, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 21508 + }, + { + "epoch": 0.21509, + "grad_norm": 0.4635090651828998, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 21509 + }, + { + "epoch": 0.2151, + "grad_norm": 0.4764658704573784, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 21510 + }, + { + "epoch": 0.21511, + "grad_norm": 0.48848869488282387, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 21511 + }, + { + "epoch": 0.21512, + "grad_norm": 0.5896650652042125, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 21512 + }, + { + "epoch": 0.21513, + "grad_norm": 0.782559478841353, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 21513 + }, + { + "epoch": 0.21514, + "grad_norm": 1.107375707917504, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 21514 + }, + { + "epoch": 0.21515, + "grad_norm": 1.1267562087728504, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 21515 + }, + { + "epoch": 0.21516, + "grad_norm": 0.9075095140535459, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 21516 + }, + { + "epoch": 0.21517, + "grad_norm": 0.7123021394755708, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 21517 + }, + { + "epoch": 0.21518, + "grad_norm": 0.6306437988468173, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 21518 + }, + { + "epoch": 0.21519, + "grad_norm": 0.6696517927769143, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 21519 + }, + { + "epoch": 0.2152, + "grad_norm": 0.6782290751702004, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 21520 + }, + { + "epoch": 0.21521, + "grad_norm": 0.6888027695439122, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 21521 + }, + { + "epoch": 0.21522, + "grad_norm": 0.8018773745446304, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 21522 + }, + { + "epoch": 0.21523, + "grad_norm": 0.8426483556681027, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 21523 + }, + { + "epoch": 0.21524, + "grad_norm": 0.7737465558517277, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 21524 + }, + { + "epoch": 0.21525, + "grad_norm": 0.7521887546647562, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 21525 + }, + { + "epoch": 0.21526, + "grad_norm": 0.7522728242439886, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 21526 + }, + { + "epoch": 0.21527, + "grad_norm": 0.6656980111679176, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 21527 + }, + { + "epoch": 0.21528, + "grad_norm": 0.6430763240594448, + "learning_rate": 0.003, + "loss": 3.975, + "step": 21528 + }, + { + "epoch": 0.21529, + "grad_norm": 0.6817506076947449, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 21529 + }, + { + "epoch": 0.2153, + "grad_norm": 0.6924867689190332, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 21530 + }, + { + "epoch": 0.21531, + "grad_norm": 0.8167673972991514, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 21531 + }, + { + "epoch": 0.21532, + "grad_norm": 0.9524539438726665, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 21532 + }, + { + "epoch": 0.21533, + "grad_norm": 1.1553086434004967, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 21533 + }, + { + "epoch": 0.21534, + "grad_norm": 0.6638121603901922, + "learning_rate": 0.003, + "loss": 3.98, + "step": 21534 + }, + { + "epoch": 0.21535, + "grad_norm": 0.7694283331521827, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 21535 + }, + { + "epoch": 0.21536, + "grad_norm": 0.7936903309933651, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 21536 + }, + { + "epoch": 0.21537, + "grad_norm": 0.884035673434823, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 21537 + }, + { + "epoch": 0.21538, + "grad_norm": 0.8955900007887853, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 21538 + }, + { + "epoch": 0.21539, + "grad_norm": 0.7742256039673923, + "learning_rate": 0.003, + "loss": 4.013, + "step": 21539 + }, + { + "epoch": 0.2154, + "grad_norm": 0.7840964339533569, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 21540 + }, + { + "epoch": 0.21541, + "grad_norm": 0.8290007420224704, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 21541 + }, + { + "epoch": 0.21542, + "grad_norm": 0.8270033436084461, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 21542 + }, + { + "epoch": 0.21543, + "grad_norm": 0.7164788703003422, + "learning_rate": 0.003, + "loss": 3.994, + "step": 21543 + }, + { + "epoch": 0.21544, + "grad_norm": 0.7539036352698089, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 21544 + }, + { + "epoch": 0.21545, + "grad_norm": 0.7928470428147327, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 21545 + }, + { + "epoch": 0.21546, + "grad_norm": 0.6876327769567612, + "learning_rate": 0.003, + "loss": 3.992, + "step": 21546 + }, + { + "epoch": 0.21547, + "grad_norm": 0.6071101469266635, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 21547 + }, + { + "epoch": 0.21548, + "grad_norm": 0.6836153151664736, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 21548 + }, + { + "epoch": 0.21549, + "grad_norm": 0.7110818284279884, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 21549 + }, + { + "epoch": 0.2155, + "grad_norm": 0.7040347973183946, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 21550 + }, + { + "epoch": 0.21551, + "grad_norm": 0.614495066725397, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 21551 + }, + { + "epoch": 0.21552, + "grad_norm": 0.6768857341706879, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 21552 + }, + { + "epoch": 0.21553, + "grad_norm": 0.8351075215463933, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 21553 + }, + { + "epoch": 0.21554, + "grad_norm": 0.997911323375554, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 21554 + }, + { + "epoch": 0.21555, + "grad_norm": 1.115741392252001, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 21555 + }, + { + "epoch": 0.21556, + "grad_norm": 0.8127206896673635, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 21556 + }, + { + "epoch": 0.21557, + "grad_norm": 0.6602112839155009, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 21557 + }, + { + "epoch": 0.21558, + "grad_norm": 0.7055786794423771, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 21558 + }, + { + "epoch": 0.21559, + "grad_norm": 0.7922725719358769, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 21559 + }, + { + "epoch": 0.2156, + "grad_norm": 0.9664966745360075, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 21560 + }, + { + "epoch": 0.21561, + "grad_norm": 0.9682927665326504, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 21561 + }, + { + "epoch": 0.21562, + "grad_norm": 0.8762452523644506, + "learning_rate": 0.003, + "loss": 4.033, + "step": 21562 + }, + { + "epoch": 0.21563, + "grad_norm": 0.7892392370905956, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 21563 + }, + { + "epoch": 0.21564, + "grad_norm": 0.7735613479265238, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 21564 + }, + { + "epoch": 0.21565, + "grad_norm": 0.6955962556641533, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 21565 + }, + { + "epoch": 0.21566, + "grad_norm": 0.7616418741976831, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 21566 + }, + { + "epoch": 0.21567, + "grad_norm": 0.7910264764717788, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 21567 + }, + { + "epoch": 0.21568, + "grad_norm": 0.760731940380159, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 21568 + }, + { + "epoch": 0.21569, + "grad_norm": 0.8746223759205888, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 21569 + }, + { + "epoch": 0.2157, + "grad_norm": 1.0531323148711496, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 21570 + }, + { + "epoch": 0.21571, + "grad_norm": 0.8214272916760731, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 21571 + }, + { + "epoch": 0.21572, + "grad_norm": 0.8036910918721004, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 21572 + }, + { + "epoch": 0.21573, + "grad_norm": 0.6852715195450121, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 21573 + }, + { + "epoch": 0.21574, + "grad_norm": 0.7177217982147056, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 21574 + }, + { + "epoch": 0.21575, + "grad_norm": 0.7615288453710415, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 21575 + }, + { + "epoch": 0.21576, + "grad_norm": 0.7670240042952748, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 21576 + }, + { + "epoch": 0.21577, + "grad_norm": 0.6893281035493783, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 21577 + }, + { + "epoch": 0.21578, + "grad_norm": 0.6778967449798808, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 21578 + }, + { + "epoch": 0.21579, + "grad_norm": 0.6764950719784444, + "learning_rate": 0.003, + "loss": 4.016, + "step": 21579 + }, + { + "epoch": 0.2158, + "grad_norm": 0.6069700372190271, + "learning_rate": 0.003, + "loss": 3.969, + "step": 21580 + }, + { + "epoch": 0.21581, + "grad_norm": 0.6556227576121595, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 21581 + }, + { + "epoch": 0.21582, + "grad_norm": 0.7646073951032787, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 21582 + }, + { + "epoch": 0.21583, + "grad_norm": 0.7284369709316454, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 21583 + }, + { + "epoch": 0.21584, + "grad_norm": 0.692593089862824, + "learning_rate": 0.003, + "loss": 4.001, + "step": 21584 + }, + { + "epoch": 0.21585, + "grad_norm": 0.7249035651257733, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 21585 + }, + { + "epoch": 0.21586, + "grad_norm": 0.756448638228449, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 21586 + }, + { + "epoch": 0.21587, + "grad_norm": 0.719273939034454, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 21587 + }, + { + "epoch": 0.21588, + "grad_norm": 0.6942947436205668, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 21588 + }, + { + "epoch": 0.21589, + "grad_norm": 0.6391367462872162, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 21589 + }, + { + "epoch": 0.2159, + "grad_norm": 0.6488610796751291, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 21590 + }, + { + "epoch": 0.21591, + "grad_norm": 0.6875675180608958, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 21591 + }, + { + "epoch": 0.21592, + "grad_norm": 0.6717451036244216, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 21592 + }, + { + "epoch": 0.21593, + "grad_norm": 0.6840008242017912, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 21593 + }, + { + "epoch": 0.21594, + "grad_norm": 0.6977277808153366, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 21594 + }, + { + "epoch": 0.21595, + "grad_norm": 0.6817117656763371, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 21595 + }, + { + "epoch": 0.21596, + "grad_norm": 0.6878876396514134, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 21596 + }, + { + "epoch": 0.21597, + "grad_norm": 0.6100894169775118, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 21597 + }, + { + "epoch": 0.21598, + "grad_norm": 0.612715323035732, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 21598 + }, + { + "epoch": 0.21599, + "grad_norm": 0.6805869575308489, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 21599 + }, + { + "epoch": 0.216, + "grad_norm": 0.7217692000342041, + "learning_rate": 0.003, + "loss": 4.016, + "step": 21600 + }, + { + "epoch": 0.21601, + "grad_norm": 0.7945710578401475, + "learning_rate": 0.003, + "loss": 3.988, + "step": 21601 + }, + { + "epoch": 0.21602, + "grad_norm": 0.8220774642743315, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 21602 + }, + { + "epoch": 0.21603, + "grad_norm": 0.942050541699878, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 21603 + }, + { + "epoch": 0.21604, + "grad_norm": 1.1243779445105908, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 21604 + }, + { + "epoch": 0.21605, + "grad_norm": 0.8542929447729344, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 21605 + }, + { + "epoch": 0.21606, + "grad_norm": 0.5170835373004551, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 21606 + }, + { + "epoch": 0.21607, + "grad_norm": 0.6958861503620332, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 21607 + }, + { + "epoch": 0.21608, + "grad_norm": 0.7975050251056339, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 21608 + }, + { + "epoch": 0.21609, + "grad_norm": 0.8210638104443414, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 21609 + }, + { + "epoch": 0.2161, + "grad_norm": 0.731027042105596, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 21610 + }, + { + "epoch": 0.21611, + "grad_norm": 0.7089563632413539, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 21611 + }, + { + "epoch": 0.21612, + "grad_norm": 0.7070323588891132, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 21612 + }, + { + "epoch": 0.21613, + "grad_norm": 0.735341461922287, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 21613 + }, + { + "epoch": 0.21614, + "grad_norm": 0.721535939919631, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 21614 + }, + { + "epoch": 0.21615, + "grad_norm": 0.759735573877958, + "learning_rate": 0.003, + "loss": 4.008, + "step": 21615 + }, + { + "epoch": 0.21616, + "grad_norm": 0.8158362119057468, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 21616 + }, + { + "epoch": 0.21617, + "grad_norm": 0.8165632217343818, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 21617 + }, + { + "epoch": 0.21618, + "grad_norm": 0.7348000991858579, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 21618 + }, + { + "epoch": 0.21619, + "grad_norm": 0.6569107199314151, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 21619 + }, + { + "epoch": 0.2162, + "grad_norm": 0.7325630813294093, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 21620 + }, + { + "epoch": 0.21621, + "grad_norm": 0.6893574660330517, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 21621 + }, + { + "epoch": 0.21622, + "grad_norm": 0.7159338620421937, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 21622 + }, + { + "epoch": 0.21623, + "grad_norm": 0.8185518348513684, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 21623 + }, + { + "epoch": 0.21624, + "grad_norm": 0.9779952946749295, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 21624 + }, + { + "epoch": 0.21625, + "grad_norm": 1.0835769617528639, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 21625 + }, + { + "epoch": 0.21626, + "grad_norm": 1.171965225413626, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 21626 + }, + { + "epoch": 0.21627, + "grad_norm": 0.871483887575747, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 21627 + }, + { + "epoch": 0.21628, + "grad_norm": 0.6948520829679857, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 21628 + }, + { + "epoch": 0.21629, + "grad_norm": 0.6439474280738605, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 21629 + }, + { + "epoch": 0.2163, + "grad_norm": 0.6331394725743058, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 21630 + }, + { + "epoch": 0.21631, + "grad_norm": 0.6392111834371074, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 21631 + }, + { + "epoch": 0.21632, + "grad_norm": 0.6148270486443549, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 21632 + }, + { + "epoch": 0.21633, + "grad_norm": 0.7429210269304053, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 21633 + }, + { + "epoch": 0.21634, + "grad_norm": 0.8147991569609716, + "learning_rate": 0.003, + "loss": 4.013, + "step": 21634 + }, + { + "epoch": 0.21635, + "grad_norm": 0.9511501047045148, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 21635 + }, + { + "epoch": 0.21636, + "grad_norm": 0.9651599623215174, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 21636 + }, + { + "epoch": 0.21637, + "grad_norm": 0.6784990439618269, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 21637 + }, + { + "epoch": 0.21638, + "grad_norm": 0.585836433280809, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 21638 + }, + { + "epoch": 0.21639, + "grad_norm": 0.6515623252419034, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 21639 + }, + { + "epoch": 0.2164, + "grad_norm": 0.7131225987062734, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 21640 + }, + { + "epoch": 0.21641, + "grad_norm": 0.7076292157366296, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 21641 + }, + { + "epoch": 0.21642, + "grad_norm": 0.610842975454428, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 21642 + }, + { + "epoch": 0.21643, + "grad_norm": 0.557942514132383, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 21643 + }, + { + "epoch": 0.21644, + "grad_norm": 0.6055125431681079, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 21644 + }, + { + "epoch": 0.21645, + "grad_norm": 0.5636207384380704, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 21645 + }, + { + "epoch": 0.21646, + "grad_norm": 0.5898178507514017, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 21646 + }, + { + "epoch": 0.21647, + "grad_norm": 0.5735239604703515, + "learning_rate": 0.003, + "loss": 3.978, + "step": 21647 + }, + { + "epoch": 0.21648, + "grad_norm": 0.6537586098045489, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 21648 + }, + { + "epoch": 0.21649, + "grad_norm": 0.7790859337582634, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 21649 + }, + { + "epoch": 0.2165, + "grad_norm": 1.0321905903904771, + "learning_rate": 0.003, + "loss": 4.019, + "step": 21650 + }, + { + "epoch": 0.21651, + "grad_norm": 1.104194805179721, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 21651 + }, + { + "epoch": 0.21652, + "grad_norm": 0.8808623689001287, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 21652 + }, + { + "epoch": 0.21653, + "grad_norm": 0.782368827651221, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 21653 + }, + { + "epoch": 0.21654, + "grad_norm": 0.6930044763831639, + "learning_rate": 0.003, + "loss": 3.97, + "step": 21654 + }, + { + "epoch": 0.21655, + "grad_norm": 0.7379683489325976, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 21655 + }, + { + "epoch": 0.21656, + "grad_norm": 0.824378599660733, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 21656 + }, + { + "epoch": 0.21657, + "grad_norm": 0.8167311875109109, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 21657 + }, + { + "epoch": 0.21658, + "grad_norm": 0.8308767310445476, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 21658 + }, + { + "epoch": 0.21659, + "grad_norm": 0.8469855789237823, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 21659 + }, + { + "epoch": 0.2166, + "grad_norm": 0.8433540889522783, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 21660 + }, + { + "epoch": 0.21661, + "grad_norm": 0.7778453622364689, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 21661 + }, + { + "epoch": 0.21662, + "grad_norm": 0.7728830005142033, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 21662 + }, + { + "epoch": 0.21663, + "grad_norm": 0.8277161582222299, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 21663 + }, + { + "epoch": 0.21664, + "grad_norm": 0.8792656826405111, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 21664 + }, + { + "epoch": 0.21665, + "grad_norm": 0.8174533726416973, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 21665 + }, + { + "epoch": 0.21666, + "grad_norm": 0.849517010320521, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 21666 + }, + { + "epoch": 0.21667, + "grad_norm": 0.7229870731993556, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 21667 + }, + { + "epoch": 0.21668, + "grad_norm": 0.6824825351221495, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 21668 + }, + { + "epoch": 0.21669, + "grad_norm": 0.7401136272485658, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 21669 + }, + { + "epoch": 0.2167, + "grad_norm": 0.8625292167363892, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 21670 + }, + { + "epoch": 0.21671, + "grad_norm": 0.988596315108751, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 21671 + }, + { + "epoch": 0.21672, + "grad_norm": 0.8717580408450686, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 21672 + }, + { + "epoch": 0.21673, + "grad_norm": 0.6427273154681948, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 21673 + }, + { + "epoch": 0.21674, + "grad_norm": 0.6621891052362727, + "learning_rate": 0.003, + "loss": 3.964, + "step": 21674 + }, + { + "epoch": 0.21675, + "grad_norm": 0.5890862120860105, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 21675 + }, + { + "epoch": 0.21676, + "grad_norm": 0.6069770900400154, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 21676 + }, + { + "epoch": 0.21677, + "grad_norm": 0.7857952308006538, + "learning_rate": 0.003, + "loss": 4.01, + "step": 21677 + }, + { + "epoch": 0.21678, + "grad_norm": 0.9420218274175279, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 21678 + }, + { + "epoch": 0.21679, + "grad_norm": 0.8925773668977259, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 21679 + }, + { + "epoch": 0.2168, + "grad_norm": 0.7963229228513289, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 21680 + }, + { + "epoch": 0.21681, + "grad_norm": 0.8740383626885583, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 21681 + }, + { + "epoch": 0.21682, + "grad_norm": 0.828427568143939, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 21682 + }, + { + "epoch": 0.21683, + "grad_norm": 0.8021044978714346, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 21683 + }, + { + "epoch": 0.21684, + "grad_norm": 0.7929263490912243, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 21684 + }, + { + "epoch": 0.21685, + "grad_norm": 0.7600552986754624, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 21685 + }, + { + "epoch": 0.21686, + "grad_norm": 0.7043368929675088, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 21686 + }, + { + "epoch": 0.21687, + "grad_norm": 0.6432459490182717, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 21687 + }, + { + "epoch": 0.21688, + "grad_norm": 0.6480919684424485, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 21688 + }, + { + "epoch": 0.21689, + "grad_norm": 0.7203214094359257, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 21689 + }, + { + "epoch": 0.2169, + "grad_norm": 0.7811359813773434, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 21690 + }, + { + "epoch": 0.21691, + "grad_norm": 0.6675964281571053, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 21691 + }, + { + "epoch": 0.21692, + "grad_norm": 0.6438838184228153, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 21692 + }, + { + "epoch": 0.21693, + "grad_norm": 0.7297877885618557, + "learning_rate": 0.003, + "loss": 3.989, + "step": 21693 + }, + { + "epoch": 0.21694, + "grad_norm": 0.718661886253666, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 21694 + }, + { + "epoch": 0.21695, + "grad_norm": 0.6060218644964609, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 21695 + }, + { + "epoch": 0.21696, + "grad_norm": 0.5612863394373355, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 21696 + }, + { + "epoch": 0.21697, + "grad_norm": 0.5120695005425213, + "learning_rate": 0.003, + "loss": 3.996, + "step": 21697 + }, + { + "epoch": 0.21698, + "grad_norm": 0.5886579638297156, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 21698 + }, + { + "epoch": 0.21699, + "grad_norm": 0.610765647872828, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 21699 + }, + { + "epoch": 0.217, + "grad_norm": 0.8551969313192836, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 21700 + }, + { + "epoch": 0.21701, + "grad_norm": 1.1668881384335883, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 21701 + }, + { + "epoch": 0.21702, + "grad_norm": 0.912056507494989, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 21702 + }, + { + "epoch": 0.21703, + "grad_norm": 0.6702653932855787, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 21703 + }, + { + "epoch": 0.21704, + "grad_norm": 0.7390069895964669, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 21704 + }, + { + "epoch": 0.21705, + "grad_norm": 0.7296156907889095, + "learning_rate": 0.003, + "loss": 3.978, + "step": 21705 + }, + { + "epoch": 0.21706, + "grad_norm": 0.7010287923326077, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 21706 + }, + { + "epoch": 0.21707, + "grad_norm": 0.7099178733323979, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 21707 + }, + { + "epoch": 0.21708, + "grad_norm": 0.653891084132593, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 21708 + }, + { + "epoch": 0.21709, + "grad_norm": 0.6199503849880239, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 21709 + }, + { + "epoch": 0.2171, + "grad_norm": 0.6852183887291864, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 21710 + }, + { + "epoch": 0.21711, + "grad_norm": 0.6399030519618577, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 21711 + }, + { + "epoch": 0.21712, + "grad_norm": 0.6267637868579334, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 21712 + }, + { + "epoch": 0.21713, + "grad_norm": 0.6027851243550085, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 21713 + }, + { + "epoch": 0.21714, + "grad_norm": 0.6171472573228935, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 21714 + }, + { + "epoch": 0.21715, + "grad_norm": 0.6746665752459146, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 21715 + }, + { + "epoch": 0.21716, + "grad_norm": 0.7171905644719286, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 21716 + }, + { + "epoch": 0.21717, + "grad_norm": 0.7144579789531911, + "learning_rate": 0.003, + "loss": 3.988, + "step": 21717 + }, + { + "epoch": 0.21718, + "grad_norm": 0.8190329189119133, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 21718 + }, + { + "epoch": 0.21719, + "grad_norm": 0.9385973638685896, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 21719 + }, + { + "epoch": 0.2172, + "grad_norm": 1.0171117653793484, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 21720 + }, + { + "epoch": 0.21721, + "grad_norm": 0.878576845627285, + "learning_rate": 0.003, + "loss": 4.019, + "step": 21721 + }, + { + "epoch": 0.21722, + "grad_norm": 0.8373304282332827, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 21722 + }, + { + "epoch": 0.21723, + "grad_norm": 0.895116725781596, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 21723 + }, + { + "epoch": 0.21724, + "grad_norm": 0.8896843852125389, + "learning_rate": 0.003, + "loss": 4.02, + "step": 21724 + }, + { + "epoch": 0.21725, + "grad_norm": 0.8644648487853722, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 21725 + }, + { + "epoch": 0.21726, + "grad_norm": 0.9010596359923649, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 21726 + }, + { + "epoch": 0.21727, + "grad_norm": 0.9030450729595231, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 21727 + }, + { + "epoch": 0.21728, + "grad_norm": 0.9638079671756375, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 21728 + }, + { + "epoch": 0.21729, + "grad_norm": 0.928136687459103, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 21729 + }, + { + "epoch": 0.2173, + "grad_norm": 0.979476129559744, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 21730 + }, + { + "epoch": 0.21731, + "grad_norm": 1.0269798683845002, + "learning_rate": 0.003, + "loss": 4.0667, + "step": 21731 + }, + { + "epoch": 0.21732, + "grad_norm": 0.8811629426395404, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 21732 + }, + { + "epoch": 0.21733, + "grad_norm": 1.0439233893924105, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 21733 + }, + { + "epoch": 0.21734, + "grad_norm": 1.000451370036425, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 21734 + }, + { + "epoch": 0.21735, + "grad_norm": 1.1112682110325547, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 21735 + }, + { + "epoch": 0.21736, + "grad_norm": 1.0387813159573893, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 21736 + }, + { + "epoch": 0.21737, + "grad_norm": 0.8883988717808624, + "learning_rate": 0.003, + "loss": 4.003, + "step": 21737 + }, + { + "epoch": 0.21738, + "grad_norm": 0.8816489015162695, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 21738 + }, + { + "epoch": 0.21739, + "grad_norm": 0.9348189507760241, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 21739 + }, + { + "epoch": 0.2174, + "grad_norm": 0.9372118164639226, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 21740 + }, + { + "epoch": 0.21741, + "grad_norm": 0.8150339990145105, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 21741 + }, + { + "epoch": 0.21742, + "grad_norm": 0.7515804304546506, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 21742 + }, + { + "epoch": 0.21743, + "grad_norm": 0.6599573807182835, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 21743 + }, + { + "epoch": 0.21744, + "grad_norm": 0.6210066722209658, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 21744 + }, + { + "epoch": 0.21745, + "grad_norm": 0.6706621801273136, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 21745 + }, + { + "epoch": 0.21746, + "grad_norm": 0.7211527434965096, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 21746 + }, + { + "epoch": 0.21747, + "grad_norm": 0.7856829065137747, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 21747 + }, + { + "epoch": 0.21748, + "grad_norm": 0.8117088560285227, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 21748 + }, + { + "epoch": 0.21749, + "grad_norm": 0.6635933863699999, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 21749 + }, + { + "epoch": 0.2175, + "grad_norm": 0.5995876843426163, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 21750 + }, + { + "epoch": 0.21751, + "grad_norm": 0.5800380094118307, + "learning_rate": 0.003, + "loss": 3.966, + "step": 21751 + }, + { + "epoch": 0.21752, + "grad_norm": 0.5272871052929949, + "learning_rate": 0.003, + "loss": 4.011, + "step": 21752 + }, + { + "epoch": 0.21753, + "grad_norm": 0.4393139597327437, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 21753 + }, + { + "epoch": 0.21754, + "grad_norm": 0.3718205489013745, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 21754 + }, + { + "epoch": 0.21755, + "grad_norm": 0.4286543109592621, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 21755 + }, + { + "epoch": 0.21756, + "grad_norm": 0.4762703118716191, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 21756 + }, + { + "epoch": 0.21757, + "grad_norm": 0.48900172434483846, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 21757 + }, + { + "epoch": 0.21758, + "grad_norm": 0.5849036438928071, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 21758 + }, + { + "epoch": 0.21759, + "grad_norm": 0.7359341785189705, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 21759 + }, + { + "epoch": 0.2176, + "grad_norm": 0.8117807336600311, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 21760 + }, + { + "epoch": 0.21761, + "grad_norm": 0.7308466681345688, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 21761 + }, + { + "epoch": 0.21762, + "grad_norm": 0.6135896020141592, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 21762 + }, + { + "epoch": 0.21763, + "grad_norm": 0.6481552863014581, + "learning_rate": 0.003, + "loss": 4.003, + "step": 21763 + }, + { + "epoch": 0.21764, + "grad_norm": 0.6503015117811051, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 21764 + }, + { + "epoch": 0.21765, + "grad_norm": 0.6846047847480842, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 21765 + }, + { + "epoch": 0.21766, + "grad_norm": 0.7093598306529065, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 21766 + }, + { + "epoch": 0.21767, + "grad_norm": 0.7476733713162782, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 21767 + }, + { + "epoch": 0.21768, + "grad_norm": 0.7147305223812175, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 21768 + }, + { + "epoch": 0.21769, + "grad_norm": 0.7433360597235835, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 21769 + }, + { + "epoch": 0.2177, + "grad_norm": 0.698054512588601, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 21770 + }, + { + "epoch": 0.21771, + "grad_norm": 0.6396738754325878, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 21771 + }, + { + "epoch": 0.21772, + "grad_norm": 0.60746452694258, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 21772 + }, + { + "epoch": 0.21773, + "grad_norm": 0.6717188408662759, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 21773 + }, + { + "epoch": 0.21774, + "grad_norm": 0.7058852244760118, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 21774 + }, + { + "epoch": 0.21775, + "grad_norm": 0.8091959121627437, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 21775 + }, + { + "epoch": 0.21776, + "grad_norm": 0.9530180973939335, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 21776 + }, + { + "epoch": 0.21777, + "grad_norm": 0.9529125520677728, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 21777 + }, + { + "epoch": 0.21778, + "grad_norm": 0.9294487155630934, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 21778 + }, + { + "epoch": 0.21779, + "grad_norm": 0.832279241816824, + "learning_rate": 0.003, + "loss": 3.986, + "step": 21779 + }, + { + "epoch": 0.2178, + "grad_norm": 0.7908931831228343, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 21780 + }, + { + "epoch": 0.21781, + "grad_norm": 0.7978953512202644, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 21781 + }, + { + "epoch": 0.21782, + "grad_norm": 0.8601042845487374, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 21782 + }, + { + "epoch": 0.21783, + "grad_norm": 0.7415543804377646, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 21783 + }, + { + "epoch": 0.21784, + "grad_norm": 0.6363327481507788, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 21784 + }, + { + "epoch": 0.21785, + "grad_norm": 0.7765511500716158, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 21785 + }, + { + "epoch": 0.21786, + "grad_norm": 0.7790314700561501, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 21786 + }, + { + "epoch": 0.21787, + "grad_norm": 0.8816716754660373, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 21787 + }, + { + "epoch": 0.21788, + "grad_norm": 0.96316779099801, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 21788 + }, + { + "epoch": 0.21789, + "grad_norm": 0.8845630292027794, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 21789 + }, + { + "epoch": 0.2179, + "grad_norm": 1.0875642947293511, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 21790 + }, + { + "epoch": 0.21791, + "grad_norm": 0.9145782431111136, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 21791 + }, + { + "epoch": 0.21792, + "grad_norm": 0.8045383468597533, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 21792 + }, + { + "epoch": 0.21793, + "grad_norm": 0.7749978468519683, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 21793 + }, + { + "epoch": 0.21794, + "grad_norm": 0.9003568757640362, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 21794 + }, + { + "epoch": 0.21795, + "grad_norm": 0.7821014458261848, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 21795 + }, + { + "epoch": 0.21796, + "grad_norm": 0.6122971179902659, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 21796 + }, + { + "epoch": 0.21797, + "grad_norm": 0.6806188757230788, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 21797 + }, + { + "epoch": 0.21798, + "grad_norm": 0.6691574724270393, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 21798 + }, + { + "epoch": 0.21799, + "grad_norm": 0.8205810630530649, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 21799 + }, + { + "epoch": 0.218, + "grad_norm": 0.9194598211371346, + "learning_rate": 0.003, + "loss": 3.998, + "step": 21800 + }, + { + "epoch": 0.21801, + "grad_norm": 0.78041280272627, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 21801 + }, + { + "epoch": 0.21802, + "grad_norm": 0.8283686843484511, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 21802 + }, + { + "epoch": 0.21803, + "grad_norm": 0.8379765061185434, + "learning_rate": 0.003, + "loss": 4.03, + "step": 21803 + }, + { + "epoch": 0.21804, + "grad_norm": 0.7113832577530264, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 21804 + }, + { + "epoch": 0.21805, + "grad_norm": 0.6921555309851021, + "learning_rate": 0.003, + "loss": 4.009, + "step": 21805 + }, + { + "epoch": 0.21806, + "grad_norm": 0.8156695666696614, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 21806 + }, + { + "epoch": 0.21807, + "grad_norm": 0.9385898793113794, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 21807 + }, + { + "epoch": 0.21808, + "grad_norm": 0.7775601480919994, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 21808 + }, + { + "epoch": 0.21809, + "grad_norm": 0.7417114686969032, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 21809 + }, + { + "epoch": 0.2181, + "grad_norm": 0.6769115965352293, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 21810 + }, + { + "epoch": 0.21811, + "grad_norm": 0.5987998059962728, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 21811 + }, + { + "epoch": 0.21812, + "grad_norm": 0.5873412365692523, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 21812 + }, + { + "epoch": 0.21813, + "grad_norm": 0.5231361737954519, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 21813 + }, + { + "epoch": 0.21814, + "grad_norm": 0.5967609396327181, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 21814 + }, + { + "epoch": 0.21815, + "grad_norm": 0.6702108342668778, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 21815 + }, + { + "epoch": 0.21816, + "grad_norm": 0.6821764716473675, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 21816 + }, + { + "epoch": 0.21817, + "grad_norm": 0.7084153793034993, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 21817 + }, + { + "epoch": 0.21818, + "grad_norm": 0.868266673811618, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 21818 + }, + { + "epoch": 0.21819, + "grad_norm": 0.9787327635690813, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 21819 + }, + { + "epoch": 0.2182, + "grad_norm": 0.9659249720718754, + "learning_rate": 0.003, + "loss": 4.005, + "step": 21820 + }, + { + "epoch": 0.21821, + "grad_norm": 0.946839189866061, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 21821 + }, + { + "epoch": 0.21822, + "grad_norm": 0.8265758491201397, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 21822 + }, + { + "epoch": 0.21823, + "grad_norm": 0.7161422479071406, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 21823 + }, + { + "epoch": 0.21824, + "grad_norm": 0.5884462013179229, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 21824 + }, + { + "epoch": 0.21825, + "grad_norm": 0.5267383981302302, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 21825 + }, + { + "epoch": 0.21826, + "grad_norm": 0.5532124632827365, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 21826 + }, + { + "epoch": 0.21827, + "grad_norm": 0.5249769901384809, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 21827 + }, + { + "epoch": 0.21828, + "grad_norm": 0.5375544044025743, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 21828 + }, + { + "epoch": 0.21829, + "grad_norm": 0.5884668850897531, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 21829 + }, + { + "epoch": 0.2183, + "grad_norm": 0.5858990110288422, + "learning_rate": 0.003, + "loss": 3.997, + "step": 21830 + }, + { + "epoch": 0.21831, + "grad_norm": 0.5481007844250675, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 21831 + }, + { + "epoch": 0.21832, + "grad_norm": 0.5084209015756598, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 21832 + }, + { + "epoch": 0.21833, + "grad_norm": 0.5619784895628318, + "learning_rate": 0.003, + "loss": 3.974, + "step": 21833 + }, + { + "epoch": 0.21834, + "grad_norm": 0.608588765261337, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 21834 + }, + { + "epoch": 0.21835, + "grad_norm": 0.610902436890255, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 21835 + }, + { + "epoch": 0.21836, + "grad_norm": 0.6616185390096542, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 21836 + }, + { + "epoch": 0.21837, + "grad_norm": 0.7769722588820848, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 21837 + }, + { + "epoch": 0.21838, + "grad_norm": 1.0072599833215505, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 21838 + }, + { + "epoch": 0.21839, + "grad_norm": 1.1329009138171646, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 21839 + }, + { + "epoch": 0.2184, + "grad_norm": 0.7204227220699175, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 21840 + }, + { + "epoch": 0.21841, + "grad_norm": 0.63551249792915, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 21841 + }, + { + "epoch": 0.21842, + "grad_norm": 0.6707966251615486, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 21842 + }, + { + "epoch": 0.21843, + "grad_norm": 0.7056431865933539, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 21843 + }, + { + "epoch": 0.21844, + "grad_norm": 0.7280915031710001, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 21844 + }, + { + "epoch": 0.21845, + "grad_norm": 0.7648089555444523, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 21845 + }, + { + "epoch": 0.21846, + "grad_norm": 0.8136168251411183, + "learning_rate": 0.003, + "loss": 3.977, + "step": 21846 + }, + { + "epoch": 0.21847, + "grad_norm": 0.9716764468111123, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 21847 + }, + { + "epoch": 0.21848, + "grad_norm": 1.0599001723700503, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 21848 + }, + { + "epoch": 0.21849, + "grad_norm": 0.7978573244678877, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 21849 + }, + { + "epoch": 0.2185, + "grad_norm": 0.8333705042433998, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 21850 + }, + { + "epoch": 0.21851, + "grad_norm": 0.9782096056607302, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 21851 + }, + { + "epoch": 0.21852, + "grad_norm": 0.984819857885364, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 21852 + }, + { + "epoch": 0.21853, + "grad_norm": 0.9396514716436675, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 21853 + }, + { + "epoch": 0.21854, + "grad_norm": 0.8346187590468996, + "learning_rate": 0.003, + "loss": 4.026, + "step": 21854 + }, + { + "epoch": 0.21855, + "grad_norm": 0.7899569797674056, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 21855 + }, + { + "epoch": 0.21856, + "grad_norm": 0.939846251675608, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 21856 + }, + { + "epoch": 0.21857, + "grad_norm": 1.1216447743149163, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 21857 + }, + { + "epoch": 0.21858, + "grad_norm": 0.9346285846027325, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 21858 + }, + { + "epoch": 0.21859, + "grad_norm": 0.8113381238461671, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 21859 + }, + { + "epoch": 0.2186, + "grad_norm": 0.6948822332904309, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 21860 + }, + { + "epoch": 0.21861, + "grad_norm": 0.6434692489238262, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 21861 + }, + { + "epoch": 0.21862, + "grad_norm": 0.7051235699159036, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 21862 + }, + { + "epoch": 0.21863, + "grad_norm": 0.7533107565515389, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 21863 + }, + { + "epoch": 0.21864, + "grad_norm": 0.8475540153327126, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 21864 + }, + { + "epoch": 0.21865, + "grad_norm": 0.9378475690183927, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 21865 + }, + { + "epoch": 0.21866, + "grad_norm": 0.8667528685314546, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 21866 + }, + { + "epoch": 0.21867, + "grad_norm": 0.7139919421699278, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 21867 + }, + { + "epoch": 0.21868, + "grad_norm": 0.67516225223979, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 21868 + }, + { + "epoch": 0.21869, + "grad_norm": 0.5955250103915073, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 21869 + }, + { + "epoch": 0.2187, + "grad_norm": 0.5469743641559364, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 21870 + }, + { + "epoch": 0.21871, + "grad_norm": 0.5686547977937167, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 21871 + }, + { + "epoch": 0.21872, + "grad_norm": 0.6856967579589992, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 21872 + }, + { + "epoch": 0.21873, + "grad_norm": 0.7168560248088387, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 21873 + }, + { + "epoch": 0.21874, + "grad_norm": 0.8331879952305444, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 21874 + }, + { + "epoch": 0.21875, + "grad_norm": 0.8013461591703351, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 21875 + }, + { + "epoch": 0.21876, + "grad_norm": 0.8331711269118282, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 21876 + }, + { + "epoch": 0.21877, + "grad_norm": 0.8577492839044096, + "learning_rate": 0.003, + "loss": 4.03, + "step": 21877 + }, + { + "epoch": 0.21878, + "grad_norm": 0.829029848071099, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 21878 + }, + { + "epoch": 0.21879, + "grad_norm": 0.7091270465192222, + "learning_rate": 0.003, + "loss": 4.004, + "step": 21879 + }, + { + "epoch": 0.2188, + "grad_norm": 0.6933314401803585, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 21880 + }, + { + "epoch": 0.21881, + "grad_norm": 0.6270809970288287, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 21881 + }, + { + "epoch": 0.21882, + "grad_norm": 0.6291395128374053, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 21882 + }, + { + "epoch": 0.21883, + "grad_norm": 0.6651148546788267, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 21883 + }, + { + "epoch": 0.21884, + "grad_norm": 0.8120782254148103, + "learning_rate": 0.003, + "loss": 4.017, + "step": 21884 + }, + { + "epoch": 0.21885, + "grad_norm": 0.8794097174085892, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 21885 + }, + { + "epoch": 0.21886, + "grad_norm": 0.9022085931889837, + "learning_rate": 0.003, + "loss": 4.0625, + "step": 21886 + }, + { + "epoch": 0.21887, + "grad_norm": 0.8814846664571767, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 21887 + }, + { + "epoch": 0.21888, + "grad_norm": 0.8460249536952954, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 21888 + }, + { + "epoch": 0.21889, + "grad_norm": 0.7696062195254149, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 21889 + }, + { + "epoch": 0.2189, + "grad_norm": 0.6013819776773771, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 21890 + }, + { + "epoch": 0.21891, + "grad_norm": 0.5851824880600511, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 21891 + }, + { + "epoch": 0.21892, + "grad_norm": 0.5994180834446364, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 21892 + }, + { + "epoch": 0.21893, + "grad_norm": 0.6481894311381435, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 21893 + }, + { + "epoch": 0.21894, + "grad_norm": 0.6660395989880424, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 21894 + }, + { + "epoch": 0.21895, + "grad_norm": 0.6475225956877946, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 21895 + }, + { + "epoch": 0.21896, + "grad_norm": 0.6659536955486085, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 21896 + }, + { + "epoch": 0.21897, + "grad_norm": 0.6895470848097377, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 21897 + }, + { + "epoch": 0.21898, + "grad_norm": 0.7532676979185, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 21898 + }, + { + "epoch": 0.21899, + "grad_norm": 0.8287148001549334, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 21899 + }, + { + "epoch": 0.219, + "grad_norm": 0.796019677400102, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 21900 + }, + { + "epoch": 0.21901, + "grad_norm": 0.7729772879149098, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 21901 + }, + { + "epoch": 0.21902, + "grad_norm": 0.6837734621302204, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 21902 + }, + { + "epoch": 0.21903, + "grad_norm": 0.6757590122100317, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 21903 + }, + { + "epoch": 0.21904, + "grad_norm": 0.6756519738557059, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 21904 + }, + { + "epoch": 0.21905, + "grad_norm": 0.661243033730855, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 21905 + }, + { + "epoch": 0.21906, + "grad_norm": 0.6774717220797228, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 21906 + }, + { + "epoch": 0.21907, + "grad_norm": 0.6852543873984404, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 21907 + }, + { + "epoch": 0.21908, + "grad_norm": 0.7175678299294688, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 21908 + }, + { + "epoch": 0.21909, + "grad_norm": 0.6787470441097162, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 21909 + }, + { + "epoch": 0.2191, + "grad_norm": 0.5705896530639447, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 21910 + }, + { + "epoch": 0.21911, + "grad_norm": 0.5817916557895816, + "learning_rate": 0.003, + "loss": 4.001, + "step": 21911 + }, + { + "epoch": 0.21912, + "grad_norm": 0.5802299027188056, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 21912 + }, + { + "epoch": 0.21913, + "grad_norm": 0.60383590107956, + "learning_rate": 0.003, + "loss": 3.992, + "step": 21913 + }, + { + "epoch": 0.21914, + "grad_norm": 0.5746429914409639, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 21914 + }, + { + "epoch": 0.21915, + "grad_norm": 0.6081939191683663, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 21915 + }, + { + "epoch": 0.21916, + "grad_norm": 0.74583333582337, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 21916 + }, + { + "epoch": 0.21917, + "grad_norm": 0.8482063702240438, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 21917 + }, + { + "epoch": 0.21918, + "grad_norm": 0.8701851713081165, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 21918 + }, + { + "epoch": 0.21919, + "grad_norm": 1.066527425170686, + "learning_rate": 0.003, + "loss": 4.0372, + "step": 21919 + }, + { + "epoch": 0.2192, + "grad_norm": 1.1201044022955988, + "learning_rate": 0.003, + "loss": 3.996, + "step": 21920 + }, + { + "epoch": 0.21921, + "grad_norm": 0.9600802917127672, + "learning_rate": 0.003, + "loss": 3.984, + "step": 21921 + }, + { + "epoch": 0.21922, + "grad_norm": 0.8251474509482762, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 21922 + }, + { + "epoch": 0.21923, + "grad_norm": 0.8099788065932878, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 21923 + }, + { + "epoch": 0.21924, + "grad_norm": 0.8808816671377209, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 21924 + }, + { + "epoch": 0.21925, + "grad_norm": 0.8966238949639297, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 21925 + }, + { + "epoch": 0.21926, + "grad_norm": 0.9323663480200468, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 21926 + }, + { + "epoch": 0.21927, + "grad_norm": 0.9588390225722895, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 21927 + }, + { + "epoch": 0.21928, + "grad_norm": 0.8890044408442661, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 21928 + }, + { + "epoch": 0.21929, + "grad_norm": 0.9157230173250038, + "learning_rate": 0.003, + "loss": 4.047, + "step": 21929 + }, + { + "epoch": 0.2193, + "grad_norm": 0.9145759116519353, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 21930 + }, + { + "epoch": 0.21931, + "grad_norm": 0.8861704926166584, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 21931 + }, + { + "epoch": 0.21932, + "grad_norm": 0.9667364932064969, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 21932 + }, + { + "epoch": 0.21933, + "grad_norm": 0.9434819957347423, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 21933 + }, + { + "epoch": 0.21934, + "grad_norm": 0.8629578577661056, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 21934 + }, + { + "epoch": 0.21935, + "grad_norm": 0.8038974463005455, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 21935 + }, + { + "epoch": 0.21936, + "grad_norm": 0.7903182570102606, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 21936 + }, + { + "epoch": 0.21937, + "grad_norm": 0.7719756176338793, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 21937 + }, + { + "epoch": 0.21938, + "grad_norm": 0.7347162315699853, + "learning_rate": 0.003, + "loss": 4.0529, + "step": 21938 + }, + { + "epoch": 0.21939, + "grad_norm": 0.7685534275406767, + "learning_rate": 0.003, + "loss": 4.008, + "step": 21939 + }, + { + "epoch": 0.2194, + "grad_norm": 0.7261515531803734, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 21940 + }, + { + "epoch": 0.21941, + "grad_norm": 0.7317917387384663, + "learning_rate": 0.003, + "loss": 4.0471, + "step": 21941 + }, + { + "epoch": 0.21942, + "grad_norm": 0.7508834475745741, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 21942 + }, + { + "epoch": 0.21943, + "grad_norm": 0.656457839672159, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 21943 + }, + { + "epoch": 0.21944, + "grad_norm": 0.5678130281378471, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 21944 + }, + { + "epoch": 0.21945, + "grad_norm": 0.5644224098619586, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 21945 + }, + { + "epoch": 0.21946, + "grad_norm": 0.4940949463662646, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 21946 + }, + { + "epoch": 0.21947, + "grad_norm": 0.5339334334941607, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 21947 + }, + { + "epoch": 0.21948, + "grad_norm": 0.571243543881665, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 21948 + }, + { + "epoch": 0.21949, + "grad_norm": 0.7340756687173055, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 21949 + }, + { + "epoch": 0.2195, + "grad_norm": 0.8592824217097081, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 21950 + }, + { + "epoch": 0.21951, + "grad_norm": 0.9849474208833018, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 21951 + }, + { + "epoch": 0.21952, + "grad_norm": 0.9138648220779905, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 21952 + }, + { + "epoch": 0.21953, + "grad_norm": 0.7638557500103293, + "learning_rate": 0.003, + "loss": 3.999, + "step": 21953 + }, + { + "epoch": 0.21954, + "grad_norm": 0.7825394392987655, + "learning_rate": 0.003, + "loss": 4.009, + "step": 21954 + }, + { + "epoch": 0.21955, + "grad_norm": 0.85557427322968, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 21955 + }, + { + "epoch": 0.21956, + "grad_norm": 0.8187058936810524, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 21956 + }, + { + "epoch": 0.21957, + "grad_norm": 0.7101030998326401, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 21957 + }, + { + "epoch": 0.21958, + "grad_norm": 0.7332891595725219, + "learning_rate": 0.003, + "loss": 3.993, + "step": 21958 + }, + { + "epoch": 0.21959, + "grad_norm": 0.6461671806843587, + "learning_rate": 0.003, + "loss": 3.989, + "step": 21959 + }, + { + "epoch": 0.2196, + "grad_norm": 0.6479135155704888, + "learning_rate": 0.003, + "loss": 4.005, + "step": 21960 + }, + { + "epoch": 0.21961, + "grad_norm": 0.7180451235747973, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 21961 + }, + { + "epoch": 0.21962, + "grad_norm": 0.7643718405485888, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 21962 + }, + { + "epoch": 0.21963, + "grad_norm": 0.7168770255323187, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 21963 + }, + { + "epoch": 0.21964, + "grad_norm": 0.7658474402404071, + "learning_rate": 0.003, + "loss": 3.976, + "step": 21964 + }, + { + "epoch": 0.21965, + "grad_norm": 0.76086197126003, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 21965 + }, + { + "epoch": 0.21966, + "grad_norm": 0.7455857394602551, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 21966 + }, + { + "epoch": 0.21967, + "grad_norm": 0.6737398452158008, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 21967 + }, + { + "epoch": 0.21968, + "grad_norm": 0.5963771244431072, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 21968 + }, + { + "epoch": 0.21969, + "grad_norm": 0.5504842368153144, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 21969 + }, + { + "epoch": 0.2197, + "grad_norm": 0.5091577711944508, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 21970 + }, + { + "epoch": 0.21971, + "grad_norm": 0.5721002648389835, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 21971 + }, + { + "epoch": 0.21972, + "grad_norm": 0.5518643335151154, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 21972 + }, + { + "epoch": 0.21973, + "grad_norm": 0.6075896508331591, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 21973 + }, + { + "epoch": 0.21974, + "grad_norm": 0.6158613751964627, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 21974 + }, + { + "epoch": 0.21975, + "grad_norm": 0.6209667032021473, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 21975 + }, + { + "epoch": 0.21976, + "grad_norm": 0.7005505820491263, + "learning_rate": 0.003, + "loss": 3.99, + "step": 21976 + }, + { + "epoch": 0.21977, + "grad_norm": 0.7924734835505968, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 21977 + }, + { + "epoch": 0.21978, + "grad_norm": 1.0545379814473055, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 21978 + }, + { + "epoch": 0.21979, + "grad_norm": 0.9435477964321273, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 21979 + }, + { + "epoch": 0.2198, + "grad_norm": 0.7720375895972718, + "learning_rate": 0.003, + "loss": 3.98, + "step": 21980 + }, + { + "epoch": 0.21981, + "grad_norm": 0.5936589064477622, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 21981 + }, + { + "epoch": 0.21982, + "grad_norm": 0.6467164366844106, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 21982 + }, + { + "epoch": 0.21983, + "grad_norm": 0.7024667248536377, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 21983 + }, + { + "epoch": 0.21984, + "grad_norm": 0.8365080308678443, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 21984 + }, + { + "epoch": 0.21985, + "grad_norm": 0.9552721202859967, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 21985 + }, + { + "epoch": 0.21986, + "grad_norm": 0.9938109214998275, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 21986 + }, + { + "epoch": 0.21987, + "grad_norm": 1.0272902016120682, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 21987 + }, + { + "epoch": 0.21988, + "grad_norm": 0.8771650254592713, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 21988 + }, + { + "epoch": 0.21989, + "grad_norm": 0.9347235802371398, + "learning_rate": 0.003, + "loss": 4.025, + "step": 21989 + }, + { + "epoch": 0.2199, + "grad_norm": 0.7365721503149512, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 21990 + }, + { + "epoch": 0.21991, + "grad_norm": 0.6711538483804355, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 21991 + }, + { + "epoch": 0.21992, + "grad_norm": 0.5882906760054394, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 21992 + }, + { + "epoch": 0.21993, + "grad_norm": 0.7568175144029445, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 21993 + }, + { + "epoch": 0.21994, + "grad_norm": 0.7327599967851494, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 21994 + }, + { + "epoch": 0.21995, + "grad_norm": 0.7841187060588687, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 21995 + }, + { + "epoch": 0.21996, + "grad_norm": 0.9012761187597955, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 21996 + }, + { + "epoch": 0.21997, + "grad_norm": 0.9121941318156871, + "learning_rate": 0.003, + "loss": 4.03, + "step": 21997 + }, + { + "epoch": 0.21998, + "grad_norm": 0.7038867967216959, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 21998 + }, + { + "epoch": 0.21999, + "grad_norm": 0.5906663970905183, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 21999 + }, + { + "epoch": 0.22, + "grad_norm": 0.695808999054914, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 22000 + }, + { + "epoch": 0.22001, + "grad_norm": 0.8440455997801025, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 22001 + }, + { + "epoch": 0.22002, + "grad_norm": 0.8453964458870865, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 22002 + }, + { + "epoch": 0.22003, + "grad_norm": 0.7430757879660453, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 22003 + }, + { + "epoch": 0.22004, + "grad_norm": 0.7967445247436853, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 22004 + }, + { + "epoch": 0.22005, + "grad_norm": 0.8200639931219765, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 22005 + }, + { + "epoch": 0.22006, + "grad_norm": 0.7824207391312857, + "learning_rate": 0.003, + "loss": 4.008, + "step": 22006 + }, + { + "epoch": 0.22007, + "grad_norm": 0.6815382131960951, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 22007 + }, + { + "epoch": 0.22008, + "grad_norm": 0.790938444636332, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 22008 + }, + { + "epoch": 0.22009, + "grad_norm": 0.7098474267563651, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 22009 + }, + { + "epoch": 0.2201, + "grad_norm": 0.7265620306433147, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 22010 + }, + { + "epoch": 0.22011, + "grad_norm": 0.7030515060008808, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 22011 + }, + { + "epoch": 0.22012, + "grad_norm": 0.7047571409033759, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 22012 + }, + { + "epoch": 0.22013, + "grad_norm": 0.7668828715333653, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 22013 + }, + { + "epoch": 0.22014, + "grad_norm": 0.8917236021289892, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 22014 + }, + { + "epoch": 0.22015, + "grad_norm": 0.8847398942968456, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 22015 + }, + { + "epoch": 0.22016, + "grad_norm": 0.7833820995044907, + "learning_rate": 0.003, + "loss": 4.02, + "step": 22016 + }, + { + "epoch": 0.22017, + "grad_norm": 0.6308943703769652, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 22017 + }, + { + "epoch": 0.22018, + "grad_norm": 0.6974200470341754, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 22018 + }, + { + "epoch": 0.22019, + "grad_norm": 0.799304842033656, + "learning_rate": 0.003, + "loss": 3.992, + "step": 22019 + }, + { + "epoch": 0.2202, + "grad_norm": 0.8779204471971735, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 22020 + }, + { + "epoch": 0.22021, + "grad_norm": 0.8570984362749139, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 22021 + }, + { + "epoch": 0.22022, + "grad_norm": 0.85003568887761, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 22022 + }, + { + "epoch": 0.22023, + "grad_norm": 0.7130859766669366, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 22023 + }, + { + "epoch": 0.22024, + "grad_norm": 0.6571306707386885, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 22024 + }, + { + "epoch": 0.22025, + "grad_norm": 0.5776994924412238, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 22025 + }, + { + "epoch": 0.22026, + "grad_norm": 0.5851487464412847, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 22026 + }, + { + "epoch": 0.22027, + "grad_norm": 0.5600598392185586, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 22027 + }, + { + "epoch": 0.22028, + "grad_norm": 0.5579367266794575, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 22028 + }, + { + "epoch": 0.22029, + "grad_norm": 0.44693514573018706, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 22029 + }, + { + "epoch": 0.2203, + "grad_norm": 0.4421232104755611, + "learning_rate": 0.003, + "loss": 4.002, + "step": 22030 + }, + { + "epoch": 0.22031, + "grad_norm": 0.5169460864166037, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 22031 + }, + { + "epoch": 0.22032, + "grad_norm": 0.7369383040442864, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 22032 + }, + { + "epoch": 0.22033, + "grad_norm": 0.9875492385753745, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 22033 + }, + { + "epoch": 0.22034, + "grad_norm": 1.169895302597046, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 22034 + }, + { + "epoch": 0.22035, + "grad_norm": 0.629442887484256, + "learning_rate": 0.003, + "loss": 4.007, + "step": 22035 + }, + { + "epoch": 0.22036, + "grad_norm": 0.7155207838935818, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 22036 + }, + { + "epoch": 0.22037, + "grad_norm": 0.9475009532192534, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 22037 + }, + { + "epoch": 0.22038, + "grad_norm": 0.8852256627707318, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 22038 + }, + { + "epoch": 0.22039, + "grad_norm": 0.7140987309290081, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 22039 + }, + { + "epoch": 0.2204, + "grad_norm": 0.6472963400251972, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 22040 + }, + { + "epoch": 0.22041, + "grad_norm": 0.6341453750791816, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 22041 + }, + { + "epoch": 0.22042, + "grad_norm": 0.6942686733482917, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 22042 + }, + { + "epoch": 0.22043, + "grad_norm": 0.7126288573043553, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 22043 + }, + { + "epoch": 0.22044, + "grad_norm": 0.7146496075116633, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 22044 + }, + { + "epoch": 0.22045, + "grad_norm": 0.6990552802259824, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 22045 + }, + { + "epoch": 0.22046, + "grad_norm": 0.643187407913674, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 22046 + }, + { + "epoch": 0.22047, + "grad_norm": 0.5989217289460931, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 22047 + }, + { + "epoch": 0.22048, + "grad_norm": 0.6073621410556866, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 22048 + }, + { + "epoch": 0.22049, + "grad_norm": 0.6631547374149641, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 22049 + }, + { + "epoch": 0.2205, + "grad_norm": 0.6694738181389146, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 22050 + }, + { + "epoch": 0.22051, + "grad_norm": 0.6424372456406638, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 22051 + }, + { + "epoch": 0.22052, + "grad_norm": 0.5174926365795308, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 22052 + }, + { + "epoch": 0.22053, + "grad_norm": 0.5358648836373169, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 22053 + }, + { + "epoch": 0.22054, + "grad_norm": 0.6171919320570968, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 22054 + }, + { + "epoch": 0.22055, + "grad_norm": 0.6476234316310461, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 22055 + }, + { + "epoch": 0.22056, + "grad_norm": 0.7846474123041219, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 22056 + }, + { + "epoch": 0.22057, + "grad_norm": 0.9028350126173306, + "learning_rate": 0.003, + "loss": 4.017, + "step": 22057 + }, + { + "epoch": 0.22058, + "grad_norm": 0.8928445562947223, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 22058 + }, + { + "epoch": 0.22059, + "grad_norm": 0.843854449282421, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 22059 + }, + { + "epoch": 0.2206, + "grad_norm": 0.8575411871812357, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 22060 + }, + { + "epoch": 0.22061, + "grad_norm": 0.8739855306434217, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 22061 + }, + { + "epoch": 0.22062, + "grad_norm": 0.9464377640638156, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 22062 + }, + { + "epoch": 0.22063, + "grad_norm": 1.137343718442177, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 22063 + }, + { + "epoch": 0.22064, + "grad_norm": 0.8376564317971438, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 22064 + }, + { + "epoch": 0.22065, + "grad_norm": 0.7043786296995681, + "learning_rate": 0.003, + "loss": 3.966, + "step": 22065 + }, + { + "epoch": 0.22066, + "grad_norm": 0.8277668939046499, + "learning_rate": 0.003, + "loss": 3.974, + "step": 22066 + }, + { + "epoch": 0.22067, + "grad_norm": 0.7981974951669328, + "learning_rate": 0.003, + "loss": 4.01, + "step": 22067 + }, + { + "epoch": 0.22068, + "grad_norm": 0.7227479236226491, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 22068 + }, + { + "epoch": 0.22069, + "grad_norm": 0.755464417411058, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 22069 + }, + { + "epoch": 0.2207, + "grad_norm": 0.8142381546499255, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 22070 + }, + { + "epoch": 0.22071, + "grad_norm": 0.9678116178921617, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 22071 + }, + { + "epoch": 0.22072, + "grad_norm": 0.9783236790299369, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 22072 + }, + { + "epoch": 0.22073, + "grad_norm": 0.9460801800876342, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 22073 + }, + { + "epoch": 0.22074, + "grad_norm": 1.1150203120944744, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 22074 + }, + { + "epoch": 0.22075, + "grad_norm": 0.8673348195945594, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 22075 + }, + { + "epoch": 0.22076, + "grad_norm": 0.6989404284527327, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 22076 + }, + { + "epoch": 0.22077, + "grad_norm": 0.7133600164095404, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 22077 + }, + { + "epoch": 0.22078, + "grad_norm": 0.6591563890473798, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 22078 + }, + { + "epoch": 0.22079, + "grad_norm": 0.6756885767452964, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 22079 + }, + { + "epoch": 0.2208, + "grad_norm": 0.7519036841022073, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 22080 + }, + { + "epoch": 0.22081, + "grad_norm": 0.802470716012958, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 22081 + }, + { + "epoch": 0.22082, + "grad_norm": 0.7802873024399, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 22082 + }, + { + "epoch": 0.22083, + "grad_norm": 0.9898832861982021, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 22083 + }, + { + "epoch": 0.22084, + "grad_norm": 0.9455178548388816, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 22084 + }, + { + "epoch": 0.22085, + "grad_norm": 0.7931452113603521, + "learning_rate": 0.003, + "loss": 4.011, + "step": 22085 + }, + { + "epoch": 0.22086, + "grad_norm": 0.7002282692581103, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 22086 + }, + { + "epoch": 0.22087, + "grad_norm": 0.6551158876247526, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 22087 + }, + { + "epoch": 0.22088, + "grad_norm": 0.674759402093935, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 22088 + }, + { + "epoch": 0.22089, + "grad_norm": 0.7089086087813566, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 22089 + }, + { + "epoch": 0.2209, + "grad_norm": 0.6705760634462763, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 22090 + }, + { + "epoch": 0.22091, + "grad_norm": 0.6375409993194727, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 22091 + }, + { + "epoch": 0.22092, + "grad_norm": 0.5323107491359, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 22092 + }, + { + "epoch": 0.22093, + "grad_norm": 0.6168957039745454, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 22093 + }, + { + "epoch": 0.22094, + "grad_norm": 0.7229920062887346, + "learning_rate": 0.003, + "loss": 3.989, + "step": 22094 + }, + { + "epoch": 0.22095, + "grad_norm": 0.8203745551385877, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 22095 + }, + { + "epoch": 0.22096, + "grad_norm": 0.8487557380685574, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 22096 + }, + { + "epoch": 0.22097, + "grad_norm": 0.8256279044503795, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 22097 + }, + { + "epoch": 0.22098, + "grad_norm": 0.8382587409696629, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 22098 + }, + { + "epoch": 0.22099, + "grad_norm": 0.7582407982329821, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 22099 + }, + { + "epoch": 0.221, + "grad_norm": 0.6659860693522524, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 22100 + }, + { + "epoch": 0.22101, + "grad_norm": 0.5893710561001384, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 22101 + }, + { + "epoch": 0.22102, + "grad_norm": 0.5868847313965134, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 22102 + }, + { + "epoch": 0.22103, + "grad_norm": 0.5682594680133796, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 22103 + }, + { + "epoch": 0.22104, + "grad_norm": 0.4898667246617674, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 22104 + }, + { + "epoch": 0.22105, + "grad_norm": 0.493782644137072, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 22105 + }, + { + "epoch": 0.22106, + "grad_norm": 0.5032771362220196, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 22106 + }, + { + "epoch": 0.22107, + "grad_norm": 0.5336158867451941, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 22107 + }, + { + "epoch": 0.22108, + "grad_norm": 0.5450732399542444, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 22108 + }, + { + "epoch": 0.22109, + "grad_norm": 0.6683944485022756, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 22109 + }, + { + "epoch": 0.2211, + "grad_norm": 0.7040170620796857, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 22110 + }, + { + "epoch": 0.22111, + "grad_norm": 0.7283172298942285, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 22111 + }, + { + "epoch": 0.22112, + "grad_norm": 0.797529898744369, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 22112 + }, + { + "epoch": 0.22113, + "grad_norm": 0.959281209828524, + "learning_rate": 0.003, + "loss": 4.01, + "step": 22113 + }, + { + "epoch": 0.22114, + "grad_norm": 1.0978940114550702, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 22114 + }, + { + "epoch": 0.22115, + "grad_norm": 0.8851116197743546, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 22115 + }, + { + "epoch": 0.22116, + "grad_norm": 0.8653264159506191, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 22116 + }, + { + "epoch": 0.22117, + "grad_norm": 0.721838814392924, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 22117 + }, + { + "epoch": 0.22118, + "grad_norm": 0.7376185994008736, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 22118 + }, + { + "epoch": 0.22119, + "grad_norm": 0.8266692826068319, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 22119 + }, + { + "epoch": 0.2212, + "grad_norm": 0.8675578077765809, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 22120 + }, + { + "epoch": 0.22121, + "grad_norm": 0.8272258645089268, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 22121 + }, + { + "epoch": 0.22122, + "grad_norm": 0.7042202932159061, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 22122 + }, + { + "epoch": 0.22123, + "grad_norm": 0.6231880743692124, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 22123 + }, + { + "epoch": 0.22124, + "grad_norm": 0.5260351333587188, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 22124 + }, + { + "epoch": 0.22125, + "grad_norm": 0.4970845388680466, + "learning_rate": 0.003, + "loss": 3.997, + "step": 22125 + }, + { + "epoch": 0.22126, + "grad_norm": 0.44880335250057346, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 22126 + }, + { + "epoch": 0.22127, + "grad_norm": 0.42638261076355877, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 22127 + }, + { + "epoch": 0.22128, + "grad_norm": 0.3903958494394582, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 22128 + }, + { + "epoch": 0.22129, + "grad_norm": 0.44749210224587627, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 22129 + }, + { + "epoch": 0.2213, + "grad_norm": 0.5506016109742737, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 22130 + }, + { + "epoch": 0.22131, + "grad_norm": 0.6861332801775706, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 22131 + }, + { + "epoch": 0.22132, + "grad_norm": 0.8204474788001203, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 22132 + }, + { + "epoch": 0.22133, + "grad_norm": 0.9856051978033933, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 22133 + }, + { + "epoch": 0.22134, + "grad_norm": 1.1010521154565487, + "learning_rate": 0.003, + "loss": 3.996, + "step": 22134 + }, + { + "epoch": 0.22135, + "grad_norm": 0.7298440245581447, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 22135 + }, + { + "epoch": 0.22136, + "grad_norm": 0.7121228042776373, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 22136 + }, + { + "epoch": 0.22137, + "grad_norm": 0.824103178726263, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 22137 + }, + { + "epoch": 0.22138, + "grad_norm": 0.7551354737002132, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 22138 + }, + { + "epoch": 0.22139, + "grad_norm": 0.6723356773163363, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 22139 + }, + { + "epoch": 0.2214, + "grad_norm": 0.7500399613370746, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 22140 + }, + { + "epoch": 0.22141, + "grad_norm": 0.8208751801814849, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 22141 + }, + { + "epoch": 0.22142, + "grad_norm": 0.880682594452394, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 22142 + }, + { + "epoch": 0.22143, + "grad_norm": 1.0861590630360896, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 22143 + }, + { + "epoch": 0.22144, + "grad_norm": 1.0744181196611016, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 22144 + }, + { + "epoch": 0.22145, + "grad_norm": 0.948573340528064, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 22145 + }, + { + "epoch": 0.22146, + "grad_norm": 0.8579923660161339, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 22146 + }, + { + "epoch": 0.22147, + "grad_norm": 0.8299594542676699, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 22147 + }, + { + "epoch": 0.22148, + "grad_norm": 0.8557258527796691, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 22148 + }, + { + "epoch": 0.22149, + "grad_norm": 1.1264500900167214, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 22149 + }, + { + "epoch": 0.2215, + "grad_norm": 1.0149350544328697, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 22150 + }, + { + "epoch": 0.22151, + "grad_norm": 1.078284887093296, + "learning_rate": 0.003, + "loss": 4.01, + "step": 22151 + }, + { + "epoch": 0.22152, + "grad_norm": 0.8157702541244342, + "learning_rate": 0.003, + "loss": 4.0518, + "step": 22152 + }, + { + "epoch": 0.22153, + "grad_norm": 0.6267794346604527, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 22153 + }, + { + "epoch": 0.22154, + "grad_norm": 0.5741120536155551, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 22154 + }, + { + "epoch": 0.22155, + "grad_norm": 0.5974493604700077, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 22155 + }, + { + "epoch": 0.22156, + "grad_norm": 0.5008396430313397, + "learning_rate": 0.003, + "loss": 4.013, + "step": 22156 + }, + { + "epoch": 0.22157, + "grad_norm": 0.5317414783085934, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 22157 + }, + { + "epoch": 0.22158, + "grad_norm": 0.5517537343340047, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 22158 + }, + { + "epoch": 0.22159, + "grad_norm": 0.5622220642693556, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 22159 + }, + { + "epoch": 0.2216, + "grad_norm": 0.5421316600893534, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 22160 + }, + { + "epoch": 0.22161, + "grad_norm": 0.7075248243513329, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 22161 + }, + { + "epoch": 0.22162, + "grad_norm": 0.8532231906998824, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 22162 + }, + { + "epoch": 0.22163, + "grad_norm": 1.0176430098242586, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 22163 + }, + { + "epoch": 0.22164, + "grad_norm": 1.0254064934722147, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 22164 + }, + { + "epoch": 0.22165, + "grad_norm": 0.7137995441362722, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 22165 + }, + { + "epoch": 0.22166, + "grad_norm": 0.6602066583238319, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 22166 + }, + { + "epoch": 0.22167, + "grad_norm": 0.7128175496980916, + "learning_rate": 0.003, + "loss": 3.994, + "step": 22167 + }, + { + "epoch": 0.22168, + "grad_norm": 0.7896215737110325, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 22168 + }, + { + "epoch": 0.22169, + "grad_norm": 0.8274882863896688, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 22169 + }, + { + "epoch": 0.2217, + "grad_norm": 0.8375130858716048, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 22170 + }, + { + "epoch": 0.22171, + "grad_norm": 0.8907889317265169, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 22171 + }, + { + "epoch": 0.22172, + "grad_norm": 0.8737480176436248, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 22172 + }, + { + "epoch": 0.22173, + "grad_norm": 0.8053472772922255, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 22173 + }, + { + "epoch": 0.22174, + "grad_norm": 0.7733706201199868, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 22174 + }, + { + "epoch": 0.22175, + "grad_norm": 0.7412520892600634, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 22175 + }, + { + "epoch": 0.22176, + "grad_norm": 0.7809110318726006, + "learning_rate": 0.003, + "loss": 4.042, + "step": 22176 + }, + { + "epoch": 0.22177, + "grad_norm": 0.7739714906858535, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 22177 + }, + { + "epoch": 0.22178, + "grad_norm": 0.8305584004018258, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 22178 + }, + { + "epoch": 0.22179, + "grad_norm": 0.8359689883022254, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 22179 + }, + { + "epoch": 0.2218, + "grad_norm": 0.7941400345069121, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 22180 + }, + { + "epoch": 0.22181, + "grad_norm": 0.732766179510451, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 22181 + }, + { + "epoch": 0.22182, + "grad_norm": 0.7569072038917667, + "learning_rate": 0.003, + "loss": 4.002, + "step": 22182 + }, + { + "epoch": 0.22183, + "grad_norm": 0.8719320554661084, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 22183 + }, + { + "epoch": 0.22184, + "grad_norm": 0.9375360494138644, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 22184 + }, + { + "epoch": 0.22185, + "grad_norm": 0.8697097444431376, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22185 + }, + { + "epoch": 0.22186, + "grad_norm": 0.7859522724013717, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 22186 + }, + { + "epoch": 0.22187, + "grad_norm": 0.7338361824100595, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 22187 + }, + { + "epoch": 0.22188, + "grad_norm": 0.6603019987191423, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 22188 + }, + { + "epoch": 0.22189, + "grad_norm": 0.6554991953893691, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 22189 + }, + { + "epoch": 0.2219, + "grad_norm": 0.7552066140367317, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 22190 + }, + { + "epoch": 0.22191, + "grad_norm": 0.7284333680546394, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 22191 + }, + { + "epoch": 0.22192, + "grad_norm": 0.6779749807872648, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 22192 + }, + { + "epoch": 0.22193, + "grad_norm": 0.6680901319201633, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 22193 + }, + { + "epoch": 0.22194, + "grad_norm": 0.7119378124496926, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 22194 + }, + { + "epoch": 0.22195, + "grad_norm": 0.7673635118067812, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 22195 + }, + { + "epoch": 0.22196, + "grad_norm": 0.8311684350194637, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 22196 + }, + { + "epoch": 0.22197, + "grad_norm": 0.884042680294808, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 22197 + }, + { + "epoch": 0.22198, + "grad_norm": 0.8780074690406335, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 22198 + }, + { + "epoch": 0.22199, + "grad_norm": 0.7705754897210945, + "learning_rate": 0.003, + "loss": 3.994, + "step": 22199 + }, + { + "epoch": 0.222, + "grad_norm": 0.7656856097685729, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 22200 + }, + { + "epoch": 0.22201, + "grad_norm": 0.8423138354364835, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 22201 + }, + { + "epoch": 0.22202, + "grad_norm": 0.8461681882295891, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 22202 + }, + { + "epoch": 0.22203, + "grad_norm": 0.7334358270773735, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 22203 + }, + { + "epoch": 0.22204, + "grad_norm": 0.7267170772143324, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 22204 + }, + { + "epoch": 0.22205, + "grad_norm": 0.7153254187069481, + "learning_rate": 0.003, + "loss": 4.02, + "step": 22205 + }, + { + "epoch": 0.22206, + "grad_norm": 0.7852125689712156, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 22206 + }, + { + "epoch": 0.22207, + "grad_norm": 0.7079625024155702, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 22207 + }, + { + "epoch": 0.22208, + "grad_norm": 0.5717553437485702, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 22208 + }, + { + "epoch": 0.22209, + "grad_norm": 0.552490512199235, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 22209 + }, + { + "epoch": 0.2221, + "grad_norm": 0.5537145112042134, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 22210 + }, + { + "epoch": 0.22211, + "grad_norm": 0.6371970018035411, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 22211 + }, + { + "epoch": 0.22212, + "grad_norm": 0.6640286564688079, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 22212 + }, + { + "epoch": 0.22213, + "grad_norm": 0.6671998291590331, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 22213 + }, + { + "epoch": 0.22214, + "grad_norm": 0.6497312002364957, + "learning_rate": 0.003, + "loss": 3.999, + "step": 22214 + }, + { + "epoch": 0.22215, + "grad_norm": 0.6904737707857304, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 22215 + }, + { + "epoch": 0.22216, + "grad_norm": 0.7489329694929404, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 22216 + }, + { + "epoch": 0.22217, + "grad_norm": 0.8229349101127434, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 22217 + }, + { + "epoch": 0.22218, + "grad_norm": 0.734059842610261, + "learning_rate": 0.003, + "loss": 4.014, + "step": 22218 + }, + { + "epoch": 0.22219, + "grad_norm": 0.6816641317382628, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 22219 + }, + { + "epoch": 0.2222, + "grad_norm": 0.6509811700379484, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 22220 + }, + { + "epoch": 0.22221, + "grad_norm": 0.5326556151261503, + "learning_rate": 0.003, + "loss": 4.004, + "step": 22221 + }, + { + "epoch": 0.22222, + "grad_norm": 0.5357440593010651, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 22222 + }, + { + "epoch": 0.22223, + "grad_norm": 0.6243625338112472, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 22223 + }, + { + "epoch": 0.22224, + "grad_norm": 0.7036286266791987, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 22224 + }, + { + "epoch": 0.22225, + "grad_norm": 0.6947158812664982, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 22225 + }, + { + "epoch": 0.22226, + "grad_norm": 0.7590232181540142, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 22226 + }, + { + "epoch": 0.22227, + "grad_norm": 0.8191261881232119, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 22227 + }, + { + "epoch": 0.22228, + "grad_norm": 0.8435478140885528, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 22228 + }, + { + "epoch": 0.22229, + "grad_norm": 1.0027992846800904, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 22229 + }, + { + "epoch": 0.2223, + "grad_norm": 0.9837153558039785, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 22230 + }, + { + "epoch": 0.22231, + "grad_norm": 0.8754398054192507, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 22231 + }, + { + "epoch": 0.22232, + "grad_norm": 0.8284192055648086, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 22232 + }, + { + "epoch": 0.22233, + "grad_norm": 0.7978943639073297, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 22233 + }, + { + "epoch": 0.22234, + "grad_norm": 0.8153693233897302, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 22234 + }, + { + "epoch": 0.22235, + "grad_norm": 0.7114250715121765, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 22235 + }, + { + "epoch": 0.22236, + "grad_norm": 0.7064300333121308, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 22236 + }, + { + "epoch": 0.22237, + "grad_norm": 0.7050972209659012, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 22237 + }, + { + "epoch": 0.22238, + "grad_norm": 0.6547921288032578, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 22238 + }, + { + "epoch": 0.22239, + "grad_norm": 0.590123166907784, + "learning_rate": 0.003, + "loss": 4.006, + "step": 22239 + }, + { + "epoch": 0.2224, + "grad_norm": 0.6053509409106911, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 22240 + }, + { + "epoch": 0.22241, + "grad_norm": 0.6066076167324461, + "learning_rate": 0.003, + "loss": 3.975, + "step": 22241 + }, + { + "epoch": 0.22242, + "grad_norm": 0.6776504921086409, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 22242 + }, + { + "epoch": 0.22243, + "grad_norm": 0.7555920024714922, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 22243 + }, + { + "epoch": 0.22244, + "grad_norm": 0.8246068770530212, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 22244 + }, + { + "epoch": 0.22245, + "grad_norm": 0.930593275976438, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 22245 + }, + { + "epoch": 0.22246, + "grad_norm": 1.112185863754824, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 22246 + }, + { + "epoch": 0.22247, + "grad_norm": 0.8556158785564902, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 22247 + }, + { + "epoch": 0.22248, + "grad_norm": 0.8126434068632176, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22248 + }, + { + "epoch": 0.22249, + "grad_norm": 0.9011601600624982, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 22249 + }, + { + "epoch": 0.2225, + "grad_norm": 0.9627385418215069, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 22250 + }, + { + "epoch": 0.22251, + "grad_norm": 0.7867234592893402, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 22251 + }, + { + "epoch": 0.22252, + "grad_norm": 0.6683528180845186, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 22252 + }, + { + "epoch": 0.22253, + "grad_norm": 0.6737152646570616, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 22253 + }, + { + "epoch": 0.22254, + "grad_norm": 0.6680518728946674, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 22254 + }, + { + "epoch": 0.22255, + "grad_norm": 0.6832489574421109, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 22255 + }, + { + "epoch": 0.22256, + "grad_norm": 0.7418709125637756, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 22256 + }, + { + "epoch": 0.22257, + "grad_norm": 0.7677978721320802, + "learning_rate": 0.003, + "loss": 4.018, + "step": 22257 + }, + { + "epoch": 0.22258, + "grad_norm": 0.7449820031837786, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 22258 + }, + { + "epoch": 0.22259, + "grad_norm": 0.6737849783018203, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 22259 + }, + { + "epoch": 0.2226, + "grad_norm": 0.758183684436134, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 22260 + }, + { + "epoch": 0.22261, + "grad_norm": 0.8215786745488033, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 22261 + }, + { + "epoch": 0.22262, + "grad_norm": 0.8092682793633745, + "learning_rate": 0.003, + "loss": 4.007, + "step": 22262 + }, + { + "epoch": 0.22263, + "grad_norm": 0.7983369152694101, + "learning_rate": 0.003, + "loss": 4.019, + "step": 22263 + }, + { + "epoch": 0.22264, + "grad_norm": 0.6142838523749142, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 22264 + }, + { + "epoch": 0.22265, + "grad_norm": 0.6257309134150747, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 22265 + }, + { + "epoch": 0.22266, + "grad_norm": 0.5536850280404204, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 22266 + }, + { + "epoch": 0.22267, + "grad_norm": 0.6003928170116507, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 22267 + }, + { + "epoch": 0.22268, + "grad_norm": 0.682893362527029, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 22268 + }, + { + "epoch": 0.22269, + "grad_norm": 0.7532983022663478, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 22269 + }, + { + "epoch": 0.2227, + "grad_norm": 0.7764380598719964, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 22270 + }, + { + "epoch": 0.22271, + "grad_norm": 0.7479913799478853, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 22271 + }, + { + "epoch": 0.22272, + "grad_norm": 0.803155884391662, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 22272 + }, + { + "epoch": 0.22273, + "grad_norm": 0.8742648763951016, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 22273 + }, + { + "epoch": 0.22274, + "grad_norm": 0.8778641786013621, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 22274 + }, + { + "epoch": 0.22275, + "grad_norm": 0.7374407196151113, + "learning_rate": 0.003, + "loss": 4.004, + "step": 22275 + }, + { + "epoch": 0.22276, + "grad_norm": 0.6233746377498592, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 22276 + }, + { + "epoch": 0.22277, + "grad_norm": 0.6580702906909545, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 22277 + }, + { + "epoch": 0.22278, + "grad_norm": 0.7078715220398302, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 22278 + }, + { + "epoch": 0.22279, + "grad_norm": 0.7358000209435193, + "learning_rate": 0.003, + "loss": 3.98, + "step": 22279 + }, + { + "epoch": 0.2228, + "grad_norm": 0.7265235451984735, + "learning_rate": 0.003, + "loss": 3.98, + "step": 22280 + }, + { + "epoch": 0.22281, + "grad_norm": 0.7951326905944285, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 22281 + }, + { + "epoch": 0.22282, + "grad_norm": 0.8815922714963162, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 22282 + }, + { + "epoch": 0.22283, + "grad_norm": 0.829002791104622, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 22283 + }, + { + "epoch": 0.22284, + "grad_norm": 0.7129537377129765, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 22284 + }, + { + "epoch": 0.22285, + "grad_norm": 0.7768406286091926, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 22285 + }, + { + "epoch": 0.22286, + "grad_norm": 0.7431336596967936, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 22286 + }, + { + "epoch": 0.22287, + "grad_norm": 0.6831252689754507, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 22287 + }, + { + "epoch": 0.22288, + "grad_norm": 0.6468608872427588, + "learning_rate": 0.003, + "loss": 3.993, + "step": 22288 + }, + { + "epoch": 0.22289, + "grad_norm": 0.6465982576832682, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 22289 + }, + { + "epoch": 0.2229, + "grad_norm": 0.6045996613119237, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 22290 + }, + { + "epoch": 0.22291, + "grad_norm": 0.5854913302638105, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 22291 + }, + { + "epoch": 0.22292, + "grad_norm": 0.5581653915283465, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 22292 + }, + { + "epoch": 0.22293, + "grad_norm": 0.6614935886406379, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 22293 + }, + { + "epoch": 0.22294, + "grad_norm": 0.8356520950287343, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 22294 + }, + { + "epoch": 0.22295, + "grad_norm": 1.0760098382237946, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 22295 + }, + { + "epoch": 0.22296, + "grad_norm": 1.0565864128175242, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 22296 + }, + { + "epoch": 0.22297, + "grad_norm": 0.8780669664831295, + "learning_rate": 0.003, + "loss": 3.981, + "step": 22297 + }, + { + "epoch": 0.22298, + "grad_norm": 0.8015395646529686, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 22298 + }, + { + "epoch": 0.22299, + "grad_norm": 0.782448002792685, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 22299 + }, + { + "epoch": 0.223, + "grad_norm": 0.7606269391523878, + "learning_rate": 0.003, + "loss": 4.011, + "step": 22300 + }, + { + "epoch": 0.22301, + "grad_norm": 0.7548105214953389, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 22301 + }, + { + "epoch": 0.22302, + "grad_norm": 0.6423563822454937, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 22302 + }, + { + "epoch": 0.22303, + "grad_norm": 0.6383885688262921, + "learning_rate": 0.003, + "loss": 3.9367, + "step": 22303 + }, + { + "epoch": 0.22304, + "grad_norm": 0.6170593644203741, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 22304 + }, + { + "epoch": 0.22305, + "grad_norm": 0.7130283803489019, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 22305 + }, + { + "epoch": 0.22306, + "grad_norm": 0.8505938824124691, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 22306 + }, + { + "epoch": 0.22307, + "grad_norm": 1.0531486486805877, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 22307 + }, + { + "epoch": 0.22308, + "grad_norm": 0.9486971831280455, + "learning_rate": 0.003, + "loss": 4.015, + "step": 22308 + }, + { + "epoch": 0.22309, + "grad_norm": 0.9645462781013745, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 22309 + }, + { + "epoch": 0.2231, + "grad_norm": 0.8563198390081272, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 22310 + }, + { + "epoch": 0.22311, + "grad_norm": 0.9286425169162371, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 22311 + }, + { + "epoch": 0.22312, + "grad_norm": 1.0073325182873636, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 22312 + }, + { + "epoch": 0.22313, + "grad_norm": 0.9483842486454943, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 22313 + }, + { + "epoch": 0.22314, + "grad_norm": 1.0245255497686807, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 22314 + }, + { + "epoch": 0.22315, + "grad_norm": 0.8493104804509216, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 22315 + }, + { + "epoch": 0.22316, + "grad_norm": 0.8461375273056162, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 22316 + }, + { + "epoch": 0.22317, + "grad_norm": 0.6607704872818302, + "learning_rate": 0.003, + "loss": 4.013, + "step": 22317 + }, + { + "epoch": 0.22318, + "grad_norm": 0.8013211433136244, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 22318 + }, + { + "epoch": 0.22319, + "grad_norm": 0.7215726388718481, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 22319 + }, + { + "epoch": 0.2232, + "grad_norm": 0.5649514832120974, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 22320 + }, + { + "epoch": 0.22321, + "grad_norm": 0.5988227908184997, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 22321 + }, + { + "epoch": 0.22322, + "grad_norm": 0.6521895287181463, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 22322 + }, + { + "epoch": 0.22323, + "grad_norm": 0.6186318319156161, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 22323 + }, + { + "epoch": 0.22324, + "grad_norm": 0.6040142487251458, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22324 + }, + { + "epoch": 0.22325, + "grad_norm": 0.7217542299974383, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 22325 + }, + { + "epoch": 0.22326, + "grad_norm": 0.8990615462719416, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 22326 + }, + { + "epoch": 0.22327, + "grad_norm": 0.990323296370568, + "learning_rate": 0.003, + "loss": 3.989, + "step": 22327 + }, + { + "epoch": 0.22328, + "grad_norm": 0.9533826479888986, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 22328 + }, + { + "epoch": 0.22329, + "grad_norm": 0.8314617456425858, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 22329 + }, + { + "epoch": 0.2233, + "grad_norm": 0.8013198006744602, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 22330 + }, + { + "epoch": 0.22331, + "grad_norm": 0.9133365831584944, + "learning_rate": 0.003, + "loss": 4.013, + "step": 22331 + }, + { + "epoch": 0.22332, + "grad_norm": 0.862495016497783, + "learning_rate": 0.003, + "loss": 4.0, + "step": 22332 + }, + { + "epoch": 0.22333, + "grad_norm": 0.9038338805550048, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 22333 + }, + { + "epoch": 0.22334, + "grad_norm": 0.9001606183243015, + "learning_rate": 0.003, + "loss": 4.0491, + "step": 22334 + }, + { + "epoch": 0.22335, + "grad_norm": 0.8803921201578901, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 22335 + }, + { + "epoch": 0.22336, + "grad_norm": 0.7438667042058457, + "learning_rate": 0.003, + "loss": 4.031, + "step": 22336 + }, + { + "epoch": 0.22337, + "grad_norm": 0.6415586535456242, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 22337 + }, + { + "epoch": 0.22338, + "grad_norm": 0.5613065789436615, + "learning_rate": 0.003, + "loss": 4.007, + "step": 22338 + }, + { + "epoch": 0.22339, + "grad_norm": 0.5778215596757237, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 22339 + }, + { + "epoch": 0.2234, + "grad_norm": 0.5127730471405653, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 22340 + }, + { + "epoch": 0.22341, + "grad_norm": 0.5550428280025473, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 22341 + }, + { + "epoch": 0.22342, + "grad_norm": 0.6002624578420032, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 22342 + }, + { + "epoch": 0.22343, + "grad_norm": 0.5949917352469525, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 22343 + }, + { + "epoch": 0.22344, + "grad_norm": 0.643591034215256, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 22344 + }, + { + "epoch": 0.22345, + "grad_norm": 0.6350238466684204, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 22345 + }, + { + "epoch": 0.22346, + "grad_norm": 0.6951669330709547, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 22346 + }, + { + "epoch": 0.22347, + "grad_norm": 0.7120077487814728, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 22347 + }, + { + "epoch": 0.22348, + "grad_norm": 0.7804306007344097, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 22348 + }, + { + "epoch": 0.22349, + "grad_norm": 0.8577997459669053, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 22349 + }, + { + "epoch": 0.2235, + "grad_norm": 0.7247807509874632, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 22350 + }, + { + "epoch": 0.22351, + "grad_norm": 0.7011506771751931, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 22351 + }, + { + "epoch": 0.22352, + "grad_norm": 0.6156805797754539, + "learning_rate": 0.003, + "loss": 4.016, + "step": 22352 + }, + { + "epoch": 0.22353, + "grad_norm": 0.7659250426308661, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 22353 + }, + { + "epoch": 0.22354, + "grad_norm": 0.9212409740908505, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 22354 + }, + { + "epoch": 0.22355, + "grad_norm": 1.0567862926761658, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 22355 + }, + { + "epoch": 0.22356, + "grad_norm": 1.1184874928544377, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 22356 + }, + { + "epoch": 0.22357, + "grad_norm": 0.7303320565601787, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 22357 + }, + { + "epoch": 0.22358, + "grad_norm": 0.723640783693516, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 22358 + }, + { + "epoch": 0.22359, + "grad_norm": 0.779423696987343, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 22359 + }, + { + "epoch": 0.2236, + "grad_norm": 0.648656345844564, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 22360 + }, + { + "epoch": 0.22361, + "grad_norm": 0.5378440302727484, + "learning_rate": 0.003, + "loss": 3.999, + "step": 22361 + }, + { + "epoch": 0.22362, + "grad_norm": 0.5162704940781239, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 22362 + }, + { + "epoch": 0.22363, + "grad_norm": 0.4731282298521864, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 22363 + }, + { + "epoch": 0.22364, + "grad_norm": 0.5422410283817081, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 22364 + }, + { + "epoch": 0.22365, + "grad_norm": 0.5813799616341834, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 22365 + }, + { + "epoch": 0.22366, + "grad_norm": 0.6224611855941768, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 22366 + }, + { + "epoch": 0.22367, + "grad_norm": 0.5899597701117099, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 22367 + }, + { + "epoch": 0.22368, + "grad_norm": 0.6228352716507954, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 22368 + }, + { + "epoch": 0.22369, + "grad_norm": 0.70581756416304, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 22369 + }, + { + "epoch": 0.2237, + "grad_norm": 0.7170218570152431, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 22370 + }, + { + "epoch": 0.22371, + "grad_norm": 0.8445455518078869, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 22371 + }, + { + "epoch": 0.22372, + "grad_norm": 1.0560973872596893, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 22372 + }, + { + "epoch": 0.22373, + "grad_norm": 1.1216821429354362, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 22373 + }, + { + "epoch": 0.22374, + "grad_norm": 0.8869535180468404, + "learning_rate": 0.003, + "loss": 4.025, + "step": 22374 + }, + { + "epoch": 0.22375, + "grad_norm": 0.9108026754590783, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 22375 + }, + { + "epoch": 0.22376, + "grad_norm": 0.7889942952920572, + "learning_rate": 0.003, + "loss": 3.996, + "step": 22376 + }, + { + "epoch": 0.22377, + "grad_norm": 0.8047629609965448, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 22377 + }, + { + "epoch": 0.22378, + "grad_norm": 0.8441015359626755, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 22378 + }, + { + "epoch": 0.22379, + "grad_norm": 0.8687831789474547, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 22379 + }, + { + "epoch": 0.2238, + "grad_norm": 0.8328568092926232, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 22380 + }, + { + "epoch": 0.22381, + "grad_norm": 0.7400682012599014, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 22381 + }, + { + "epoch": 0.22382, + "grad_norm": 0.6265925139423428, + "learning_rate": 0.003, + "loss": 4.013, + "step": 22382 + }, + { + "epoch": 0.22383, + "grad_norm": 0.6738378446182705, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 22383 + }, + { + "epoch": 0.22384, + "grad_norm": 0.7202391730735314, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 22384 + }, + { + "epoch": 0.22385, + "grad_norm": 0.8188762655298129, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 22385 + }, + { + "epoch": 0.22386, + "grad_norm": 0.919508040361057, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 22386 + }, + { + "epoch": 0.22387, + "grad_norm": 1.0238988828452786, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 22387 + }, + { + "epoch": 0.22388, + "grad_norm": 0.7923159134112981, + "learning_rate": 0.003, + "loss": 4.03, + "step": 22388 + }, + { + "epoch": 0.22389, + "grad_norm": 0.6829526606840076, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 22389 + }, + { + "epoch": 0.2239, + "grad_norm": 0.7615033726543202, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 22390 + }, + { + "epoch": 0.22391, + "grad_norm": 0.7798991829761511, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 22391 + }, + { + "epoch": 0.22392, + "grad_norm": 0.7292716242995769, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 22392 + }, + { + "epoch": 0.22393, + "grad_norm": 0.6614671442692002, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 22393 + }, + { + "epoch": 0.22394, + "grad_norm": 0.6876815370424093, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 22394 + }, + { + "epoch": 0.22395, + "grad_norm": 0.6866658673492847, + "learning_rate": 0.003, + "loss": 4.017, + "step": 22395 + }, + { + "epoch": 0.22396, + "grad_norm": 0.688728189347793, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 22396 + }, + { + "epoch": 0.22397, + "grad_norm": 0.6567000671844009, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 22397 + }, + { + "epoch": 0.22398, + "grad_norm": 0.5375353935755056, + "learning_rate": 0.003, + "loss": 3.966, + "step": 22398 + }, + { + "epoch": 0.22399, + "grad_norm": 0.48737657850201216, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 22399 + }, + { + "epoch": 0.224, + "grad_norm": 0.5831386708583228, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 22400 + }, + { + "epoch": 0.22401, + "grad_norm": 0.6298118450133323, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 22401 + }, + { + "epoch": 0.22402, + "grad_norm": 0.740527282044012, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 22402 + }, + { + "epoch": 0.22403, + "grad_norm": 0.8723634267908114, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 22403 + }, + { + "epoch": 0.22404, + "grad_norm": 0.9774601635870569, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 22404 + }, + { + "epoch": 0.22405, + "grad_norm": 0.9652933712874885, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 22405 + }, + { + "epoch": 0.22406, + "grad_norm": 0.876632472083471, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 22406 + }, + { + "epoch": 0.22407, + "grad_norm": 0.8265963571713426, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 22407 + }, + { + "epoch": 0.22408, + "grad_norm": 0.8306847879904058, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 22408 + }, + { + "epoch": 0.22409, + "grad_norm": 0.636235260608773, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 22409 + }, + { + "epoch": 0.2241, + "grad_norm": 0.6872051332908815, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 22410 + }, + { + "epoch": 0.22411, + "grad_norm": 0.7590981659711677, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 22411 + }, + { + "epoch": 0.22412, + "grad_norm": 0.7933189923469889, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 22412 + }, + { + "epoch": 0.22413, + "grad_norm": 0.8543972502509617, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 22413 + }, + { + "epoch": 0.22414, + "grad_norm": 1.114440437854117, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 22414 + }, + { + "epoch": 0.22415, + "grad_norm": 1.0549805267935524, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 22415 + }, + { + "epoch": 0.22416, + "grad_norm": 0.7770877808043389, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 22416 + }, + { + "epoch": 0.22417, + "grad_norm": 0.728821897554459, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 22417 + }, + { + "epoch": 0.22418, + "grad_norm": 0.7896117346229218, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 22418 + }, + { + "epoch": 0.22419, + "grad_norm": 0.7469239338409811, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 22419 + }, + { + "epoch": 0.2242, + "grad_norm": 0.7198232107748622, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 22420 + }, + { + "epoch": 0.22421, + "grad_norm": 0.749412999546654, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 22421 + }, + { + "epoch": 0.22422, + "grad_norm": 0.7889637146719929, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 22422 + }, + { + "epoch": 0.22423, + "grad_norm": 0.6883336965540722, + "learning_rate": 0.003, + "loss": 4.011, + "step": 22423 + }, + { + "epoch": 0.22424, + "grad_norm": 0.7015423711846998, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 22424 + }, + { + "epoch": 0.22425, + "grad_norm": 0.8535032909253535, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 22425 + }, + { + "epoch": 0.22426, + "grad_norm": 1.0035816808822215, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 22426 + }, + { + "epoch": 0.22427, + "grad_norm": 1.026796800861743, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 22427 + }, + { + "epoch": 0.22428, + "grad_norm": 0.8380384890302895, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 22428 + }, + { + "epoch": 0.22429, + "grad_norm": 0.8198084776718808, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 22429 + }, + { + "epoch": 0.2243, + "grad_norm": 0.7246637275437651, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 22430 + }, + { + "epoch": 0.22431, + "grad_norm": 0.7166662947806084, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 22431 + }, + { + "epoch": 0.22432, + "grad_norm": 0.7835080150707694, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22432 + }, + { + "epoch": 0.22433, + "grad_norm": 0.7403519129814988, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 22433 + }, + { + "epoch": 0.22434, + "grad_norm": 0.6416479256153174, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 22434 + }, + { + "epoch": 0.22435, + "grad_norm": 0.5700064873496316, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 22435 + }, + { + "epoch": 0.22436, + "grad_norm": 0.5073474248213582, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 22436 + }, + { + "epoch": 0.22437, + "grad_norm": 0.48448711080230183, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 22437 + }, + { + "epoch": 0.22438, + "grad_norm": 0.47527858696934305, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 22438 + }, + { + "epoch": 0.22439, + "grad_norm": 0.48237055753213687, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 22439 + }, + { + "epoch": 0.2244, + "grad_norm": 0.508403770188866, + "learning_rate": 0.003, + "loss": 4.011, + "step": 22440 + }, + { + "epoch": 0.22441, + "grad_norm": 0.5606845973060485, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 22441 + }, + { + "epoch": 0.22442, + "grad_norm": 0.6638914223961677, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 22442 + }, + { + "epoch": 0.22443, + "grad_norm": 0.674952623285602, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 22443 + }, + { + "epoch": 0.22444, + "grad_norm": 0.6770601383335705, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 22444 + }, + { + "epoch": 0.22445, + "grad_norm": 0.7409174449130481, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 22445 + }, + { + "epoch": 0.22446, + "grad_norm": 0.7885210158375724, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 22446 + }, + { + "epoch": 0.22447, + "grad_norm": 0.8532602429400907, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 22447 + }, + { + "epoch": 0.22448, + "grad_norm": 0.8905978391815984, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 22448 + }, + { + "epoch": 0.22449, + "grad_norm": 0.8235520386562306, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 22449 + }, + { + "epoch": 0.2245, + "grad_norm": 0.7786943898249786, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 22450 + }, + { + "epoch": 0.22451, + "grad_norm": 0.8778584594279184, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 22451 + }, + { + "epoch": 0.22452, + "grad_norm": 0.974104957228737, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 22452 + }, + { + "epoch": 0.22453, + "grad_norm": 1.0815518202039731, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 22453 + }, + { + "epoch": 0.22454, + "grad_norm": 0.9773922709817603, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 22454 + }, + { + "epoch": 0.22455, + "grad_norm": 0.8775634904218896, + "learning_rate": 0.003, + "loss": 3.971, + "step": 22455 + }, + { + "epoch": 0.22456, + "grad_norm": 0.8368329498642146, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 22456 + }, + { + "epoch": 0.22457, + "grad_norm": 0.7372450884840538, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 22457 + }, + { + "epoch": 0.22458, + "grad_norm": 0.7156219891607458, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 22458 + }, + { + "epoch": 0.22459, + "grad_norm": 0.7260481489354553, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 22459 + }, + { + "epoch": 0.2246, + "grad_norm": 0.8072308881930222, + "learning_rate": 0.003, + "loss": 4.019, + "step": 22460 + }, + { + "epoch": 0.22461, + "grad_norm": 0.7716792572724853, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 22461 + }, + { + "epoch": 0.22462, + "grad_norm": 0.6799137000248898, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 22462 + }, + { + "epoch": 0.22463, + "grad_norm": 0.7263433191376755, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 22463 + }, + { + "epoch": 0.22464, + "grad_norm": 0.8322083747857328, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 22464 + }, + { + "epoch": 0.22465, + "grad_norm": 0.9961114961215198, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 22465 + }, + { + "epoch": 0.22466, + "grad_norm": 1.0882270270793897, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 22466 + }, + { + "epoch": 0.22467, + "grad_norm": 0.80434977628652, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 22467 + }, + { + "epoch": 0.22468, + "grad_norm": 0.741486026071878, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 22468 + }, + { + "epoch": 0.22469, + "grad_norm": 0.7407155829850272, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 22469 + }, + { + "epoch": 0.2247, + "grad_norm": 0.6614321891400998, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 22470 + }, + { + "epoch": 0.22471, + "grad_norm": 0.5984205552281395, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 22471 + }, + { + "epoch": 0.22472, + "grad_norm": 0.5841968773116348, + "learning_rate": 0.003, + "loss": 4.0313, + "step": 22472 + }, + { + "epoch": 0.22473, + "grad_norm": 0.6260664909065826, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 22473 + }, + { + "epoch": 0.22474, + "grad_norm": 0.6239930196595647, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 22474 + }, + { + "epoch": 0.22475, + "grad_norm": 0.6493370690358368, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 22475 + }, + { + "epoch": 0.22476, + "grad_norm": 0.8199318469843546, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 22476 + }, + { + "epoch": 0.22477, + "grad_norm": 1.0664454057842156, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 22477 + }, + { + "epoch": 0.22478, + "grad_norm": 0.7921856659057593, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 22478 + }, + { + "epoch": 0.22479, + "grad_norm": 0.5642208395164611, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 22479 + }, + { + "epoch": 0.2248, + "grad_norm": 0.6673444029222201, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 22480 + }, + { + "epoch": 0.22481, + "grad_norm": 0.743413237745133, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 22481 + }, + { + "epoch": 0.22482, + "grad_norm": 0.7353720351163973, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 22482 + }, + { + "epoch": 0.22483, + "grad_norm": 0.7016289134800254, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 22483 + }, + { + "epoch": 0.22484, + "grad_norm": 0.6852730561748013, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 22484 + }, + { + "epoch": 0.22485, + "grad_norm": 0.6889643058977005, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 22485 + }, + { + "epoch": 0.22486, + "grad_norm": 0.6822024603152922, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 22486 + }, + { + "epoch": 0.22487, + "grad_norm": 0.7782586575034508, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 22487 + }, + { + "epoch": 0.22488, + "grad_norm": 0.7448970908093846, + "learning_rate": 0.003, + "loss": 4.019, + "step": 22488 + }, + { + "epoch": 0.22489, + "grad_norm": 0.7189611226715533, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 22489 + }, + { + "epoch": 0.2249, + "grad_norm": 0.7210927214601185, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 22490 + }, + { + "epoch": 0.22491, + "grad_norm": 0.8477260709993183, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 22491 + }, + { + "epoch": 0.22492, + "grad_norm": 0.819733518616579, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 22492 + }, + { + "epoch": 0.22493, + "grad_norm": 0.697907690650963, + "learning_rate": 0.003, + "loss": 3.962, + "step": 22493 + }, + { + "epoch": 0.22494, + "grad_norm": 0.6242389553126454, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 22494 + }, + { + "epoch": 0.22495, + "grad_norm": 0.5529516417694273, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 22495 + }, + { + "epoch": 0.22496, + "grad_norm": 0.6121482001498908, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 22496 + }, + { + "epoch": 0.22497, + "grad_norm": 0.6773770098471831, + "learning_rate": 0.003, + "loss": 3.998, + "step": 22497 + }, + { + "epoch": 0.22498, + "grad_norm": 0.7412715272036118, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 22498 + }, + { + "epoch": 0.22499, + "grad_norm": 0.8637705658509885, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 22499 + }, + { + "epoch": 0.225, + "grad_norm": 0.977052571343058, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 22500 + }, + { + "epoch": 0.22501, + "grad_norm": 0.9886078577108498, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 22501 + }, + { + "epoch": 0.22502, + "grad_norm": 0.8670834397493586, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 22502 + }, + { + "epoch": 0.22503, + "grad_norm": 0.6754314917633857, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 22503 + }, + { + "epoch": 0.22504, + "grad_norm": 0.8645493982973446, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 22504 + }, + { + "epoch": 0.22505, + "grad_norm": 0.8990328978822053, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 22505 + }, + { + "epoch": 0.22506, + "grad_norm": 0.8260923242253777, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 22506 + }, + { + "epoch": 0.22507, + "grad_norm": 0.8336922263048032, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 22507 + }, + { + "epoch": 0.22508, + "grad_norm": 0.9509372303614535, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 22508 + }, + { + "epoch": 0.22509, + "grad_norm": 0.9233049575238654, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 22509 + }, + { + "epoch": 0.2251, + "grad_norm": 0.9257095591771148, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 22510 + }, + { + "epoch": 0.22511, + "grad_norm": 1.0396613454254524, + "learning_rate": 0.003, + "loss": 3.99, + "step": 22511 + }, + { + "epoch": 0.22512, + "grad_norm": 0.9789003946619236, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 22512 + }, + { + "epoch": 0.22513, + "grad_norm": 0.932669619883955, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 22513 + }, + { + "epoch": 0.22514, + "grad_norm": 0.8041592277479777, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 22514 + }, + { + "epoch": 0.22515, + "grad_norm": 0.717052262389654, + "learning_rate": 0.003, + "loss": 4.011, + "step": 22515 + }, + { + "epoch": 0.22516, + "grad_norm": 0.739116082994375, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 22516 + }, + { + "epoch": 0.22517, + "grad_norm": 0.8116470421441548, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 22517 + }, + { + "epoch": 0.22518, + "grad_norm": 0.8725820404561764, + "learning_rate": 0.003, + "loss": 4.005, + "step": 22518 + }, + { + "epoch": 0.22519, + "grad_norm": 0.8879489190205514, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 22519 + }, + { + "epoch": 0.2252, + "grad_norm": 0.9604043083510011, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 22520 + }, + { + "epoch": 0.22521, + "grad_norm": 1.0357401723949928, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 22521 + }, + { + "epoch": 0.22522, + "grad_norm": 0.8123939004383205, + "learning_rate": 0.003, + "loss": 4.012, + "step": 22522 + }, + { + "epoch": 0.22523, + "grad_norm": 0.7447462319251135, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 22523 + }, + { + "epoch": 0.22524, + "grad_norm": 0.832925445667763, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 22524 + }, + { + "epoch": 0.22525, + "grad_norm": 0.8301104691315058, + "learning_rate": 0.003, + "loss": 4.003, + "step": 22525 + }, + { + "epoch": 0.22526, + "grad_norm": 0.9416420370351521, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 22526 + }, + { + "epoch": 0.22527, + "grad_norm": 0.9786259693792503, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 22527 + }, + { + "epoch": 0.22528, + "grad_norm": 0.8732192613206099, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 22528 + }, + { + "epoch": 0.22529, + "grad_norm": 0.6664664009839965, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 22529 + }, + { + "epoch": 0.2253, + "grad_norm": 0.705327965660991, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 22530 + }, + { + "epoch": 0.22531, + "grad_norm": 0.6218940188258794, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 22531 + }, + { + "epoch": 0.22532, + "grad_norm": 0.5846470182440031, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 22532 + }, + { + "epoch": 0.22533, + "grad_norm": 0.6477748940420012, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 22533 + }, + { + "epoch": 0.22534, + "grad_norm": 0.6430688666251444, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 22534 + }, + { + "epoch": 0.22535, + "grad_norm": 0.6397412158640715, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 22535 + }, + { + "epoch": 0.22536, + "grad_norm": 0.6532454254605905, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 22536 + }, + { + "epoch": 0.22537, + "grad_norm": 0.6917886790198319, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 22537 + }, + { + "epoch": 0.22538, + "grad_norm": 0.6413676606324162, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 22538 + }, + { + "epoch": 0.22539, + "grad_norm": 0.6130961029036581, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 22539 + }, + { + "epoch": 0.2254, + "grad_norm": 0.6354030909405154, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 22540 + }, + { + "epoch": 0.22541, + "grad_norm": 0.6200102508247365, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 22541 + }, + { + "epoch": 0.22542, + "grad_norm": 0.6682682596300477, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 22542 + }, + { + "epoch": 0.22543, + "grad_norm": 0.7026575204574528, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 22543 + }, + { + "epoch": 0.22544, + "grad_norm": 0.7872880971012476, + "learning_rate": 0.003, + "loss": 3.978, + "step": 22544 + }, + { + "epoch": 0.22545, + "grad_norm": 0.7313217048328445, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 22545 + }, + { + "epoch": 0.22546, + "grad_norm": 0.5796342864952283, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 22546 + }, + { + "epoch": 0.22547, + "grad_norm": 0.49084249775083555, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 22547 + }, + { + "epoch": 0.22548, + "grad_norm": 0.5278095587486528, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 22548 + }, + { + "epoch": 0.22549, + "grad_norm": 0.6148474358480994, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 22549 + }, + { + "epoch": 0.2255, + "grad_norm": 0.6549157659471305, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 22550 + }, + { + "epoch": 0.22551, + "grad_norm": 0.7016260799362696, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 22551 + }, + { + "epoch": 0.22552, + "grad_norm": 0.800439631731123, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 22552 + }, + { + "epoch": 0.22553, + "grad_norm": 0.9346612782678286, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 22553 + }, + { + "epoch": 0.22554, + "grad_norm": 0.893469211764797, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 22554 + }, + { + "epoch": 0.22555, + "grad_norm": 0.7583153081346318, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 22555 + }, + { + "epoch": 0.22556, + "grad_norm": 0.8481268647926049, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 22556 + }, + { + "epoch": 0.22557, + "grad_norm": 0.7917896301747602, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 22557 + }, + { + "epoch": 0.22558, + "grad_norm": 0.8685475627581483, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 22558 + }, + { + "epoch": 0.22559, + "grad_norm": 0.9854709388670626, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 22559 + }, + { + "epoch": 0.2256, + "grad_norm": 0.9261828931564656, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 22560 + }, + { + "epoch": 0.22561, + "grad_norm": 0.7900209867962065, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 22561 + }, + { + "epoch": 0.22562, + "grad_norm": 0.7636279706038992, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 22562 + }, + { + "epoch": 0.22563, + "grad_norm": 0.7503999694230828, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 22563 + }, + { + "epoch": 0.22564, + "grad_norm": 0.6847930824743871, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 22564 + }, + { + "epoch": 0.22565, + "grad_norm": 0.8047508160206843, + "learning_rate": 0.003, + "loss": 4.03, + "step": 22565 + }, + { + "epoch": 0.22566, + "grad_norm": 0.8514747609164485, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 22566 + }, + { + "epoch": 0.22567, + "grad_norm": 0.767676631284306, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 22567 + }, + { + "epoch": 0.22568, + "grad_norm": 0.7121202866154357, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 22568 + }, + { + "epoch": 0.22569, + "grad_norm": 0.7837483000218511, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 22569 + }, + { + "epoch": 0.2257, + "grad_norm": 0.8465505745934139, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 22570 + }, + { + "epoch": 0.22571, + "grad_norm": 0.9248435401758979, + "learning_rate": 0.003, + "loss": 4.006, + "step": 22571 + }, + { + "epoch": 0.22572, + "grad_norm": 0.8294208447134657, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 22572 + }, + { + "epoch": 0.22573, + "grad_norm": 0.7879908842998289, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 22573 + }, + { + "epoch": 0.22574, + "grad_norm": 0.7833344080050826, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 22574 + }, + { + "epoch": 0.22575, + "grad_norm": 0.637142349768145, + "learning_rate": 0.003, + "loss": 4.006, + "step": 22575 + }, + { + "epoch": 0.22576, + "grad_norm": 0.5455903058639102, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 22576 + }, + { + "epoch": 0.22577, + "grad_norm": 0.5825251236910254, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 22577 + }, + { + "epoch": 0.22578, + "grad_norm": 0.6314482942814564, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 22578 + }, + { + "epoch": 0.22579, + "grad_norm": 0.6109787953893051, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 22579 + }, + { + "epoch": 0.2258, + "grad_norm": 0.5114190599718357, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 22580 + }, + { + "epoch": 0.22581, + "grad_norm": 0.450589046642832, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 22581 + }, + { + "epoch": 0.22582, + "grad_norm": 0.43664238611709855, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 22582 + }, + { + "epoch": 0.22583, + "grad_norm": 0.4123478425911067, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 22583 + }, + { + "epoch": 0.22584, + "grad_norm": 0.4148749752734901, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 22584 + }, + { + "epoch": 0.22585, + "grad_norm": 0.40808885684167484, + "learning_rate": 0.003, + "loss": 4.011, + "step": 22585 + }, + { + "epoch": 0.22586, + "grad_norm": 0.5174876828853513, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 22586 + }, + { + "epoch": 0.22587, + "grad_norm": 0.7269029966670119, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 22587 + }, + { + "epoch": 0.22588, + "grad_norm": 1.0623765107562222, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 22588 + }, + { + "epoch": 0.22589, + "grad_norm": 1.1044886249142265, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 22589 + }, + { + "epoch": 0.2259, + "grad_norm": 0.8672622131577911, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 22590 + }, + { + "epoch": 0.22591, + "grad_norm": 0.7742661506301132, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 22591 + }, + { + "epoch": 0.22592, + "grad_norm": 0.802300545447302, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 22592 + }, + { + "epoch": 0.22593, + "grad_norm": 0.7248761174411897, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 22593 + }, + { + "epoch": 0.22594, + "grad_norm": 0.7310292210974919, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 22594 + }, + { + "epoch": 0.22595, + "grad_norm": 0.7639005792311827, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 22595 + }, + { + "epoch": 0.22596, + "grad_norm": 0.9003093701438271, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 22596 + }, + { + "epoch": 0.22597, + "grad_norm": 0.9264016968649862, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 22597 + }, + { + "epoch": 0.22598, + "grad_norm": 0.8819344264013732, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 22598 + }, + { + "epoch": 0.22599, + "grad_norm": 0.8641081364803778, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 22599 + }, + { + "epoch": 0.226, + "grad_norm": 1.0506778984492822, + "learning_rate": 0.003, + "loss": 4.019, + "step": 22600 + }, + { + "epoch": 0.22601, + "grad_norm": 1.0192241656334518, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 22601 + }, + { + "epoch": 0.22602, + "grad_norm": 0.9671938067779788, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 22602 + }, + { + "epoch": 0.22603, + "grad_norm": 0.9409531827980432, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 22603 + }, + { + "epoch": 0.22604, + "grad_norm": 1.0469600300059285, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 22604 + }, + { + "epoch": 0.22605, + "grad_norm": 0.9821750962693918, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 22605 + }, + { + "epoch": 0.22606, + "grad_norm": 1.0079469948652202, + "learning_rate": 0.003, + "loss": 4.0442, + "step": 22606 + }, + { + "epoch": 0.22607, + "grad_norm": 1.0854170683610143, + "learning_rate": 0.003, + "loss": 4.029, + "step": 22607 + }, + { + "epoch": 0.22608, + "grad_norm": 0.9605185478935105, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 22608 + }, + { + "epoch": 0.22609, + "grad_norm": 0.8767427629291316, + "learning_rate": 0.003, + "loss": 4.051, + "step": 22609 + }, + { + "epoch": 0.2261, + "grad_norm": 0.7851541437915184, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 22610 + }, + { + "epoch": 0.22611, + "grad_norm": 0.7802136738398305, + "learning_rate": 0.003, + "loss": 3.98, + "step": 22611 + }, + { + "epoch": 0.22612, + "grad_norm": 0.7460684502323357, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 22612 + }, + { + "epoch": 0.22613, + "grad_norm": 0.7288819483940361, + "learning_rate": 0.003, + "loss": 4.0449, + "step": 22613 + }, + { + "epoch": 0.22614, + "grad_norm": 0.7579453252302605, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 22614 + }, + { + "epoch": 0.22615, + "grad_norm": 0.6986261346753035, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 22615 + }, + { + "epoch": 0.22616, + "grad_norm": 0.6477781818131898, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 22616 + }, + { + "epoch": 0.22617, + "grad_norm": 0.6828910058649335, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 22617 + }, + { + "epoch": 0.22618, + "grad_norm": 0.6833774437834151, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 22618 + }, + { + "epoch": 0.22619, + "grad_norm": 0.5069771763853067, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 22619 + }, + { + "epoch": 0.2262, + "grad_norm": 0.48157232917697557, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 22620 + }, + { + "epoch": 0.22621, + "grad_norm": 0.4987028471399979, + "learning_rate": 0.003, + "loss": 3.989, + "step": 22621 + }, + { + "epoch": 0.22622, + "grad_norm": 0.5264337595969121, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 22622 + }, + { + "epoch": 0.22623, + "grad_norm": 0.5306491614706427, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 22623 + }, + { + "epoch": 0.22624, + "grad_norm": 0.49294575109901606, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 22624 + }, + { + "epoch": 0.22625, + "grad_norm": 0.5366896198074036, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 22625 + }, + { + "epoch": 0.22626, + "grad_norm": 0.5842432791869548, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 22626 + }, + { + "epoch": 0.22627, + "grad_norm": 0.6444020882744582, + "learning_rate": 0.003, + "loss": 3.989, + "step": 22627 + }, + { + "epoch": 0.22628, + "grad_norm": 0.8906662784570227, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 22628 + }, + { + "epoch": 0.22629, + "grad_norm": 1.201804442771889, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 22629 + }, + { + "epoch": 0.2263, + "grad_norm": 0.6715366909761195, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 22630 + }, + { + "epoch": 0.22631, + "grad_norm": 0.714049657858352, + "learning_rate": 0.003, + "loss": 4.016, + "step": 22631 + }, + { + "epoch": 0.22632, + "grad_norm": 0.8079459749562677, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 22632 + }, + { + "epoch": 0.22633, + "grad_norm": 0.7666196005295447, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 22633 + }, + { + "epoch": 0.22634, + "grad_norm": 0.6986028958046777, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 22634 + }, + { + "epoch": 0.22635, + "grad_norm": 0.6991698771956105, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 22635 + }, + { + "epoch": 0.22636, + "grad_norm": 0.6209288486739503, + "learning_rate": 0.003, + "loss": 3.995, + "step": 22636 + }, + { + "epoch": 0.22637, + "grad_norm": 0.6508279046921469, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 22637 + }, + { + "epoch": 0.22638, + "grad_norm": 0.7137889639861856, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 22638 + }, + { + "epoch": 0.22639, + "grad_norm": 0.6865061998634012, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 22639 + }, + { + "epoch": 0.2264, + "grad_norm": 0.6668067023967985, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 22640 + }, + { + "epoch": 0.22641, + "grad_norm": 0.7098223527621257, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 22641 + }, + { + "epoch": 0.22642, + "grad_norm": 0.6560625394674194, + "learning_rate": 0.003, + "loss": 3.992, + "step": 22642 + }, + { + "epoch": 0.22643, + "grad_norm": 0.5572338126712396, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 22643 + }, + { + "epoch": 0.22644, + "grad_norm": 0.5538791142843001, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 22644 + }, + { + "epoch": 0.22645, + "grad_norm": 0.574129556478329, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 22645 + }, + { + "epoch": 0.22646, + "grad_norm": 0.6395225264722296, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 22646 + }, + { + "epoch": 0.22647, + "grad_norm": 0.6075832589231626, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 22647 + }, + { + "epoch": 0.22648, + "grad_norm": 0.4849558527989291, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 22648 + }, + { + "epoch": 0.22649, + "grad_norm": 0.44421288946648124, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 22649 + }, + { + "epoch": 0.2265, + "grad_norm": 0.5666314629660768, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 22650 + }, + { + "epoch": 0.22651, + "grad_norm": 0.7558707217406732, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 22651 + }, + { + "epoch": 0.22652, + "grad_norm": 0.9318964848885707, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 22652 + }, + { + "epoch": 0.22653, + "grad_norm": 1.040380445361711, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 22653 + }, + { + "epoch": 0.22654, + "grad_norm": 0.9932935839447256, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 22654 + }, + { + "epoch": 0.22655, + "grad_norm": 0.9332187545734434, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 22655 + }, + { + "epoch": 0.22656, + "grad_norm": 0.8911429108439112, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 22656 + }, + { + "epoch": 0.22657, + "grad_norm": 0.8745364130394142, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 22657 + }, + { + "epoch": 0.22658, + "grad_norm": 0.868791433291614, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 22658 + }, + { + "epoch": 0.22659, + "grad_norm": 0.9613574510744753, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 22659 + }, + { + "epoch": 0.2266, + "grad_norm": 1.0634605360549205, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 22660 + }, + { + "epoch": 0.22661, + "grad_norm": 1.0180575832527794, + "learning_rate": 0.003, + "loss": 4.0423, + "step": 22661 + }, + { + "epoch": 0.22662, + "grad_norm": 1.1391837305001025, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 22662 + }, + { + "epoch": 0.22663, + "grad_norm": 0.852939665513751, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 22663 + }, + { + "epoch": 0.22664, + "grad_norm": 0.8472508437875075, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 22664 + }, + { + "epoch": 0.22665, + "grad_norm": 0.8866053234136446, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 22665 + }, + { + "epoch": 0.22666, + "grad_norm": 0.7809635338621516, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 22666 + }, + { + "epoch": 0.22667, + "grad_norm": 0.8942514273656881, + "learning_rate": 0.003, + "loss": 4.008, + "step": 22667 + }, + { + "epoch": 0.22668, + "grad_norm": 0.9422756775920295, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 22668 + }, + { + "epoch": 0.22669, + "grad_norm": 0.8205690783633957, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 22669 + }, + { + "epoch": 0.2267, + "grad_norm": 0.882815967635574, + "learning_rate": 0.003, + "loss": 4.01, + "step": 22670 + }, + { + "epoch": 0.22671, + "grad_norm": 0.8389028493774299, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 22671 + }, + { + "epoch": 0.22672, + "grad_norm": 0.847341390356432, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 22672 + }, + { + "epoch": 0.22673, + "grad_norm": 0.8086661720292744, + "learning_rate": 0.003, + "loss": 3.977, + "step": 22673 + }, + { + "epoch": 0.22674, + "grad_norm": 0.8278040241852577, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 22674 + }, + { + "epoch": 0.22675, + "grad_norm": 0.9469823660054406, + "learning_rate": 0.003, + "loss": 4.0446, + "step": 22675 + }, + { + "epoch": 0.22676, + "grad_norm": 0.8626122999059631, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 22676 + }, + { + "epoch": 0.22677, + "grad_norm": 0.8623437431667066, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 22677 + }, + { + "epoch": 0.22678, + "grad_norm": 0.9328144180130987, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 22678 + }, + { + "epoch": 0.22679, + "grad_norm": 0.8686037848380999, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 22679 + }, + { + "epoch": 0.2268, + "grad_norm": 0.8156982392100295, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 22680 + }, + { + "epoch": 0.22681, + "grad_norm": 0.6962405366116252, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 22681 + }, + { + "epoch": 0.22682, + "grad_norm": 0.625813614688267, + "learning_rate": 0.003, + "loss": 4.006, + "step": 22682 + }, + { + "epoch": 0.22683, + "grad_norm": 0.6354177830988871, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 22683 + }, + { + "epoch": 0.22684, + "grad_norm": 0.5587346523725433, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 22684 + }, + { + "epoch": 0.22685, + "grad_norm": 0.553336912619847, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 22685 + }, + { + "epoch": 0.22686, + "grad_norm": 0.583247733372453, + "learning_rate": 0.003, + "loss": 3.981, + "step": 22686 + }, + { + "epoch": 0.22687, + "grad_norm": 0.76197369950693, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 22687 + }, + { + "epoch": 0.22688, + "grad_norm": 0.9515320714697402, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 22688 + }, + { + "epoch": 0.22689, + "grad_norm": 0.8715589939600175, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 22689 + }, + { + "epoch": 0.2269, + "grad_norm": 0.663169651118655, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 22690 + }, + { + "epoch": 0.22691, + "grad_norm": 0.5549875722356021, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 22691 + }, + { + "epoch": 0.22692, + "grad_norm": 0.6695332483676788, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 22692 + }, + { + "epoch": 0.22693, + "grad_norm": 0.6848362836414673, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 22693 + }, + { + "epoch": 0.22694, + "grad_norm": 0.6071263368063606, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 22694 + }, + { + "epoch": 0.22695, + "grad_norm": 0.6421514077456835, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 22695 + }, + { + "epoch": 0.22696, + "grad_norm": 0.6545127094135431, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 22696 + }, + { + "epoch": 0.22697, + "grad_norm": 0.5543977374028638, + "learning_rate": 0.003, + "loss": 4.016, + "step": 22697 + }, + { + "epoch": 0.22698, + "grad_norm": 0.48223185308021194, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 22698 + }, + { + "epoch": 0.22699, + "grad_norm": 0.4806217346689214, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 22699 + }, + { + "epoch": 0.227, + "grad_norm": 0.4786821420396655, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 22700 + }, + { + "epoch": 0.22701, + "grad_norm": 0.5741908232675801, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 22701 + }, + { + "epoch": 0.22702, + "grad_norm": 0.6040360783866373, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 22702 + }, + { + "epoch": 0.22703, + "grad_norm": 0.6037325114042851, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 22703 + }, + { + "epoch": 0.22704, + "grad_norm": 0.6187864669634895, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 22704 + }, + { + "epoch": 0.22705, + "grad_norm": 0.6522530778346195, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 22705 + }, + { + "epoch": 0.22706, + "grad_norm": 0.6389316828687981, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 22706 + }, + { + "epoch": 0.22707, + "grad_norm": 0.5712045531686754, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 22707 + }, + { + "epoch": 0.22708, + "grad_norm": 0.7285386513586384, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 22708 + }, + { + "epoch": 0.22709, + "grad_norm": 0.8600638466673659, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 22709 + }, + { + "epoch": 0.2271, + "grad_norm": 0.9720643334815988, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 22710 + }, + { + "epoch": 0.22711, + "grad_norm": 1.214309223035382, + "learning_rate": 0.003, + "loss": 4.004, + "step": 22711 + }, + { + "epoch": 0.22712, + "grad_norm": 0.7792785550232337, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 22712 + }, + { + "epoch": 0.22713, + "grad_norm": 0.6101510507664502, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 22713 + }, + { + "epoch": 0.22714, + "grad_norm": 0.7184587714513144, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 22714 + }, + { + "epoch": 0.22715, + "grad_norm": 0.7848703925562167, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 22715 + }, + { + "epoch": 0.22716, + "grad_norm": 0.7980679220252158, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 22716 + }, + { + "epoch": 0.22717, + "grad_norm": 0.6836137842103448, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 22717 + }, + { + "epoch": 0.22718, + "grad_norm": 0.6681298913311627, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 22718 + }, + { + "epoch": 0.22719, + "grad_norm": 0.6681173552716717, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 22719 + }, + { + "epoch": 0.2272, + "grad_norm": 0.7811588765220844, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 22720 + }, + { + "epoch": 0.22721, + "grad_norm": 0.7579586784505247, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 22721 + }, + { + "epoch": 0.22722, + "grad_norm": 0.5993966609403294, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 22722 + }, + { + "epoch": 0.22723, + "grad_norm": 0.6623561579737576, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 22723 + }, + { + "epoch": 0.22724, + "grad_norm": 0.6638712118048304, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 22724 + }, + { + "epoch": 0.22725, + "grad_norm": 0.7321691527166023, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 22725 + }, + { + "epoch": 0.22726, + "grad_norm": 0.7547053024934567, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 22726 + }, + { + "epoch": 0.22727, + "grad_norm": 0.8651083243537319, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 22727 + }, + { + "epoch": 0.22728, + "grad_norm": 0.9338629227617653, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 22728 + }, + { + "epoch": 0.22729, + "grad_norm": 0.9479037697582136, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 22729 + }, + { + "epoch": 0.2273, + "grad_norm": 1.0818191777388744, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 22730 + }, + { + "epoch": 0.22731, + "grad_norm": 0.8762766274041155, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 22731 + }, + { + "epoch": 0.22732, + "grad_norm": 0.7614168562602789, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 22732 + }, + { + "epoch": 0.22733, + "grad_norm": 0.6650265660778039, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 22733 + }, + { + "epoch": 0.22734, + "grad_norm": 0.7176608944264959, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 22734 + }, + { + "epoch": 0.22735, + "grad_norm": 0.7654209467162899, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 22735 + }, + { + "epoch": 0.22736, + "grad_norm": 0.9621625037508562, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 22736 + }, + { + "epoch": 0.22737, + "grad_norm": 1.0552429124237788, + "learning_rate": 0.003, + "loss": 4.011, + "step": 22737 + }, + { + "epoch": 0.22738, + "grad_norm": 0.8947991375792034, + "learning_rate": 0.003, + "loss": 4.0382, + "step": 22738 + }, + { + "epoch": 0.22739, + "grad_norm": 0.8682497665210865, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 22739 + }, + { + "epoch": 0.2274, + "grad_norm": 0.7115196822666884, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 22740 + }, + { + "epoch": 0.22741, + "grad_norm": 0.6740998855078567, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 22741 + }, + { + "epoch": 0.22742, + "grad_norm": 0.6621186807742948, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 22742 + }, + { + "epoch": 0.22743, + "grad_norm": 0.7028651290236644, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 22743 + }, + { + "epoch": 0.22744, + "grad_norm": 0.7555585346376706, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 22744 + }, + { + "epoch": 0.22745, + "grad_norm": 0.7528579567254485, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 22745 + }, + { + "epoch": 0.22746, + "grad_norm": 0.7420836536796951, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 22746 + }, + { + "epoch": 0.22747, + "grad_norm": 0.6938822326357301, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 22747 + }, + { + "epoch": 0.22748, + "grad_norm": 0.9087273898178284, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 22748 + }, + { + "epoch": 0.22749, + "grad_norm": 0.9798884165376156, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 22749 + }, + { + "epoch": 0.2275, + "grad_norm": 1.1533018101052726, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 22750 + }, + { + "epoch": 0.22751, + "grad_norm": 1.0661694724380033, + "learning_rate": 0.003, + "loss": 4.012, + "step": 22751 + }, + { + "epoch": 0.22752, + "grad_norm": 0.7052653007881202, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 22752 + }, + { + "epoch": 0.22753, + "grad_norm": 0.6074627038895961, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 22753 + }, + { + "epoch": 0.22754, + "grad_norm": 0.5999375911078667, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 22754 + }, + { + "epoch": 0.22755, + "grad_norm": 0.5888493950803809, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 22755 + }, + { + "epoch": 0.22756, + "grad_norm": 0.5917944589684301, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 22756 + }, + { + "epoch": 0.22757, + "grad_norm": 0.579001825052004, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 22757 + }, + { + "epoch": 0.22758, + "grad_norm": 0.591070533274092, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 22758 + }, + { + "epoch": 0.22759, + "grad_norm": 0.5762621142033109, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 22759 + }, + { + "epoch": 0.2276, + "grad_norm": 0.5377622967344864, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 22760 + }, + { + "epoch": 0.22761, + "grad_norm": 0.5759657456546746, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 22761 + }, + { + "epoch": 0.22762, + "grad_norm": 0.6480163895859515, + "learning_rate": 0.003, + "loss": 4.023, + "step": 22762 + }, + { + "epoch": 0.22763, + "grad_norm": 0.7006789310713918, + "learning_rate": 0.003, + "loss": 4.007, + "step": 22763 + }, + { + "epoch": 0.22764, + "grad_norm": 0.6413897357306794, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 22764 + }, + { + "epoch": 0.22765, + "grad_norm": 0.6036870547463824, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 22765 + }, + { + "epoch": 0.22766, + "grad_norm": 0.632662590042417, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 22766 + }, + { + "epoch": 0.22767, + "grad_norm": 0.7601863137646737, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 22767 + }, + { + "epoch": 0.22768, + "grad_norm": 0.8225510079336076, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 22768 + }, + { + "epoch": 0.22769, + "grad_norm": 0.8033825192953095, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 22769 + }, + { + "epoch": 0.2277, + "grad_norm": 0.8562047625064612, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 22770 + }, + { + "epoch": 0.22771, + "grad_norm": 0.9302360088404079, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 22771 + }, + { + "epoch": 0.22772, + "grad_norm": 0.8910011836854165, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 22772 + }, + { + "epoch": 0.22773, + "grad_norm": 0.6375948939174361, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 22773 + }, + { + "epoch": 0.22774, + "grad_norm": 0.6581721521266831, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 22774 + }, + { + "epoch": 0.22775, + "grad_norm": 0.7384144090061283, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 22775 + }, + { + "epoch": 0.22776, + "grad_norm": 0.7309457804162454, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 22776 + }, + { + "epoch": 0.22777, + "grad_norm": 0.7454172859295869, + "learning_rate": 0.003, + "loss": 3.997, + "step": 22777 + }, + { + "epoch": 0.22778, + "grad_norm": 0.7780442973921022, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 22778 + }, + { + "epoch": 0.22779, + "grad_norm": 0.8303571249606351, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 22779 + }, + { + "epoch": 0.2278, + "grad_norm": 1.0246195464013237, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 22780 + }, + { + "epoch": 0.22781, + "grad_norm": 0.9481222512748069, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 22781 + }, + { + "epoch": 0.22782, + "grad_norm": 0.9695332365464543, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 22782 + }, + { + "epoch": 0.22783, + "grad_norm": 0.9555102036272847, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 22783 + }, + { + "epoch": 0.22784, + "grad_norm": 0.8911926928612898, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 22784 + }, + { + "epoch": 0.22785, + "grad_norm": 0.7698486782630193, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 22785 + }, + { + "epoch": 0.22786, + "grad_norm": 0.7211328457147901, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 22786 + }, + { + "epoch": 0.22787, + "grad_norm": 0.8073826399531063, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 22787 + }, + { + "epoch": 0.22788, + "grad_norm": 0.8642426279705305, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 22788 + }, + { + "epoch": 0.22789, + "grad_norm": 0.9095930557042402, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 22789 + }, + { + "epoch": 0.2279, + "grad_norm": 0.7956226108569149, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 22790 + }, + { + "epoch": 0.22791, + "grad_norm": 0.7099793004825958, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 22791 + }, + { + "epoch": 0.22792, + "grad_norm": 0.7818771737269793, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 22792 + }, + { + "epoch": 0.22793, + "grad_norm": 0.624846255204313, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 22793 + }, + { + "epoch": 0.22794, + "grad_norm": 0.6715397502426942, + "learning_rate": 0.003, + "loss": 3.998, + "step": 22794 + }, + { + "epoch": 0.22795, + "grad_norm": 0.653040050747009, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 22795 + }, + { + "epoch": 0.22796, + "grad_norm": 0.7280451899769068, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 22796 + }, + { + "epoch": 0.22797, + "grad_norm": 0.8222313372147919, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 22797 + }, + { + "epoch": 0.22798, + "grad_norm": 0.8663307508855099, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 22798 + }, + { + "epoch": 0.22799, + "grad_norm": 0.9407704139655766, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 22799 + }, + { + "epoch": 0.228, + "grad_norm": 1.026928611773055, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 22800 + }, + { + "epoch": 0.22801, + "grad_norm": 0.7997088300880091, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 22801 + }, + { + "epoch": 0.22802, + "grad_norm": 0.6859061118438088, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 22802 + }, + { + "epoch": 0.22803, + "grad_norm": 0.7408871475351876, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 22803 + }, + { + "epoch": 0.22804, + "grad_norm": 0.8335762835550706, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 22804 + }, + { + "epoch": 0.22805, + "grad_norm": 0.8398900421164578, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 22805 + }, + { + "epoch": 0.22806, + "grad_norm": 0.8414465039647213, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 22806 + }, + { + "epoch": 0.22807, + "grad_norm": 0.8752950934304009, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 22807 + }, + { + "epoch": 0.22808, + "grad_norm": 0.9237619260900455, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 22808 + }, + { + "epoch": 0.22809, + "grad_norm": 0.8831057924518385, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 22809 + }, + { + "epoch": 0.2281, + "grad_norm": 0.7181492030563533, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 22810 + }, + { + "epoch": 0.22811, + "grad_norm": 0.7345941240068794, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 22811 + }, + { + "epoch": 0.22812, + "grad_norm": 0.7277629414140034, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 22812 + }, + { + "epoch": 0.22813, + "grad_norm": 0.6047787772811198, + "learning_rate": 0.003, + "loss": 3.98, + "step": 22813 + }, + { + "epoch": 0.22814, + "grad_norm": 0.6614461698726722, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 22814 + }, + { + "epoch": 0.22815, + "grad_norm": 0.6554097771526453, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 22815 + }, + { + "epoch": 0.22816, + "grad_norm": 0.6333034257002369, + "learning_rate": 0.003, + "loss": 4.013, + "step": 22816 + }, + { + "epoch": 0.22817, + "grad_norm": 0.6325115516294476, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 22817 + }, + { + "epoch": 0.22818, + "grad_norm": 0.7455881292522059, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 22818 + }, + { + "epoch": 0.22819, + "grad_norm": 0.6774377840749052, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 22819 + }, + { + "epoch": 0.2282, + "grad_norm": 0.6335999357179946, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 22820 + }, + { + "epoch": 0.22821, + "grad_norm": 0.5676815513625012, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 22821 + }, + { + "epoch": 0.22822, + "grad_norm": 0.5399377763561373, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 22822 + }, + { + "epoch": 0.22823, + "grad_norm": 0.5504329496378643, + "learning_rate": 0.003, + "loss": 3.986, + "step": 22823 + }, + { + "epoch": 0.22824, + "grad_norm": 0.5234356430458693, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 22824 + }, + { + "epoch": 0.22825, + "grad_norm": 0.6056809455539456, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 22825 + }, + { + "epoch": 0.22826, + "grad_norm": 0.7164094748923573, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 22826 + }, + { + "epoch": 0.22827, + "grad_norm": 0.8296825549320939, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 22827 + }, + { + "epoch": 0.22828, + "grad_norm": 0.8679540454356481, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 22828 + }, + { + "epoch": 0.22829, + "grad_norm": 0.8469268586690786, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 22829 + }, + { + "epoch": 0.2283, + "grad_norm": 0.7642927515290029, + "learning_rate": 0.003, + "loss": 3.997, + "step": 22830 + }, + { + "epoch": 0.22831, + "grad_norm": 0.6931593445712981, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 22831 + }, + { + "epoch": 0.22832, + "grad_norm": 0.6722104926896562, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 22832 + }, + { + "epoch": 0.22833, + "grad_norm": 0.791822915284931, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 22833 + }, + { + "epoch": 0.22834, + "grad_norm": 0.8340974856187494, + "learning_rate": 0.003, + "loss": 4.021, + "step": 22834 + }, + { + "epoch": 0.22835, + "grad_norm": 0.8831407625510378, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 22835 + }, + { + "epoch": 0.22836, + "grad_norm": 0.8091440533203766, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 22836 + }, + { + "epoch": 0.22837, + "grad_norm": 0.716489027732392, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 22837 + }, + { + "epoch": 0.22838, + "grad_norm": 0.6297832306585904, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 22838 + }, + { + "epoch": 0.22839, + "grad_norm": 0.5851109519646929, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 22839 + }, + { + "epoch": 0.2284, + "grad_norm": 0.518260607414325, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 22840 + }, + { + "epoch": 0.22841, + "grad_norm": 0.5700504214349338, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 22841 + }, + { + "epoch": 0.22842, + "grad_norm": 0.6002322005857612, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 22842 + }, + { + "epoch": 0.22843, + "grad_norm": 0.754949411683011, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 22843 + }, + { + "epoch": 0.22844, + "grad_norm": 0.8807842528520267, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 22844 + }, + { + "epoch": 0.22845, + "grad_norm": 0.9003986424085737, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 22845 + }, + { + "epoch": 0.22846, + "grad_norm": 0.7355730209927843, + "learning_rate": 0.003, + "loss": 3.984, + "step": 22846 + }, + { + "epoch": 0.22847, + "grad_norm": 0.7338002585302497, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 22847 + }, + { + "epoch": 0.22848, + "grad_norm": 0.7978257361123261, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 22848 + }, + { + "epoch": 0.22849, + "grad_norm": 0.8229045971082031, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 22849 + }, + { + "epoch": 0.2285, + "grad_norm": 0.7831404962637342, + "learning_rate": 0.003, + "loss": 3.999, + "step": 22850 + }, + { + "epoch": 0.22851, + "grad_norm": 0.8876985553798392, + "learning_rate": 0.003, + "loss": 3.993, + "step": 22851 + }, + { + "epoch": 0.22852, + "grad_norm": 0.7672787387974402, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 22852 + }, + { + "epoch": 0.22853, + "grad_norm": 0.7423516061756352, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 22853 + }, + { + "epoch": 0.22854, + "grad_norm": 0.7957853358870686, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 22854 + }, + { + "epoch": 0.22855, + "grad_norm": 0.8936665964512039, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 22855 + }, + { + "epoch": 0.22856, + "grad_norm": 0.9408424640619528, + "learning_rate": 0.003, + "loss": 4.0425, + "step": 22856 + }, + { + "epoch": 0.22857, + "grad_norm": 0.9299690960498221, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 22857 + }, + { + "epoch": 0.22858, + "grad_norm": 0.748787035141101, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 22858 + }, + { + "epoch": 0.22859, + "grad_norm": 0.7974609015068357, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 22859 + }, + { + "epoch": 0.2286, + "grad_norm": 0.8503775102873501, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 22860 + }, + { + "epoch": 0.22861, + "grad_norm": 0.7829904731670801, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 22861 + }, + { + "epoch": 0.22862, + "grad_norm": 0.8120050659949063, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 22862 + }, + { + "epoch": 0.22863, + "grad_norm": 0.8320563283929416, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 22863 + }, + { + "epoch": 0.22864, + "grad_norm": 0.8270823504564249, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 22864 + }, + { + "epoch": 0.22865, + "grad_norm": 0.9421437853720384, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 22865 + }, + { + "epoch": 0.22866, + "grad_norm": 0.9492544923885538, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 22866 + }, + { + "epoch": 0.22867, + "grad_norm": 0.9639972625679915, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 22867 + }, + { + "epoch": 0.22868, + "grad_norm": 0.9528737368835686, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 22868 + }, + { + "epoch": 0.22869, + "grad_norm": 0.8747491945153706, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 22869 + }, + { + "epoch": 0.2287, + "grad_norm": 0.6847690224035201, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 22870 + }, + { + "epoch": 0.22871, + "grad_norm": 0.706269732557421, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 22871 + }, + { + "epoch": 0.22872, + "grad_norm": 0.6810035512228381, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 22872 + }, + { + "epoch": 0.22873, + "grad_norm": 0.7109780233766005, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 22873 + }, + { + "epoch": 0.22874, + "grad_norm": 0.6893191260962264, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 22874 + }, + { + "epoch": 0.22875, + "grad_norm": 0.6244494577754074, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 22875 + }, + { + "epoch": 0.22876, + "grad_norm": 0.5377158694278087, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 22876 + }, + { + "epoch": 0.22877, + "grad_norm": 0.43764980924223457, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 22877 + }, + { + "epoch": 0.22878, + "grad_norm": 0.5412439326947098, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 22878 + }, + { + "epoch": 0.22879, + "grad_norm": 0.5660718553127629, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 22879 + }, + { + "epoch": 0.2288, + "grad_norm": 0.6477007847048568, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 22880 + }, + { + "epoch": 0.22881, + "grad_norm": 0.6393563984344451, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 22881 + }, + { + "epoch": 0.22882, + "grad_norm": 0.777583862223286, + "learning_rate": 0.003, + "loss": 3.999, + "step": 22882 + }, + { + "epoch": 0.22883, + "grad_norm": 0.9183434758325114, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 22883 + }, + { + "epoch": 0.22884, + "grad_norm": 0.9452010257070995, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 22884 + }, + { + "epoch": 0.22885, + "grad_norm": 0.8010514280226858, + "learning_rate": 0.003, + "loss": 3.997, + "step": 22885 + }, + { + "epoch": 0.22886, + "grad_norm": 0.7546110610486247, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 22886 + }, + { + "epoch": 0.22887, + "grad_norm": 0.848967593028541, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 22887 + }, + { + "epoch": 0.22888, + "grad_norm": 0.7885453850356119, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 22888 + }, + { + "epoch": 0.22889, + "grad_norm": 0.763741237422184, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 22889 + }, + { + "epoch": 0.2289, + "grad_norm": 0.7923632901018216, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 22890 + }, + { + "epoch": 0.22891, + "grad_norm": 0.8138596080115095, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 22891 + }, + { + "epoch": 0.22892, + "grad_norm": 0.7882120239682286, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 22892 + }, + { + "epoch": 0.22893, + "grad_norm": 0.7460947384734385, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 22893 + }, + { + "epoch": 0.22894, + "grad_norm": 0.7360920515416278, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 22894 + }, + { + "epoch": 0.22895, + "grad_norm": 0.7468639916913005, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 22895 + }, + { + "epoch": 0.22896, + "grad_norm": 0.7337770345723237, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 22896 + }, + { + "epoch": 0.22897, + "grad_norm": 0.7507217631370945, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 22897 + }, + { + "epoch": 0.22898, + "grad_norm": 0.7442497595348619, + "learning_rate": 0.003, + "loss": 3.983, + "step": 22898 + }, + { + "epoch": 0.22899, + "grad_norm": 0.8599777959649892, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 22899 + }, + { + "epoch": 0.229, + "grad_norm": 0.9923619851327129, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 22900 + }, + { + "epoch": 0.22901, + "grad_norm": 1.1319143299233443, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 22901 + }, + { + "epoch": 0.22902, + "grad_norm": 0.8041917826170554, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 22902 + }, + { + "epoch": 0.22903, + "grad_norm": 0.7238682829411925, + "learning_rate": 0.003, + "loss": 4.028, + "step": 22903 + }, + { + "epoch": 0.22904, + "grad_norm": 0.7985803360116025, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 22904 + }, + { + "epoch": 0.22905, + "grad_norm": 0.7961571197223262, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 22905 + }, + { + "epoch": 0.22906, + "grad_norm": 0.7980300947623346, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 22906 + }, + { + "epoch": 0.22907, + "grad_norm": 0.6972412652797453, + "learning_rate": 0.003, + "loss": 4.008, + "step": 22907 + }, + { + "epoch": 0.22908, + "grad_norm": 0.6826005676616897, + "learning_rate": 0.003, + "loss": 4.004, + "step": 22908 + }, + { + "epoch": 0.22909, + "grad_norm": 0.6441061399090826, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 22909 + }, + { + "epoch": 0.2291, + "grad_norm": 0.7242239601167967, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 22910 + }, + { + "epoch": 0.22911, + "grad_norm": 0.7812386571205873, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 22911 + }, + { + "epoch": 0.22912, + "grad_norm": 0.7554160137377134, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 22912 + }, + { + "epoch": 0.22913, + "grad_norm": 0.7372000085951794, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 22913 + }, + { + "epoch": 0.22914, + "grad_norm": 0.6933171691836245, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 22914 + }, + { + "epoch": 0.22915, + "grad_norm": 0.5617112210942524, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 22915 + }, + { + "epoch": 0.22916, + "grad_norm": 0.5083893892070983, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 22916 + }, + { + "epoch": 0.22917, + "grad_norm": 0.5170641604058251, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 22917 + }, + { + "epoch": 0.22918, + "grad_norm": 0.5269404159799539, + "learning_rate": 0.003, + "loss": 4.011, + "step": 22918 + }, + { + "epoch": 0.22919, + "grad_norm": 0.5886073445237835, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 22919 + }, + { + "epoch": 0.2292, + "grad_norm": 0.6188249758998293, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 22920 + }, + { + "epoch": 0.22921, + "grad_norm": 0.6305351147756024, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 22921 + }, + { + "epoch": 0.22922, + "grad_norm": 0.6512307741025783, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 22922 + }, + { + "epoch": 0.22923, + "grad_norm": 0.6388473026826896, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 22923 + }, + { + "epoch": 0.22924, + "grad_norm": 0.6010982047986861, + "learning_rate": 0.003, + "loss": 3.972, + "step": 22924 + }, + { + "epoch": 0.22925, + "grad_norm": 0.5835410529554005, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 22925 + }, + { + "epoch": 0.22926, + "grad_norm": 0.6500029153493979, + "learning_rate": 0.003, + "loss": 4.014, + "step": 22926 + }, + { + "epoch": 0.22927, + "grad_norm": 0.7847204503522849, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 22927 + }, + { + "epoch": 0.22928, + "grad_norm": 0.8420644744093517, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 22928 + }, + { + "epoch": 0.22929, + "grad_norm": 0.8425177074665102, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 22929 + }, + { + "epoch": 0.2293, + "grad_norm": 0.7387918060702379, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 22930 + }, + { + "epoch": 0.22931, + "grad_norm": 0.7182126590059609, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 22931 + }, + { + "epoch": 0.22932, + "grad_norm": 0.7488917437252003, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 22932 + }, + { + "epoch": 0.22933, + "grad_norm": 0.8648696553924385, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 22933 + }, + { + "epoch": 0.22934, + "grad_norm": 1.0331650847330105, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 22934 + }, + { + "epoch": 0.22935, + "grad_norm": 0.9204062086673132, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 22935 + }, + { + "epoch": 0.22936, + "grad_norm": 0.8854257241881544, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 22936 + }, + { + "epoch": 0.22937, + "grad_norm": 0.7439929146848856, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 22937 + }, + { + "epoch": 0.22938, + "grad_norm": 0.5528700007161388, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 22938 + }, + { + "epoch": 0.22939, + "grad_norm": 0.6495643574745311, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 22939 + }, + { + "epoch": 0.2294, + "grad_norm": 0.6655720982112963, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 22940 + }, + { + "epoch": 0.22941, + "grad_norm": 0.7254681511635923, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 22941 + }, + { + "epoch": 0.22942, + "grad_norm": 0.7064630836262875, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 22942 + }, + { + "epoch": 0.22943, + "grad_norm": 0.755669391171973, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 22943 + }, + { + "epoch": 0.22944, + "grad_norm": 0.8250062091062988, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 22944 + }, + { + "epoch": 0.22945, + "grad_norm": 0.9602726625574294, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 22945 + }, + { + "epoch": 0.22946, + "grad_norm": 1.0800028433999425, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 22946 + }, + { + "epoch": 0.22947, + "grad_norm": 1.0457107265224104, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 22947 + }, + { + "epoch": 0.22948, + "grad_norm": 0.859917281699288, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 22948 + }, + { + "epoch": 0.22949, + "grad_norm": 0.7953798128373513, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 22949 + }, + { + "epoch": 0.2295, + "grad_norm": 0.6816978360816288, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 22950 + }, + { + "epoch": 0.22951, + "grad_norm": 0.7266292202464208, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 22951 + }, + { + "epoch": 0.22952, + "grad_norm": 0.7681320721034812, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 22952 + }, + { + "epoch": 0.22953, + "grad_norm": 0.8062129019985542, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 22953 + }, + { + "epoch": 0.22954, + "grad_norm": 0.8489924223963913, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 22954 + }, + { + "epoch": 0.22955, + "grad_norm": 0.6623873408079273, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 22955 + }, + { + "epoch": 0.22956, + "grad_norm": 0.674903023053855, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 22956 + }, + { + "epoch": 0.22957, + "grad_norm": 0.684906535441546, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 22957 + }, + { + "epoch": 0.22958, + "grad_norm": 0.7322733834378926, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 22958 + }, + { + "epoch": 0.22959, + "grad_norm": 0.7809397528558542, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 22959 + }, + { + "epoch": 0.2296, + "grad_norm": 0.9753952181627543, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 22960 + }, + { + "epoch": 0.22961, + "grad_norm": 1.1729924280689636, + "learning_rate": 0.003, + "loss": 4.01, + "step": 22961 + }, + { + "epoch": 0.22962, + "grad_norm": 0.7881161320228455, + "learning_rate": 0.003, + "loss": 4.018, + "step": 22962 + }, + { + "epoch": 0.22963, + "grad_norm": 0.6597074963562686, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 22963 + }, + { + "epoch": 0.22964, + "grad_norm": 0.7570345960689352, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 22964 + }, + { + "epoch": 0.22965, + "grad_norm": 0.7183717056069254, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 22965 + }, + { + "epoch": 0.22966, + "grad_norm": 0.7764632866800013, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 22966 + }, + { + "epoch": 0.22967, + "grad_norm": 0.7899793156502434, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 22967 + }, + { + "epoch": 0.22968, + "grad_norm": 0.7867430292421994, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 22968 + }, + { + "epoch": 0.22969, + "grad_norm": 0.7440371768785915, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 22969 + }, + { + "epoch": 0.2297, + "grad_norm": 0.8619566774703338, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 22970 + }, + { + "epoch": 0.22971, + "grad_norm": 0.9438846755146199, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 22971 + }, + { + "epoch": 0.22972, + "grad_norm": 0.9619872416737365, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 22972 + }, + { + "epoch": 0.22973, + "grad_norm": 0.8901694139273261, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 22973 + }, + { + "epoch": 0.22974, + "grad_norm": 0.8474443727676124, + "learning_rate": 0.003, + "loss": 4.029, + "step": 22974 + }, + { + "epoch": 0.22975, + "grad_norm": 0.6955501147370217, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 22975 + }, + { + "epoch": 0.22976, + "grad_norm": 0.6980498146935564, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 22976 + }, + { + "epoch": 0.22977, + "grad_norm": 0.6427469183664, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 22977 + }, + { + "epoch": 0.22978, + "grad_norm": 0.5969607297233596, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 22978 + }, + { + "epoch": 0.22979, + "grad_norm": 0.6710907359047272, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 22979 + }, + { + "epoch": 0.2298, + "grad_norm": 0.8104151879865027, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 22980 + }, + { + "epoch": 0.22981, + "grad_norm": 1.078374163032046, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 22981 + }, + { + "epoch": 0.22982, + "grad_norm": 1.1510676630895398, + "learning_rate": 0.003, + "loss": 4.013, + "step": 22982 + }, + { + "epoch": 0.22983, + "grad_norm": 0.6967775943636922, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 22983 + }, + { + "epoch": 0.22984, + "grad_norm": 0.7946459439410548, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 22984 + }, + { + "epoch": 0.22985, + "grad_norm": 0.7653215642480301, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 22985 + }, + { + "epoch": 0.22986, + "grad_norm": 0.7677743670000683, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 22986 + }, + { + "epoch": 0.22987, + "grad_norm": 0.7056060749306293, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 22987 + }, + { + "epoch": 0.22988, + "grad_norm": 0.6927291714685474, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 22988 + }, + { + "epoch": 0.22989, + "grad_norm": 0.6166709232550341, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 22989 + }, + { + "epoch": 0.2299, + "grad_norm": 0.6242312844674911, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 22990 + }, + { + "epoch": 0.22991, + "grad_norm": 0.5415134735837003, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 22991 + }, + { + "epoch": 0.22992, + "grad_norm": 0.5367288926974337, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 22992 + }, + { + "epoch": 0.22993, + "grad_norm": 0.5670637316265583, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 22993 + }, + { + "epoch": 0.22994, + "grad_norm": 0.5565997680860723, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 22994 + }, + { + "epoch": 0.22995, + "grad_norm": 0.5521477901693864, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 22995 + }, + { + "epoch": 0.22996, + "grad_norm": 0.5762709936438036, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 22996 + }, + { + "epoch": 0.22997, + "grad_norm": 0.5708052000323154, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 22997 + }, + { + "epoch": 0.22998, + "grad_norm": 0.500657971276503, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 22998 + }, + { + "epoch": 0.22999, + "grad_norm": 0.4983985919611865, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 22999 + }, + { + "epoch": 0.23, + "grad_norm": 0.4768941572266223, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 23000 + }, + { + "epoch": 0.23001, + "grad_norm": 0.4950553646957972, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 23001 + }, + { + "epoch": 0.23002, + "grad_norm": 0.5399030725962106, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 23002 + }, + { + "epoch": 0.23003, + "grad_norm": 0.7041499749157173, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 23003 + }, + { + "epoch": 0.23004, + "grad_norm": 0.9100904890405483, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 23004 + }, + { + "epoch": 0.23005, + "grad_norm": 0.9897853056691226, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 23005 + }, + { + "epoch": 0.23006, + "grad_norm": 0.9313110555392011, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 23006 + }, + { + "epoch": 0.23007, + "grad_norm": 0.8249268162231657, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 23007 + }, + { + "epoch": 0.23008, + "grad_norm": 0.7871152798892155, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 23008 + }, + { + "epoch": 0.23009, + "grad_norm": 0.7251352635532894, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 23009 + }, + { + "epoch": 0.2301, + "grad_norm": 0.7960173917226556, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 23010 + }, + { + "epoch": 0.23011, + "grad_norm": 0.8185024894526041, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 23011 + }, + { + "epoch": 0.23012, + "grad_norm": 0.8210165741314343, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 23012 + }, + { + "epoch": 0.23013, + "grad_norm": 0.9102517248958281, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 23013 + }, + { + "epoch": 0.23014, + "grad_norm": 0.9514435244775995, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 23014 + }, + { + "epoch": 0.23015, + "grad_norm": 0.9524043653038727, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 23015 + }, + { + "epoch": 0.23016, + "grad_norm": 1.0477301899502118, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 23016 + }, + { + "epoch": 0.23017, + "grad_norm": 1.0800543619221858, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 23017 + }, + { + "epoch": 0.23018, + "grad_norm": 0.8942975351595046, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 23018 + }, + { + "epoch": 0.23019, + "grad_norm": 0.9658559860394639, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 23019 + }, + { + "epoch": 0.2302, + "grad_norm": 0.8879042398435851, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 23020 + }, + { + "epoch": 0.23021, + "grad_norm": 1.0369445229860446, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 23021 + }, + { + "epoch": 0.23022, + "grad_norm": 1.0076648414516718, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 23022 + }, + { + "epoch": 0.23023, + "grad_norm": 1.0383070621301191, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 23023 + }, + { + "epoch": 0.23024, + "grad_norm": 0.8326609312123907, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 23024 + }, + { + "epoch": 0.23025, + "grad_norm": 0.697905652873314, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 23025 + }, + { + "epoch": 0.23026, + "grad_norm": 0.7526189097279193, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 23026 + }, + { + "epoch": 0.23027, + "grad_norm": 0.7882208516008916, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 23027 + }, + { + "epoch": 0.23028, + "grad_norm": 0.8404361828528434, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 23028 + }, + { + "epoch": 0.23029, + "grad_norm": 0.9688797012248935, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23029 + }, + { + "epoch": 0.2303, + "grad_norm": 0.8875957744713316, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23030 + }, + { + "epoch": 0.23031, + "grad_norm": 0.7788914995021389, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 23031 + }, + { + "epoch": 0.23032, + "grad_norm": 0.6911607614860636, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 23032 + }, + { + "epoch": 0.23033, + "grad_norm": 0.6556445942753296, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 23033 + }, + { + "epoch": 0.23034, + "grad_norm": 0.576929039577413, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 23034 + }, + { + "epoch": 0.23035, + "grad_norm": 0.6323183668352879, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 23035 + }, + { + "epoch": 0.23036, + "grad_norm": 0.6264068356413045, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 23036 + }, + { + "epoch": 0.23037, + "grad_norm": 0.5978795365088645, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 23037 + }, + { + "epoch": 0.23038, + "grad_norm": 0.5924566238225667, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 23038 + }, + { + "epoch": 0.23039, + "grad_norm": 0.5621145943369175, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 23039 + }, + { + "epoch": 0.2304, + "grad_norm": 0.5677511080813687, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 23040 + }, + { + "epoch": 0.23041, + "grad_norm": 0.5361872490403545, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 23041 + }, + { + "epoch": 0.23042, + "grad_norm": 0.5846243971314664, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 23042 + }, + { + "epoch": 0.23043, + "grad_norm": 0.5333121798082419, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 23043 + }, + { + "epoch": 0.23044, + "grad_norm": 0.5052396341085862, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 23044 + }, + { + "epoch": 0.23045, + "grad_norm": 0.5116952866393689, + "learning_rate": 0.003, + "loss": 3.984, + "step": 23045 + }, + { + "epoch": 0.23046, + "grad_norm": 0.5145895853842356, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 23046 + }, + { + "epoch": 0.23047, + "grad_norm": 0.5886192655621496, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 23047 + }, + { + "epoch": 0.23048, + "grad_norm": 0.6860999643858405, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 23048 + }, + { + "epoch": 0.23049, + "grad_norm": 0.9254486978564838, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 23049 + }, + { + "epoch": 0.2305, + "grad_norm": 1.1144969173364723, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 23050 + }, + { + "epoch": 0.23051, + "grad_norm": 0.892415240041208, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 23051 + }, + { + "epoch": 0.23052, + "grad_norm": 0.7346423646036593, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 23052 + }, + { + "epoch": 0.23053, + "grad_norm": 0.8679930542785779, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 23053 + }, + { + "epoch": 0.23054, + "grad_norm": 0.8678870284369946, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 23054 + }, + { + "epoch": 0.23055, + "grad_norm": 0.905954024680382, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 23055 + }, + { + "epoch": 0.23056, + "grad_norm": 0.8315693908669427, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 23056 + }, + { + "epoch": 0.23057, + "grad_norm": 0.8491756140263574, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 23057 + }, + { + "epoch": 0.23058, + "grad_norm": 0.8701867358388592, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 23058 + }, + { + "epoch": 0.23059, + "grad_norm": 0.9223954288071465, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 23059 + }, + { + "epoch": 0.2306, + "grad_norm": 0.9481275789008853, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 23060 + }, + { + "epoch": 0.23061, + "grad_norm": 1.0480256216861348, + "learning_rate": 0.003, + "loss": 4.0474, + "step": 23061 + }, + { + "epoch": 0.23062, + "grad_norm": 0.8074538506592871, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 23062 + }, + { + "epoch": 0.23063, + "grad_norm": 0.7092979254442221, + "learning_rate": 0.003, + "loss": 4.019, + "step": 23063 + }, + { + "epoch": 0.23064, + "grad_norm": 0.7047672710537607, + "learning_rate": 0.003, + "loss": 4.006, + "step": 23064 + }, + { + "epoch": 0.23065, + "grad_norm": 0.6296307641520436, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 23065 + }, + { + "epoch": 0.23066, + "grad_norm": 0.7208632648485351, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 23066 + }, + { + "epoch": 0.23067, + "grad_norm": 0.8231227116212201, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 23067 + }, + { + "epoch": 0.23068, + "grad_norm": 0.8561083790082544, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 23068 + }, + { + "epoch": 0.23069, + "grad_norm": 0.81280104390844, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 23069 + }, + { + "epoch": 0.2307, + "grad_norm": 0.712191902311694, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 23070 + }, + { + "epoch": 0.23071, + "grad_norm": 0.6566493500098699, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 23071 + }, + { + "epoch": 0.23072, + "grad_norm": 0.6651362821145634, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 23072 + }, + { + "epoch": 0.23073, + "grad_norm": 0.5903022648398403, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 23073 + }, + { + "epoch": 0.23074, + "grad_norm": 0.5761374949577742, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 23074 + }, + { + "epoch": 0.23075, + "grad_norm": 0.5375319760646369, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 23075 + }, + { + "epoch": 0.23076, + "grad_norm": 0.5509710143864402, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 23076 + }, + { + "epoch": 0.23077, + "grad_norm": 0.6242269204529831, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 23077 + }, + { + "epoch": 0.23078, + "grad_norm": 0.597413315703945, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 23078 + }, + { + "epoch": 0.23079, + "grad_norm": 0.5477858249738959, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 23079 + }, + { + "epoch": 0.2308, + "grad_norm": 0.5539714345737451, + "learning_rate": 0.003, + "loss": 4.011, + "step": 23080 + }, + { + "epoch": 0.23081, + "grad_norm": 0.6346094310297955, + "learning_rate": 0.003, + "loss": 4.009, + "step": 23081 + }, + { + "epoch": 0.23082, + "grad_norm": 0.5133488189256595, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 23082 + }, + { + "epoch": 0.23083, + "grad_norm": 0.46479405090404435, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 23083 + }, + { + "epoch": 0.23084, + "grad_norm": 0.4942620938209905, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 23084 + }, + { + "epoch": 0.23085, + "grad_norm": 0.5290511728024487, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 23085 + }, + { + "epoch": 0.23086, + "grad_norm": 0.5795614009906637, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 23086 + }, + { + "epoch": 0.23087, + "grad_norm": 0.7716668345004262, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 23087 + }, + { + "epoch": 0.23088, + "grad_norm": 1.0858759637308801, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 23088 + }, + { + "epoch": 0.23089, + "grad_norm": 1.1625810078987064, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 23089 + }, + { + "epoch": 0.2309, + "grad_norm": 0.7610004974054336, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 23090 + }, + { + "epoch": 0.23091, + "grad_norm": 0.7364907645621396, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 23091 + }, + { + "epoch": 0.23092, + "grad_norm": 0.7608374385502165, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 23092 + }, + { + "epoch": 0.23093, + "grad_norm": 0.8660494686276343, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 23093 + }, + { + "epoch": 0.23094, + "grad_norm": 0.88805360094501, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 23094 + }, + { + "epoch": 0.23095, + "grad_norm": 0.8072129394418877, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 23095 + }, + { + "epoch": 0.23096, + "grad_norm": 0.7506953646663492, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 23096 + }, + { + "epoch": 0.23097, + "grad_norm": 0.717649461396818, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 23097 + }, + { + "epoch": 0.23098, + "grad_norm": 0.7834240859885924, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 23098 + }, + { + "epoch": 0.23099, + "grad_norm": 0.7519492941392869, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 23099 + }, + { + "epoch": 0.231, + "grad_norm": 0.7889046163275738, + "learning_rate": 0.003, + "loss": 4.012, + "step": 23100 + }, + { + "epoch": 0.23101, + "grad_norm": 0.8176025986998736, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 23101 + }, + { + "epoch": 0.23102, + "grad_norm": 0.7174060797557433, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 23102 + }, + { + "epoch": 0.23103, + "grad_norm": 0.711583626004303, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 23103 + }, + { + "epoch": 0.23104, + "grad_norm": 0.7018383131736606, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23104 + }, + { + "epoch": 0.23105, + "grad_norm": 0.7418581591603935, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 23105 + }, + { + "epoch": 0.23106, + "grad_norm": 0.7881343387613695, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 23106 + }, + { + "epoch": 0.23107, + "grad_norm": 0.8177614301359946, + "learning_rate": 0.003, + "loss": 4.023, + "step": 23107 + }, + { + "epoch": 0.23108, + "grad_norm": 0.8404416155310184, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 23108 + }, + { + "epoch": 0.23109, + "grad_norm": 0.8535444790066516, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 23109 + }, + { + "epoch": 0.2311, + "grad_norm": 0.8015370785197542, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 23110 + }, + { + "epoch": 0.23111, + "grad_norm": 0.7954576811755907, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 23111 + }, + { + "epoch": 0.23112, + "grad_norm": 0.8503012371984365, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 23112 + }, + { + "epoch": 0.23113, + "grad_norm": 0.9727202963821562, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 23113 + }, + { + "epoch": 0.23114, + "grad_norm": 0.8966045749290709, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 23114 + }, + { + "epoch": 0.23115, + "grad_norm": 0.8927204574845131, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 23115 + }, + { + "epoch": 0.23116, + "grad_norm": 0.826998840199722, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 23116 + }, + { + "epoch": 0.23117, + "grad_norm": 0.8667232556950019, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 23117 + }, + { + "epoch": 0.23118, + "grad_norm": 0.8175994499497958, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 23118 + }, + { + "epoch": 0.23119, + "grad_norm": 0.9008621629798854, + "learning_rate": 0.003, + "loss": 4.025, + "step": 23119 + }, + { + "epoch": 0.2312, + "grad_norm": 0.8441086622555987, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 23120 + }, + { + "epoch": 0.23121, + "grad_norm": 0.8440519487626549, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 23121 + }, + { + "epoch": 0.23122, + "grad_norm": 0.9530097731369659, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 23122 + }, + { + "epoch": 0.23123, + "grad_norm": 0.9078573939418239, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 23123 + }, + { + "epoch": 0.23124, + "grad_norm": 0.8626410937826903, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 23124 + }, + { + "epoch": 0.23125, + "grad_norm": 0.7403683370303646, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 23125 + }, + { + "epoch": 0.23126, + "grad_norm": 0.7242850513071853, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 23126 + }, + { + "epoch": 0.23127, + "grad_norm": 0.7036725455399531, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 23127 + }, + { + "epoch": 0.23128, + "grad_norm": 0.6222583722891116, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 23128 + }, + { + "epoch": 0.23129, + "grad_norm": 0.5326359659448782, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 23129 + }, + { + "epoch": 0.2313, + "grad_norm": 0.5139032318308697, + "learning_rate": 0.003, + "loss": 3.985, + "step": 23130 + }, + { + "epoch": 0.23131, + "grad_norm": 0.4641398657915859, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 23131 + }, + { + "epoch": 0.23132, + "grad_norm": 0.4578982071912616, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 23132 + }, + { + "epoch": 0.23133, + "grad_norm": 0.39914614305742274, + "learning_rate": 0.003, + "loss": 3.988, + "step": 23133 + }, + { + "epoch": 0.23134, + "grad_norm": 0.4225273504350179, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 23134 + }, + { + "epoch": 0.23135, + "grad_norm": 0.501852047578079, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 23135 + }, + { + "epoch": 0.23136, + "grad_norm": 0.6064163084216992, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 23136 + }, + { + "epoch": 0.23137, + "grad_norm": 0.7326107092232491, + "learning_rate": 0.003, + "loss": 3.977, + "step": 23137 + }, + { + "epoch": 0.23138, + "grad_norm": 0.8432330201053696, + "learning_rate": 0.003, + "loss": 3.966, + "step": 23138 + }, + { + "epoch": 0.23139, + "grad_norm": 0.9347729617168146, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 23139 + }, + { + "epoch": 0.2314, + "grad_norm": 0.9609930106334671, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 23140 + }, + { + "epoch": 0.23141, + "grad_norm": 0.9043205408039903, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 23141 + }, + { + "epoch": 0.23142, + "grad_norm": 0.8704720709517718, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 23142 + }, + { + "epoch": 0.23143, + "grad_norm": 0.8242648389233147, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 23143 + }, + { + "epoch": 0.23144, + "grad_norm": 0.8571130192362955, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 23144 + }, + { + "epoch": 0.23145, + "grad_norm": 0.9475936601159444, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23145 + }, + { + "epoch": 0.23146, + "grad_norm": 0.937010840281327, + "learning_rate": 0.003, + "loss": 4.0477, + "step": 23146 + }, + { + "epoch": 0.23147, + "grad_norm": 0.9491793151218615, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 23147 + }, + { + "epoch": 0.23148, + "grad_norm": 0.9232706719270499, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 23148 + }, + { + "epoch": 0.23149, + "grad_norm": 0.9169839864125402, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 23149 + }, + { + "epoch": 0.2315, + "grad_norm": 0.8805148104379595, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 23150 + }, + { + "epoch": 0.23151, + "grad_norm": 0.9572294862020274, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 23151 + }, + { + "epoch": 0.23152, + "grad_norm": 1.1511361109456424, + "learning_rate": 0.003, + "loss": 4.001, + "step": 23152 + }, + { + "epoch": 0.23153, + "grad_norm": 0.9101870081519269, + "learning_rate": 0.003, + "loss": 4.0506, + "step": 23153 + }, + { + "epoch": 0.23154, + "grad_norm": 0.9550587574921128, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 23154 + }, + { + "epoch": 0.23155, + "grad_norm": 1.0951027999288359, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 23155 + }, + { + "epoch": 0.23156, + "grad_norm": 1.4565067453816167, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 23156 + }, + { + "epoch": 0.23157, + "grad_norm": 0.9987854743601013, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 23157 + }, + { + "epoch": 0.23158, + "grad_norm": 1.20704318470679, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 23158 + }, + { + "epoch": 0.23159, + "grad_norm": 0.9144044551933334, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 23159 + }, + { + "epoch": 0.2316, + "grad_norm": 0.9797875383777195, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 23160 + }, + { + "epoch": 0.23161, + "grad_norm": 0.9883174689062325, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 23161 + }, + { + "epoch": 0.23162, + "grad_norm": 0.9294172151202174, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 23162 + }, + { + "epoch": 0.23163, + "grad_norm": 0.8392577719964938, + "learning_rate": 0.003, + "loss": 4.0479, + "step": 23163 + }, + { + "epoch": 0.23164, + "grad_norm": 1.0410983078133809, + "learning_rate": 0.003, + "loss": 4.045, + "step": 23164 + }, + { + "epoch": 0.23165, + "grad_norm": 1.1391947829565816, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 23165 + }, + { + "epoch": 0.23166, + "grad_norm": 0.7856131955658411, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 23166 + }, + { + "epoch": 0.23167, + "grad_norm": 0.6898980668533548, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 23167 + }, + { + "epoch": 0.23168, + "grad_norm": 0.8492649516647472, + "learning_rate": 0.003, + "loss": 3.986, + "step": 23168 + }, + { + "epoch": 0.23169, + "grad_norm": 1.0985778547730767, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 23169 + }, + { + "epoch": 0.2317, + "grad_norm": 0.7892781801748583, + "learning_rate": 0.003, + "loss": 4.02, + "step": 23170 + }, + { + "epoch": 0.23171, + "grad_norm": 0.7745024993518462, + "learning_rate": 0.003, + "loss": 4.02, + "step": 23171 + }, + { + "epoch": 0.23172, + "grad_norm": 0.9359253013259682, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 23172 + }, + { + "epoch": 0.23173, + "grad_norm": 0.9354804693862855, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 23173 + }, + { + "epoch": 0.23174, + "grad_norm": 0.9030023823742261, + "learning_rate": 0.003, + "loss": 4.005, + "step": 23174 + }, + { + "epoch": 0.23175, + "grad_norm": 1.0265116716183627, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 23175 + }, + { + "epoch": 0.23176, + "grad_norm": 0.9583322358370927, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 23176 + }, + { + "epoch": 0.23177, + "grad_norm": 0.9241809933901254, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 23177 + }, + { + "epoch": 0.23178, + "grad_norm": 0.9640511689264657, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 23178 + }, + { + "epoch": 0.23179, + "grad_norm": 0.9849191175391946, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 23179 + }, + { + "epoch": 0.2318, + "grad_norm": 0.915979607790926, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 23180 + }, + { + "epoch": 0.23181, + "grad_norm": 0.8546194705619378, + "learning_rate": 0.003, + "loss": 4.04, + "step": 23181 + }, + { + "epoch": 0.23182, + "grad_norm": 0.700044825327342, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 23182 + }, + { + "epoch": 0.23183, + "grad_norm": 0.6836088342400838, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 23183 + }, + { + "epoch": 0.23184, + "grad_norm": 0.6390165860135797, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 23184 + }, + { + "epoch": 0.23185, + "grad_norm": 0.6871936069538322, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 23185 + }, + { + "epoch": 0.23186, + "grad_norm": 0.7499575169854344, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 23186 + }, + { + "epoch": 0.23187, + "grad_norm": 0.8445829370066011, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 23187 + }, + { + "epoch": 0.23188, + "grad_norm": 0.7982789237373877, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 23188 + }, + { + "epoch": 0.23189, + "grad_norm": 0.5919335235749368, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 23189 + }, + { + "epoch": 0.2319, + "grad_norm": 0.5331516434951623, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 23190 + }, + { + "epoch": 0.23191, + "grad_norm": 0.5325113112575288, + "learning_rate": 0.003, + "loss": 4.003, + "step": 23191 + }, + { + "epoch": 0.23192, + "grad_norm": 0.5965669644213215, + "learning_rate": 0.003, + "loss": 4.017, + "step": 23192 + }, + { + "epoch": 0.23193, + "grad_norm": 0.6055955424686271, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 23193 + }, + { + "epoch": 0.23194, + "grad_norm": 0.5989429655236139, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 23194 + }, + { + "epoch": 0.23195, + "grad_norm": 0.5778282682321929, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 23195 + }, + { + "epoch": 0.23196, + "grad_norm": 0.604223374258081, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 23196 + }, + { + "epoch": 0.23197, + "grad_norm": 0.6377589246921734, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 23197 + }, + { + "epoch": 0.23198, + "grad_norm": 0.7113168022612695, + "learning_rate": 0.003, + "loss": 3.974, + "step": 23198 + }, + { + "epoch": 0.23199, + "grad_norm": 0.8039971116473356, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 23199 + }, + { + "epoch": 0.232, + "grad_norm": 0.8026811117545458, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 23200 + }, + { + "epoch": 0.23201, + "grad_norm": 0.6182088990865213, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 23201 + }, + { + "epoch": 0.23202, + "grad_norm": 0.4451957548021637, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 23202 + }, + { + "epoch": 0.23203, + "grad_norm": 0.577481352490466, + "learning_rate": 0.003, + "loss": 4.001, + "step": 23203 + }, + { + "epoch": 0.23204, + "grad_norm": 0.7501259380616632, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 23204 + }, + { + "epoch": 0.23205, + "grad_norm": 0.7894369764744075, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 23205 + }, + { + "epoch": 0.23206, + "grad_norm": 0.6004380756653895, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 23206 + }, + { + "epoch": 0.23207, + "grad_norm": 0.5767051907627523, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 23207 + }, + { + "epoch": 0.23208, + "grad_norm": 0.6597694923297199, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 23208 + }, + { + "epoch": 0.23209, + "grad_norm": 0.7074402551492744, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 23209 + }, + { + "epoch": 0.2321, + "grad_norm": 0.7104675777344109, + "learning_rate": 0.003, + "loss": 3.998, + "step": 23210 + }, + { + "epoch": 0.23211, + "grad_norm": 0.6561726153571328, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 23211 + }, + { + "epoch": 0.23212, + "grad_norm": 0.6174960860511063, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 23212 + }, + { + "epoch": 0.23213, + "grad_norm": 0.6581972848872604, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 23213 + }, + { + "epoch": 0.23214, + "grad_norm": 0.6656845250473615, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 23214 + }, + { + "epoch": 0.23215, + "grad_norm": 0.7133626965946434, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 23215 + }, + { + "epoch": 0.23216, + "grad_norm": 0.7631752927374785, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 23216 + }, + { + "epoch": 0.23217, + "grad_norm": 0.7524219590854874, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 23217 + }, + { + "epoch": 0.23218, + "grad_norm": 0.7559892552564925, + "learning_rate": 0.003, + "loss": 3.983, + "step": 23218 + }, + { + "epoch": 0.23219, + "grad_norm": 0.7670721300211629, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 23219 + }, + { + "epoch": 0.2322, + "grad_norm": 0.7983775473677912, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 23220 + }, + { + "epoch": 0.23221, + "grad_norm": 0.7544689860781596, + "learning_rate": 0.003, + "loss": 3.986, + "step": 23221 + }, + { + "epoch": 0.23222, + "grad_norm": 0.7379020790916087, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 23222 + }, + { + "epoch": 0.23223, + "grad_norm": 0.710160840343017, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 23223 + }, + { + "epoch": 0.23224, + "grad_norm": 0.7189151742845659, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 23224 + }, + { + "epoch": 0.23225, + "grad_norm": 0.7477422559881592, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 23225 + }, + { + "epoch": 0.23226, + "grad_norm": 0.7659806465020987, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 23226 + }, + { + "epoch": 0.23227, + "grad_norm": 0.8812292704073704, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 23227 + }, + { + "epoch": 0.23228, + "grad_norm": 0.9640967734240824, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 23228 + }, + { + "epoch": 0.23229, + "grad_norm": 1.0404518443722286, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 23229 + }, + { + "epoch": 0.2323, + "grad_norm": 0.8551410522675945, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 23230 + }, + { + "epoch": 0.23231, + "grad_norm": 0.7145741193431396, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 23231 + }, + { + "epoch": 0.23232, + "grad_norm": 0.7881224424465976, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 23232 + }, + { + "epoch": 0.23233, + "grad_norm": 0.7463867568863415, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 23233 + }, + { + "epoch": 0.23234, + "grad_norm": 0.8091609773409304, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 23234 + }, + { + "epoch": 0.23235, + "grad_norm": 0.8976846356407264, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 23235 + }, + { + "epoch": 0.23236, + "grad_norm": 0.975168737229428, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 23236 + }, + { + "epoch": 0.23237, + "grad_norm": 1.0776421284814894, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 23237 + }, + { + "epoch": 0.23238, + "grad_norm": 0.8568089596399906, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 23238 + }, + { + "epoch": 0.23239, + "grad_norm": 0.6975325260486451, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 23239 + }, + { + "epoch": 0.2324, + "grad_norm": 0.6750077418452585, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 23240 + }, + { + "epoch": 0.23241, + "grad_norm": 0.615716386485276, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 23241 + }, + { + "epoch": 0.23242, + "grad_norm": 0.6998941218322817, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 23242 + }, + { + "epoch": 0.23243, + "grad_norm": 0.6831136454541066, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 23243 + }, + { + "epoch": 0.23244, + "grad_norm": 0.6669509256114706, + "learning_rate": 0.003, + "loss": 3.993, + "step": 23244 + }, + { + "epoch": 0.23245, + "grad_norm": 0.567722783441989, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 23245 + }, + { + "epoch": 0.23246, + "grad_norm": 0.5094516402098099, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 23246 + }, + { + "epoch": 0.23247, + "grad_norm": 0.5119503373391145, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 23247 + }, + { + "epoch": 0.23248, + "grad_norm": 0.583928923842963, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 23248 + }, + { + "epoch": 0.23249, + "grad_norm": 0.672807052329938, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 23249 + }, + { + "epoch": 0.2325, + "grad_norm": 0.8987236131619909, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 23250 + }, + { + "epoch": 0.23251, + "grad_norm": 1.07549600138619, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 23251 + }, + { + "epoch": 0.23252, + "grad_norm": 0.7884661864523065, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 23252 + }, + { + "epoch": 0.23253, + "grad_norm": 0.786107754887661, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 23253 + }, + { + "epoch": 0.23254, + "grad_norm": 0.8231684805812857, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 23254 + }, + { + "epoch": 0.23255, + "grad_norm": 0.8134893078524126, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 23255 + }, + { + "epoch": 0.23256, + "grad_norm": 0.8633532181556088, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 23256 + }, + { + "epoch": 0.23257, + "grad_norm": 0.9015147624622114, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 23257 + }, + { + "epoch": 0.23258, + "grad_norm": 0.734738799409338, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 23258 + }, + { + "epoch": 0.23259, + "grad_norm": 0.7551633314720682, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 23259 + }, + { + "epoch": 0.2326, + "grad_norm": 0.820632138655338, + "learning_rate": 0.003, + "loss": 4.022, + "step": 23260 + }, + { + "epoch": 0.23261, + "grad_norm": 0.8207303865431055, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 23261 + }, + { + "epoch": 0.23262, + "grad_norm": 0.7889622440831087, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 23262 + }, + { + "epoch": 0.23263, + "grad_norm": 0.8150707929334008, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 23263 + }, + { + "epoch": 0.23264, + "grad_norm": 0.7756777696540681, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 23264 + }, + { + "epoch": 0.23265, + "grad_norm": 0.7766733980012417, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 23265 + }, + { + "epoch": 0.23266, + "grad_norm": 0.7620616694385478, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 23266 + }, + { + "epoch": 0.23267, + "grad_norm": 0.7751041762620364, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 23267 + }, + { + "epoch": 0.23268, + "grad_norm": 0.734578626246554, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 23268 + }, + { + "epoch": 0.23269, + "grad_norm": 0.6996121716400142, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 23269 + }, + { + "epoch": 0.2327, + "grad_norm": 0.6338505434514259, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 23270 + }, + { + "epoch": 0.23271, + "grad_norm": 0.60324114944252, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 23271 + }, + { + "epoch": 0.23272, + "grad_norm": 0.6896872720725028, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 23272 + }, + { + "epoch": 0.23273, + "grad_norm": 0.6823334682841183, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 23273 + }, + { + "epoch": 0.23274, + "grad_norm": 0.6169862364527872, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 23274 + }, + { + "epoch": 0.23275, + "grad_norm": 0.5899600219043409, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 23275 + }, + { + "epoch": 0.23276, + "grad_norm": 0.5281029334219159, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 23276 + }, + { + "epoch": 0.23277, + "grad_norm": 0.5835621719651592, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 23277 + }, + { + "epoch": 0.23278, + "grad_norm": 0.6861259577984113, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 23278 + }, + { + "epoch": 0.23279, + "grad_norm": 0.7884309489234882, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 23279 + }, + { + "epoch": 0.2328, + "grad_norm": 0.7587929025087151, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 23280 + }, + { + "epoch": 0.23281, + "grad_norm": 0.8010579957695564, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 23281 + }, + { + "epoch": 0.23282, + "grad_norm": 0.8477735971219988, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 23282 + }, + { + "epoch": 0.23283, + "grad_norm": 0.8099495040691249, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 23283 + }, + { + "epoch": 0.23284, + "grad_norm": 0.8534662150417193, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 23284 + }, + { + "epoch": 0.23285, + "grad_norm": 0.7653514144485919, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 23285 + }, + { + "epoch": 0.23286, + "grad_norm": 0.664497007245853, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 23286 + }, + { + "epoch": 0.23287, + "grad_norm": 0.6961722269737782, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 23287 + }, + { + "epoch": 0.23288, + "grad_norm": 0.7816749333551879, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 23288 + }, + { + "epoch": 0.23289, + "grad_norm": 0.8126268780527534, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23289 + }, + { + "epoch": 0.2329, + "grad_norm": 0.800813129071899, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 23290 + }, + { + "epoch": 0.23291, + "grad_norm": 0.7887703703541533, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 23291 + }, + { + "epoch": 0.23292, + "grad_norm": 0.7072233249347745, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 23292 + }, + { + "epoch": 0.23293, + "grad_norm": 0.6720341156264767, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 23293 + }, + { + "epoch": 0.23294, + "grad_norm": 0.7292020548451911, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 23294 + }, + { + "epoch": 0.23295, + "grad_norm": 0.9499467131760284, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 23295 + }, + { + "epoch": 0.23296, + "grad_norm": 1.1921579493329406, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 23296 + }, + { + "epoch": 0.23297, + "grad_norm": 0.8143382952388387, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 23297 + }, + { + "epoch": 0.23298, + "grad_norm": 0.7543357928981732, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 23298 + }, + { + "epoch": 0.23299, + "grad_norm": 0.8085489683729976, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 23299 + }, + { + "epoch": 0.233, + "grad_norm": 0.765557100018101, + "learning_rate": 0.003, + "loss": 4.0369, + "step": 23300 + }, + { + "epoch": 0.23301, + "grad_norm": 0.8110168979287864, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 23301 + }, + { + "epoch": 0.23302, + "grad_norm": 0.7914754032280669, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 23302 + }, + { + "epoch": 0.23303, + "grad_norm": 0.7505271228316193, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 23303 + }, + { + "epoch": 0.23304, + "grad_norm": 0.7229506411504528, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 23304 + }, + { + "epoch": 0.23305, + "grad_norm": 0.7544248652816602, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 23305 + }, + { + "epoch": 0.23306, + "grad_norm": 0.6622177789054038, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 23306 + }, + { + "epoch": 0.23307, + "grad_norm": 0.7385620837364563, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 23307 + }, + { + "epoch": 0.23308, + "grad_norm": 0.680012851438445, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 23308 + }, + { + "epoch": 0.23309, + "grad_norm": 0.6604758814412957, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 23309 + }, + { + "epoch": 0.2331, + "grad_norm": 0.746106633648859, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 23310 + }, + { + "epoch": 0.23311, + "grad_norm": 0.7835829831949958, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 23311 + }, + { + "epoch": 0.23312, + "grad_norm": 0.7839510694297803, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 23312 + }, + { + "epoch": 0.23313, + "grad_norm": 0.7826390482894882, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 23313 + }, + { + "epoch": 0.23314, + "grad_norm": 0.8052843088016862, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 23314 + }, + { + "epoch": 0.23315, + "grad_norm": 0.7628987437363668, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 23315 + }, + { + "epoch": 0.23316, + "grad_norm": 0.7801144664604572, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 23316 + }, + { + "epoch": 0.23317, + "grad_norm": 0.6988390996856768, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 23317 + }, + { + "epoch": 0.23318, + "grad_norm": 0.6889246689211186, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 23318 + }, + { + "epoch": 0.23319, + "grad_norm": 0.630593515021545, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 23319 + }, + { + "epoch": 0.2332, + "grad_norm": 0.7212769880931271, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 23320 + }, + { + "epoch": 0.23321, + "grad_norm": 0.7659952771506717, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 23321 + }, + { + "epoch": 0.23322, + "grad_norm": 0.7038900012143467, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 23322 + }, + { + "epoch": 0.23323, + "grad_norm": 0.5479673537349914, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 23323 + }, + { + "epoch": 0.23324, + "grad_norm": 0.5355238176322019, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 23324 + }, + { + "epoch": 0.23325, + "grad_norm": 0.47330132370681904, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 23325 + }, + { + "epoch": 0.23326, + "grad_norm": 0.46528432298902506, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 23326 + }, + { + "epoch": 0.23327, + "grad_norm": 0.5567787344509721, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 23327 + }, + { + "epoch": 0.23328, + "grad_norm": 0.8653537163431851, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 23328 + }, + { + "epoch": 0.23329, + "grad_norm": 1.188117580138035, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 23329 + }, + { + "epoch": 0.2333, + "grad_norm": 0.6978005003720412, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 23330 + }, + { + "epoch": 0.23331, + "grad_norm": 0.48282256052687705, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 23331 + }, + { + "epoch": 0.23332, + "grad_norm": 0.6190993295584776, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 23332 + }, + { + "epoch": 0.23333, + "grad_norm": 0.7306911166342481, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 23333 + }, + { + "epoch": 0.23334, + "grad_norm": 0.7845234563552897, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 23334 + }, + { + "epoch": 0.23335, + "grad_norm": 0.7306540913221193, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 23335 + }, + { + "epoch": 0.23336, + "grad_norm": 0.6855019560237553, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 23336 + }, + { + "epoch": 0.23337, + "grad_norm": 0.7906447092531778, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 23337 + }, + { + "epoch": 0.23338, + "grad_norm": 0.8262230246488591, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 23338 + }, + { + "epoch": 0.23339, + "grad_norm": 0.8476219829895328, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23339 + }, + { + "epoch": 0.2334, + "grad_norm": 0.7851364432318593, + "learning_rate": 0.003, + "loss": 4.015, + "step": 23340 + }, + { + "epoch": 0.23341, + "grad_norm": 0.7629414417342198, + "learning_rate": 0.003, + "loss": 4.003, + "step": 23341 + }, + { + "epoch": 0.23342, + "grad_norm": 0.882407649453423, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 23342 + }, + { + "epoch": 0.23343, + "grad_norm": 0.9895732303245759, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 23343 + }, + { + "epoch": 0.23344, + "grad_norm": 1.0478145824780543, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 23344 + }, + { + "epoch": 0.23345, + "grad_norm": 1.011425010786582, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 23345 + }, + { + "epoch": 0.23346, + "grad_norm": 1.0098519774107149, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 23346 + }, + { + "epoch": 0.23347, + "grad_norm": 1.0469657776631665, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 23347 + }, + { + "epoch": 0.23348, + "grad_norm": 0.9244428346072695, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 23348 + }, + { + "epoch": 0.23349, + "grad_norm": 0.9643862417140472, + "learning_rate": 0.003, + "loss": 3.977, + "step": 23349 + }, + { + "epoch": 0.2335, + "grad_norm": 0.9621946893960917, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 23350 + }, + { + "epoch": 0.23351, + "grad_norm": 0.8764322949974355, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 23351 + }, + { + "epoch": 0.23352, + "grad_norm": 1.0756769286307017, + "learning_rate": 0.003, + "loss": 4.055, + "step": 23352 + }, + { + "epoch": 0.23353, + "grad_norm": 0.9896785353535708, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 23353 + }, + { + "epoch": 0.23354, + "grad_norm": 1.002447982089954, + "learning_rate": 0.003, + "loss": 4.0504, + "step": 23354 + }, + { + "epoch": 0.23355, + "grad_norm": 0.9798620437472746, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 23355 + }, + { + "epoch": 0.23356, + "grad_norm": 1.072192395006125, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 23356 + }, + { + "epoch": 0.23357, + "grad_norm": 0.9418960811150219, + "learning_rate": 0.003, + "loss": 4.0453, + "step": 23357 + }, + { + "epoch": 0.23358, + "grad_norm": 1.1221195901540635, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 23358 + }, + { + "epoch": 0.23359, + "grad_norm": 1.0426275235369824, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 23359 + }, + { + "epoch": 0.2336, + "grad_norm": 1.0839631697905698, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 23360 + }, + { + "epoch": 0.23361, + "grad_norm": 1.0373456796695077, + "learning_rate": 0.003, + "loss": 4.0615, + "step": 23361 + }, + { + "epoch": 0.23362, + "grad_norm": 0.9233389372205856, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 23362 + }, + { + "epoch": 0.23363, + "grad_norm": 0.8947984955749553, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 23363 + }, + { + "epoch": 0.23364, + "grad_norm": 0.9131153560412809, + "learning_rate": 0.003, + "loss": 4.0498, + "step": 23364 + }, + { + "epoch": 0.23365, + "grad_norm": 0.9519488162676292, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 23365 + }, + { + "epoch": 0.23366, + "grad_norm": 0.8750642285855926, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 23366 + }, + { + "epoch": 0.23367, + "grad_norm": 0.6619683687798414, + "learning_rate": 0.003, + "loss": 4.0466, + "step": 23367 + }, + { + "epoch": 0.23368, + "grad_norm": 0.6035787759784076, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23368 + }, + { + "epoch": 0.23369, + "grad_norm": 0.5487358468727876, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 23369 + }, + { + "epoch": 0.2337, + "grad_norm": 0.5836747495621315, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 23370 + }, + { + "epoch": 0.23371, + "grad_norm": 0.6180865198642819, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 23371 + }, + { + "epoch": 0.23372, + "grad_norm": 0.6140073932709372, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 23372 + }, + { + "epoch": 0.23373, + "grad_norm": 0.5433888791263164, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 23373 + }, + { + "epoch": 0.23374, + "grad_norm": 0.5729301346002176, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 23374 + }, + { + "epoch": 0.23375, + "grad_norm": 0.6233734396146792, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 23375 + }, + { + "epoch": 0.23376, + "grad_norm": 0.6427280335692623, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 23376 + }, + { + "epoch": 0.23377, + "grad_norm": 0.7207761236080084, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 23377 + }, + { + "epoch": 0.23378, + "grad_norm": 0.5925294164987894, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 23378 + }, + { + "epoch": 0.23379, + "grad_norm": 0.4965765100229236, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 23379 + }, + { + "epoch": 0.2338, + "grad_norm": 0.569011391130101, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 23380 + }, + { + "epoch": 0.23381, + "grad_norm": 0.5718557285153271, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 23381 + }, + { + "epoch": 0.23382, + "grad_norm": 0.5930112950935916, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 23382 + }, + { + "epoch": 0.23383, + "grad_norm": 0.836439787641066, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 23383 + }, + { + "epoch": 0.23384, + "grad_norm": 1.1543741080817622, + "learning_rate": 0.003, + "loss": 3.984, + "step": 23384 + }, + { + "epoch": 0.23385, + "grad_norm": 0.9088491521037071, + "learning_rate": 0.003, + "loss": 3.995, + "step": 23385 + }, + { + "epoch": 0.23386, + "grad_norm": 0.699257367063053, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 23386 + }, + { + "epoch": 0.23387, + "grad_norm": 0.7217080664704, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 23387 + }, + { + "epoch": 0.23388, + "grad_norm": 0.7373739150508184, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 23388 + }, + { + "epoch": 0.23389, + "grad_norm": 0.6695172772259024, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 23389 + }, + { + "epoch": 0.2339, + "grad_norm": 0.7670516440381694, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 23390 + }, + { + "epoch": 0.23391, + "grad_norm": 0.8087028041324127, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 23391 + }, + { + "epoch": 0.23392, + "grad_norm": 0.8659634294439719, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 23392 + }, + { + "epoch": 0.23393, + "grad_norm": 0.7928322619088312, + "learning_rate": 0.003, + "loss": 4.0346, + "step": 23393 + }, + { + "epoch": 0.23394, + "grad_norm": 0.7758148532636804, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 23394 + }, + { + "epoch": 0.23395, + "grad_norm": 0.7564366405757561, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 23395 + }, + { + "epoch": 0.23396, + "grad_norm": 0.7053048909115516, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 23396 + }, + { + "epoch": 0.23397, + "grad_norm": 0.6664535673261938, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 23397 + }, + { + "epoch": 0.23398, + "grad_norm": 0.7257403365991838, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 23398 + }, + { + "epoch": 0.23399, + "grad_norm": 0.7582928915400303, + "learning_rate": 0.003, + "loss": 4.026, + "step": 23399 + }, + { + "epoch": 0.234, + "grad_norm": 0.7505974628202597, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 23400 + }, + { + "epoch": 0.23401, + "grad_norm": 0.7157112609760063, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 23401 + }, + { + "epoch": 0.23402, + "grad_norm": 0.5458371706328785, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 23402 + }, + { + "epoch": 0.23403, + "grad_norm": 0.5928563032820902, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 23403 + }, + { + "epoch": 0.23404, + "grad_norm": 0.49088972110287366, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 23404 + }, + { + "epoch": 0.23405, + "grad_norm": 0.5063126197175183, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 23405 + }, + { + "epoch": 0.23406, + "grad_norm": 0.49152908167241033, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 23406 + }, + { + "epoch": 0.23407, + "grad_norm": 0.5468404364323138, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 23407 + }, + { + "epoch": 0.23408, + "grad_norm": 0.6630721633619499, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 23408 + }, + { + "epoch": 0.23409, + "grad_norm": 0.7581589707100734, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 23409 + }, + { + "epoch": 0.2341, + "grad_norm": 0.7599027952054631, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 23410 + }, + { + "epoch": 0.23411, + "grad_norm": 0.7632908091990799, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 23411 + }, + { + "epoch": 0.23412, + "grad_norm": 0.6870573587054671, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 23412 + }, + { + "epoch": 0.23413, + "grad_norm": 0.6873667683523372, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 23413 + }, + { + "epoch": 0.23414, + "grad_norm": 0.6865393052628631, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 23414 + }, + { + "epoch": 0.23415, + "grad_norm": 0.6646310449064504, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 23415 + }, + { + "epoch": 0.23416, + "grad_norm": 0.6918988503104359, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 23416 + }, + { + "epoch": 0.23417, + "grad_norm": 0.7084627142845003, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 23417 + }, + { + "epoch": 0.23418, + "grad_norm": 0.7036350509981745, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 23418 + }, + { + "epoch": 0.23419, + "grad_norm": 0.6415008872586259, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 23419 + }, + { + "epoch": 0.2342, + "grad_norm": 0.6081841103868024, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 23420 + }, + { + "epoch": 0.23421, + "grad_norm": 0.6213563708500699, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 23421 + }, + { + "epoch": 0.23422, + "grad_norm": 0.6523772809947164, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 23422 + }, + { + "epoch": 0.23423, + "grad_norm": 0.7396367613338968, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 23423 + }, + { + "epoch": 0.23424, + "grad_norm": 0.9650718091072756, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 23424 + }, + { + "epoch": 0.23425, + "grad_norm": 1.1146739265066774, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 23425 + }, + { + "epoch": 0.23426, + "grad_norm": 0.8218722518636996, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 23426 + }, + { + "epoch": 0.23427, + "grad_norm": 0.7919253730884585, + "learning_rate": 0.003, + "loss": 3.999, + "step": 23427 + }, + { + "epoch": 0.23428, + "grad_norm": 0.750782114487077, + "learning_rate": 0.003, + "loss": 3.972, + "step": 23428 + }, + { + "epoch": 0.23429, + "grad_norm": 0.7044509805743951, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 23429 + }, + { + "epoch": 0.2343, + "grad_norm": 0.7376618328067989, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 23430 + }, + { + "epoch": 0.23431, + "grad_norm": 0.7894322869205277, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 23431 + }, + { + "epoch": 0.23432, + "grad_norm": 0.8468798934738131, + "learning_rate": 0.003, + "loss": 3.992, + "step": 23432 + }, + { + "epoch": 0.23433, + "grad_norm": 0.8904895534648348, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 23433 + }, + { + "epoch": 0.23434, + "grad_norm": 0.9755831598830741, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 23434 + }, + { + "epoch": 0.23435, + "grad_norm": 1.0200588753488868, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 23435 + }, + { + "epoch": 0.23436, + "grad_norm": 0.7465605234437344, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 23436 + }, + { + "epoch": 0.23437, + "grad_norm": 0.6885317838684123, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 23437 + }, + { + "epoch": 0.23438, + "grad_norm": 0.6433345073761422, + "learning_rate": 0.003, + "loss": 4.0496, + "step": 23438 + }, + { + "epoch": 0.23439, + "grad_norm": 0.7077269718606977, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 23439 + }, + { + "epoch": 0.2344, + "grad_norm": 0.7968296847577062, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 23440 + }, + { + "epoch": 0.23441, + "grad_norm": 0.7876750889762439, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 23441 + }, + { + "epoch": 0.23442, + "grad_norm": 0.8090577010253391, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 23442 + }, + { + "epoch": 0.23443, + "grad_norm": 0.8004298669359688, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 23443 + }, + { + "epoch": 0.23444, + "grad_norm": 0.6809346365853046, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 23444 + }, + { + "epoch": 0.23445, + "grad_norm": 0.5284151144626134, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 23445 + }, + { + "epoch": 0.23446, + "grad_norm": 0.5871526518916403, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 23446 + }, + { + "epoch": 0.23447, + "grad_norm": 0.6535748189803838, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 23447 + }, + { + "epoch": 0.23448, + "grad_norm": 0.7026999388949483, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 23448 + }, + { + "epoch": 0.23449, + "grad_norm": 0.7587960711940542, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 23449 + }, + { + "epoch": 0.2345, + "grad_norm": 0.7935021075921329, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 23450 + }, + { + "epoch": 0.23451, + "grad_norm": 0.783619636252718, + "learning_rate": 0.003, + "loss": 4.009, + "step": 23451 + }, + { + "epoch": 0.23452, + "grad_norm": 0.7668134689798383, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 23452 + }, + { + "epoch": 0.23453, + "grad_norm": 0.8065220854995045, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 23453 + }, + { + "epoch": 0.23454, + "grad_norm": 0.7610473581039647, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 23454 + }, + { + "epoch": 0.23455, + "grad_norm": 0.7842457650924786, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 23455 + }, + { + "epoch": 0.23456, + "grad_norm": 0.8180487848314619, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 23456 + }, + { + "epoch": 0.23457, + "grad_norm": 0.7934396032490355, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 23457 + }, + { + "epoch": 0.23458, + "grad_norm": 0.8509397009473596, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 23458 + }, + { + "epoch": 0.23459, + "grad_norm": 0.8396153883711887, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 23459 + }, + { + "epoch": 0.2346, + "grad_norm": 0.7327298394525147, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 23460 + }, + { + "epoch": 0.23461, + "grad_norm": 0.7021971537362837, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 23461 + }, + { + "epoch": 0.23462, + "grad_norm": 0.5985785187364926, + "learning_rate": 0.003, + "loss": 4.013, + "step": 23462 + }, + { + "epoch": 0.23463, + "grad_norm": 0.5697412369952529, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 23463 + }, + { + "epoch": 0.23464, + "grad_norm": 0.5984429195567441, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 23464 + }, + { + "epoch": 0.23465, + "grad_norm": 0.639401483334042, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 23465 + }, + { + "epoch": 0.23466, + "grad_norm": 0.802306525755477, + "learning_rate": 0.003, + "loss": 3.996, + "step": 23466 + }, + { + "epoch": 0.23467, + "grad_norm": 0.9288402735299527, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 23467 + }, + { + "epoch": 0.23468, + "grad_norm": 0.8809270286682039, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 23468 + }, + { + "epoch": 0.23469, + "grad_norm": 0.8212962252184237, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 23469 + }, + { + "epoch": 0.2347, + "grad_norm": 0.9052550299720321, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 23470 + }, + { + "epoch": 0.23471, + "grad_norm": 0.8426478865697477, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 23471 + }, + { + "epoch": 0.23472, + "grad_norm": 0.9137896703438767, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 23472 + }, + { + "epoch": 0.23473, + "grad_norm": 0.8858337746481139, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 23473 + }, + { + "epoch": 0.23474, + "grad_norm": 0.7519462299144642, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 23474 + }, + { + "epoch": 0.23475, + "grad_norm": 0.8005952851309412, + "learning_rate": 0.003, + "loss": 3.987, + "step": 23475 + }, + { + "epoch": 0.23476, + "grad_norm": 0.9418847560592758, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 23476 + }, + { + "epoch": 0.23477, + "grad_norm": 1.0762121286740862, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 23477 + }, + { + "epoch": 0.23478, + "grad_norm": 0.8810999146418931, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 23478 + }, + { + "epoch": 0.23479, + "grad_norm": 0.7904644631117052, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 23479 + }, + { + "epoch": 0.2348, + "grad_norm": 0.7735082450234219, + "learning_rate": 0.003, + "loss": 4.0648, + "step": 23480 + }, + { + "epoch": 0.23481, + "grad_norm": 0.7751615998459566, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 23481 + }, + { + "epoch": 0.23482, + "grad_norm": 0.721281137521928, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 23482 + }, + { + "epoch": 0.23483, + "grad_norm": 0.8467866045296681, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 23483 + }, + { + "epoch": 0.23484, + "grad_norm": 0.8857140008053487, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 23484 + }, + { + "epoch": 0.23485, + "grad_norm": 0.8910614106042847, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 23485 + }, + { + "epoch": 0.23486, + "grad_norm": 0.7923032588129798, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 23486 + }, + { + "epoch": 0.23487, + "grad_norm": 0.7458472811221643, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 23487 + }, + { + "epoch": 0.23488, + "grad_norm": 0.6541167947704067, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 23488 + }, + { + "epoch": 0.23489, + "grad_norm": 0.7359295041528059, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 23489 + }, + { + "epoch": 0.2349, + "grad_norm": 0.8886947746759716, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 23490 + }, + { + "epoch": 0.23491, + "grad_norm": 1.0244693194097823, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 23491 + }, + { + "epoch": 0.23492, + "grad_norm": 0.9539177068099123, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 23492 + }, + { + "epoch": 0.23493, + "grad_norm": 0.7735591309259756, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 23493 + }, + { + "epoch": 0.23494, + "grad_norm": 0.6971434495292426, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 23494 + }, + { + "epoch": 0.23495, + "grad_norm": 0.6768390839711574, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 23495 + }, + { + "epoch": 0.23496, + "grad_norm": 0.6276843134632396, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 23496 + }, + { + "epoch": 0.23497, + "grad_norm": 0.6081875798474428, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 23497 + }, + { + "epoch": 0.23498, + "grad_norm": 0.5997192783687194, + "learning_rate": 0.003, + "loss": 3.981, + "step": 23498 + }, + { + "epoch": 0.23499, + "grad_norm": 0.5679011435998159, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 23499 + }, + { + "epoch": 0.235, + "grad_norm": 0.5855250413443822, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 23500 + }, + { + "epoch": 0.23501, + "grad_norm": 0.5832425055137256, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 23501 + }, + { + "epoch": 0.23502, + "grad_norm": 0.5922927351279519, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 23502 + }, + { + "epoch": 0.23503, + "grad_norm": 0.6246239784606173, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 23503 + }, + { + "epoch": 0.23504, + "grad_norm": 0.684127127663136, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 23504 + }, + { + "epoch": 0.23505, + "grad_norm": 0.7558166248824877, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 23505 + }, + { + "epoch": 0.23506, + "grad_norm": 0.835198762497784, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 23506 + }, + { + "epoch": 0.23507, + "grad_norm": 0.614121593037165, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 23507 + }, + { + "epoch": 0.23508, + "grad_norm": 0.5727528218891564, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 23508 + }, + { + "epoch": 0.23509, + "grad_norm": 0.600200553932663, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 23509 + }, + { + "epoch": 0.2351, + "grad_norm": 0.6084413684797249, + "learning_rate": 0.003, + "loss": 3.989, + "step": 23510 + }, + { + "epoch": 0.23511, + "grad_norm": 0.606009589702263, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 23511 + }, + { + "epoch": 0.23512, + "grad_norm": 0.5079464311076524, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 23512 + }, + { + "epoch": 0.23513, + "grad_norm": 0.5274545958756583, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 23513 + }, + { + "epoch": 0.23514, + "grad_norm": 0.5099661845635294, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 23514 + }, + { + "epoch": 0.23515, + "grad_norm": 0.5101041581742908, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 23515 + }, + { + "epoch": 0.23516, + "grad_norm": 0.5415649852464889, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 23516 + }, + { + "epoch": 0.23517, + "grad_norm": 0.6244444750088904, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 23517 + }, + { + "epoch": 0.23518, + "grad_norm": 0.7119088827857405, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 23518 + }, + { + "epoch": 0.23519, + "grad_norm": 0.645803870207179, + "learning_rate": 0.003, + "loss": 3.991, + "step": 23519 + }, + { + "epoch": 0.2352, + "grad_norm": 0.6276469682181063, + "learning_rate": 0.003, + "loss": 4.027, + "step": 23520 + }, + { + "epoch": 0.23521, + "grad_norm": 0.6703273826134106, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 23521 + }, + { + "epoch": 0.23522, + "grad_norm": 0.7665656288329569, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 23522 + }, + { + "epoch": 0.23523, + "grad_norm": 0.6919188236990672, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 23523 + }, + { + "epoch": 0.23524, + "grad_norm": 0.7118524114145884, + "learning_rate": 0.003, + "loss": 4.0456, + "step": 23524 + }, + { + "epoch": 0.23525, + "grad_norm": 0.7211359292235232, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23525 + }, + { + "epoch": 0.23526, + "grad_norm": 0.842412764312349, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 23526 + }, + { + "epoch": 0.23527, + "grad_norm": 0.8417096294932443, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 23527 + }, + { + "epoch": 0.23528, + "grad_norm": 0.9792234135859954, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 23528 + }, + { + "epoch": 0.23529, + "grad_norm": 1.1167419375906065, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 23529 + }, + { + "epoch": 0.2353, + "grad_norm": 0.9590409635117962, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 23530 + }, + { + "epoch": 0.23531, + "grad_norm": 0.8575024281194078, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 23531 + }, + { + "epoch": 0.23532, + "grad_norm": 0.8079787473142701, + "learning_rate": 0.003, + "loss": 4.038, + "step": 23532 + }, + { + "epoch": 0.23533, + "grad_norm": 0.8493479573311169, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 23533 + }, + { + "epoch": 0.23534, + "grad_norm": 1.0221637869725435, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 23534 + }, + { + "epoch": 0.23535, + "grad_norm": 1.1534448740154686, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 23535 + }, + { + "epoch": 0.23536, + "grad_norm": 0.8533616138973563, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 23536 + }, + { + "epoch": 0.23537, + "grad_norm": 0.8416860330529194, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 23537 + }, + { + "epoch": 0.23538, + "grad_norm": 0.8127416652946139, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 23538 + }, + { + "epoch": 0.23539, + "grad_norm": 0.7910595573366823, + "learning_rate": 0.003, + "loss": 4.006, + "step": 23539 + }, + { + "epoch": 0.2354, + "grad_norm": 0.8466965658827951, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 23540 + }, + { + "epoch": 0.23541, + "grad_norm": 0.8795815951414664, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 23541 + }, + { + "epoch": 0.23542, + "grad_norm": 0.8386414109985587, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 23542 + }, + { + "epoch": 0.23543, + "grad_norm": 0.8728564408250862, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 23543 + }, + { + "epoch": 0.23544, + "grad_norm": 0.8300105733524238, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 23544 + }, + { + "epoch": 0.23545, + "grad_norm": 0.8572485019235754, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 23545 + }, + { + "epoch": 0.23546, + "grad_norm": 0.8830000346069191, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 23546 + }, + { + "epoch": 0.23547, + "grad_norm": 0.928379897162833, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 23547 + }, + { + "epoch": 0.23548, + "grad_norm": 1.1190612014986898, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 23548 + }, + { + "epoch": 0.23549, + "grad_norm": 0.9993982163822056, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 23549 + }, + { + "epoch": 0.2355, + "grad_norm": 0.8016424164349296, + "learning_rate": 0.003, + "loss": 3.999, + "step": 23550 + }, + { + "epoch": 0.23551, + "grad_norm": 0.7118536389037959, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 23551 + }, + { + "epoch": 0.23552, + "grad_norm": 0.7261761157031248, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 23552 + }, + { + "epoch": 0.23553, + "grad_norm": 0.6873284374020181, + "learning_rate": 0.003, + "loss": 4.005, + "step": 23553 + }, + { + "epoch": 0.23554, + "grad_norm": 0.7538469211535742, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 23554 + }, + { + "epoch": 0.23555, + "grad_norm": 0.7710192929527541, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 23555 + }, + { + "epoch": 0.23556, + "grad_norm": 0.7042405918550373, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 23556 + }, + { + "epoch": 0.23557, + "grad_norm": 0.6225406413113755, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 23557 + }, + { + "epoch": 0.23558, + "grad_norm": 0.6216297341917665, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 23558 + }, + { + "epoch": 0.23559, + "grad_norm": 0.6305048187824606, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 23559 + }, + { + "epoch": 0.2356, + "grad_norm": 0.6033645157734304, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 23560 + }, + { + "epoch": 0.23561, + "grad_norm": 0.6565646870108126, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 23561 + }, + { + "epoch": 0.23562, + "grad_norm": 0.6848610304140766, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 23562 + }, + { + "epoch": 0.23563, + "grad_norm": 0.6281918675510586, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 23563 + }, + { + "epoch": 0.23564, + "grad_norm": 0.588675399856567, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 23564 + }, + { + "epoch": 0.23565, + "grad_norm": 0.5154703245508051, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 23565 + }, + { + "epoch": 0.23566, + "grad_norm": 0.549612387920211, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 23566 + }, + { + "epoch": 0.23567, + "grad_norm": 0.5438825722459979, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 23567 + }, + { + "epoch": 0.23568, + "grad_norm": 0.7000701439938087, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 23568 + }, + { + "epoch": 0.23569, + "grad_norm": 0.838643426838729, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 23569 + }, + { + "epoch": 0.2357, + "grad_norm": 0.8116948963462421, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 23570 + }, + { + "epoch": 0.23571, + "grad_norm": 0.705963341282948, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23571 + }, + { + "epoch": 0.23572, + "grad_norm": 0.6735935520650879, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 23572 + }, + { + "epoch": 0.23573, + "grad_norm": 0.696971043491987, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 23573 + }, + { + "epoch": 0.23574, + "grad_norm": 0.7407425660036709, + "learning_rate": 0.003, + "loss": 3.972, + "step": 23574 + }, + { + "epoch": 0.23575, + "grad_norm": 0.8011410689647485, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 23575 + }, + { + "epoch": 0.23576, + "grad_norm": 0.8111098608556531, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 23576 + }, + { + "epoch": 0.23577, + "grad_norm": 0.8379789641751886, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 23577 + }, + { + "epoch": 0.23578, + "grad_norm": 0.8472477592630518, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 23578 + }, + { + "epoch": 0.23579, + "grad_norm": 0.9136782418010359, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 23579 + }, + { + "epoch": 0.2358, + "grad_norm": 0.8466400590445006, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23580 + }, + { + "epoch": 0.23581, + "grad_norm": 0.6669128151331404, + "learning_rate": 0.003, + "loss": 3.99, + "step": 23581 + }, + { + "epoch": 0.23582, + "grad_norm": 0.6549738092812283, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 23582 + }, + { + "epoch": 0.23583, + "grad_norm": 0.7057786662298701, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 23583 + }, + { + "epoch": 0.23584, + "grad_norm": 0.5971427849826163, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 23584 + }, + { + "epoch": 0.23585, + "grad_norm": 0.6063125232638535, + "learning_rate": 0.003, + "loss": 3.981, + "step": 23585 + }, + { + "epoch": 0.23586, + "grad_norm": 0.6717809514254603, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 23586 + }, + { + "epoch": 0.23587, + "grad_norm": 0.8142949614009036, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 23587 + }, + { + "epoch": 0.23588, + "grad_norm": 0.9350471133811122, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 23588 + }, + { + "epoch": 0.23589, + "grad_norm": 1.1565172520879163, + "learning_rate": 0.003, + "loss": 4.015, + "step": 23589 + }, + { + "epoch": 0.2359, + "grad_norm": 0.9498666638155767, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 23590 + }, + { + "epoch": 0.23591, + "grad_norm": 0.8297158429584156, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23591 + }, + { + "epoch": 0.23592, + "grad_norm": 0.7470849150726755, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 23592 + }, + { + "epoch": 0.23593, + "grad_norm": 0.6914775837398386, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 23593 + }, + { + "epoch": 0.23594, + "grad_norm": 0.6022818900985678, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 23594 + }, + { + "epoch": 0.23595, + "grad_norm": 0.6483246950558806, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 23595 + }, + { + "epoch": 0.23596, + "grad_norm": 0.7241313862441073, + "learning_rate": 0.003, + "loss": 4.007, + "step": 23596 + }, + { + "epoch": 0.23597, + "grad_norm": 0.7849620115816001, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 23597 + }, + { + "epoch": 0.23598, + "grad_norm": 0.6678936297515146, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 23598 + }, + { + "epoch": 0.23599, + "grad_norm": 0.6775226174188892, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 23599 + }, + { + "epoch": 0.236, + "grad_norm": 0.7038023006727729, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 23600 + }, + { + "epoch": 0.23601, + "grad_norm": 0.7329760322126214, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 23601 + }, + { + "epoch": 0.23602, + "grad_norm": 0.78265599537015, + "learning_rate": 0.003, + "loss": 3.995, + "step": 23602 + }, + { + "epoch": 0.23603, + "grad_norm": 0.7887055968927874, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 23603 + }, + { + "epoch": 0.23604, + "grad_norm": 0.7731507394540142, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 23604 + }, + { + "epoch": 0.23605, + "grad_norm": 0.801093259439491, + "learning_rate": 0.003, + "loss": 4.019, + "step": 23605 + }, + { + "epoch": 0.23606, + "grad_norm": 0.8502861223069603, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 23606 + }, + { + "epoch": 0.23607, + "grad_norm": 0.8867382581228999, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 23607 + }, + { + "epoch": 0.23608, + "grad_norm": 0.8414840794885582, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 23608 + }, + { + "epoch": 0.23609, + "grad_norm": 0.9017385175497991, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 23609 + }, + { + "epoch": 0.2361, + "grad_norm": 0.8688249686193444, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 23610 + }, + { + "epoch": 0.23611, + "grad_norm": 0.920388648416139, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 23611 + }, + { + "epoch": 0.23612, + "grad_norm": 0.9315934324620894, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 23612 + }, + { + "epoch": 0.23613, + "grad_norm": 0.9590674836572646, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 23613 + }, + { + "epoch": 0.23614, + "grad_norm": 0.7950277960909655, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 23614 + }, + { + "epoch": 0.23615, + "grad_norm": 0.7079592758564446, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 23615 + }, + { + "epoch": 0.23616, + "grad_norm": 0.7992043874004295, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 23616 + }, + { + "epoch": 0.23617, + "grad_norm": 0.9178737907796148, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 23617 + }, + { + "epoch": 0.23618, + "grad_norm": 0.7430415514522993, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 23618 + }, + { + "epoch": 0.23619, + "grad_norm": 0.654023441213039, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 23619 + }, + { + "epoch": 0.2362, + "grad_norm": 0.5856115323030354, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 23620 + }, + { + "epoch": 0.23621, + "grad_norm": 0.6219702019391126, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 23621 + }, + { + "epoch": 0.23622, + "grad_norm": 0.5989725189275678, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 23622 + }, + { + "epoch": 0.23623, + "grad_norm": 0.5260788785537412, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 23623 + }, + { + "epoch": 0.23624, + "grad_norm": 0.48337853728607255, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 23624 + }, + { + "epoch": 0.23625, + "grad_norm": 0.4831793098830929, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 23625 + }, + { + "epoch": 0.23626, + "grad_norm": 0.4949124239019567, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 23626 + }, + { + "epoch": 0.23627, + "grad_norm": 0.5275887930043053, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 23627 + }, + { + "epoch": 0.23628, + "grad_norm": 0.5639148204241371, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 23628 + }, + { + "epoch": 0.23629, + "grad_norm": 0.5188038294239291, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 23629 + }, + { + "epoch": 0.2363, + "grad_norm": 0.5517561051276016, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 23630 + }, + { + "epoch": 0.23631, + "grad_norm": 0.5668360679500111, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 23631 + }, + { + "epoch": 0.23632, + "grad_norm": 0.713454123579158, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 23632 + }, + { + "epoch": 0.23633, + "grad_norm": 0.8380150299190284, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 23633 + }, + { + "epoch": 0.23634, + "grad_norm": 0.9613339722252894, + "learning_rate": 0.003, + "loss": 4.032, + "step": 23634 + }, + { + "epoch": 0.23635, + "grad_norm": 1.0844887256837352, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 23635 + }, + { + "epoch": 0.23636, + "grad_norm": 0.7621591936523737, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 23636 + }, + { + "epoch": 0.23637, + "grad_norm": 0.639042291849212, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 23637 + }, + { + "epoch": 0.23638, + "grad_norm": 0.5105555890967313, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 23638 + }, + { + "epoch": 0.23639, + "grad_norm": 0.5713119266077463, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 23639 + }, + { + "epoch": 0.2364, + "grad_norm": 0.5724456702733842, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 23640 + }, + { + "epoch": 0.23641, + "grad_norm": 0.604435998831999, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 23641 + }, + { + "epoch": 0.23642, + "grad_norm": 0.6123952822082495, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 23642 + }, + { + "epoch": 0.23643, + "grad_norm": 0.7323139122398629, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 23643 + }, + { + "epoch": 0.23644, + "grad_norm": 0.8337910032610829, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 23644 + }, + { + "epoch": 0.23645, + "grad_norm": 0.8843763360388549, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 23645 + }, + { + "epoch": 0.23646, + "grad_norm": 0.9759106871321527, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 23646 + }, + { + "epoch": 0.23647, + "grad_norm": 1.1829399665806823, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 23647 + }, + { + "epoch": 0.23648, + "grad_norm": 0.8421081602230156, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 23648 + }, + { + "epoch": 0.23649, + "grad_norm": 0.8368659114229041, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 23649 + }, + { + "epoch": 0.2365, + "grad_norm": 1.0393116536314346, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 23650 + }, + { + "epoch": 0.23651, + "grad_norm": 1.0234949414828833, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 23651 + }, + { + "epoch": 0.23652, + "grad_norm": 0.9439855178351407, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 23652 + }, + { + "epoch": 0.23653, + "grad_norm": 0.9831210437330115, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 23653 + }, + { + "epoch": 0.23654, + "grad_norm": 0.9669775959512343, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 23654 + }, + { + "epoch": 0.23655, + "grad_norm": 0.8967125523966073, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 23655 + }, + { + "epoch": 0.23656, + "grad_norm": 0.7424902065245768, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 23656 + }, + { + "epoch": 0.23657, + "grad_norm": 0.6656198824021996, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 23657 + }, + { + "epoch": 0.23658, + "grad_norm": 0.6799748721803383, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 23658 + }, + { + "epoch": 0.23659, + "grad_norm": 0.7682975843628458, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 23659 + }, + { + "epoch": 0.2366, + "grad_norm": 0.8571651948489388, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 23660 + }, + { + "epoch": 0.23661, + "grad_norm": 0.7537524448001239, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 23661 + }, + { + "epoch": 0.23662, + "grad_norm": 0.5970095843739188, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 23662 + }, + { + "epoch": 0.23663, + "grad_norm": 0.5598365645705075, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 23663 + }, + { + "epoch": 0.23664, + "grad_norm": 0.5068405647385132, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 23664 + }, + { + "epoch": 0.23665, + "grad_norm": 0.48409903149244127, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 23665 + }, + { + "epoch": 0.23666, + "grad_norm": 0.4923764660710827, + "learning_rate": 0.003, + "loss": 4.0532, + "step": 23666 + }, + { + "epoch": 0.23667, + "grad_norm": 0.582352835567841, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 23667 + }, + { + "epoch": 0.23668, + "grad_norm": 0.6784427432189937, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 23668 + }, + { + "epoch": 0.23669, + "grad_norm": 0.787240140498857, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 23669 + }, + { + "epoch": 0.2367, + "grad_norm": 0.87403841885239, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 23670 + }, + { + "epoch": 0.23671, + "grad_norm": 0.8973902569711266, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 23671 + }, + { + "epoch": 0.23672, + "grad_norm": 0.7842338039324769, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 23672 + }, + { + "epoch": 0.23673, + "grad_norm": 0.6029971612843196, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 23673 + }, + { + "epoch": 0.23674, + "grad_norm": 0.5889962908767064, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 23674 + }, + { + "epoch": 0.23675, + "grad_norm": 0.5235025682554958, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 23675 + }, + { + "epoch": 0.23676, + "grad_norm": 0.618401482296784, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 23676 + }, + { + "epoch": 0.23677, + "grad_norm": 0.7007005559142718, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 23677 + }, + { + "epoch": 0.23678, + "grad_norm": 0.6718998981107819, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 23678 + }, + { + "epoch": 0.23679, + "grad_norm": 0.603328059546957, + "learning_rate": 0.003, + "loss": 3.984, + "step": 23679 + }, + { + "epoch": 0.2368, + "grad_norm": 0.643834291676446, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 23680 + }, + { + "epoch": 0.23681, + "grad_norm": 0.7239959384321664, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 23681 + }, + { + "epoch": 0.23682, + "grad_norm": 0.7754075328314686, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 23682 + }, + { + "epoch": 0.23683, + "grad_norm": 0.8223174571119736, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 23683 + }, + { + "epoch": 0.23684, + "grad_norm": 0.9806493342615991, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 23684 + }, + { + "epoch": 0.23685, + "grad_norm": 1.0798369088921018, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 23685 + }, + { + "epoch": 0.23686, + "grad_norm": 0.8048979623377831, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 23686 + }, + { + "epoch": 0.23687, + "grad_norm": 0.8718403347973005, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 23687 + }, + { + "epoch": 0.23688, + "grad_norm": 0.7963203018710344, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 23688 + }, + { + "epoch": 0.23689, + "grad_norm": 0.7732993320606031, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 23689 + }, + { + "epoch": 0.2369, + "grad_norm": 0.6985570270054743, + "learning_rate": 0.003, + "loss": 4.022, + "step": 23690 + }, + { + "epoch": 0.23691, + "grad_norm": 0.7387954429729636, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 23691 + }, + { + "epoch": 0.23692, + "grad_norm": 0.7841804707605473, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 23692 + }, + { + "epoch": 0.23693, + "grad_norm": 0.8884205865885968, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 23693 + }, + { + "epoch": 0.23694, + "grad_norm": 0.9945796019267602, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 23694 + }, + { + "epoch": 0.23695, + "grad_norm": 0.9553802994441154, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 23695 + }, + { + "epoch": 0.23696, + "grad_norm": 0.7846889864971754, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 23696 + }, + { + "epoch": 0.23697, + "grad_norm": 0.7287281035063877, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 23697 + }, + { + "epoch": 0.23698, + "grad_norm": 0.75091678188314, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 23698 + }, + { + "epoch": 0.23699, + "grad_norm": 0.7786381115787511, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 23699 + }, + { + "epoch": 0.237, + "grad_norm": 0.7077688924165233, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 23700 + }, + { + "epoch": 0.23701, + "grad_norm": 0.8137873478922891, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 23701 + }, + { + "epoch": 0.23702, + "grad_norm": 1.0248666383062728, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 23702 + }, + { + "epoch": 0.23703, + "grad_norm": 1.0283894889063927, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 23703 + }, + { + "epoch": 0.23704, + "grad_norm": 0.8664100287029264, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 23704 + }, + { + "epoch": 0.23705, + "grad_norm": 0.8159889774006083, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 23705 + }, + { + "epoch": 0.23706, + "grad_norm": 0.6831666915483511, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 23706 + }, + { + "epoch": 0.23707, + "grad_norm": 0.7815061239495567, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 23707 + }, + { + "epoch": 0.23708, + "grad_norm": 0.8866483346318265, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 23708 + }, + { + "epoch": 0.23709, + "grad_norm": 0.8781683866083723, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 23709 + }, + { + "epoch": 0.2371, + "grad_norm": 0.853875905568221, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 23710 + }, + { + "epoch": 0.23711, + "grad_norm": 0.6846193958679458, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 23711 + }, + { + "epoch": 0.23712, + "grad_norm": 0.6447390291169907, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 23712 + }, + { + "epoch": 0.23713, + "grad_norm": 0.6577264352111567, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 23713 + }, + { + "epoch": 0.23714, + "grad_norm": 0.7041684220921144, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 23714 + }, + { + "epoch": 0.23715, + "grad_norm": 0.846644331639384, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 23715 + }, + { + "epoch": 0.23716, + "grad_norm": 1.0808109386858076, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 23716 + }, + { + "epoch": 0.23717, + "grad_norm": 1.0601144490673762, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 23717 + }, + { + "epoch": 0.23718, + "grad_norm": 0.9014369397105333, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 23718 + }, + { + "epoch": 0.23719, + "grad_norm": 0.7952727247983432, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 23719 + }, + { + "epoch": 0.2372, + "grad_norm": 0.8105581944957254, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 23720 + }, + { + "epoch": 0.23721, + "grad_norm": 0.8295988895406041, + "learning_rate": 0.003, + "loss": 3.976, + "step": 23721 + }, + { + "epoch": 0.23722, + "grad_norm": 0.8125089274282982, + "learning_rate": 0.003, + "loss": 4.007, + "step": 23722 + }, + { + "epoch": 0.23723, + "grad_norm": 0.832995207586792, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 23723 + }, + { + "epoch": 0.23724, + "grad_norm": 0.8230064007952005, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 23724 + }, + { + "epoch": 0.23725, + "grad_norm": 0.8148303111835239, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 23725 + }, + { + "epoch": 0.23726, + "grad_norm": 0.7362585378802879, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 23726 + }, + { + "epoch": 0.23727, + "grad_norm": 0.778953782171132, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 23727 + }, + { + "epoch": 0.23728, + "grad_norm": 0.799433214409683, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 23728 + }, + { + "epoch": 0.23729, + "grad_norm": 0.7201028317079521, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 23729 + }, + { + "epoch": 0.2373, + "grad_norm": 0.7599822624475111, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 23730 + }, + { + "epoch": 0.23731, + "grad_norm": 0.7042288496459767, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 23731 + }, + { + "epoch": 0.23732, + "grad_norm": 0.68552356392103, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 23732 + }, + { + "epoch": 0.23733, + "grad_norm": 0.630565157116945, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 23733 + }, + { + "epoch": 0.23734, + "grad_norm": 0.4863634013546654, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 23734 + }, + { + "epoch": 0.23735, + "grad_norm": 0.5091623639098505, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 23735 + }, + { + "epoch": 0.23736, + "grad_norm": 0.5133649574657773, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 23736 + }, + { + "epoch": 0.23737, + "grad_norm": 0.43447204779118404, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 23737 + }, + { + "epoch": 0.23738, + "grad_norm": 0.43586240029275924, + "learning_rate": 0.003, + "loss": 3.994, + "step": 23738 + }, + { + "epoch": 0.23739, + "grad_norm": 0.44811162063867904, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 23739 + }, + { + "epoch": 0.2374, + "grad_norm": 0.4869898901376951, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 23740 + }, + { + "epoch": 0.23741, + "grad_norm": 0.48298020939053754, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 23741 + }, + { + "epoch": 0.23742, + "grad_norm": 0.5969275004300314, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 23742 + }, + { + "epoch": 0.23743, + "grad_norm": 0.7058650618156859, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 23743 + }, + { + "epoch": 0.23744, + "grad_norm": 0.9128053357511708, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 23744 + }, + { + "epoch": 0.23745, + "grad_norm": 1.059645990607614, + "learning_rate": 0.003, + "loss": 4.002, + "step": 23745 + }, + { + "epoch": 0.23746, + "grad_norm": 0.7636697399555624, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 23746 + }, + { + "epoch": 0.23747, + "grad_norm": 0.84159131156507, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 23747 + }, + { + "epoch": 0.23748, + "grad_norm": 0.8971164207925933, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 23748 + }, + { + "epoch": 0.23749, + "grad_norm": 0.8142575066246822, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 23749 + }, + { + "epoch": 0.2375, + "grad_norm": 0.8118051161292765, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 23750 + }, + { + "epoch": 0.23751, + "grad_norm": 0.7605788638987604, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 23751 + }, + { + "epoch": 0.23752, + "grad_norm": 0.7717693455127109, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 23752 + }, + { + "epoch": 0.23753, + "grad_norm": 0.8742426940141962, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 23753 + }, + { + "epoch": 0.23754, + "grad_norm": 0.7815046627852975, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 23754 + }, + { + "epoch": 0.23755, + "grad_norm": 0.7326196034088124, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 23755 + }, + { + "epoch": 0.23756, + "grad_norm": 0.7115572088852112, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 23756 + }, + { + "epoch": 0.23757, + "grad_norm": 0.7080389777468726, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 23757 + }, + { + "epoch": 0.23758, + "grad_norm": 0.7901868766478624, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 23758 + }, + { + "epoch": 0.23759, + "grad_norm": 0.7920653137584274, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 23759 + }, + { + "epoch": 0.2376, + "grad_norm": 0.7860064051828293, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 23760 + }, + { + "epoch": 0.23761, + "grad_norm": 0.8130142084378513, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 23761 + }, + { + "epoch": 0.23762, + "grad_norm": 0.7936303777437959, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 23762 + }, + { + "epoch": 0.23763, + "grad_norm": 0.9131863889450108, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 23763 + }, + { + "epoch": 0.23764, + "grad_norm": 0.9856526351022414, + "learning_rate": 0.003, + "loss": 3.972, + "step": 23764 + }, + { + "epoch": 0.23765, + "grad_norm": 0.9697435177349035, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 23765 + }, + { + "epoch": 0.23766, + "grad_norm": 0.9615424122671168, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 23766 + }, + { + "epoch": 0.23767, + "grad_norm": 0.8747776373759631, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 23767 + }, + { + "epoch": 0.23768, + "grad_norm": 0.8382052688824172, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 23768 + }, + { + "epoch": 0.23769, + "grad_norm": 0.8828766647484594, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 23769 + }, + { + "epoch": 0.2377, + "grad_norm": 0.8750762617503275, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 23770 + }, + { + "epoch": 0.23771, + "grad_norm": 0.9244060069082264, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 23771 + }, + { + "epoch": 0.23772, + "grad_norm": 0.8862030103078172, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 23772 + }, + { + "epoch": 0.23773, + "grad_norm": 0.896315670370995, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 23773 + }, + { + "epoch": 0.23774, + "grad_norm": 0.8308648930724408, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 23774 + }, + { + "epoch": 0.23775, + "grad_norm": 0.8242484780144127, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 23775 + }, + { + "epoch": 0.23776, + "grad_norm": 0.8361920906690653, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 23776 + }, + { + "epoch": 0.23777, + "grad_norm": 0.8093497489497241, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 23777 + }, + { + "epoch": 0.23778, + "grad_norm": 0.9493233004271007, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 23778 + }, + { + "epoch": 0.23779, + "grad_norm": 1.0039266279260903, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 23779 + }, + { + "epoch": 0.2378, + "grad_norm": 0.9954156586925269, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 23780 + }, + { + "epoch": 0.23781, + "grad_norm": 0.8095494052946898, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 23781 + }, + { + "epoch": 0.23782, + "grad_norm": 0.7373765177426802, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 23782 + }, + { + "epoch": 0.23783, + "grad_norm": 0.6691206751698829, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 23783 + }, + { + "epoch": 0.23784, + "grad_norm": 0.5936570081078172, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 23784 + }, + { + "epoch": 0.23785, + "grad_norm": 0.6293217340270579, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 23785 + }, + { + "epoch": 0.23786, + "grad_norm": 0.6577781863555315, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 23786 + }, + { + "epoch": 0.23787, + "grad_norm": 0.7369246225956075, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 23787 + }, + { + "epoch": 0.23788, + "grad_norm": 0.7065016664990171, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 23788 + }, + { + "epoch": 0.23789, + "grad_norm": 0.658096292154981, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 23789 + }, + { + "epoch": 0.2379, + "grad_norm": 0.7294223168866366, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 23790 + }, + { + "epoch": 0.23791, + "grad_norm": 0.726625000330788, + "learning_rate": 0.003, + "loss": 3.979, + "step": 23791 + }, + { + "epoch": 0.23792, + "grad_norm": 0.6457362301406437, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 23792 + }, + { + "epoch": 0.23793, + "grad_norm": 0.5674887264116283, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 23793 + }, + { + "epoch": 0.23794, + "grad_norm": 0.505633625103957, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 23794 + }, + { + "epoch": 0.23795, + "grad_norm": 0.4940033314877615, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 23795 + }, + { + "epoch": 0.23796, + "grad_norm": 0.5562771286099057, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 23796 + }, + { + "epoch": 0.23797, + "grad_norm": 0.6454685421739675, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 23797 + }, + { + "epoch": 0.23798, + "grad_norm": 0.8267418340519522, + "learning_rate": 0.003, + "loss": 3.974, + "step": 23798 + }, + { + "epoch": 0.23799, + "grad_norm": 1.0208197776041064, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 23799 + }, + { + "epoch": 0.238, + "grad_norm": 0.9966399445217513, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 23800 + }, + { + "epoch": 0.23801, + "grad_norm": 0.7354349827352598, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 23801 + }, + { + "epoch": 0.23802, + "grad_norm": 0.8039851078805113, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 23802 + }, + { + "epoch": 0.23803, + "grad_norm": 0.9972295056154277, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 23803 + }, + { + "epoch": 0.23804, + "grad_norm": 0.9300776006331374, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 23804 + }, + { + "epoch": 0.23805, + "grad_norm": 0.9433221531543531, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 23805 + }, + { + "epoch": 0.23806, + "grad_norm": 0.9531575503006097, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 23806 + }, + { + "epoch": 0.23807, + "grad_norm": 0.8590094905233862, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 23807 + }, + { + "epoch": 0.23808, + "grad_norm": 0.7583564508504179, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 23808 + }, + { + "epoch": 0.23809, + "grad_norm": 0.7805859513839474, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 23809 + }, + { + "epoch": 0.2381, + "grad_norm": 0.7650776106493504, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 23810 + }, + { + "epoch": 0.23811, + "grad_norm": 0.707033332254749, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 23811 + }, + { + "epoch": 0.23812, + "grad_norm": 0.7131777368579582, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 23812 + }, + { + "epoch": 0.23813, + "grad_norm": 0.7369742945528243, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 23813 + }, + { + "epoch": 0.23814, + "grad_norm": 0.8081242592987758, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 23814 + }, + { + "epoch": 0.23815, + "grad_norm": 0.7377200838999262, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 23815 + }, + { + "epoch": 0.23816, + "grad_norm": 0.7328598374063674, + "learning_rate": 0.003, + "loss": 3.992, + "step": 23816 + }, + { + "epoch": 0.23817, + "grad_norm": 0.6599267829731522, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 23817 + }, + { + "epoch": 0.23818, + "grad_norm": 0.639411615149748, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 23818 + }, + { + "epoch": 0.23819, + "grad_norm": 0.5903867471290757, + "learning_rate": 0.003, + "loss": 4.009, + "step": 23819 + }, + { + "epoch": 0.2382, + "grad_norm": 0.5966303980008132, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 23820 + }, + { + "epoch": 0.23821, + "grad_norm": 0.6700001985398344, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 23821 + }, + { + "epoch": 0.23822, + "grad_norm": 0.6766824741148336, + "learning_rate": 0.003, + "loss": 4.014, + "step": 23822 + }, + { + "epoch": 0.23823, + "grad_norm": 0.6419370443191525, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 23823 + }, + { + "epoch": 0.23824, + "grad_norm": 0.6750407966168529, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 23824 + }, + { + "epoch": 0.23825, + "grad_norm": 0.7947693115150174, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 23825 + }, + { + "epoch": 0.23826, + "grad_norm": 0.8504260449659654, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 23826 + }, + { + "epoch": 0.23827, + "grad_norm": 0.7165169992116185, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 23827 + }, + { + "epoch": 0.23828, + "grad_norm": 0.5749116035837187, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 23828 + }, + { + "epoch": 0.23829, + "grad_norm": 0.545514188961954, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 23829 + }, + { + "epoch": 0.2383, + "grad_norm": 0.5502497983395033, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 23830 + }, + { + "epoch": 0.23831, + "grad_norm": 0.611063275569749, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 23831 + }, + { + "epoch": 0.23832, + "grad_norm": 0.7123407501547808, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 23832 + }, + { + "epoch": 0.23833, + "grad_norm": 0.8627509371714234, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 23833 + }, + { + "epoch": 0.23834, + "grad_norm": 0.9333968946131149, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 23834 + }, + { + "epoch": 0.23835, + "grad_norm": 0.9415833507604159, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 23835 + }, + { + "epoch": 0.23836, + "grad_norm": 0.7977737904173907, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 23836 + }, + { + "epoch": 0.23837, + "grad_norm": 0.6166653562233742, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 23837 + }, + { + "epoch": 0.23838, + "grad_norm": 0.5944721624301791, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 23838 + }, + { + "epoch": 0.23839, + "grad_norm": 0.680513230586265, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 23839 + }, + { + "epoch": 0.2384, + "grad_norm": 0.666757004492497, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 23840 + }, + { + "epoch": 0.23841, + "grad_norm": 0.6750162804630755, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 23841 + }, + { + "epoch": 0.23842, + "grad_norm": 0.642279663720618, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 23842 + }, + { + "epoch": 0.23843, + "grad_norm": 0.6788113131107502, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 23843 + }, + { + "epoch": 0.23844, + "grad_norm": 0.6676154587488508, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 23844 + }, + { + "epoch": 0.23845, + "grad_norm": 0.7460669489212165, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 23845 + }, + { + "epoch": 0.23846, + "grad_norm": 0.8355770692715725, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 23846 + }, + { + "epoch": 0.23847, + "grad_norm": 0.8072755223999634, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 23847 + }, + { + "epoch": 0.23848, + "grad_norm": 0.8354731880139568, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 23848 + }, + { + "epoch": 0.23849, + "grad_norm": 0.8092323674759812, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 23849 + }, + { + "epoch": 0.2385, + "grad_norm": 0.7433231615004948, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 23850 + }, + { + "epoch": 0.23851, + "grad_norm": 0.6474131039192237, + "learning_rate": 0.003, + "loss": 4.011, + "step": 23851 + }, + { + "epoch": 0.23852, + "grad_norm": 0.6875801456713408, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 23852 + }, + { + "epoch": 0.23853, + "grad_norm": 0.6598498272114046, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 23853 + }, + { + "epoch": 0.23854, + "grad_norm": 0.6284854607443849, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 23854 + }, + { + "epoch": 0.23855, + "grad_norm": 0.6743988122846927, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 23855 + }, + { + "epoch": 0.23856, + "grad_norm": 0.7164804171005341, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 23856 + }, + { + "epoch": 0.23857, + "grad_norm": 0.8432540991446981, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 23857 + }, + { + "epoch": 0.23858, + "grad_norm": 0.8406246167930537, + "learning_rate": 0.003, + "loss": 4.0472, + "step": 23858 + }, + { + "epoch": 0.23859, + "grad_norm": 0.7987086492444638, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 23859 + }, + { + "epoch": 0.2386, + "grad_norm": 0.7718908607894098, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 23860 + }, + { + "epoch": 0.23861, + "grad_norm": 0.7062700769900517, + "learning_rate": 0.003, + "loss": 3.966, + "step": 23861 + }, + { + "epoch": 0.23862, + "grad_norm": 0.7111122996966945, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 23862 + }, + { + "epoch": 0.23863, + "grad_norm": 0.9190836490289983, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 23863 + }, + { + "epoch": 0.23864, + "grad_norm": 1.011683795745693, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 23864 + }, + { + "epoch": 0.23865, + "grad_norm": 0.9414564193631754, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 23865 + }, + { + "epoch": 0.23866, + "grad_norm": 0.993575048864142, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 23866 + }, + { + "epoch": 0.23867, + "grad_norm": 0.9807725558451031, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 23867 + }, + { + "epoch": 0.23868, + "grad_norm": 0.8969677612496518, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 23868 + }, + { + "epoch": 0.23869, + "grad_norm": 0.8650525732129097, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 23869 + }, + { + "epoch": 0.2387, + "grad_norm": 0.834611317607101, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 23870 + }, + { + "epoch": 0.23871, + "grad_norm": 0.8238802761384012, + "learning_rate": 0.003, + "loss": 3.998, + "step": 23871 + }, + { + "epoch": 0.23872, + "grad_norm": 0.7876059843143993, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 23872 + }, + { + "epoch": 0.23873, + "grad_norm": 0.8153606346942678, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 23873 + }, + { + "epoch": 0.23874, + "grad_norm": 0.9134720944518973, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 23874 + }, + { + "epoch": 0.23875, + "grad_norm": 0.6910197471932078, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 23875 + }, + { + "epoch": 0.23876, + "grad_norm": 0.7008647548238857, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 23876 + }, + { + "epoch": 0.23877, + "grad_norm": 0.625964450489806, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 23877 + }, + { + "epoch": 0.23878, + "grad_norm": 0.6286858390782497, + "learning_rate": 0.003, + "loss": 3.973, + "step": 23878 + }, + { + "epoch": 0.23879, + "grad_norm": 0.6565955997001796, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 23879 + }, + { + "epoch": 0.2388, + "grad_norm": 0.6398471306757608, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 23880 + }, + { + "epoch": 0.23881, + "grad_norm": 0.6420861055763166, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 23881 + }, + { + "epoch": 0.23882, + "grad_norm": 0.6504804164097325, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 23882 + }, + { + "epoch": 0.23883, + "grad_norm": 0.7272961293860913, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 23883 + }, + { + "epoch": 0.23884, + "grad_norm": 0.690377247805414, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 23884 + }, + { + "epoch": 0.23885, + "grad_norm": 0.6551955827074732, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 23885 + }, + { + "epoch": 0.23886, + "grad_norm": 0.641911529672074, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 23886 + }, + { + "epoch": 0.23887, + "grad_norm": 0.6477113265844737, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 23887 + }, + { + "epoch": 0.23888, + "grad_norm": 0.6687111334627668, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 23888 + }, + { + "epoch": 0.23889, + "grad_norm": 0.6694206640474094, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 23889 + }, + { + "epoch": 0.2389, + "grad_norm": 0.6645179412142291, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 23890 + }, + { + "epoch": 0.23891, + "grad_norm": 0.7994794796882575, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 23891 + }, + { + "epoch": 0.23892, + "grad_norm": 0.9589336045203374, + "learning_rate": 0.003, + "loss": 3.98, + "step": 23892 + }, + { + "epoch": 0.23893, + "grad_norm": 0.973032686017359, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 23893 + }, + { + "epoch": 0.23894, + "grad_norm": 0.9181191218541005, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 23894 + }, + { + "epoch": 0.23895, + "grad_norm": 0.8535404536686939, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 23895 + }, + { + "epoch": 0.23896, + "grad_norm": 0.8234733480592296, + "learning_rate": 0.003, + "loss": 3.982, + "step": 23896 + }, + { + "epoch": 0.23897, + "grad_norm": 0.7870815723039661, + "learning_rate": 0.003, + "loss": 3.998, + "step": 23897 + }, + { + "epoch": 0.23898, + "grad_norm": 0.7989028864067768, + "learning_rate": 0.003, + "loss": 4.012, + "step": 23898 + }, + { + "epoch": 0.23899, + "grad_norm": 0.7437174697267941, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 23899 + }, + { + "epoch": 0.239, + "grad_norm": 0.7521371799614301, + "learning_rate": 0.003, + "loss": 3.982, + "step": 23900 + }, + { + "epoch": 0.23901, + "grad_norm": 0.6521516938641987, + "learning_rate": 0.003, + "loss": 4.001, + "step": 23901 + }, + { + "epoch": 0.23902, + "grad_norm": 0.5667581348170099, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 23902 + }, + { + "epoch": 0.23903, + "grad_norm": 0.5591858275792014, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 23903 + }, + { + "epoch": 0.23904, + "grad_norm": 0.544160028671356, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 23904 + }, + { + "epoch": 0.23905, + "grad_norm": 0.5308188989173538, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 23905 + }, + { + "epoch": 0.23906, + "grad_norm": 0.5903717643407742, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 23906 + }, + { + "epoch": 0.23907, + "grad_norm": 0.6201154011296746, + "learning_rate": 0.003, + "loss": 3.993, + "step": 23907 + }, + { + "epoch": 0.23908, + "grad_norm": 0.6906900499677665, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 23908 + }, + { + "epoch": 0.23909, + "grad_norm": 0.831273807127482, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 23909 + }, + { + "epoch": 0.2391, + "grad_norm": 1.0385557278394366, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 23910 + }, + { + "epoch": 0.23911, + "grad_norm": 1.1361598562083346, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 23911 + }, + { + "epoch": 0.23912, + "grad_norm": 0.8100719022692846, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 23912 + }, + { + "epoch": 0.23913, + "grad_norm": 0.7121294553535595, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 23913 + }, + { + "epoch": 0.23914, + "grad_norm": 0.749745432591235, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 23914 + }, + { + "epoch": 0.23915, + "grad_norm": 0.7009744570271929, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 23915 + }, + { + "epoch": 0.23916, + "grad_norm": 0.7763390910038728, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 23916 + }, + { + "epoch": 0.23917, + "grad_norm": 0.8461656843246878, + "learning_rate": 0.003, + "loss": 3.984, + "step": 23917 + }, + { + "epoch": 0.23918, + "grad_norm": 0.760836826032013, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 23918 + }, + { + "epoch": 0.23919, + "grad_norm": 0.7678874198316314, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 23919 + }, + { + "epoch": 0.2392, + "grad_norm": 1.040560144224798, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 23920 + }, + { + "epoch": 0.23921, + "grad_norm": 1.0703366455976273, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 23921 + }, + { + "epoch": 0.23922, + "grad_norm": 0.9112065526019051, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 23922 + }, + { + "epoch": 0.23923, + "grad_norm": 0.8480296015324008, + "learning_rate": 0.003, + "loss": 3.977, + "step": 23923 + }, + { + "epoch": 0.23924, + "grad_norm": 0.8111027029489909, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 23924 + }, + { + "epoch": 0.23925, + "grad_norm": 0.6629941170361768, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 23925 + }, + { + "epoch": 0.23926, + "grad_norm": 0.7110884307973473, + "learning_rate": 0.003, + "loss": 4.025, + "step": 23926 + }, + { + "epoch": 0.23927, + "grad_norm": 0.695890880692668, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 23927 + }, + { + "epoch": 0.23928, + "grad_norm": 0.7388100065978233, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 23928 + }, + { + "epoch": 0.23929, + "grad_norm": 0.8145346598066155, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 23929 + }, + { + "epoch": 0.2393, + "grad_norm": 0.7806058046412615, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 23930 + }, + { + "epoch": 0.23931, + "grad_norm": 0.7631870385816301, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 23931 + }, + { + "epoch": 0.23932, + "grad_norm": 0.7486067533409083, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 23932 + }, + { + "epoch": 0.23933, + "grad_norm": 0.7534757524242569, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 23933 + }, + { + "epoch": 0.23934, + "grad_norm": 0.9122900882876861, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 23934 + }, + { + "epoch": 0.23935, + "grad_norm": 0.9339306754794128, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 23935 + }, + { + "epoch": 0.23936, + "grad_norm": 0.9186469896782484, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 23936 + }, + { + "epoch": 0.23937, + "grad_norm": 0.8503757506085428, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 23937 + }, + { + "epoch": 0.23938, + "grad_norm": 0.8666198366788003, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 23938 + }, + { + "epoch": 0.23939, + "grad_norm": 0.929976005424698, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 23939 + }, + { + "epoch": 0.2394, + "grad_norm": 0.9244699211329702, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 23940 + }, + { + "epoch": 0.23941, + "grad_norm": 0.772950190667028, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 23941 + }, + { + "epoch": 0.23942, + "grad_norm": 0.7873452664763216, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 23942 + }, + { + "epoch": 0.23943, + "grad_norm": 0.6661692375315629, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 23943 + }, + { + "epoch": 0.23944, + "grad_norm": 0.5977701656959477, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 23944 + }, + { + "epoch": 0.23945, + "grad_norm": 0.5228498626436572, + "learning_rate": 0.003, + "loss": 3.984, + "step": 23945 + }, + { + "epoch": 0.23946, + "grad_norm": 0.4775424610812799, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 23946 + }, + { + "epoch": 0.23947, + "grad_norm": 0.4685735812227734, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 23947 + }, + { + "epoch": 0.23948, + "grad_norm": 0.44312214285635415, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 23948 + }, + { + "epoch": 0.23949, + "grad_norm": 0.4607359143446728, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 23949 + }, + { + "epoch": 0.2395, + "grad_norm": 0.47575508992785553, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 23950 + }, + { + "epoch": 0.23951, + "grad_norm": 0.5252658560247647, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 23951 + }, + { + "epoch": 0.23952, + "grad_norm": 0.7046004470908873, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 23952 + }, + { + "epoch": 0.23953, + "grad_norm": 1.0340650496626338, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 23953 + }, + { + "epoch": 0.23954, + "grad_norm": 1.2753446306559832, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 23954 + }, + { + "epoch": 0.23955, + "grad_norm": 0.5573857542919062, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 23955 + }, + { + "epoch": 0.23956, + "grad_norm": 0.8430015298840388, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 23956 + }, + { + "epoch": 0.23957, + "grad_norm": 0.811714879626114, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 23957 + }, + { + "epoch": 0.23958, + "grad_norm": 0.6802521603930519, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 23958 + }, + { + "epoch": 0.23959, + "grad_norm": 0.7218911908178742, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 23959 + }, + { + "epoch": 0.2396, + "grad_norm": 0.7913186690750192, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 23960 + }, + { + "epoch": 0.23961, + "grad_norm": 0.8168405744697927, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 23961 + }, + { + "epoch": 0.23962, + "grad_norm": 0.7951756410086119, + "learning_rate": 0.003, + "loss": 4.013, + "step": 23962 + }, + { + "epoch": 0.23963, + "grad_norm": 0.8532455541322195, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 23963 + }, + { + "epoch": 0.23964, + "grad_norm": 0.8486741834363123, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 23964 + }, + { + "epoch": 0.23965, + "grad_norm": 0.8248186812880677, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 23965 + }, + { + "epoch": 0.23966, + "grad_norm": 0.823109542493972, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 23966 + }, + { + "epoch": 0.23967, + "grad_norm": 0.8569224546471954, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 23967 + }, + { + "epoch": 0.23968, + "grad_norm": 0.8213224658404465, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 23968 + }, + { + "epoch": 0.23969, + "grad_norm": 0.7814822962024282, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 23969 + }, + { + "epoch": 0.2397, + "grad_norm": 0.7379578332178139, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 23970 + }, + { + "epoch": 0.23971, + "grad_norm": 0.7377413544603122, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 23971 + }, + { + "epoch": 0.23972, + "grad_norm": 0.7170816048043747, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 23972 + }, + { + "epoch": 0.23973, + "grad_norm": 0.7046989810589374, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 23973 + }, + { + "epoch": 0.23974, + "grad_norm": 0.616881933609788, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 23974 + }, + { + "epoch": 0.23975, + "grad_norm": 0.556576810669866, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 23975 + }, + { + "epoch": 0.23976, + "grad_norm": 0.4533454211913066, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 23976 + }, + { + "epoch": 0.23977, + "grad_norm": 0.4890153664579566, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 23977 + }, + { + "epoch": 0.23978, + "grad_norm": 0.47742685773695004, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 23978 + }, + { + "epoch": 0.23979, + "grad_norm": 0.47576920991442734, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 23979 + }, + { + "epoch": 0.2398, + "grad_norm": 0.5157548673494556, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 23980 + }, + { + "epoch": 0.23981, + "grad_norm": 0.6256242352991529, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 23981 + }, + { + "epoch": 0.23982, + "grad_norm": 0.7508394873051406, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 23982 + }, + { + "epoch": 0.23983, + "grad_norm": 0.9154078606055388, + "learning_rate": 0.003, + "loss": 3.981, + "step": 23983 + }, + { + "epoch": 0.23984, + "grad_norm": 0.9780818960149791, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 23984 + }, + { + "epoch": 0.23985, + "grad_norm": 1.006524649656554, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 23985 + }, + { + "epoch": 0.23986, + "grad_norm": 0.7611417996906081, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 23986 + }, + { + "epoch": 0.23987, + "grad_norm": 0.6453957700728218, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 23987 + }, + { + "epoch": 0.23988, + "grad_norm": 0.7424466666996759, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 23988 + }, + { + "epoch": 0.23989, + "grad_norm": 0.7813316349442596, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 23989 + }, + { + "epoch": 0.2399, + "grad_norm": 0.8362697407530026, + "learning_rate": 0.003, + "loss": 4.029, + "step": 23990 + }, + { + "epoch": 0.23991, + "grad_norm": 0.8860413390262658, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 23991 + }, + { + "epoch": 0.23992, + "grad_norm": 0.8435592249141431, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 23992 + }, + { + "epoch": 0.23993, + "grad_norm": 0.8970086926363815, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 23993 + }, + { + "epoch": 0.23994, + "grad_norm": 0.8245071921189246, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 23994 + }, + { + "epoch": 0.23995, + "grad_norm": 0.7506962661227303, + "learning_rate": 0.003, + "loss": 3.988, + "step": 23995 + }, + { + "epoch": 0.23996, + "grad_norm": 0.7499684327202267, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 23996 + }, + { + "epoch": 0.23997, + "grad_norm": 0.7496302382100183, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 23997 + }, + { + "epoch": 0.23998, + "grad_norm": 0.7623920470487434, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 23998 + }, + { + "epoch": 0.23999, + "grad_norm": 0.8970669175542473, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 23999 + }, + { + "epoch": 0.24, + "grad_norm": 1.020656100001593, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 24000 + }, + { + "epoch": 0.24001, + "grad_norm": 1.0423207126370058, + "learning_rate": 0.003, + "loss": 4.006, + "step": 24001 + }, + { + "epoch": 0.24002, + "grad_norm": 0.8528742981169922, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 24002 + }, + { + "epoch": 0.24003, + "grad_norm": 0.8552645263480185, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 24003 + }, + { + "epoch": 0.24004, + "grad_norm": 0.854409175893156, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 24004 + }, + { + "epoch": 0.24005, + "grad_norm": 0.8738947469821885, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 24005 + }, + { + "epoch": 0.24006, + "grad_norm": 0.8833326943558871, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 24006 + }, + { + "epoch": 0.24007, + "grad_norm": 0.823968372301891, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 24007 + }, + { + "epoch": 0.24008, + "grad_norm": 0.7279812383737583, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 24008 + }, + { + "epoch": 0.24009, + "grad_norm": 0.722212627933679, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 24009 + }, + { + "epoch": 0.2401, + "grad_norm": 0.7501989602651149, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 24010 + }, + { + "epoch": 0.24011, + "grad_norm": 0.7722470786557656, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 24011 + }, + { + "epoch": 0.24012, + "grad_norm": 0.642003448678568, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 24012 + }, + { + "epoch": 0.24013, + "grad_norm": 0.6419018600294584, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 24013 + }, + { + "epoch": 0.24014, + "grad_norm": 0.6109245835008564, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 24014 + }, + { + "epoch": 0.24015, + "grad_norm": 0.5399059907976015, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 24015 + }, + { + "epoch": 0.24016, + "grad_norm": 0.515694096907296, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 24016 + }, + { + "epoch": 0.24017, + "grad_norm": 0.5585723148478853, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 24017 + }, + { + "epoch": 0.24018, + "grad_norm": 0.5793055417946846, + "learning_rate": 0.003, + "loss": 3.997, + "step": 24018 + }, + { + "epoch": 0.24019, + "grad_norm": 0.6642541851401289, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 24019 + }, + { + "epoch": 0.2402, + "grad_norm": 0.7794556369654774, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 24020 + }, + { + "epoch": 0.24021, + "grad_norm": 0.8874130383437128, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 24021 + }, + { + "epoch": 0.24022, + "grad_norm": 0.8861480281314438, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 24022 + }, + { + "epoch": 0.24023, + "grad_norm": 0.7564442773502039, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 24023 + }, + { + "epoch": 0.24024, + "grad_norm": 0.6716273282233918, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 24024 + }, + { + "epoch": 0.24025, + "grad_norm": 0.5634956695740835, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 24025 + }, + { + "epoch": 0.24026, + "grad_norm": 0.51658370041928, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 24026 + }, + { + "epoch": 0.24027, + "grad_norm": 0.6687251860687317, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 24027 + }, + { + "epoch": 0.24028, + "grad_norm": 0.6700158844806846, + "learning_rate": 0.003, + "loss": 4.02, + "step": 24028 + }, + { + "epoch": 0.24029, + "grad_norm": 0.7125072825750725, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 24029 + }, + { + "epoch": 0.2403, + "grad_norm": 0.7410517549876009, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 24030 + }, + { + "epoch": 0.24031, + "grad_norm": 0.7558709421450535, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 24031 + }, + { + "epoch": 0.24032, + "grad_norm": 0.6971738999142753, + "learning_rate": 0.003, + "loss": 3.9488, + "step": 24032 + }, + { + "epoch": 0.24033, + "grad_norm": 0.6753129230111727, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 24033 + }, + { + "epoch": 0.24034, + "grad_norm": 0.6628066730200963, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 24034 + }, + { + "epoch": 0.24035, + "grad_norm": 0.7091699493198822, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 24035 + }, + { + "epoch": 0.24036, + "grad_norm": 0.5672795689670297, + "learning_rate": 0.003, + "loss": 3.988, + "step": 24036 + }, + { + "epoch": 0.24037, + "grad_norm": 0.5081327349597472, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 24037 + }, + { + "epoch": 0.24038, + "grad_norm": 0.5959451170429604, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 24038 + }, + { + "epoch": 0.24039, + "grad_norm": 0.6396881510930963, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 24039 + }, + { + "epoch": 0.2404, + "grad_norm": 0.7407230962941884, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 24040 + }, + { + "epoch": 0.24041, + "grad_norm": 0.9910918544769504, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 24041 + }, + { + "epoch": 0.24042, + "grad_norm": 1.2257750794153865, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 24042 + }, + { + "epoch": 0.24043, + "grad_norm": 0.543351996706451, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 24043 + }, + { + "epoch": 0.24044, + "grad_norm": 0.972095271384092, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 24044 + }, + { + "epoch": 0.24045, + "grad_norm": 1.1246826020474232, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 24045 + }, + { + "epoch": 0.24046, + "grad_norm": 0.7311208939261543, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 24046 + }, + { + "epoch": 0.24047, + "grad_norm": 0.8730678693967522, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 24047 + }, + { + "epoch": 0.24048, + "grad_norm": 0.8737947214926297, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 24048 + }, + { + "epoch": 0.24049, + "grad_norm": 0.845770573432941, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 24049 + }, + { + "epoch": 0.2405, + "grad_norm": 0.8149209685579656, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 24050 + }, + { + "epoch": 0.24051, + "grad_norm": 0.8273697183220782, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 24051 + }, + { + "epoch": 0.24052, + "grad_norm": 0.7733939304092595, + "learning_rate": 0.003, + "loss": 3.998, + "step": 24052 + }, + { + "epoch": 0.24053, + "grad_norm": 0.8025187737227368, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 24053 + }, + { + "epoch": 0.24054, + "grad_norm": 0.8738779269968968, + "learning_rate": 0.003, + "loss": 4.006, + "step": 24054 + }, + { + "epoch": 0.24055, + "grad_norm": 0.8957758790157988, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 24055 + }, + { + "epoch": 0.24056, + "grad_norm": 0.809395093122552, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 24056 + }, + { + "epoch": 0.24057, + "grad_norm": 0.6579536359540826, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 24057 + }, + { + "epoch": 0.24058, + "grad_norm": 0.724372751165546, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 24058 + }, + { + "epoch": 0.24059, + "grad_norm": 0.7352697502285058, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 24059 + }, + { + "epoch": 0.2406, + "grad_norm": 0.6722536906774985, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 24060 + }, + { + "epoch": 0.24061, + "grad_norm": 0.5971774892985988, + "learning_rate": 0.003, + "loss": 3.996, + "step": 24061 + }, + { + "epoch": 0.24062, + "grad_norm": 0.5889478554622933, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 24062 + }, + { + "epoch": 0.24063, + "grad_norm": 0.5089512293821424, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 24063 + }, + { + "epoch": 0.24064, + "grad_norm": 0.5655827717016358, + "learning_rate": 0.003, + "loss": 3.958, + "step": 24064 + }, + { + "epoch": 0.24065, + "grad_norm": 0.5601007194984932, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 24065 + }, + { + "epoch": 0.24066, + "grad_norm": 0.5255602496742149, + "learning_rate": 0.003, + "loss": 3.999, + "step": 24066 + }, + { + "epoch": 0.24067, + "grad_norm": 0.5730610655374239, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 24067 + }, + { + "epoch": 0.24068, + "grad_norm": 0.5988930947392489, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 24068 + }, + { + "epoch": 0.24069, + "grad_norm": 0.5838987895399841, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 24069 + }, + { + "epoch": 0.2407, + "grad_norm": 0.5602125005537432, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 24070 + }, + { + "epoch": 0.24071, + "grad_norm": 0.7022053718030866, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 24071 + }, + { + "epoch": 0.24072, + "grad_norm": 0.7876691127111125, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 24072 + }, + { + "epoch": 0.24073, + "grad_norm": 0.6783905975906085, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 24073 + }, + { + "epoch": 0.24074, + "grad_norm": 0.5118146398267368, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 24074 + }, + { + "epoch": 0.24075, + "grad_norm": 0.47547676599562383, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 24075 + }, + { + "epoch": 0.24076, + "grad_norm": 0.5045382274501793, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 24076 + }, + { + "epoch": 0.24077, + "grad_norm": 0.5467348448080196, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 24077 + }, + { + "epoch": 0.24078, + "grad_norm": 0.6600635193825464, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 24078 + }, + { + "epoch": 0.24079, + "grad_norm": 0.7617316501970102, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 24079 + }, + { + "epoch": 0.2408, + "grad_norm": 0.8445889300913397, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 24080 + }, + { + "epoch": 0.24081, + "grad_norm": 0.8612733996999142, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 24081 + }, + { + "epoch": 0.24082, + "grad_norm": 0.9891697000220616, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 24082 + }, + { + "epoch": 0.24083, + "grad_norm": 1.151579009659419, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 24083 + }, + { + "epoch": 0.24084, + "grad_norm": 0.9805511095162168, + "learning_rate": 0.003, + "loss": 4.0464, + "step": 24084 + }, + { + "epoch": 0.24085, + "grad_norm": 1.080520350889555, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 24085 + }, + { + "epoch": 0.24086, + "grad_norm": 0.9641020922688512, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 24086 + }, + { + "epoch": 0.24087, + "grad_norm": 1.1514243017033272, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 24087 + }, + { + "epoch": 0.24088, + "grad_norm": 1.0349495342829342, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 24088 + }, + { + "epoch": 0.24089, + "grad_norm": 1.1389998646163344, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 24089 + }, + { + "epoch": 0.2409, + "grad_norm": 0.9109622714389345, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 24090 + }, + { + "epoch": 0.24091, + "grad_norm": 0.827920585093496, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 24091 + }, + { + "epoch": 0.24092, + "grad_norm": 0.6890961228230968, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 24092 + }, + { + "epoch": 0.24093, + "grad_norm": 0.7305129044935654, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 24093 + }, + { + "epoch": 0.24094, + "grad_norm": 0.6869428925515852, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 24094 + }, + { + "epoch": 0.24095, + "grad_norm": 0.6237500666440063, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 24095 + }, + { + "epoch": 0.24096, + "grad_norm": 0.7103710361815881, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 24096 + }, + { + "epoch": 0.24097, + "grad_norm": 0.7446310935175752, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 24097 + }, + { + "epoch": 0.24098, + "grad_norm": 0.7285781360289291, + "learning_rate": 0.003, + "loss": 4.004, + "step": 24098 + }, + { + "epoch": 0.24099, + "grad_norm": 0.7106491539082339, + "learning_rate": 0.003, + "loss": 4.005, + "step": 24099 + }, + { + "epoch": 0.241, + "grad_norm": 0.6010783474972887, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 24100 + }, + { + "epoch": 0.24101, + "grad_norm": 0.5957179049470901, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 24101 + }, + { + "epoch": 0.24102, + "grad_norm": 0.6211812389149051, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 24102 + }, + { + "epoch": 0.24103, + "grad_norm": 0.586000976148174, + "learning_rate": 0.003, + "loss": 4.011, + "step": 24103 + }, + { + "epoch": 0.24104, + "grad_norm": 0.6061011034100309, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 24104 + }, + { + "epoch": 0.24105, + "grad_norm": 0.597213450612653, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 24105 + }, + { + "epoch": 0.24106, + "grad_norm": 0.6035342610191885, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 24106 + }, + { + "epoch": 0.24107, + "grad_norm": 0.7004680030828957, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 24107 + }, + { + "epoch": 0.24108, + "grad_norm": 0.7425002737779036, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24108 + }, + { + "epoch": 0.24109, + "grad_norm": 0.7108903139388929, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 24109 + }, + { + "epoch": 0.2411, + "grad_norm": 0.8757386402028341, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 24110 + }, + { + "epoch": 0.24111, + "grad_norm": 1.0435541625979419, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 24111 + }, + { + "epoch": 0.24112, + "grad_norm": 0.7696593941293828, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 24112 + }, + { + "epoch": 0.24113, + "grad_norm": 0.6718866830527415, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 24113 + }, + { + "epoch": 0.24114, + "grad_norm": 0.7508011725441233, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 24114 + }, + { + "epoch": 0.24115, + "grad_norm": 0.9982060592511913, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 24115 + }, + { + "epoch": 0.24116, + "grad_norm": 1.0552288605280231, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 24116 + }, + { + "epoch": 0.24117, + "grad_norm": 0.7802907467779532, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 24117 + }, + { + "epoch": 0.24118, + "grad_norm": 0.7809365348336853, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 24118 + }, + { + "epoch": 0.24119, + "grad_norm": 0.8251680453453644, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 24119 + }, + { + "epoch": 0.2412, + "grad_norm": 0.7983898403666152, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 24120 + }, + { + "epoch": 0.24121, + "grad_norm": 0.7970939181082574, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 24121 + }, + { + "epoch": 0.24122, + "grad_norm": 0.7709096495685845, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 24122 + }, + { + "epoch": 0.24123, + "grad_norm": 0.8187192615712799, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 24123 + }, + { + "epoch": 0.24124, + "grad_norm": 0.9676431170363078, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24124 + }, + { + "epoch": 0.24125, + "grad_norm": 0.843632195690662, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 24125 + }, + { + "epoch": 0.24126, + "grad_norm": 0.8510994031934741, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 24126 + }, + { + "epoch": 0.24127, + "grad_norm": 0.8276320817687864, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 24127 + }, + { + "epoch": 0.24128, + "grad_norm": 0.7158057131635549, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 24128 + }, + { + "epoch": 0.24129, + "grad_norm": 0.797450625169956, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 24129 + }, + { + "epoch": 0.2413, + "grad_norm": 0.7933704499977551, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 24130 + }, + { + "epoch": 0.24131, + "grad_norm": 0.8430949139697897, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 24131 + }, + { + "epoch": 0.24132, + "grad_norm": 0.9115821399930295, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 24132 + }, + { + "epoch": 0.24133, + "grad_norm": 0.8029301297728856, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 24133 + }, + { + "epoch": 0.24134, + "grad_norm": 0.8269384838773213, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 24134 + }, + { + "epoch": 0.24135, + "grad_norm": 0.8181213734755717, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 24135 + }, + { + "epoch": 0.24136, + "grad_norm": 0.7304048230987228, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 24136 + }, + { + "epoch": 0.24137, + "grad_norm": 0.7568382464209019, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 24137 + }, + { + "epoch": 0.24138, + "grad_norm": 0.7139860948404793, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 24138 + }, + { + "epoch": 0.24139, + "grad_norm": 0.6263057633539952, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 24139 + }, + { + "epoch": 0.2414, + "grad_norm": 0.6613408548143578, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 24140 + }, + { + "epoch": 0.24141, + "grad_norm": 0.6508689805687434, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 24141 + }, + { + "epoch": 0.24142, + "grad_norm": 0.634026849720622, + "learning_rate": 0.003, + "loss": 4.007, + "step": 24142 + }, + { + "epoch": 0.24143, + "grad_norm": 0.6309830603322314, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 24143 + }, + { + "epoch": 0.24144, + "grad_norm": 0.664217010598794, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 24144 + }, + { + "epoch": 0.24145, + "grad_norm": 0.8216612520005557, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 24145 + }, + { + "epoch": 0.24146, + "grad_norm": 0.8517427776602273, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 24146 + }, + { + "epoch": 0.24147, + "grad_norm": 0.7343127792749985, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 24147 + }, + { + "epoch": 0.24148, + "grad_norm": 0.7198944777246762, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 24148 + }, + { + "epoch": 0.24149, + "grad_norm": 0.681377505126136, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 24149 + }, + { + "epoch": 0.2415, + "grad_norm": 0.7644018388016312, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 24150 + }, + { + "epoch": 0.24151, + "grad_norm": 0.9414328428151322, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 24151 + }, + { + "epoch": 0.24152, + "grad_norm": 0.9259760755746284, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 24152 + }, + { + "epoch": 0.24153, + "grad_norm": 0.7399789960872518, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 24153 + }, + { + "epoch": 0.24154, + "grad_norm": 0.6998023864408226, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 24154 + }, + { + "epoch": 0.24155, + "grad_norm": 0.7165822078979329, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 24155 + }, + { + "epoch": 0.24156, + "grad_norm": 0.6539420171507424, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 24156 + }, + { + "epoch": 0.24157, + "grad_norm": 0.5884104440188506, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 24157 + }, + { + "epoch": 0.24158, + "grad_norm": 0.5967848317143547, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 24158 + }, + { + "epoch": 0.24159, + "grad_norm": 0.6801981548225974, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 24159 + }, + { + "epoch": 0.2416, + "grad_norm": 0.6853515602860734, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 24160 + }, + { + "epoch": 0.24161, + "grad_norm": 0.7446539882502896, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 24161 + }, + { + "epoch": 0.24162, + "grad_norm": 0.8072380552239535, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 24162 + }, + { + "epoch": 0.24163, + "grad_norm": 0.8264644550000174, + "learning_rate": 0.003, + "loss": 4.005, + "step": 24163 + }, + { + "epoch": 0.24164, + "grad_norm": 0.7435421726423038, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 24164 + }, + { + "epoch": 0.24165, + "grad_norm": 0.6363038414034341, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 24165 + }, + { + "epoch": 0.24166, + "grad_norm": 0.6253685259407658, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 24166 + }, + { + "epoch": 0.24167, + "grad_norm": 0.7005368406667769, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 24167 + }, + { + "epoch": 0.24168, + "grad_norm": 0.817926268120287, + "learning_rate": 0.003, + "loss": 4.0402, + "step": 24168 + }, + { + "epoch": 0.24169, + "grad_norm": 0.7547393990813, + "learning_rate": 0.003, + "loss": 3.981, + "step": 24169 + }, + { + "epoch": 0.2417, + "grad_norm": 0.7616851463876199, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 24170 + }, + { + "epoch": 0.24171, + "grad_norm": 0.7700902394169843, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 24171 + }, + { + "epoch": 0.24172, + "grad_norm": 0.7630100242605735, + "learning_rate": 0.003, + "loss": 3.982, + "step": 24172 + }, + { + "epoch": 0.24173, + "grad_norm": 0.7926672541192732, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 24173 + }, + { + "epoch": 0.24174, + "grad_norm": 0.8367518392084674, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 24174 + }, + { + "epoch": 0.24175, + "grad_norm": 0.9485271940782738, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 24175 + }, + { + "epoch": 0.24176, + "grad_norm": 1.0810442293565388, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 24176 + }, + { + "epoch": 0.24177, + "grad_norm": 0.8835573233264984, + "learning_rate": 0.003, + "loss": 4.007, + "step": 24177 + }, + { + "epoch": 0.24178, + "grad_norm": 0.8065413054920164, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 24178 + }, + { + "epoch": 0.24179, + "grad_norm": 0.7368379365035536, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 24179 + }, + { + "epoch": 0.2418, + "grad_norm": 0.6734095510094229, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 24180 + }, + { + "epoch": 0.24181, + "grad_norm": 0.7702829916277691, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 24181 + }, + { + "epoch": 0.24182, + "grad_norm": 1.0120410864344047, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 24182 + }, + { + "epoch": 0.24183, + "grad_norm": 0.8771139084408537, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 24183 + }, + { + "epoch": 0.24184, + "grad_norm": 0.6564341832441527, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 24184 + }, + { + "epoch": 0.24185, + "grad_norm": 0.6386505460316072, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 24185 + }, + { + "epoch": 0.24186, + "grad_norm": 0.6829265639795596, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 24186 + }, + { + "epoch": 0.24187, + "grad_norm": 0.7203618234006983, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 24187 + }, + { + "epoch": 0.24188, + "grad_norm": 0.7354609489946845, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 24188 + }, + { + "epoch": 0.24189, + "grad_norm": 0.7390652932559163, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 24189 + }, + { + "epoch": 0.2419, + "grad_norm": 0.7047478778652896, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 24190 + }, + { + "epoch": 0.24191, + "grad_norm": 0.7602494093159966, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 24191 + }, + { + "epoch": 0.24192, + "grad_norm": 0.8362048216613068, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 24192 + }, + { + "epoch": 0.24193, + "grad_norm": 0.7885359269443478, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 24193 + }, + { + "epoch": 0.24194, + "grad_norm": 0.6915335871391616, + "learning_rate": 0.003, + "loss": 3.984, + "step": 24194 + }, + { + "epoch": 0.24195, + "grad_norm": 0.6748382463715741, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 24195 + }, + { + "epoch": 0.24196, + "grad_norm": 0.680847247615716, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 24196 + }, + { + "epoch": 0.24197, + "grad_norm": 0.7550463396853254, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 24197 + }, + { + "epoch": 0.24198, + "grad_norm": 0.9335147634621319, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 24198 + }, + { + "epoch": 0.24199, + "grad_norm": 1.1197330507357923, + "learning_rate": 0.003, + "loss": 4.003, + "step": 24199 + }, + { + "epoch": 0.242, + "grad_norm": 0.8699063588120436, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 24200 + }, + { + "epoch": 0.24201, + "grad_norm": 0.7222858715518576, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 24201 + }, + { + "epoch": 0.24202, + "grad_norm": 0.7495169279558597, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 24202 + }, + { + "epoch": 0.24203, + "grad_norm": 0.8261900023568763, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 24203 + }, + { + "epoch": 0.24204, + "grad_norm": 0.8470353807809953, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 24204 + }, + { + "epoch": 0.24205, + "grad_norm": 0.8198215485561167, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 24205 + }, + { + "epoch": 0.24206, + "grad_norm": 0.6621171102983028, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 24206 + }, + { + "epoch": 0.24207, + "grad_norm": 0.6271518168696866, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 24207 + }, + { + "epoch": 0.24208, + "grad_norm": 0.665639918846977, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 24208 + }, + { + "epoch": 0.24209, + "grad_norm": 0.6985548363585046, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 24209 + }, + { + "epoch": 0.2421, + "grad_norm": 0.653871513771924, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 24210 + }, + { + "epoch": 0.24211, + "grad_norm": 0.6098982522183068, + "learning_rate": 0.003, + "loss": 4.016, + "step": 24211 + }, + { + "epoch": 0.24212, + "grad_norm": 0.6129041149806522, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 24212 + }, + { + "epoch": 0.24213, + "grad_norm": 0.6589011501898161, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 24213 + }, + { + "epoch": 0.24214, + "grad_norm": 0.689363517319121, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 24214 + }, + { + "epoch": 0.24215, + "grad_norm": 0.72072482461247, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 24215 + }, + { + "epoch": 0.24216, + "grad_norm": 0.638037894484661, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 24216 + }, + { + "epoch": 0.24217, + "grad_norm": 0.6600408920404702, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 24217 + }, + { + "epoch": 0.24218, + "grad_norm": 0.7453191881689881, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 24218 + }, + { + "epoch": 0.24219, + "grad_norm": 0.8088810594649215, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 24219 + }, + { + "epoch": 0.2422, + "grad_norm": 0.8685594034163333, + "learning_rate": 0.003, + "loss": 4.003, + "step": 24220 + }, + { + "epoch": 0.24221, + "grad_norm": 0.9239803881445621, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 24221 + }, + { + "epoch": 0.24222, + "grad_norm": 0.9767637540525607, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 24222 + }, + { + "epoch": 0.24223, + "grad_norm": 0.9877198057860255, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 24223 + }, + { + "epoch": 0.24224, + "grad_norm": 1.026604460963359, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 24224 + }, + { + "epoch": 0.24225, + "grad_norm": 0.9950185366863094, + "learning_rate": 0.003, + "loss": 4.011, + "step": 24225 + }, + { + "epoch": 0.24226, + "grad_norm": 0.7228661499113214, + "learning_rate": 0.003, + "loss": 3.998, + "step": 24226 + }, + { + "epoch": 0.24227, + "grad_norm": 0.601028321479248, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 24227 + }, + { + "epoch": 0.24228, + "grad_norm": 0.6805490293070807, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24228 + }, + { + "epoch": 0.24229, + "grad_norm": 0.6934499688183656, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 24229 + }, + { + "epoch": 0.2423, + "grad_norm": 0.70086636771627, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 24230 + }, + { + "epoch": 0.24231, + "grad_norm": 0.5055901769262802, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 24231 + }, + { + "epoch": 0.24232, + "grad_norm": 0.5589755355042156, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 24232 + }, + { + "epoch": 0.24233, + "grad_norm": 0.5005716301252556, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 24233 + }, + { + "epoch": 0.24234, + "grad_norm": 0.5428795378959602, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 24234 + }, + { + "epoch": 0.24235, + "grad_norm": 0.5273603985962717, + "learning_rate": 0.003, + "loss": 3.95, + "step": 24235 + }, + { + "epoch": 0.24236, + "grad_norm": 0.5736996320338713, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 24236 + }, + { + "epoch": 0.24237, + "grad_norm": 0.6896767964288144, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 24237 + }, + { + "epoch": 0.24238, + "grad_norm": 0.84566776162022, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 24238 + }, + { + "epoch": 0.24239, + "grad_norm": 1.0773597546870528, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 24239 + }, + { + "epoch": 0.2424, + "grad_norm": 1.0017626241070183, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 24240 + }, + { + "epoch": 0.24241, + "grad_norm": 0.8469806785079741, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 24241 + }, + { + "epoch": 0.24242, + "grad_norm": 0.8660469725710617, + "learning_rate": 0.003, + "loss": 4.0, + "step": 24242 + }, + { + "epoch": 0.24243, + "grad_norm": 0.903056883843222, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 24243 + }, + { + "epoch": 0.24244, + "grad_norm": 0.8279417955367794, + "learning_rate": 0.003, + "loss": 3.986, + "step": 24244 + }, + { + "epoch": 0.24245, + "grad_norm": 0.6954893698142578, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 24245 + }, + { + "epoch": 0.24246, + "grad_norm": 0.6436296167440901, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 24246 + }, + { + "epoch": 0.24247, + "grad_norm": 0.6646245176026871, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 24247 + }, + { + "epoch": 0.24248, + "grad_norm": 0.6477045515741046, + "learning_rate": 0.003, + "loss": 4.018, + "step": 24248 + }, + { + "epoch": 0.24249, + "grad_norm": 0.7216195746229606, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 24249 + }, + { + "epoch": 0.2425, + "grad_norm": 0.8780590465066853, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 24250 + }, + { + "epoch": 0.24251, + "grad_norm": 0.9080682916575662, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 24251 + }, + { + "epoch": 0.24252, + "grad_norm": 0.8924579485225973, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 24252 + }, + { + "epoch": 0.24253, + "grad_norm": 0.8473206364765133, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 24253 + }, + { + "epoch": 0.24254, + "grad_norm": 0.903952480771942, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 24254 + }, + { + "epoch": 0.24255, + "grad_norm": 0.832018057336135, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 24255 + }, + { + "epoch": 0.24256, + "grad_norm": 0.8235165596249794, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 24256 + }, + { + "epoch": 0.24257, + "grad_norm": 0.8506743571088992, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 24257 + }, + { + "epoch": 0.24258, + "grad_norm": 1.0319782344029653, + "learning_rate": 0.003, + "loss": 4.002, + "step": 24258 + }, + { + "epoch": 0.24259, + "grad_norm": 0.9265409062089379, + "learning_rate": 0.003, + "loss": 3.966, + "step": 24259 + }, + { + "epoch": 0.2426, + "grad_norm": 0.7744728664210316, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 24260 + }, + { + "epoch": 0.24261, + "grad_norm": 0.7753439668065004, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 24261 + }, + { + "epoch": 0.24262, + "grad_norm": 0.80730098617791, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 24262 + }, + { + "epoch": 0.24263, + "grad_norm": 0.6791339441604458, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 24263 + }, + { + "epoch": 0.24264, + "grad_norm": 0.6520950085302307, + "learning_rate": 0.003, + "loss": 4.0412, + "step": 24264 + }, + { + "epoch": 0.24265, + "grad_norm": 0.6917948377352907, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 24265 + }, + { + "epoch": 0.24266, + "grad_norm": 0.7836506340771867, + "learning_rate": 0.003, + "loss": 3.984, + "step": 24266 + }, + { + "epoch": 0.24267, + "grad_norm": 0.8664632733802304, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 24267 + }, + { + "epoch": 0.24268, + "grad_norm": 0.8400642057450893, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 24268 + }, + { + "epoch": 0.24269, + "grad_norm": 0.7925017680126403, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 24269 + }, + { + "epoch": 0.2427, + "grad_norm": 0.8464847097449597, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 24270 + }, + { + "epoch": 0.24271, + "grad_norm": 0.7362967657620502, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 24271 + }, + { + "epoch": 0.24272, + "grad_norm": 0.6505792647644958, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 24272 + }, + { + "epoch": 0.24273, + "grad_norm": 0.576615179102861, + "learning_rate": 0.003, + "loss": 3.985, + "step": 24273 + }, + { + "epoch": 0.24274, + "grad_norm": 0.6227724522333713, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 24274 + }, + { + "epoch": 0.24275, + "grad_norm": 0.6206113737051029, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 24275 + }, + { + "epoch": 0.24276, + "grad_norm": 0.5868503864684009, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 24276 + }, + { + "epoch": 0.24277, + "grad_norm": 0.7123086487806979, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 24277 + }, + { + "epoch": 0.24278, + "grad_norm": 0.7651734875216618, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 24278 + }, + { + "epoch": 0.24279, + "grad_norm": 0.7782693327345355, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 24279 + }, + { + "epoch": 0.2428, + "grad_norm": 0.7573200151366131, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 24280 + }, + { + "epoch": 0.24281, + "grad_norm": 0.7532633422842524, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 24281 + }, + { + "epoch": 0.24282, + "grad_norm": 0.7316308246648769, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 24282 + }, + { + "epoch": 0.24283, + "grad_norm": 0.6816279596862532, + "learning_rate": 0.003, + "loss": 3.988, + "step": 24283 + }, + { + "epoch": 0.24284, + "grad_norm": 0.6521598831091914, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 24284 + }, + { + "epoch": 0.24285, + "grad_norm": 0.6785943183801664, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 24285 + }, + { + "epoch": 0.24286, + "grad_norm": 0.6912864356630857, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 24286 + }, + { + "epoch": 0.24287, + "grad_norm": 0.6568706897812329, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 24287 + }, + { + "epoch": 0.24288, + "grad_norm": 0.5933550390667717, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 24288 + }, + { + "epoch": 0.24289, + "grad_norm": 0.6135529016064254, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 24289 + }, + { + "epoch": 0.2429, + "grad_norm": 0.5774254620862578, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 24290 + }, + { + "epoch": 0.24291, + "grad_norm": 0.6652270696148462, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 24291 + }, + { + "epoch": 0.24292, + "grad_norm": 0.8106993426171091, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 24292 + }, + { + "epoch": 0.24293, + "grad_norm": 0.9531821767809145, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 24293 + }, + { + "epoch": 0.24294, + "grad_norm": 0.9426551564027627, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 24294 + }, + { + "epoch": 0.24295, + "grad_norm": 0.9220042766048802, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 24295 + }, + { + "epoch": 0.24296, + "grad_norm": 0.9312131168698242, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 24296 + }, + { + "epoch": 0.24297, + "grad_norm": 0.9209440513068637, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 24297 + }, + { + "epoch": 0.24298, + "grad_norm": 0.8986455248653099, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 24298 + }, + { + "epoch": 0.24299, + "grad_norm": 0.9291363166337879, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 24299 + }, + { + "epoch": 0.243, + "grad_norm": 1.06952829899101, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 24300 + }, + { + "epoch": 0.24301, + "grad_norm": 1.158050447782962, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 24301 + }, + { + "epoch": 0.24302, + "grad_norm": 0.8548862728303486, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 24302 + }, + { + "epoch": 0.24303, + "grad_norm": 0.7086360849614879, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 24303 + }, + { + "epoch": 0.24304, + "grad_norm": 0.6580672715696299, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 24304 + }, + { + "epoch": 0.24305, + "grad_norm": 0.6808582380042348, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 24305 + }, + { + "epoch": 0.24306, + "grad_norm": 0.6256151215260181, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 24306 + }, + { + "epoch": 0.24307, + "grad_norm": 0.6681458327780908, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 24307 + }, + { + "epoch": 0.24308, + "grad_norm": 0.6959173360502937, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 24308 + }, + { + "epoch": 0.24309, + "grad_norm": 0.6269044529339461, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 24309 + }, + { + "epoch": 0.2431, + "grad_norm": 0.689614117963366, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 24310 + }, + { + "epoch": 0.24311, + "grad_norm": 0.7714264884194585, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 24311 + }, + { + "epoch": 0.24312, + "grad_norm": 0.7611718102826717, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 24312 + }, + { + "epoch": 0.24313, + "grad_norm": 0.8272953460957265, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 24313 + }, + { + "epoch": 0.24314, + "grad_norm": 0.8859422825195057, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 24314 + }, + { + "epoch": 0.24315, + "grad_norm": 0.8702492370228264, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 24315 + }, + { + "epoch": 0.24316, + "grad_norm": 0.7547713399897908, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 24316 + }, + { + "epoch": 0.24317, + "grad_norm": 0.8897078522622529, + "learning_rate": 0.003, + "loss": 4.0487, + "step": 24317 + }, + { + "epoch": 0.24318, + "grad_norm": 0.812609968435527, + "learning_rate": 0.003, + "loss": 4.003, + "step": 24318 + }, + { + "epoch": 0.24319, + "grad_norm": 0.8459268723788286, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 24319 + }, + { + "epoch": 0.2432, + "grad_norm": 0.7577513779024289, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 24320 + }, + { + "epoch": 0.24321, + "grad_norm": 0.8263213202078001, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 24321 + }, + { + "epoch": 0.24322, + "grad_norm": 0.8213197453030757, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 24322 + }, + { + "epoch": 0.24323, + "grad_norm": 0.8364893258632133, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 24323 + }, + { + "epoch": 0.24324, + "grad_norm": 0.8559087830856948, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 24324 + }, + { + "epoch": 0.24325, + "grad_norm": 0.7972278026231231, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 24325 + }, + { + "epoch": 0.24326, + "grad_norm": 0.6951974377811738, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 24326 + }, + { + "epoch": 0.24327, + "grad_norm": 0.6212171233575617, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 24327 + }, + { + "epoch": 0.24328, + "grad_norm": 0.6523632835863575, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 24328 + }, + { + "epoch": 0.24329, + "grad_norm": 0.5143863049768107, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 24329 + }, + { + "epoch": 0.2433, + "grad_norm": 0.47799592153798964, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 24330 + }, + { + "epoch": 0.24331, + "grad_norm": 0.42789268537544195, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 24331 + }, + { + "epoch": 0.24332, + "grad_norm": 0.48854052055532443, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 24332 + }, + { + "epoch": 0.24333, + "grad_norm": 0.499781881138666, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 24333 + }, + { + "epoch": 0.24334, + "grad_norm": 0.46975126189262767, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 24334 + }, + { + "epoch": 0.24335, + "grad_norm": 0.43970162768095405, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 24335 + }, + { + "epoch": 0.24336, + "grad_norm": 0.46247643965553065, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 24336 + }, + { + "epoch": 0.24337, + "grad_norm": 0.545666828280887, + "learning_rate": 0.003, + "loss": 3.975, + "step": 24337 + }, + { + "epoch": 0.24338, + "grad_norm": 0.7394710219817996, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 24338 + }, + { + "epoch": 0.24339, + "grad_norm": 1.0375332140381597, + "learning_rate": 0.003, + "loss": 4.006, + "step": 24339 + }, + { + "epoch": 0.2434, + "grad_norm": 1.1747826261004017, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 24340 + }, + { + "epoch": 0.24341, + "grad_norm": 0.8228978411585787, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 24341 + }, + { + "epoch": 0.24342, + "grad_norm": 0.7735042070382286, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 24342 + }, + { + "epoch": 0.24343, + "grad_norm": 0.9916023762637394, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 24343 + }, + { + "epoch": 0.24344, + "grad_norm": 1.1671516460461537, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 24344 + }, + { + "epoch": 0.24345, + "grad_norm": 0.7608648034562627, + "learning_rate": 0.003, + "loss": 3.998, + "step": 24345 + }, + { + "epoch": 0.24346, + "grad_norm": 0.7596119645134649, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 24346 + }, + { + "epoch": 0.24347, + "grad_norm": 0.7896534535268632, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 24347 + }, + { + "epoch": 0.24348, + "grad_norm": 0.746967723565307, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 24348 + }, + { + "epoch": 0.24349, + "grad_norm": 0.7087455326333781, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 24349 + }, + { + "epoch": 0.2435, + "grad_norm": 0.7147862925429367, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 24350 + }, + { + "epoch": 0.24351, + "grad_norm": 0.6776549214117864, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 24351 + }, + { + "epoch": 0.24352, + "grad_norm": 0.5727062649898955, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 24352 + }, + { + "epoch": 0.24353, + "grad_norm": 0.443372875162221, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 24353 + }, + { + "epoch": 0.24354, + "grad_norm": 0.5376932621375903, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 24354 + }, + { + "epoch": 0.24355, + "grad_norm": 0.6354844781357415, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 24355 + }, + { + "epoch": 0.24356, + "grad_norm": 0.6931207936600442, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 24356 + }, + { + "epoch": 0.24357, + "grad_norm": 0.7126294711643809, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 24357 + }, + { + "epoch": 0.24358, + "grad_norm": 0.6160140444654476, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 24358 + }, + { + "epoch": 0.24359, + "grad_norm": 0.570749918059756, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 24359 + }, + { + "epoch": 0.2436, + "grad_norm": 0.6804978493656427, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 24360 + }, + { + "epoch": 0.24361, + "grad_norm": 0.6023505931123233, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 24361 + }, + { + "epoch": 0.24362, + "grad_norm": 0.6355892014589234, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 24362 + }, + { + "epoch": 0.24363, + "grad_norm": 0.7347244283581685, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 24363 + }, + { + "epoch": 0.24364, + "grad_norm": 0.8423399153353219, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 24364 + }, + { + "epoch": 0.24365, + "grad_norm": 0.9581713207265998, + "learning_rate": 0.003, + "loss": 3.989, + "step": 24365 + }, + { + "epoch": 0.24366, + "grad_norm": 0.9232249926392847, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 24366 + }, + { + "epoch": 0.24367, + "grad_norm": 0.8558904146834831, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 24367 + }, + { + "epoch": 0.24368, + "grad_norm": 0.8914042537415134, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 24368 + }, + { + "epoch": 0.24369, + "grad_norm": 0.9163019053465412, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 24369 + }, + { + "epoch": 0.2437, + "grad_norm": 1.0560487320938448, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 24370 + }, + { + "epoch": 0.24371, + "grad_norm": 1.0467952219438486, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 24371 + }, + { + "epoch": 0.24372, + "grad_norm": 0.9025721533140636, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 24372 + }, + { + "epoch": 0.24373, + "grad_norm": 0.7151825461144947, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 24373 + }, + { + "epoch": 0.24374, + "grad_norm": 0.8205837329381268, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 24374 + }, + { + "epoch": 0.24375, + "grad_norm": 1.0411299999877308, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 24375 + }, + { + "epoch": 0.24376, + "grad_norm": 0.9711144789996176, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 24376 + }, + { + "epoch": 0.24377, + "grad_norm": 0.9621415342804289, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 24377 + }, + { + "epoch": 0.24378, + "grad_norm": 0.9053615165800916, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 24378 + }, + { + "epoch": 0.24379, + "grad_norm": 0.8140825995996244, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 24379 + }, + { + "epoch": 0.2438, + "grad_norm": 0.8577821943554325, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 24380 + }, + { + "epoch": 0.24381, + "grad_norm": 0.8646379730084739, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 24381 + }, + { + "epoch": 0.24382, + "grad_norm": 0.9185713034877642, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 24382 + }, + { + "epoch": 0.24383, + "grad_norm": 1.0115824935003674, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 24383 + }, + { + "epoch": 0.24384, + "grad_norm": 1.0201715912277156, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 24384 + }, + { + "epoch": 0.24385, + "grad_norm": 1.0559280960481618, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 24385 + }, + { + "epoch": 0.24386, + "grad_norm": 0.9513156805264469, + "learning_rate": 0.003, + "loss": 4.038, + "step": 24386 + }, + { + "epoch": 0.24387, + "grad_norm": 0.8828727027609995, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 24387 + }, + { + "epoch": 0.24388, + "grad_norm": 0.9800373493314609, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 24388 + }, + { + "epoch": 0.24389, + "grad_norm": 1.1200630147763604, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 24389 + }, + { + "epoch": 0.2439, + "grad_norm": 0.8859465614070522, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 24390 + }, + { + "epoch": 0.24391, + "grad_norm": 0.8452890837566147, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 24391 + }, + { + "epoch": 0.24392, + "grad_norm": 0.7585174806525494, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 24392 + }, + { + "epoch": 0.24393, + "grad_norm": 0.7465034343020537, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 24393 + }, + { + "epoch": 0.24394, + "grad_norm": 0.6805442461213327, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 24394 + }, + { + "epoch": 0.24395, + "grad_norm": 0.6908147344049159, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 24395 + }, + { + "epoch": 0.24396, + "grad_norm": 0.6866619736048345, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 24396 + }, + { + "epoch": 0.24397, + "grad_norm": 0.6201056162357937, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 24397 + }, + { + "epoch": 0.24398, + "grad_norm": 0.5781594795498025, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 24398 + }, + { + "epoch": 0.24399, + "grad_norm": 0.5576522258701485, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 24399 + }, + { + "epoch": 0.244, + "grad_norm": 0.531244120990939, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 24400 + }, + { + "epoch": 0.24401, + "grad_norm": 0.5347302558518304, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 24401 + }, + { + "epoch": 0.24402, + "grad_norm": 0.48592755022524015, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 24402 + }, + { + "epoch": 0.24403, + "grad_norm": 0.5038281419071442, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 24403 + }, + { + "epoch": 0.24404, + "grad_norm": 0.5297191028368554, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 24404 + }, + { + "epoch": 0.24405, + "grad_norm": 0.4910812909687015, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 24405 + }, + { + "epoch": 0.24406, + "grad_norm": 0.5032662698362157, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 24406 + }, + { + "epoch": 0.24407, + "grad_norm": 0.6487659748069835, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 24407 + }, + { + "epoch": 0.24408, + "grad_norm": 0.821649122556901, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 24408 + }, + { + "epoch": 0.24409, + "grad_norm": 0.8290923099774942, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 24409 + }, + { + "epoch": 0.2441, + "grad_norm": 0.7615123922753188, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 24410 + }, + { + "epoch": 0.24411, + "grad_norm": 0.6430988638749607, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 24411 + }, + { + "epoch": 0.24412, + "grad_norm": 0.5847296405740432, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 24412 + }, + { + "epoch": 0.24413, + "grad_norm": 0.5813166269793457, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 24413 + }, + { + "epoch": 0.24414, + "grad_norm": 0.5633979196708282, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 24414 + }, + { + "epoch": 0.24415, + "grad_norm": 0.5557277823801746, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 24415 + }, + { + "epoch": 0.24416, + "grad_norm": 0.5757485856311152, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 24416 + }, + { + "epoch": 0.24417, + "grad_norm": 0.6366188024456962, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 24417 + }, + { + "epoch": 0.24418, + "grad_norm": 0.7147359408671697, + "learning_rate": 0.003, + "loss": 3.977, + "step": 24418 + }, + { + "epoch": 0.24419, + "grad_norm": 0.7104083594963204, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 24419 + }, + { + "epoch": 0.2442, + "grad_norm": 0.6147685858767533, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 24420 + }, + { + "epoch": 0.24421, + "grad_norm": 0.6205495493491064, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 24421 + }, + { + "epoch": 0.24422, + "grad_norm": 0.6193090541440565, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 24422 + }, + { + "epoch": 0.24423, + "grad_norm": 0.6626638847806997, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 24423 + }, + { + "epoch": 0.24424, + "grad_norm": 0.6107804002688108, + "learning_rate": 0.003, + "loss": 4.002, + "step": 24424 + }, + { + "epoch": 0.24425, + "grad_norm": 0.7237055845526715, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 24425 + }, + { + "epoch": 0.24426, + "grad_norm": 0.7553286091431437, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 24426 + }, + { + "epoch": 0.24427, + "grad_norm": 0.7797070210886168, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24427 + }, + { + "epoch": 0.24428, + "grad_norm": 0.8212651778892751, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 24428 + }, + { + "epoch": 0.24429, + "grad_norm": 0.8366697585420835, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 24429 + }, + { + "epoch": 0.2443, + "grad_norm": 0.8550110583603721, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 24430 + }, + { + "epoch": 0.24431, + "grad_norm": 0.8728188226975748, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 24431 + }, + { + "epoch": 0.24432, + "grad_norm": 0.8923291889518552, + "learning_rate": 0.003, + "loss": 4.005, + "step": 24432 + }, + { + "epoch": 0.24433, + "grad_norm": 0.8214604458731006, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 24433 + }, + { + "epoch": 0.24434, + "grad_norm": 0.9518241679625419, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24434 + }, + { + "epoch": 0.24435, + "grad_norm": 1.0280967719057865, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 24435 + }, + { + "epoch": 0.24436, + "grad_norm": 1.0319883306969697, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 24436 + }, + { + "epoch": 0.24437, + "grad_norm": 1.0376661901856412, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 24437 + }, + { + "epoch": 0.24438, + "grad_norm": 1.021861518725118, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 24438 + }, + { + "epoch": 0.24439, + "grad_norm": 0.9650405001784792, + "learning_rate": 0.003, + "loss": 4.014, + "step": 24439 + }, + { + "epoch": 0.2444, + "grad_norm": 1.1332301830525597, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 24440 + }, + { + "epoch": 0.24441, + "grad_norm": 0.9443907508687323, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 24441 + }, + { + "epoch": 0.24442, + "grad_norm": 0.8646036005399882, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24442 + }, + { + "epoch": 0.24443, + "grad_norm": 0.9963974816648308, + "learning_rate": 0.003, + "loss": 4.025, + "step": 24443 + }, + { + "epoch": 0.24444, + "grad_norm": 1.078734351023378, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 24444 + }, + { + "epoch": 0.24445, + "grad_norm": 0.9573235665922197, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 24445 + }, + { + "epoch": 0.24446, + "grad_norm": 0.9595341994989401, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 24446 + }, + { + "epoch": 0.24447, + "grad_norm": 0.945932421147736, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 24447 + }, + { + "epoch": 0.24448, + "grad_norm": 0.9345996799113993, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 24448 + }, + { + "epoch": 0.24449, + "grad_norm": 0.8236354697207443, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 24449 + }, + { + "epoch": 0.2445, + "grad_norm": 0.8322112283463148, + "learning_rate": 0.003, + "loss": 3.988, + "step": 24450 + }, + { + "epoch": 0.24451, + "grad_norm": 0.8432235258447361, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 24451 + }, + { + "epoch": 0.24452, + "grad_norm": 0.9027771624262061, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 24452 + }, + { + "epoch": 0.24453, + "grad_norm": 0.8676955958797145, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 24453 + }, + { + "epoch": 0.24454, + "grad_norm": 0.6764806780423838, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 24454 + }, + { + "epoch": 0.24455, + "grad_norm": 0.6040511180422973, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 24455 + }, + { + "epoch": 0.24456, + "grad_norm": 0.5313409015922032, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 24456 + }, + { + "epoch": 0.24457, + "grad_norm": 0.554803227872254, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 24457 + }, + { + "epoch": 0.24458, + "grad_norm": 0.6086178578305156, + "learning_rate": 0.003, + "loss": 4.023, + "step": 24458 + }, + { + "epoch": 0.24459, + "grad_norm": 0.647028093674204, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 24459 + }, + { + "epoch": 0.2446, + "grad_norm": 0.5464786506914572, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 24460 + }, + { + "epoch": 0.24461, + "grad_norm": 0.515785538092233, + "learning_rate": 0.003, + "loss": 4.001, + "step": 24461 + }, + { + "epoch": 0.24462, + "grad_norm": 0.5118127094126541, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 24462 + }, + { + "epoch": 0.24463, + "grad_norm": 0.5287032531212329, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 24463 + }, + { + "epoch": 0.24464, + "grad_norm": 0.49046634366721625, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 24464 + }, + { + "epoch": 0.24465, + "grad_norm": 0.49272862199551865, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 24465 + }, + { + "epoch": 0.24466, + "grad_norm": 0.5079835028307796, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 24466 + }, + { + "epoch": 0.24467, + "grad_norm": 0.5681794129145556, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 24467 + }, + { + "epoch": 0.24468, + "grad_norm": 0.6840845567247262, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 24468 + }, + { + "epoch": 0.24469, + "grad_norm": 0.7763946459789619, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 24469 + }, + { + "epoch": 0.2447, + "grad_norm": 0.836967382859626, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 24470 + }, + { + "epoch": 0.24471, + "grad_norm": 0.7686189581083027, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 24471 + }, + { + "epoch": 0.24472, + "grad_norm": 0.6927165538088647, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 24472 + }, + { + "epoch": 0.24473, + "grad_norm": 0.6381047634539214, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 24473 + }, + { + "epoch": 0.24474, + "grad_norm": 0.5614773640576607, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 24474 + }, + { + "epoch": 0.24475, + "grad_norm": 0.5708178168387483, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 24475 + }, + { + "epoch": 0.24476, + "grad_norm": 0.5817733717385605, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 24476 + }, + { + "epoch": 0.24477, + "grad_norm": 0.5174637898817614, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 24477 + }, + { + "epoch": 0.24478, + "grad_norm": 0.5583599507856941, + "learning_rate": 0.003, + "loss": 3.992, + "step": 24478 + }, + { + "epoch": 0.24479, + "grad_norm": 0.6100439753468572, + "learning_rate": 0.003, + "loss": 3.987, + "step": 24479 + }, + { + "epoch": 0.2448, + "grad_norm": 0.7058453839653569, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 24480 + }, + { + "epoch": 0.24481, + "grad_norm": 0.8479288245144541, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 24481 + }, + { + "epoch": 0.24482, + "grad_norm": 1.0331516868547144, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 24482 + }, + { + "epoch": 0.24483, + "grad_norm": 1.0855811250507836, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 24483 + }, + { + "epoch": 0.24484, + "grad_norm": 0.8097525995850442, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 24484 + }, + { + "epoch": 0.24485, + "grad_norm": 0.7244482293862292, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 24485 + }, + { + "epoch": 0.24486, + "grad_norm": 0.6816976598498639, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 24486 + }, + { + "epoch": 0.24487, + "grad_norm": 0.6792238924200007, + "learning_rate": 0.003, + "loss": 3.982, + "step": 24487 + }, + { + "epoch": 0.24488, + "grad_norm": 0.6617822002781931, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 24488 + }, + { + "epoch": 0.24489, + "grad_norm": 0.6103823498128264, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 24489 + }, + { + "epoch": 0.2449, + "grad_norm": 0.5876592341868683, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 24490 + }, + { + "epoch": 0.24491, + "grad_norm": 0.6178788917578387, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 24491 + }, + { + "epoch": 0.24492, + "grad_norm": 0.6837919265883603, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 24492 + }, + { + "epoch": 0.24493, + "grad_norm": 0.8168851878379425, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 24493 + }, + { + "epoch": 0.24494, + "grad_norm": 1.049415764683888, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 24494 + }, + { + "epoch": 0.24495, + "grad_norm": 1.2195765160263794, + "learning_rate": 0.003, + "loss": 4.019, + "step": 24495 + }, + { + "epoch": 0.24496, + "grad_norm": 0.878422063230412, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 24496 + }, + { + "epoch": 0.24497, + "grad_norm": 0.8726074260674076, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 24497 + }, + { + "epoch": 0.24498, + "grad_norm": 0.8724431193844969, + "learning_rate": 0.003, + "loss": 4.015, + "step": 24498 + }, + { + "epoch": 0.24499, + "grad_norm": 0.7785185912226471, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 24499 + }, + { + "epoch": 0.245, + "grad_norm": 0.7635989214702766, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 24500 + }, + { + "epoch": 0.24501, + "grad_norm": 0.7848562711230735, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 24501 + }, + { + "epoch": 0.24502, + "grad_norm": 0.8904051560872784, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 24502 + }, + { + "epoch": 0.24503, + "grad_norm": 0.9718327807717635, + "learning_rate": 0.003, + "loss": 3.99, + "step": 24503 + }, + { + "epoch": 0.24504, + "grad_norm": 1.0435647923140094, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 24504 + }, + { + "epoch": 0.24505, + "grad_norm": 0.8785997918607819, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 24505 + }, + { + "epoch": 0.24506, + "grad_norm": 0.8231473285033983, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 24506 + }, + { + "epoch": 0.24507, + "grad_norm": 0.7172893010621129, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 24507 + }, + { + "epoch": 0.24508, + "grad_norm": 0.6467456811541077, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 24508 + }, + { + "epoch": 0.24509, + "grad_norm": 0.6018398086309517, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 24509 + }, + { + "epoch": 0.2451, + "grad_norm": 0.6127601155056417, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 24510 + }, + { + "epoch": 0.24511, + "grad_norm": 0.597958392207749, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 24511 + }, + { + "epoch": 0.24512, + "grad_norm": 0.5857353735669945, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 24512 + }, + { + "epoch": 0.24513, + "grad_norm": 0.5317476741159245, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 24513 + }, + { + "epoch": 0.24514, + "grad_norm": 0.506268105946075, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 24514 + }, + { + "epoch": 0.24515, + "grad_norm": 0.4934701342234527, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 24515 + }, + { + "epoch": 0.24516, + "grad_norm": 0.4853756645459337, + "learning_rate": 0.003, + "loss": 3.981, + "step": 24516 + }, + { + "epoch": 0.24517, + "grad_norm": 0.47063448549455555, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 24517 + }, + { + "epoch": 0.24518, + "grad_norm": 0.5246243256262068, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 24518 + }, + { + "epoch": 0.24519, + "grad_norm": 0.7509057893269778, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 24519 + }, + { + "epoch": 0.2452, + "grad_norm": 1.0178102867893026, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 24520 + }, + { + "epoch": 0.24521, + "grad_norm": 1.0535610488686562, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 24521 + }, + { + "epoch": 0.24522, + "grad_norm": 0.8061209348031608, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 24522 + }, + { + "epoch": 0.24523, + "grad_norm": 0.8604594475703321, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 24523 + }, + { + "epoch": 0.24524, + "grad_norm": 0.9105546891267252, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 24524 + }, + { + "epoch": 0.24525, + "grad_norm": 1.0514768130475542, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 24525 + }, + { + "epoch": 0.24526, + "grad_norm": 0.9485173732750339, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 24526 + }, + { + "epoch": 0.24527, + "grad_norm": 0.8715849550554411, + "learning_rate": 0.003, + "loss": 4.036, + "step": 24527 + }, + { + "epoch": 0.24528, + "grad_norm": 0.8321339414969523, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 24528 + }, + { + "epoch": 0.24529, + "grad_norm": 0.8524909695023495, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 24529 + }, + { + "epoch": 0.2453, + "grad_norm": 0.7888929665341401, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 24530 + }, + { + "epoch": 0.24531, + "grad_norm": 0.8076695126505452, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 24531 + }, + { + "epoch": 0.24532, + "grad_norm": 0.9217124876636709, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 24532 + }, + { + "epoch": 0.24533, + "grad_norm": 0.9015244083712665, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 24533 + }, + { + "epoch": 0.24534, + "grad_norm": 1.0187835820532178, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 24534 + }, + { + "epoch": 0.24535, + "grad_norm": 1.1023190536466345, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 24535 + }, + { + "epoch": 0.24536, + "grad_norm": 0.9307175809622132, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 24536 + }, + { + "epoch": 0.24537, + "grad_norm": 0.8219316381888304, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 24537 + }, + { + "epoch": 0.24538, + "grad_norm": 0.6582764959682806, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 24538 + }, + { + "epoch": 0.24539, + "grad_norm": 0.6918033358576219, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 24539 + }, + { + "epoch": 0.2454, + "grad_norm": 0.7565071238853364, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 24540 + }, + { + "epoch": 0.24541, + "grad_norm": 0.7942227202738503, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 24541 + }, + { + "epoch": 0.24542, + "grad_norm": 0.8050223865769077, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 24542 + }, + { + "epoch": 0.24543, + "grad_norm": 0.7502423562875331, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 24543 + }, + { + "epoch": 0.24544, + "grad_norm": 0.6488224455487933, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 24544 + }, + { + "epoch": 0.24545, + "grad_norm": 0.6398916239466742, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 24545 + }, + { + "epoch": 0.24546, + "grad_norm": 0.6781384376491241, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 24546 + }, + { + "epoch": 0.24547, + "grad_norm": 0.6710156759933499, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 24547 + }, + { + "epoch": 0.24548, + "grad_norm": 0.720215039178746, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24548 + }, + { + "epoch": 0.24549, + "grad_norm": 0.7889127083882975, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 24549 + }, + { + "epoch": 0.2455, + "grad_norm": 0.8273379618543665, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 24550 + }, + { + "epoch": 0.24551, + "grad_norm": 0.7754638002278464, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 24551 + }, + { + "epoch": 0.24552, + "grad_norm": 0.6506034833825083, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 24552 + }, + { + "epoch": 0.24553, + "grad_norm": 0.6339517768406032, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 24553 + }, + { + "epoch": 0.24554, + "grad_norm": 0.6087151478734606, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 24554 + }, + { + "epoch": 0.24555, + "grad_norm": 0.653285597225138, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 24555 + }, + { + "epoch": 0.24556, + "grad_norm": 0.6546681739522685, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 24556 + }, + { + "epoch": 0.24557, + "grad_norm": 0.6520206935632917, + "learning_rate": 0.003, + "loss": 3.96, + "step": 24557 + }, + { + "epoch": 0.24558, + "grad_norm": 0.7014861641397595, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 24558 + }, + { + "epoch": 0.24559, + "grad_norm": 0.6882872493859875, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 24559 + }, + { + "epoch": 0.2456, + "grad_norm": 0.5901246497753453, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 24560 + }, + { + "epoch": 0.24561, + "grad_norm": 0.5534021523034115, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 24561 + }, + { + "epoch": 0.24562, + "grad_norm": 0.528098639816165, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 24562 + }, + { + "epoch": 0.24563, + "grad_norm": 0.5466447647405612, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 24563 + }, + { + "epoch": 0.24564, + "grad_norm": 0.5360183806473179, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 24564 + }, + { + "epoch": 0.24565, + "grad_norm": 0.5568367844072202, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 24565 + }, + { + "epoch": 0.24566, + "grad_norm": 0.5201663730223077, + "learning_rate": 0.003, + "loss": 3.979, + "step": 24566 + }, + { + "epoch": 0.24567, + "grad_norm": 0.51985695463206, + "learning_rate": 0.003, + "loss": 3.993, + "step": 24567 + }, + { + "epoch": 0.24568, + "grad_norm": 0.5093778986559206, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 24568 + }, + { + "epoch": 0.24569, + "grad_norm": 0.5607594905731004, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 24569 + }, + { + "epoch": 0.2457, + "grad_norm": 0.5384296112654382, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 24570 + }, + { + "epoch": 0.24571, + "grad_norm": 0.5982654066801866, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 24571 + }, + { + "epoch": 0.24572, + "grad_norm": 0.7641304328709486, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 24572 + }, + { + "epoch": 0.24573, + "grad_norm": 1.0279736976090683, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 24573 + }, + { + "epoch": 0.24574, + "grad_norm": 1.171710935165813, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 24574 + }, + { + "epoch": 0.24575, + "grad_norm": 0.899200245909421, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 24575 + }, + { + "epoch": 0.24576, + "grad_norm": 0.8702782141984101, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 24576 + }, + { + "epoch": 0.24577, + "grad_norm": 0.8923118733205175, + "learning_rate": 0.003, + "loss": 4.014, + "step": 24577 + }, + { + "epoch": 0.24578, + "grad_norm": 0.9619156806028313, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 24578 + }, + { + "epoch": 0.24579, + "grad_norm": 0.9276282329032539, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24579 + }, + { + "epoch": 0.2458, + "grad_norm": 0.820043003820165, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 24580 + }, + { + "epoch": 0.24581, + "grad_norm": 0.779935538513049, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 24581 + }, + { + "epoch": 0.24582, + "grad_norm": 0.8920604737794047, + "learning_rate": 0.003, + "loss": 4.0537, + "step": 24582 + }, + { + "epoch": 0.24583, + "grad_norm": 0.9539883875196687, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 24583 + }, + { + "epoch": 0.24584, + "grad_norm": 0.9575996502520642, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 24584 + }, + { + "epoch": 0.24585, + "grad_norm": 1.017880481276888, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 24585 + }, + { + "epoch": 0.24586, + "grad_norm": 0.8859956360059015, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 24586 + }, + { + "epoch": 0.24587, + "grad_norm": 0.8289240463200986, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 24587 + }, + { + "epoch": 0.24588, + "grad_norm": 0.836356355271981, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 24588 + }, + { + "epoch": 0.24589, + "grad_norm": 0.9501856352278514, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 24589 + }, + { + "epoch": 0.2459, + "grad_norm": 1.0538305875933236, + "learning_rate": 0.003, + "loss": 4.023, + "step": 24590 + }, + { + "epoch": 0.24591, + "grad_norm": 1.0371668824222684, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 24591 + }, + { + "epoch": 0.24592, + "grad_norm": 0.9153378063253566, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 24592 + }, + { + "epoch": 0.24593, + "grad_norm": 0.8317795717532149, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 24593 + }, + { + "epoch": 0.24594, + "grad_norm": 0.6560234129805467, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 24594 + }, + { + "epoch": 0.24595, + "grad_norm": 0.8021715492510437, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 24595 + }, + { + "epoch": 0.24596, + "grad_norm": 1.025909893719223, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 24596 + }, + { + "epoch": 0.24597, + "grad_norm": 0.936576904359579, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 24597 + }, + { + "epoch": 0.24598, + "grad_norm": 0.7481935870840966, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 24598 + }, + { + "epoch": 0.24599, + "grad_norm": 0.6343285706889956, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 24599 + }, + { + "epoch": 0.246, + "grad_norm": 0.6153007514949865, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 24600 + }, + { + "epoch": 0.24601, + "grad_norm": 0.5783240689109459, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 24601 + }, + { + "epoch": 0.24602, + "grad_norm": 0.5799527278608125, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 24602 + }, + { + "epoch": 0.24603, + "grad_norm": 0.6560628708450834, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 24603 + }, + { + "epoch": 0.24604, + "grad_norm": 0.6244299531744028, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 24604 + }, + { + "epoch": 0.24605, + "grad_norm": 0.5391354597461793, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 24605 + }, + { + "epoch": 0.24606, + "grad_norm": 0.5079241799630632, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 24606 + }, + { + "epoch": 0.24607, + "grad_norm": 0.5646184218343389, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 24607 + }, + { + "epoch": 0.24608, + "grad_norm": 0.5692536948211165, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 24608 + }, + { + "epoch": 0.24609, + "grad_norm": 0.55860082879006, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 24609 + }, + { + "epoch": 0.2461, + "grad_norm": 0.4934432365278093, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 24610 + }, + { + "epoch": 0.24611, + "grad_norm": 0.5246015494657927, + "learning_rate": 0.003, + "loss": 3.964, + "step": 24611 + }, + { + "epoch": 0.24612, + "grad_norm": 0.590539003623739, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 24612 + }, + { + "epoch": 0.24613, + "grad_norm": 0.6056765486900674, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 24613 + }, + { + "epoch": 0.24614, + "grad_norm": 0.756725896646354, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 24614 + }, + { + "epoch": 0.24615, + "grad_norm": 0.9648877640437337, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 24615 + }, + { + "epoch": 0.24616, + "grad_norm": 0.8897404578298371, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 24616 + }, + { + "epoch": 0.24617, + "grad_norm": 0.8297820545018033, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 24617 + }, + { + "epoch": 0.24618, + "grad_norm": 0.8265203765204494, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 24618 + }, + { + "epoch": 0.24619, + "grad_norm": 0.8042854556312783, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 24619 + }, + { + "epoch": 0.2462, + "grad_norm": 0.7487939140126807, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 24620 + }, + { + "epoch": 0.24621, + "grad_norm": 0.6589949118511342, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 24621 + }, + { + "epoch": 0.24622, + "grad_norm": 0.5439296275363137, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 24622 + }, + { + "epoch": 0.24623, + "grad_norm": 0.5842120493679843, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 24623 + }, + { + "epoch": 0.24624, + "grad_norm": 0.7739435888964256, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 24624 + }, + { + "epoch": 0.24625, + "grad_norm": 0.9749254111987201, + "learning_rate": 0.003, + "loss": 4.007, + "step": 24625 + }, + { + "epoch": 0.24626, + "grad_norm": 1.0406671064290185, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 24626 + }, + { + "epoch": 0.24627, + "grad_norm": 0.91763867225215, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 24627 + }, + { + "epoch": 0.24628, + "grad_norm": 0.7818378987139964, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 24628 + }, + { + "epoch": 0.24629, + "grad_norm": 0.6478441605475529, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 24629 + }, + { + "epoch": 0.2463, + "grad_norm": 0.5789673256126843, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 24630 + }, + { + "epoch": 0.24631, + "grad_norm": 0.5168295662161987, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 24631 + }, + { + "epoch": 0.24632, + "grad_norm": 0.5251238908799595, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 24632 + }, + { + "epoch": 0.24633, + "grad_norm": 0.4895563338647295, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 24633 + }, + { + "epoch": 0.24634, + "grad_norm": 0.4430556444568868, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 24634 + }, + { + "epoch": 0.24635, + "grad_norm": 0.48046351714071306, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 24635 + }, + { + "epoch": 0.24636, + "grad_norm": 0.5658548531359199, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 24636 + }, + { + "epoch": 0.24637, + "grad_norm": 0.7618707743964545, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 24637 + }, + { + "epoch": 0.24638, + "grad_norm": 0.9004142081846068, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 24638 + }, + { + "epoch": 0.24639, + "grad_norm": 0.8633473042408246, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 24639 + }, + { + "epoch": 0.2464, + "grad_norm": 0.8282510459599591, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 24640 + }, + { + "epoch": 0.24641, + "grad_norm": 0.9481752763418717, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 24641 + }, + { + "epoch": 0.24642, + "grad_norm": 0.9670806752731053, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 24642 + }, + { + "epoch": 0.24643, + "grad_norm": 1.0333222950609147, + "learning_rate": 0.003, + "loss": 4.0395, + "step": 24643 + }, + { + "epoch": 0.24644, + "grad_norm": 0.9613888903307465, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 24644 + }, + { + "epoch": 0.24645, + "grad_norm": 0.990287869859596, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 24645 + }, + { + "epoch": 0.24646, + "grad_norm": 0.9340758536261448, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 24646 + }, + { + "epoch": 0.24647, + "grad_norm": 0.8116688959258037, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 24647 + }, + { + "epoch": 0.24648, + "grad_norm": 0.8956396261063264, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 24648 + }, + { + "epoch": 0.24649, + "grad_norm": 0.7846579675974674, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24649 + }, + { + "epoch": 0.2465, + "grad_norm": 0.7391069584590197, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 24650 + }, + { + "epoch": 0.24651, + "grad_norm": 0.8386866243566639, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 24651 + }, + { + "epoch": 0.24652, + "grad_norm": 0.9246404890035875, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 24652 + }, + { + "epoch": 0.24653, + "grad_norm": 0.7809817754297365, + "learning_rate": 0.003, + "loss": 3.992, + "step": 24653 + }, + { + "epoch": 0.24654, + "grad_norm": 0.792027899825234, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 24654 + }, + { + "epoch": 0.24655, + "grad_norm": 0.7877848600970965, + "learning_rate": 0.003, + "loss": 4.0514, + "step": 24655 + }, + { + "epoch": 0.24656, + "grad_norm": 0.8761235080951728, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 24656 + }, + { + "epoch": 0.24657, + "grad_norm": 0.8910156271384846, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 24657 + }, + { + "epoch": 0.24658, + "grad_norm": 0.9303466730804126, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 24658 + }, + { + "epoch": 0.24659, + "grad_norm": 0.9268609181947611, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 24659 + }, + { + "epoch": 0.2466, + "grad_norm": 0.9891028399265088, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 24660 + }, + { + "epoch": 0.24661, + "grad_norm": 0.9160446349231723, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 24661 + }, + { + "epoch": 0.24662, + "grad_norm": 0.8294118621039599, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 24662 + }, + { + "epoch": 0.24663, + "grad_norm": 0.6843724622859945, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 24663 + }, + { + "epoch": 0.24664, + "grad_norm": 0.6791309113380951, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 24664 + }, + { + "epoch": 0.24665, + "grad_norm": 0.6811781754455876, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 24665 + }, + { + "epoch": 0.24666, + "grad_norm": 0.7135457678343664, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 24666 + }, + { + "epoch": 0.24667, + "grad_norm": 0.7336177174849858, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 24667 + }, + { + "epoch": 0.24668, + "grad_norm": 0.7945911276580229, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 24668 + }, + { + "epoch": 0.24669, + "grad_norm": 0.756496868197791, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 24669 + }, + { + "epoch": 0.2467, + "grad_norm": 0.6545847778660131, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 24670 + }, + { + "epoch": 0.24671, + "grad_norm": 0.6017932318255034, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 24671 + }, + { + "epoch": 0.24672, + "grad_norm": 0.5262568850915205, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 24672 + }, + { + "epoch": 0.24673, + "grad_norm": 0.49493331125869555, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 24673 + }, + { + "epoch": 0.24674, + "grad_norm": 0.479227592954038, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 24674 + }, + { + "epoch": 0.24675, + "grad_norm": 0.4961702270091104, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 24675 + }, + { + "epoch": 0.24676, + "grad_norm": 0.5342976892943719, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 24676 + }, + { + "epoch": 0.24677, + "grad_norm": 0.5654755344368255, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 24677 + }, + { + "epoch": 0.24678, + "grad_norm": 0.6716869963790831, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 24678 + }, + { + "epoch": 0.24679, + "grad_norm": 0.9207499946485873, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 24679 + }, + { + "epoch": 0.2468, + "grad_norm": 1.134916279276091, + "learning_rate": 0.003, + "loss": 3.989, + "step": 24680 + }, + { + "epoch": 0.24681, + "grad_norm": 0.5813903890530597, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 24681 + }, + { + "epoch": 0.24682, + "grad_norm": 0.6239069307509005, + "learning_rate": 0.003, + "loss": 3.971, + "step": 24682 + }, + { + "epoch": 0.24683, + "grad_norm": 0.8296354264267551, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 24683 + }, + { + "epoch": 0.24684, + "grad_norm": 0.7222247524542016, + "learning_rate": 0.003, + "loss": 3.992, + "step": 24684 + }, + { + "epoch": 0.24685, + "grad_norm": 0.6724162393434335, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 24685 + }, + { + "epoch": 0.24686, + "grad_norm": 0.8450454858137431, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 24686 + }, + { + "epoch": 0.24687, + "grad_norm": 0.8758618153773844, + "learning_rate": 0.003, + "loss": 4.013, + "step": 24687 + }, + { + "epoch": 0.24688, + "grad_norm": 0.8055633957980266, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 24688 + }, + { + "epoch": 0.24689, + "grad_norm": 0.7994897466870846, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 24689 + }, + { + "epoch": 0.2469, + "grad_norm": 0.7270642914693114, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 24690 + }, + { + "epoch": 0.24691, + "grad_norm": 0.7358797097403278, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 24691 + }, + { + "epoch": 0.24692, + "grad_norm": 0.6982920275327831, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 24692 + }, + { + "epoch": 0.24693, + "grad_norm": 0.6419966375701632, + "learning_rate": 0.003, + "loss": 4.007, + "step": 24693 + }, + { + "epoch": 0.24694, + "grad_norm": 0.6442162360794579, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 24694 + }, + { + "epoch": 0.24695, + "grad_norm": 0.6427010239056187, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 24695 + }, + { + "epoch": 0.24696, + "grad_norm": 0.6953461955317922, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 24696 + }, + { + "epoch": 0.24697, + "grad_norm": 0.736140313733856, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 24697 + }, + { + "epoch": 0.24698, + "grad_norm": 0.7099470365231338, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 24698 + }, + { + "epoch": 0.24699, + "grad_norm": 0.7401387391954619, + "learning_rate": 0.003, + "loss": 4.011, + "step": 24699 + }, + { + "epoch": 0.247, + "grad_norm": 0.8706376138628652, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 24700 + }, + { + "epoch": 0.24701, + "grad_norm": 0.9649625749148956, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 24701 + }, + { + "epoch": 0.24702, + "grad_norm": 1.0138695575341907, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 24702 + }, + { + "epoch": 0.24703, + "grad_norm": 0.7426207033625528, + "learning_rate": 0.003, + "loss": 3.983, + "step": 24703 + }, + { + "epoch": 0.24704, + "grad_norm": 0.7814870816737897, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 24704 + }, + { + "epoch": 0.24705, + "grad_norm": 0.9057218062745811, + "learning_rate": 0.003, + "loss": 4.018, + "step": 24705 + }, + { + "epoch": 0.24706, + "grad_norm": 0.8368135071592309, + "learning_rate": 0.003, + "loss": 4.019, + "step": 24706 + }, + { + "epoch": 0.24707, + "grad_norm": 0.7961499244632043, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 24707 + }, + { + "epoch": 0.24708, + "grad_norm": 0.7852278818034003, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 24708 + }, + { + "epoch": 0.24709, + "grad_norm": 0.7463176915477423, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 24709 + }, + { + "epoch": 0.2471, + "grad_norm": 0.7500157958476933, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 24710 + }, + { + "epoch": 0.24711, + "grad_norm": 0.7763918112031533, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 24711 + }, + { + "epoch": 0.24712, + "grad_norm": 0.9156287792043233, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 24712 + }, + { + "epoch": 0.24713, + "grad_norm": 0.9593723936401716, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 24713 + }, + { + "epoch": 0.24714, + "grad_norm": 0.7267625538018005, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 24714 + }, + { + "epoch": 0.24715, + "grad_norm": 0.6479035068079172, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 24715 + }, + { + "epoch": 0.24716, + "grad_norm": 0.6452793689943173, + "learning_rate": 0.003, + "loss": 3.978, + "step": 24716 + }, + { + "epoch": 0.24717, + "grad_norm": 0.7194912528204654, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 24717 + }, + { + "epoch": 0.24718, + "grad_norm": 0.6612465701767688, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 24718 + }, + { + "epoch": 0.24719, + "grad_norm": 0.6353328085123198, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 24719 + }, + { + "epoch": 0.2472, + "grad_norm": 0.7120535602461623, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 24720 + }, + { + "epoch": 0.24721, + "grad_norm": 0.8028703415717942, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 24721 + }, + { + "epoch": 0.24722, + "grad_norm": 0.8827150347116637, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 24722 + }, + { + "epoch": 0.24723, + "grad_norm": 0.932379499958041, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 24723 + }, + { + "epoch": 0.24724, + "grad_norm": 0.8660444696553425, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 24724 + }, + { + "epoch": 0.24725, + "grad_norm": 0.8910636643522115, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 24725 + }, + { + "epoch": 0.24726, + "grad_norm": 1.1096899460900087, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 24726 + }, + { + "epoch": 0.24727, + "grad_norm": 0.9072438294711878, + "learning_rate": 0.003, + "loss": 3.986, + "step": 24727 + }, + { + "epoch": 0.24728, + "grad_norm": 0.7672590995258338, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 24728 + }, + { + "epoch": 0.24729, + "grad_norm": 0.7104176312561308, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 24729 + }, + { + "epoch": 0.2473, + "grad_norm": 0.7704755270809207, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 24730 + }, + { + "epoch": 0.24731, + "grad_norm": 0.7394437491518319, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 24731 + }, + { + "epoch": 0.24732, + "grad_norm": 0.6813602685230977, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 24732 + }, + { + "epoch": 0.24733, + "grad_norm": 0.6687386382505931, + "learning_rate": 0.003, + "loss": 4.006, + "step": 24733 + }, + { + "epoch": 0.24734, + "grad_norm": 0.7254108213451024, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 24734 + }, + { + "epoch": 0.24735, + "grad_norm": 0.6758641081480242, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 24735 + }, + { + "epoch": 0.24736, + "grad_norm": 0.6142459276107947, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 24736 + }, + { + "epoch": 0.24737, + "grad_norm": 0.5640013456819891, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 24737 + }, + { + "epoch": 0.24738, + "grad_norm": 0.6092360480203135, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 24738 + }, + { + "epoch": 0.24739, + "grad_norm": 0.6100834708682206, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 24739 + }, + { + "epoch": 0.2474, + "grad_norm": 0.6097716007637372, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 24740 + }, + { + "epoch": 0.24741, + "grad_norm": 0.5395775545137621, + "learning_rate": 0.003, + "loss": 3.982, + "step": 24741 + }, + { + "epoch": 0.24742, + "grad_norm": 0.5624446762231256, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 24742 + }, + { + "epoch": 0.24743, + "grad_norm": 0.6739425147780884, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 24743 + }, + { + "epoch": 0.24744, + "grad_norm": 0.758836243606955, + "learning_rate": 0.003, + "loss": 3.973, + "step": 24744 + }, + { + "epoch": 0.24745, + "grad_norm": 0.8345948215322861, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 24745 + }, + { + "epoch": 0.24746, + "grad_norm": 1.0981573803627211, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 24746 + }, + { + "epoch": 0.24747, + "grad_norm": 0.9967496974130757, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 24747 + }, + { + "epoch": 0.24748, + "grad_norm": 0.8126117741119616, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 24748 + }, + { + "epoch": 0.24749, + "grad_norm": 0.9336898434897593, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 24749 + }, + { + "epoch": 0.2475, + "grad_norm": 0.9865495165047188, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 24750 + }, + { + "epoch": 0.24751, + "grad_norm": 0.9842184228820471, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 24751 + }, + { + "epoch": 0.24752, + "grad_norm": 1.1078207957622177, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 24752 + }, + { + "epoch": 0.24753, + "grad_norm": 1.053413379649574, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 24753 + }, + { + "epoch": 0.24754, + "grad_norm": 0.8691503508287581, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 24754 + }, + { + "epoch": 0.24755, + "grad_norm": 0.8233369982331415, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 24755 + }, + { + "epoch": 0.24756, + "grad_norm": 0.824454304288463, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 24756 + }, + { + "epoch": 0.24757, + "grad_norm": 0.7659575206414759, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 24757 + }, + { + "epoch": 0.24758, + "grad_norm": 0.8381526461469414, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 24758 + }, + { + "epoch": 0.24759, + "grad_norm": 0.8046003556538527, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 24759 + }, + { + "epoch": 0.2476, + "grad_norm": 0.6664865080986674, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 24760 + }, + { + "epoch": 0.24761, + "grad_norm": 0.5952794339161708, + "learning_rate": 0.003, + "loss": 3.99, + "step": 24761 + }, + { + "epoch": 0.24762, + "grad_norm": 0.5873041840708813, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 24762 + }, + { + "epoch": 0.24763, + "grad_norm": 0.6919275632071442, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 24763 + }, + { + "epoch": 0.24764, + "grad_norm": 0.6793013158973645, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 24764 + }, + { + "epoch": 0.24765, + "grad_norm": 0.6779552492013282, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 24765 + }, + { + "epoch": 0.24766, + "grad_norm": 0.6160468113845834, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 24766 + }, + { + "epoch": 0.24767, + "grad_norm": 0.6197654830367167, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 24767 + }, + { + "epoch": 0.24768, + "grad_norm": 0.6323431721021713, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 24768 + }, + { + "epoch": 0.24769, + "grad_norm": 0.6788681685034332, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 24769 + }, + { + "epoch": 0.2477, + "grad_norm": 0.7092804525585372, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 24770 + }, + { + "epoch": 0.24771, + "grad_norm": 0.7397246388057974, + "learning_rate": 0.003, + "loss": 4.0, + "step": 24771 + }, + { + "epoch": 0.24772, + "grad_norm": 0.6658441678384962, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 24772 + }, + { + "epoch": 0.24773, + "grad_norm": 0.7130599260348339, + "learning_rate": 0.003, + "loss": 3.982, + "step": 24773 + }, + { + "epoch": 0.24774, + "grad_norm": 0.6845416594555065, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 24774 + }, + { + "epoch": 0.24775, + "grad_norm": 0.74360292567571, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 24775 + }, + { + "epoch": 0.24776, + "grad_norm": 0.8445893931446187, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 24776 + }, + { + "epoch": 0.24777, + "grad_norm": 0.9519894359516868, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 24777 + }, + { + "epoch": 0.24778, + "grad_norm": 1.050529391198966, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 24778 + }, + { + "epoch": 0.24779, + "grad_norm": 0.8995290363769491, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 24779 + }, + { + "epoch": 0.2478, + "grad_norm": 0.8613210738013147, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 24780 + }, + { + "epoch": 0.24781, + "grad_norm": 0.8260259341296079, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 24781 + }, + { + "epoch": 0.24782, + "grad_norm": 0.7741041833183798, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 24782 + }, + { + "epoch": 0.24783, + "grad_norm": 0.7486386956975187, + "learning_rate": 0.003, + "loss": 4.027, + "step": 24783 + }, + { + "epoch": 0.24784, + "grad_norm": 0.8058020655088537, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 24784 + }, + { + "epoch": 0.24785, + "grad_norm": 0.7132083256344538, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 24785 + }, + { + "epoch": 0.24786, + "grad_norm": 0.6767775842351011, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 24786 + }, + { + "epoch": 0.24787, + "grad_norm": 0.6906221942353122, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 24787 + }, + { + "epoch": 0.24788, + "grad_norm": 0.7265022259762991, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 24788 + }, + { + "epoch": 0.24789, + "grad_norm": 0.7504998090457833, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 24789 + }, + { + "epoch": 0.2479, + "grad_norm": 0.805822171484891, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 24790 + }, + { + "epoch": 0.24791, + "grad_norm": 0.7444416837279535, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 24791 + }, + { + "epoch": 0.24792, + "grad_norm": 0.7614468730046011, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 24792 + }, + { + "epoch": 0.24793, + "grad_norm": 0.7223029224903142, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 24793 + }, + { + "epoch": 0.24794, + "grad_norm": 0.6958144133591978, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 24794 + }, + { + "epoch": 0.24795, + "grad_norm": 0.6884517680835571, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 24795 + }, + { + "epoch": 0.24796, + "grad_norm": 0.6619121157873161, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 24796 + }, + { + "epoch": 0.24797, + "grad_norm": 0.7035965857816479, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 24797 + }, + { + "epoch": 0.24798, + "grad_norm": 0.7567736959420365, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 24798 + }, + { + "epoch": 0.24799, + "grad_norm": 0.8563269003679965, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 24799 + }, + { + "epoch": 0.248, + "grad_norm": 0.9097598350027634, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 24800 + }, + { + "epoch": 0.24801, + "grad_norm": 0.8817262139742067, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 24801 + }, + { + "epoch": 0.24802, + "grad_norm": 0.7660433050377115, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 24802 + }, + { + "epoch": 0.24803, + "grad_norm": 0.6812976079616089, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 24803 + }, + { + "epoch": 0.24804, + "grad_norm": 0.7526663857199462, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 24804 + }, + { + "epoch": 0.24805, + "grad_norm": 0.8727789564900402, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 24805 + }, + { + "epoch": 0.24806, + "grad_norm": 0.8674460854328385, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 24806 + }, + { + "epoch": 0.24807, + "grad_norm": 0.7904285379774957, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 24807 + }, + { + "epoch": 0.24808, + "grad_norm": 0.7333640123710898, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 24808 + }, + { + "epoch": 0.24809, + "grad_norm": 0.7450911930650128, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 24809 + }, + { + "epoch": 0.2481, + "grad_norm": 0.6459135991755526, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 24810 + }, + { + "epoch": 0.24811, + "grad_norm": 0.6630320545330188, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 24811 + }, + { + "epoch": 0.24812, + "grad_norm": 0.7270404287419607, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 24812 + }, + { + "epoch": 0.24813, + "grad_norm": 0.7119432831463348, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 24813 + }, + { + "epoch": 0.24814, + "grad_norm": 0.6860568757780948, + "learning_rate": 0.003, + "loss": 4.003, + "step": 24814 + }, + { + "epoch": 0.24815, + "grad_norm": 0.7751388543002, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 24815 + }, + { + "epoch": 0.24816, + "grad_norm": 1.0114222968862385, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 24816 + }, + { + "epoch": 0.24817, + "grad_norm": 1.137822726006878, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 24817 + }, + { + "epoch": 0.24818, + "grad_norm": 0.6586243389149928, + "learning_rate": 0.003, + "loss": 4.002, + "step": 24818 + }, + { + "epoch": 0.24819, + "grad_norm": 0.6238953722536459, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 24819 + }, + { + "epoch": 0.2482, + "grad_norm": 0.7229852044641515, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 24820 + }, + { + "epoch": 0.24821, + "grad_norm": 0.6833433211964285, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 24821 + }, + { + "epoch": 0.24822, + "grad_norm": 0.5754830268255314, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 24822 + }, + { + "epoch": 0.24823, + "grad_norm": 0.5112456378523734, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 24823 + }, + { + "epoch": 0.24824, + "grad_norm": 0.5072464430939612, + "learning_rate": 0.003, + "loss": 3.956, + "step": 24824 + }, + { + "epoch": 0.24825, + "grad_norm": 0.5024257555989691, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 24825 + }, + { + "epoch": 0.24826, + "grad_norm": 0.5854194888446946, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 24826 + }, + { + "epoch": 0.24827, + "grad_norm": 0.7067225680892854, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 24827 + }, + { + "epoch": 0.24828, + "grad_norm": 0.7574566676399644, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 24828 + }, + { + "epoch": 0.24829, + "grad_norm": 0.826059310493233, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 24829 + }, + { + "epoch": 0.2483, + "grad_norm": 0.8281512702213365, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 24830 + }, + { + "epoch": 0.24831, + "grad_norm": 0.7931103851200417, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 24831 + }, + { + "epoch": 0.24832, + "grad_norm": 0.8221783265276664, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 24832 + }, + { + "epoch": 0.24833, + "grad_norm": 0.7758752087077427, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 24833 + }, + { + "epoch": 0.24834, + "grad_norm": 0.7503559317808217, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 24834 + }, + { + "epoch": 0.24835, + "grad_norm": 0.7509261905366124, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 24835 + }, + { + "epoch": 0.24836, + "grad_norm": 0.7492936082608564, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 24836 + }, + { + "epoch": 0.24837, + "grad_norm": 0.8287498403862312, + "learning_rate": 0.003, + "loss": 4.011, + "step": 24837 + }, + { + "epoch": 0.24838, + "grad_norm": 0.7742051187486616, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 24838 + }, + { + "epoch": 0.24839, + "grad_norm": 0.7537340230807204, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 24839 + }, + { + "epoch": 0.2484, + "grad_norm": 0.7407634030353978, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 24840 + }, + { + "epoch": 0.24841, + "grad_norm": 0.8538206446115529, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 24841 + }, + { + "epoch": 0.24842, + "grad_norm": 1.0369637364989153, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 24842 + }, + { + "epoch": 0.24843, + "grad_norm": 1.1580178242826535, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 24843 + }, + { + "epoch": 0.24844, + "grad_norm": 0.8904699609792087, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 24844 + }, + { + "epoch": 0.24845, + "grad_norm": 0.8643143852094233, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 24845 + }, + { + "epoch": 0.24846, + "grad_norm": 0.7666807882075897, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 24846 + }, + { + "epoch": 0.24847, + "grad_norm": 0.72675273865454, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 24847 + }, + { + "epoch": 0.24848, + "grad_norm": 0.717989726116239, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 24848 + }, + { + "epoch": 0.24849, + "grad_norm": 0.7365545459657905, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 24849 + }, + { + "epoch": 0.2485, + "grad_norm": 0.8195813734989255, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 24850 + }, + { + "epoch": 0.24851, + "grad_norm": 0.7746841674850736, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 24851 + }, + { + "epoch": 0.24852, + "grad_norm": 0.7128005185903262, + "learning_rate": 0.003, + "loss": 4.014, + "step": 24852 + }, + { + "epoch": 0.24853, + "grad_norm": 0.6508503679942031, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 24853 + }, + { + "epoch": 0.24854, + "grad_norm": 0.6161193080083968, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 24854 + }, + { + "epoch": 0.24855, + "grad_norm": 0.6272237004667809, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 24855 + }, + { + "epoch": 0.24856, + "grad_norm": 0.644776139383404, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 24856 + }, + { + "epoch": 0.24857, + "grad_norm": 0.7267402514759279, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 24857 + }, + { + "epoch": 0.24858, + "grad_norm": 0.8076306455143638, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 24858 + }, + { + "epoch": 0.24859, + "grad_norm": 0.8042364929521044, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 24859 + }, + { + "epoch": 0.2486, + "grad_norm": 0.801435630255965, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 24860 + }, + { + "epoch": 0.24861, + "grad_norm": 0.8751018585487709, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 24861 + }, + { + "epoch": 0.24862, + "grad_norm": 0.8700673605331015, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 24862 + }, + { + "epoch": 0.24863, + "grad_norm": 0.7314149493557224, + "learning_rate": 0.003, + "loss": 4.007, + "step": 24863 + }, + { + "epoch": 0.24864, + "grad_norm": 0.7112705236903821, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 24864 + }, + { + "epoch": 0.24865, + "grad_norm": 0.6739496632272034, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 24865 + }, + { + "epoch": 0.24866, + "grad_norm": 0.5678820190189605, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 24866 + }, + { + "epoch": 0.24867, + "grad_norm": 0.5527067181406348, + "learning_rate": 0.003, + "loss": 3.982, + "step": 24867 + }, + { + "epoch": 0.24868, + "grad_norm": 0.7027327752870818, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 24868 + }, + { + "epoch": 0.24869, + "grad_norm": 0.9354931226801562, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 24869 + }, + { + "epoch": 0.2487, + "grad_norm": 1.1615861445120856, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 24870 + }, + { + "epoch": 0.24871, + "grad_norm": 0.6753401047414013, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 24871 + }, + { + "epoch": 0.24872, + "grad_norm": 0.6651025113431295, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 24872 + }, + { + "epoch": 0.24873, + "grad_norm": 0.6801254725289613, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 24873 + }, + { + "epoch": 0.24874, + "grad_norm": 0.6398172600489702, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 24874 + }, + { + "epoch": 0.24875, + "grad_norm": 0.5813497066453517, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 24875 + }, + { + "epoch": 0.24876, + "grad_norm": 0.5570165837519214, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 24876 + }, + { + "epoch": 0.24877, + "grad_norm": 0.5028712611083063, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 24877 + }, + { + "epoch": 0.24878, + "grad_norm": 0.5714492160860553, + "learning_rate": 0.003, + "loss": 3.969, + "step": 24878 + }, + { + "epoch": 0.24879, + "grad_norm": 0.6183126137405731, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 24879 + }, + { + "epoch": 0.2488, + "grad_norm": 0.6216659179833393, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 24880 + }, + { + "epoch": 0.24881, + "grad_norm": 0.6572994539429968, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 24881 + }, + { + "epoch": 0.24882, + "grad_norm": 0.6738793252218268, + "learning_rate": 0.003, + "loss": 3.972, + "step": 24882 + }, + { + "epoch": 0.24883, + "grad_norm": 0.7356851951532333, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 24883 + }, + { + "epoch": 0.24884, + "grad_norm": 0.7919632959749184, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 24884 + }, + { + "epoch": 0.24885, + "grad_norm": 0.7876336664715042, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 24885 + }, + { + "epoch": 0.24886, + "grad_norm": 0.8660144325875903, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 24886 + }, + { + "epoch": 0.24887, + "grad_norm": 0.9273934279035242, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 24887 + }, + { + "epoch": 0.24888, + "grad_norm": 0.8171837255932605, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 24888 + }, + { + "epoch": 0.24889, + "grad_norm": 0.7937651226864451, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 24889 + }, + { + "epoch": 0.2489, + "grad_norm": 0.7936044803847645, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 24890 + }, + { + "epoch": 0.24891, + "grad_norm": 0.8111154252084162, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 24891 + }, + { + "epoch": 0.24892, + "grad_norm": 0.8820298317926532, + "learning_rate": 0.003, + "loss": 4.0482, + "step": 24892 + }, + { + "epoch": 0.24893, + "grad_norm": 0.9401317938607567, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 24893 + }, + { + "epoch": 0.24894, + "grad_norm": 0.9524775247979766, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 24894 + }, + { + "epoch": 0.24895, + "grad_norm": 1.0639214315706838, + "learning_rate": 0.003, + "loss": 4.029, + "step": 24895 + }, + { + "epoch": 0.24896, + "grad_norm": 0.905240249115481, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 24896 + }, + { + "epoch": 0.24897, + "grad_norm": 0.8322084250229351, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 24897 + }, + { + "epoch": 0.24898, + "grad_norm": 0.8298870900394372, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 24898 + }, + { + "epoch": 0.24899, + "grad_norm": 0.7366360771443524, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 24899 + }, + { + "epoch": 0.249, + "grad_norm": 0.7178771899432024, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 24900 + }, + { + "epoch": 0.24901, + "grad_norm": 0.6709396282435286, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 24901 + }, + { + "epoch": 0.24902, + "grad_norm": 0.5894074463284142, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 24902 + }, + { + "epoch": 0.24903, + "grad_norm": 0.5194775897299665, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 24903 + }, + { + "epoch": 0.24904, + "grad_norm": 0.5386837772584023, + "learning_rate": 0.003, + "loss": 3.982, + "step": 24904 + }, + { + "epoch": 0.24905, + "grad_norm": 0.6006171988033323, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 24905 + }, + { + "epoch": 0.24906, + "grad_norm": 0.6811834898647189, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 24906 + }, + { + "epoch": 0.24907, + "grad_norm": 0.8627987141694372, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 24907 + }, + { + "epoch": 0.24908, + "grad_norm": 1.1219893161609802, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 24908 + }, + { + "epoch": 0.24909, + "grad_norm": 0.8643748912303476, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 24909 + }, + { + "epoch": 0.2491, + "grad_norm": 0.7894973177858463, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 24910 + }, + { + "epoch": 0.24911, + "grad_norm": 0.7634428163402655, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 24911 + }, + { + "epoch": 0.24912, + "grad_norm": 0.713472811718414, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 24912 + }, + { + "epoch": 0.24913, + "grad_norm": 0.7222508907617733, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 24913 + }, + { + "epoch": 0.24914, + "grad_norm": 0.7408475058134404, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 24914 + }, + { + "epoch": 0.24915, + "grad_norm": 0.6464003272680945, + "learning_rate": 0.003, + "loss": 4.03, + "step": 24915 + }, + { + "epoch": 0.24916, + "grad_norm": 0.6397258595752645, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 24916 + }, + { + "epoch": 0.24917, + "grad_norm": 0.7282583557892852, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 24917 + }, + { + "epoch": 0.24918, + "grad_norm": 0.8336888862974141, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 24918 + }, + { + "epoch": 0.24919, + "grad_norm": 1.0105884347901206, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 24919 + }, + { + "epoch": 0.2492, + "grad_norm": 1.1129610379011712, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 24920 + }, + { + "epoch": 0.24921, + "grad_norm": 0.9331928427472199, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 24921 + }, + { + "epoch": 0.24922, + "grad_norm": 0.8484129713467307, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 24922 + }, + { + "epoch": 0.24923, + "grad_norm": 0.9298612989568074, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 24923 + }, + { + "epoch": 0.24924, + "grad_norm": 0.9651046247589864, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 24924 + }, + { + "epoch": 0.24925, + "grad_norm": 0.803598801810715, + "learning_rate": 0.003, + "loss": 4.002, + "step": 24925 + }, + { + "epoch": 0.24926, + "grad_norm": 0.7797697120251206, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 24926 + }, + { + "epoch": 0.24927, + "grad_norm": 0.936690647322902, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 24927 + }, + { + "epoch": 0.24928, + "grad_norm": 1.099077263176571, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 24928 + }, + { + "epoch": 0.24929, + "grad_norm": 1.0269513767639906, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 24929 + }, + { + "epoch": 0.2493, + "grad_norm": 0.8157007967955145, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 24930 + }, + { + "epoch": 0.24931, + "grad_norm": 0.7215541348793926, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 24931 + }, + { + "epoch": 0.24932, + "grad_norm": 0.6427307816297427, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 24932 + }, + { + "epoch": 0.24933, + "grad_norm": 0.5978223161609281, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 24933 + }, + { + "epoch": 0.24934, + "grad_norm": 0.5704205661309036, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 24934 + }, + { + "epoch": 0.24935, + "grad_norm": 0.5708107602195589, + "learning_rate": 0.003, + "loss": 4.012, + "step": 24935 + }, + { + "epoch": 0.24936, + "grad_norm": 0.6098504652247851, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 24936 + }, + { + "epoch": 0.24937, + "grad_norm": 0.6155891003155247, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 24937 + }, + { + "epoch": 0.24938, + "grad_norm": 0.6724590330480235, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 24938 + }, + { + "epoch": 0.24939, + "grad_norm": 0.6600177147957237, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 24939 + }, + { + "epoch": 0.2494, + "grad_norm": 0.6709770345016721, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 24940 + }, + { + "epoch": 0.24941, + "grad_norm": 0.6583131242924962, + "learning_rate": 0.003, + "loss": 3.992, + "step": 24941 + }, + { + "epoch": 0.24942, + "grad_norm": 0.7065636614428535, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 24942 + }, + { + "epoch": 0.24943, + "grad_norm": 0.7260733514711806, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 24943 + }, + { + "epoch": 0.24944, + "grad_norm": 0.587078322671691, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 24944 + }, + { + "epoch": 0.24945, + "grad_norm": 0.4872326276614992, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 24945 + }, + { + "epoch": 0.24946, + "grad_norm": 0.5993197433491632, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 24946 + }, + { + "epoch": 0.24947, + "grad_norm": 0.7368639077513064, + "learning_rate": 0.003, + "loss": 3.988, + "step": 24947 + }, + { + "epoch": 0.24948, + "grad_norm": 0.9922286754661165, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 24948 + }, + { + "epoch": 0.24949, + "grad_norm": 0.9651890062950222, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 24949 + }, + { + "epoch": 0.2495, + "grad_norm": 0.7146897594917286, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 24950 + }, + { + "epoch": 0.24951, + "grad_norm": 0.6481227103225075, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 24951 + }, + { + "epoch": 0.24952, + "grad_norm": 0.6713746313308224, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 24952 + }, + { + "epoch": 0.24953, + "grad_norm": 0.5908540677100418, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 24953 + }, + { + "epoch": 0.24954, + "grad_norm": 0.584627189974253, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 24954 + }, + { + "epoch": 0.24955, + "grad_norm": 0.6528476063622836, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 24955 + }, + { + "epoch": 0.24956, + "grad_norm": 0.702202491739534, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 24956 + }, + { + "epoch": 0.24957, + "grad_norm": 0.6579780212461367, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 24957 + }, + { + "epoch": 0.24958, + "grad_norm": 0.751203009491025, + "learning_rate": 0.003, + "loss": 4.01, + "step": 24958 + }, + { + "epoch": 0.24959, + "grad_norm": 0.9162762070594106, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 24959 + }, + { + "epoch": 0.2496, + "grad_norm": 1.0680114592021228, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 24960 + }, + { + "epoch": 0.24961, + "grad_norm": 0.9107752798721737, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 24961 + }, + { + "epoch": 0.24962, + "grad_norm": 0.8144882259029249, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 24962 + }, + { + "epoch": 0.24963, + "grad_norm": 0.6173589148977396, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 24963 + }, + { + "epoch": 0.24964, + "grad_norm": 0.593805278498741, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 24964 + }, + { + "epoch": 0.24965, + "grad_norm": 0.6223654166990317, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 24965 + }, + { + "epoch": 0.24966, + "grad_norm": 0.5729172629153567, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 24966 + }, + { + "epoch": 0.24967, + "grad_norm": 0.5348854219110553, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 24967 + }, + { + "epoch": 0.24968, + "grad_norm": 0.5518850567373988, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 24968 + }, + { + "epoch": 0.24969, + "grad_norm": 0.5768573450098397, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 24969 + }, + { + "epoch": 0.2497, + "grad_norm": 0.6773156653010566, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 24970 + }, + { + "epoch": 0.24971, + "grad_norm": 0.7907427004415946, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 24971 + }, + { + "epoch": 0.24972, + "grad_norm": 0.9055836089771682, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 24972 + }, + { + "epoch": 0.24973, + "grad_norm": 0.9157088230262936, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 24973 + }, + { + "epoch": 0.24974, + "grad_norm": 0.7783122591576922, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 24974 + }, + { + "epoch": 0.24975, + "grad_norm": 0.7000325192455242, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 24975 + }, + { + "epoch": 0.24976, + "grad_norm": 0.7769816366799092, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 24976 + }, + { + "epoch": 0.24977, + "grad_norm": 0.814719558505342, + "learning_rate": 0.003, + "loss": 4.004, + "step": 24977 + }, + { + "epoch": 0.24978, + "grad_norm": 0.7048493594340757, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 24978 + }, + { + "epoch": 0.24979, + "grad_norm": 0.7487168622525326, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 24979 + }, + { + "epoch": 0.2498, + "grad_norm": 0.8636902581062224, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 24980 + }, + { + "epoch": 0.24981, + "grad_norm": 0.8778641780421933, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 24981 + }, + { + "epoch": 0.24982, + "grad_norm": 1.1180717631207864, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 24982 + }, + { + "epoch": 0.24983, + "grad_norm": 1.0394375801510467, + "learning_rate": 0.003, + "loss": 3.993, + "step": 24983 + }, + { + "epoch": 0.24984, + "grad_norm": 0.8353737762037999, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 24984 + }, + { + "epoch": 0.24985, + "grad_norm": 0.8452363956129066, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 24985 + }, + { + "epoch": 0.24986, + "grad_norm": 0.9880658678463923, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 24986 + }, + { + "epoch": 0.24987, + "grad_norm": 0.9560919475625882, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 24987 + }, + { + "epoch": 0.24988, + "grad_norm": 1.0449688896078937, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 24988 + }, + { + "epoch": 0.24989, + "grad_norm": 0.8912247959355567, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 24989 + }, + { + "epoch": 0.2499, + "grad_norm": 0.943505961973583, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 24990 + }, + { + "epoch": 0.24991, + "grad_norm": 0.8542420504807667, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 24991 + }, + { + "epoch": 0.24992, + "grad_norm": 0.8040659835830393, + "learning_rate": 0.003, + "loss": 4.0406, + "step": 24992 + }, + { + "epoch": 0.24993, + "grad_norm": 0.7930789730069427, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 24993 + }, + { + "epoch": 0.24994, + "grad_norm": 0.7397485908085293, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 24994 + }, + { + "epoch": 0.24995, + "grad_norm": 0.7438031495361794, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 24995 + }, + { + "epoch": 0.24996, + "grad_norm": 0.866728931337775, + "learning_rate": 0.003, + "loss": 4.012, + "step": 24996 + }, + { + "epoch": 0.24997, + "grad_norm": 1.0457604580440067, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 24997 + }, + { + "epoch": 0.24998, + "grad_norm": 1.1373741113483162, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 24998 + }, + { + "epoch": 0.24999, + "grad_norm": 0.7317379762568438, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 24999 + }, + { + "epoch": 0.25, + "grad_norm": 0.6590633393614885, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 25000 + }, + { + "epoch": 0.25001, + "grad_norm": 0.6595313121791389, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 25001 + }, + { + "epoch": 0.25002, + "grad_norm": 0.6779180969163103, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 25002 + }, + { + "epoch": 0.25003, + "grad_norm": 0.6482351272674579, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 25003 + }, + { + "epoch": 0.25004, + "grad_norm": 0.5822919690237915, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 25004 + }, + { + "epoch": 0.25005, + "grad_norm": 0.6521831600353845, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 25005 + }, + { + "epoch": 0.25006, + "grad_norm": 0.7259954551816502, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 25006 + }, + { + "epoch": 0.25007, + "grad_norm": 0.6742345882288789, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 25007 + }, + { + "epoch": 0.25008, + "grad_norm": 0.5696751882418559, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 25008 + }, + { + "epoch": 0.25009, + "grad_norm": 0.5586813076894819, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 25009 + }, + { + "epoch": 0.2501, + "grad_norm": 0.5638351656923131, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 25010 + }, + { + "epoch": 0.25011, + "grad_norm": 0.5570618356436909, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 25011 + }, + { + "epoch": 0.25012, + "grad_norm": 0.6011201482238158, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 25012 + }, + { + "epoch": 0.25013, + "grad_norm": 0.6962894625752573, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 25013 + }, + { + "epoch": 0.25014, + "grad_norm": 0.7098614526774345, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 25014 + }, + { + "epoch": 0.25015, + "grad_norm": 0.7076330926638311, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 25015 + }, + { + "epoch": 0.25016, + "grad_norm": 0.6943938960726996, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 25016 + }, + { + "epoch": 0.25017, + "grad_norm": 0.7202696802039519, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 25017 + }, + { + "epoch": 0.25018, + "grad_norm": 0.7260904860594289, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 25018 + }, + { + "epoch": 0.25019, + "grad_norm": 0.7720851792546698, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 25019 + }, + { + "epoch": 0.2502, + "grad_norm": 0.7533442646415303, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 25020 + }, + { + "epoch": 0.25021, + "grad_norm": 0.6625570456989979, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 25021 + }, + { + "epoch": 0.25022, + "grad_norm": 0.6557352433979404, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 25022 + }, + { + "epoch": 0.25023, + "grad_norm": 0.6611979664672962, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 25023 + }, + { + "epoch": 0.25024, + "grad_norm": 0.6682498716578794, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 25024 + }, + { + "epoch": 0.25025, + "grad_norm": 0.6903556424287116, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 25025 + }, + { + "epoch": 0.25026, + "grad_norm": 0.7118629574912311, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 25026 + }, + { + "epoch": 0.25027, + "grad_norm": 0.839958662335139, + "learning_rate": 0.003, + "loss": 4.005, + "step": 25027 + }, + { + "epoch": 0.25028, + "grad_norm": 0.9091694503746534, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 25028 + }, + { + "epoch": 0.25029, + "grad_norm": 0.8860321655863509, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 25029 + }, + { + "epoch": 0.2503, + "grad_norm": 0.9032518550857938, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 25030 + }, + { + "epoch": 0.25031, + "grad_norm": 0.8393665232220355, + "learning_rate": 0.003, + "loss": 4.015, + "step": 25031 + }, + { + "epoch": 0.25032, + "grad_norm": 0.8157896207020217, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 25032 + }, + { + "epoch": 0.25033, + "grad_norm": 0.9027163895825299, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 25033 + }, + { + "epoch": 0.25034, + "grad_norm": 0.8801328344024659, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 25034 + }, + { + "epoch": 0.25035, + "grad_norm": 0.9402498670598741, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 25035 + }, + { + "epoch": 0.25036, + "grad_norm": 1.0855024212362625, + "learning_rate": 0.003, + "loss": 4.021, + "step": 25036 + }, + { + "epoch": 0.25037, + "grad_norm": 0.8974119355818403, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 25037 + }, + { + "epoch": 0.25038, + "grad_norm": 0.7703851955193783, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 25038 + }, + { + "epoch": 0.25039, + "grad_norm": 0.7997156480532596, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 25039 + }, + { + "epoch": 0.2504, + "grad_norm": 0.8217911029098349, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 25040 + }, + { + "epoch": 0.25041, + "grad_norm": 0.7928130612197543, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 25041 + }, + { + "epoch": 0.25042, + "grad_norm": 0.8778778001726257, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 25042 + }, + { + "epoch": 0.25043, + "grad_norm": 1.0042159612680202, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 25043 + }, + { + "epoch": 0.25044, + "grad_norm": 1.2063431948636552, + "learning_rate": 0.003, + "loss": 4.0419, + "step": 25044 + }, + { + "epoch": 0.25045, + "grad_norm": 0.7324843334053264, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 25045 + }, + { + "epoch": 0.25046, + "grad_norm": 0.7157556631255482, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 25046 + }, + { + "epoch": 0.25047, + "grad_norm": 0.6757874467936802, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 25047 + }, + { + "epoch": 0.25048, + "grad_norm": 0.808571543860653, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 25048 + }, + { + "epoch": 0.25049, + "grad_norm": 0.7851950068886392, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 25049 + }, + { + "epoch": 0.2505, + "grad_norm": 0.8187069362034363, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 25050 + }, + { + "epoch": 0.25051, + "grad_norm": 0.7151164059313635, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 25051 + }, + { + "epoch": 0.25052, + "grad_norm": 0.691121804116509, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 25052 + }, + { + "epoch": 0.25053, + "grad_norm": 0.8282257594975653, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 25053 + }, + { + "epoch": 0.25054, + "grad_norm": 0.9463381502384001, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 25054 + }, + { + "epoch": 0.25055, + "grad_norm": 1.0333154592189902, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 25055 + }, + { + "epoch": 0.25056, + "grad_norm": 0.9133796220331937, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 25056 + }, + { + "epoch": 0.25057, + "grad_norm": 0.710134733102437, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 25057 + }, + { + "epoch": 0.25058, + "grad_norm": 0.6408172866742105, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 25058 + }, + { + "epoch": 0.25059, + "grad_norm": 0.6423082119403657, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 25059 + }, + { + "epoch": 0.2506, + "grad_norm": 0.5931415784986092, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 25060 + }, + { + "epoch": 0.25061, + "grad_norm": 0.5332260157561655, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 25061 + }, + { + "epoch": 0.25062, + "grad_norm": 0.5180001785512238, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 25062 + }, + { + "epoch": 0.25063, + "grad_norm": 0.5443845366190481, + "learning_rate": 0.003, + "loss": 3.997, + "step": 25063 + }, + { + "epoch": 0.25064, + "grad_norm": 0.5258843408146148, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 25064 + }, + { + "epoch": 0.25065, + "grad_norm": 0.5145230333551493, + "learning_rate": 0.003, + "loss": 3.995, + "step": 25065 + }, + { + "epoch": 0.25066, + "grad_norm": 0.5133333583615929, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 25066 + }, + { + "epoch": 0.25067, + "grad_norm": 0.5501062711658106, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 25067 + }, + { + "epoch": 0.25068, + "grad_norm": 0.5697932732913426, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 25068 + }, + { + "epoch": 0.25069, + "grad_norm": 0.6387496730310136, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 25069 + }, + { + "epoch": 0.2507, + "grad_norm": 0.6041413472583674, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 25070 + }, + { + "epoch": 0.25071, + "grad_norm": 0.5095872427313082, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 25071 + }, + { + "epoch": 0.25072, + "grad_norm": 0.5422491144211689, + "learning_rate": 0.003, + "loss": 3.982, + "step": 25072 + }, + { + "epoch": 0.25073, + "grad_norm": 0.591205847486999, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 25073 + }, + { + "epoch": 0.25074, + "grad_norm": 0.6027365875692261, + "learning_rate": 0.003, + "loss": 4.018, + "step": 25074 + }, + { + "epoch": 0.25075, + "grad_norm": 0.6930153981793578, + "learning_rate": 0.003, + "loss": 3.95, + "step": 25075 + }, + { + "epoch": 0.25076, + "grad_norm": 0.7316809209111611, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 25076 + }, + { + "epoch": 0.25077, + "grad_norm": 0.688470747470412, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 25077 + }, + { + "epoch": 0.25078, + "grad_norm": 0.7601162663121298, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 25078 + }, + { + "epoch": 0.25079, + "grad_norm": 0.8850310349332744, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 25079 + }, + { + "epoch": 0.2508, + "grad_norm": 1.1006312944692347, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 25080 + }, + { + "epoch": 0.25081, + "grad_norm": 1.0968434314494862, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 25081 + }, + { + "epoch": 0.25082, + "grad_norm": 0.746812478078696, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 25082 + }, + { + "epoch": 0.25083, + "grad_norm": 0.6838035178151961, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 25083 + }, + { + "epoch": 0.25084, + "grad_norm": 0.8044698841373851, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 25084 + }, + { + "epoch": 0.25085, + "grad_norm": 0.8658239421893738, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 25085 + }, + { + "epoch": 0.25086, + "grad_norm": 0.749581056576559, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 25086 + }, + { + "epoch": 0.25087, + "grad_norm": 0.7205668428822145, + "learning_rate": 0.003, + "loss": 4.011, + "step": 25087 + }, + { + "epoch": 0.25088, + "grad_norm": 0.7692720804983115, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 25088 + }, + { + "epoch": 0.25089, + "grad_norm": 0.7896188554899279, + "learning_rate": 0.003, + "loss": 4.014, + "step": 25089 + }, + { + "epoch": 0.2509, + "grad_norm": 0.8538475350934602, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 25090 + }, + { + "epoch": 0.25091, + "grad_norm": 0.8359362560819068, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 25091 + }, + { + "epoch": 0.25092, + "grad_norm": 0.8082531189599019, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 25092 + }, + { + "epoch": 0.25093, + "grad_norm": 0.6740327825563561, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 25093 + }, + { + "epoch": 0.25094, + "grad_norm": 0.6155839996353657, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 25094 + }, + { + "epoch": 0.25095, + "grad_norm": 0.6958374178041277, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 25095 + }, + { + "epoch": 0.25096, + "grad_norm": 0.8042173592630407, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 25096 + }, + { + "epoch": 0.25097, + "grad_norm": 0.8265579720909408, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 25097 + }, + { + "epoch": 0.25098, + "grad_norm": 0.7237292178562261, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 25098 + }, + { + "epoch": 0.25099, + "grad_norm": 0.6876839080631375, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 25099 + }, + { + "epoch": 0.251, + "grad_norm": 0.664200607691341, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 25100 + }, + { + "epoch": 0.25101, + "grad_norm": 0.7167485399833422, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 25101 + }, + { + "epoch": 0.25102, + "grad_norm": 0.6983929047933901, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 25102 + }, + { + "epoch": 0.25103, + "grad_norm": 0.8313519224529612, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 25103 + }, + { + "epoch": 0.25104, + "grad_norm": 1.0311894279263591, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 25104 + }, + { + "epoch": 0.25105, + "grad_norm": 0.9256573256764946, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 25105 + }, + { + "epoch": 0.25106, + "grad_norm": 0.913672462686042, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 25106 + }, + { + "epoch": 0.25107, + "grad_norm": 0.9416277803499072, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 25107 + }, + { + "epoch": 0.25108, + "grad_norm": 0.8851823366063145, + "learning_rate": 0.003, + "loss": 3.988, + "step": 25108 + }, + { + "epoch": 0.25109, + "grad_norm": 0.767705650452199, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 25109 + }, + { + "epoch": 0.2511, + "grad_norm": 0.7636191276258905, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 25110 + }, + { + "epoch": 0.25111, + "grad_norm": 0.8348851360495775, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 25111 + }, + { + "epoch": 0.25112, + "grad_norm": 0.8081405366196774, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 25112 + }, + { + "epoch": 0.25113, + "grad_norm": 0.857594005929864, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 25113 + }, + { + "epoch": 0.25114, + "grad_norm": 0.8956057299089404, + "learning_rate": 0.003, + "loss": 4.012, + "step": 25114 + }, + { + "epoch": 0.25115, + "grad_norm": 0.7668624433501585, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 25115 + }, + { + "epoch": 0.25116, + "grad_norm": 0.6789532435799747, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 25116 + }, + { + "epoch": 0.25117, + "grad_norm": 0.7266209076795384, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 25117 + }, + { + "epoch": 0.25118, + "grad_norm": 0.6899254851332951, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 25118 + }, + { + "epoch": 0.25119, + "grad_norm": 0.6865746210094782, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 25119 + }, + { + "epoch": 0.2512, + "grad_norm": 0.7012964192170156, + "learning_rate": 0.003, + "loss": 3.988, + "step": 25120 + }, + { + "epoch": 0.25121, + "grad_norm": 0.7793495780357999, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 25121 + }, + { + "epoch": 0.25122, + "grad_norm": 0.7568553350914992, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 25122 + }, + { + "epoch": 0.25123, + "grad_norm": 0.6519271202719115, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 25123 + }, + { + "epoch": 0.25124, + "grad_norm": 0.6974159737105387, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 25124 + }, + { + "epoch": 0.25125, + "grad_norm": 0.7509300109787347, + "learning_rate": 0.003, + "loss": 4.009, + "step": 25125 + }, + { + "epoch": 0.25126, + "grad_norm": 0.7123550710110638, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 25126 + }, + { + "epoch": 0.25127, + "grad_norm": 0.7345618097851071, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 25127 + }, + { + "epoch": 0.25128, + "grad_norm": 0.796929053662913, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 25128 + }, + { + "epoch": 0.25129, + "grad_norm": 0.7532838989112327, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 25129 + }, + { + "epoch": 0.2513, + "grad_norm": 0.7589356908938353, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 25130 + }, + { + "epoch": 0.25131, + "grad_norm": 0.7344536834988793, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 25131 + }, + { + "epoch": 0.25132, + "grad_norm": 0.7357966987441438, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 25132 + }, + { + "epoch": 0.25133, + "grad_norm": 0.8126367119158123, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 25133 + }, + { + "epoch": 0.25134, + "grad_norm": 0.8322103248234392, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 25134 + }, + { + "epoch": 0.25135, + "grad_norm": 0.9083419588750272, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 25135 + }, + { + "epoch": 0.25136, + "grad_norm": 0.994956580415079, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 25136 + }, + { + "epoch": 0.25137, + "grad_norm": 1.1227099577336983, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 25137 + }, + { + "epoch": 0.25138, + "grad_norm": 0.8930175172911572, + "learning_rate": 0.003, + "loss": 4.011, + "step": 25138 + }, + { + "epoch": 0.25139, + "grad_norm": 0.9166659175711329, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 25139 + }, + { + "epoch": 0.2514, + "grad_norm": 1.063444744961362, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 25140 + }, + { + "epoch": 0.25141, + "grad_norm": 0.9116957086864614, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 25141 + }, + { + "epoch": 0.25142, + "grad_norm": 0.7775469235250918, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 25142 + }, + { + "epoch": 0.25143, + "grad_norm": 0.7846490377874302, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 25143 + }, + { + "epoch": 0.25144, + "grad_norm": 0.800791759289016, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 25144 + }, + { + "epoch": 0.25145, + "grad_norm": 0.8089543252587941, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 25145 + }, + { + "epoch": 0.25146, + "grad_norm": 0.7619747439015466, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 25146 + }, + { + "epoch": 0.25147, + "grad_norm": 0.8500312933077642, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 25147 + }, + { + "epoch": 0.25148, + "grad_norm": 0.810582725075197, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 25148 + }, + { + "epoch": 0.25149, + "grad_norm": 0.8417558004669867, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 25149 + }, + { + "epoch": 0.2515, + "grad_norm": 0.9369632646988516, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 25150 + }, + { + "epoch": 0.25151, + "grad_norm": 0.8905143617666227, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 25151 + }, + { + "epoch": 0.25152, + "grad_norm": 0.6677572956151608, + "learning_rate": 0.003, + "loss": 4.018, + "step": 25152 + }, + { + "epoch": 0.25153, + "grad_norm": 0.664499642703004, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 25153 + }, + { + "epoch": 0.25154, + "grad_norm": 0.7228484850047441, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 25154 + }, + { + "epoch": 0.25155, + "grad_norm": 0.6764416676241108, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 25155 + }, + { + "epoch": 0.25156, + "grad_norm": 0.5649496554844876, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 25156 + }, + { + "epoch": 0.25157, + "grad_norm": 0.5737121785245274, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 25157 + }, + { + "epoch": 0.25158, + "grad_norm": 0.4817458287017246, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 25158 + }, + { + "epoch": 0.25159, + "grad_norm": 0.5299142977515188, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 25159 + }, + { + "epoch": 0.2516, + "grad_norm": 0.5930888642584311, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 25160 + }, + { + "epoch": 0.25161, + "grad_norm": 0.5644513245273287, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 25161 + }, + { + "epoch": 0.25162, + "grad_norm": 0.5896134721395422, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 25162 + }, + { + "epoch": 0.25163, + "grad_norm": 0.6645392045121052, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 25163 + }, + { + "epoch": 0.25164, + "grad_norm": 0.6340881670840077, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 25164 + }, + { + "epoch": 0.25165, + "grad_norm": 0.6446306452794334, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 25165 + }, + { + "epoch": 0.25166, + "grad_norm": 0.6752891643332603, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 25166 + }, + { + "epoch": 0.25167, + "grad_norm": 0.6715160052537871, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 25167 + }, + { + "epoch": 0.25168, + "grad_norm": 0.6961777228612474, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 25168 + }, + { + "epoch": 0.25169, + "grad_norm": 0.8873478918449447, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 25169 + }, + { + "epoch": 0.2517, + "grad_norm": 1.0139076453374245, + "learning_rate": 0.003, + "loss": 3.998, + "step": 25170 + }, + { + "epoch": 0.25171, + "grad_norm": 0.9220854958018314, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 25171 + }, + { + "epoch": 0.25172, + "grad_norm": 0.8319263573866729, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 25172 + }, + { + "epoch": 0.25173, + "grad_norm": 0.8418367262637235, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 25173 + }, + { + "epoch": 0.25174, + "grad_norm": 0.9402671451457574, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 25174 + }, + { + "epoch": 0.25175, + "grad_norm": 1.041129711204674, + "learning_rate": 0.003, + "loss": 4.01, + "step": 25175 + }, + { + "epoch": 0.25176, + "grad_norm": 0.9166783409759165, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 25176 + }, + { + "epoch": 0.25177, + "grad_norm": 0.9346553440016402, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 25177 + }, + { + "epoch": 0.25178, + "grad_norm": 0.9095223810945273, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 25178 + }, + { + "epoch": 0.25179, + "grad_norm": 0.8342823961676588, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 25179 + }, + { + "epoch": 0.2518, + "grad_norm": 0.7585928659676002, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 25180 + }, + { + "epoch": 0.25181, + "grad_norm": 0.8413508566889722, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 25181 + }, + { + "epoch": 0.25182, + "grad_norm": 0.8001671326595572, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 25182 + }, + { + "epoch": 0.25183, + "grad_norm": 0.7014619556409842, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 25183 + }, + { + "epoch": 0.25184, + "grad_norm": 0.6396906585206569, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 25184 + }, + { + "epoch": 0.25185, + "grad_norm": 0.5848375632507696, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 25185 + }, + { + "epoch": 0.25186, + "grad_norm": 0.612505513126313, + "learning_rate": 0.003, + "loss": 3.989, + "step": 25186 + }, + { + "epoch": 0.25187, + "grad_norm": 0.7249115481740743, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 25187 + }, + { + "epoch": 0.25188, + "grad_norm": 0.6646609425544401, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 25188 + }, + { + "epoch": 0.25189, + "grad_norm": 0.7404549608102429, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 25189 + }, + { + "epoch": 0.2519, + "grad_norm": 0.7892934493195495, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 25190 + }, + { + "epoch": 0.25191, + "grad_norm": 0.7547433054480116, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 25191 + }, + { + "epoch": 0.25192, + "grad_norm": 0.8039352914263719, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 25192 + }, + { + "epoch": 0.25193, + "grad_norm": 0.7436522457208125, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 25193 + }, + { + "epoch": 0.25194, + "grad_norm": 0.7082428507332125, + "learning_rate": 0.003, + "loss": 3.966, + "step": 25194 + }, + { + "epoch": 0.25195, + "grad_norm": 0.5327463144283092, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 25195 + }, + { + "epoch": 0.25196, + "grad_norm": 0.49417595968164674, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 25196 + }, + { + "epoch": 0.25197, + "grad_norm": 0.4834902563604595, + "learning_rate": 0.003, + "loss": 3.975, + "step": 25197 + }, + { + "epoch": 0.25198, + "grad_norm": 0.5404145809988572, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 25198 + }, + { + "epoch": 0.25199, + "grad_norm": 0.6700778584409186, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 25199 + }, + { + "epoch": 0.252, + "grad_norm": 0.7826385950486214, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 25200 + }, + { + "epoch": 0.25201, + "grad_norm": 0.8673962878159668, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 25201 + }, + { + "epoch": 0.25202, + "grad_norm": 0.8698774958340085, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 25202 + }, + { + "epoch": 0.25203, + "grad_norm": 0.734305948474466, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 25203 + }, + { + "epoch": 0.25204, + "grad_norm": 0.6596706543465524, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 25204 + }, + { + "epoch": 0.25205, + "grad_norm": 0.6560273001050969, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 25205 + }, + { + "epoch": 0.25206, + "grad_norm": 0.698500229774561, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 25206 + }, + { + "epoch": 0.25207, + "grad_norm": 0.7728648213160793, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 25207 + }, + { + "epoch": 0.25208, + "grad_norm": 0.8281596172872394, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 25208 + }, + { + "epoch": 0.25209, + "grad_norm": 0.9187662569407676, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 25209 + }, + { + "epoch": 0.2521, + "grad_norm": 0.9781527432219449, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 25210 + }, + { + "epoch": 0.25211, + "grad_norm": 0.9573994041587269, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 25211 + }, + { + "epoch": 0.25212, + "grad_norm": 0.7643241160038564, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 25212 + }, + { + "epoch": 0.25213, + "grad_norm": 0.6289950789044811, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 25213 + }, + { + "epoch": 0.25214, + "grad_norm": 0.5590531969421603, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 25214 + }, + { + "epoch": 0.25215, + "grad_norm": 0.5605798003398247, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 25215 + }, + { + "epoch": 0.25216, + "grad_norm": 0.4903228613498588, + "learning_rate": 0.003, + "loss": 4.001, + "step": 25216 + }, + { + "epoch": 0.25217, + "grad_norm": 0.5025149867111092, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 25217 + }, + { + "epoch": 0.25218, + "grad_norm": 0.6286511221399568, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 25218 + }, + { + "epoch": 0.25219, + "grad_norm": 0.7558332414123139, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 25219 + }, + { + "epoch": 0.2522, + "grad_norm": 0.7807622335656664, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 25220 + }, + { + "epoch": 0.25221, + "grad_norm": 0.7250935075046447, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 25221 + }, + { + "epoch": 0.25222, + "grad_norm": 0.8169031069003577, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 25222 + }, + { + "epoch": 0.25223, + "grad_norm": 0.7096939131712644, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 25223 + }, + { + "epoch": 0.25224, + "grad_norm": 0.6955780600105501, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 25224 + }, + { + "epoch": 0.25225, + "grad_norm": 0.8359170272996088, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 25225 + }, + { + "epoch": 0.25226, + "grad_norm": 0.9683025170821172, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 25226 + }, + { + "epoch": 0.25227, + "grad_norm": 0.9739747002145932, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 25227 + }, + { + "epoch": 0.25228, + "grad_norm": 0.9376235606620854, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 25228 + }, + { + "epoch": 0.25229, + "grad_norm": 0.9958478729219525, + "learning_rate": 0.003, + "loss": 4.009, + "step": 25229 + }, + { + "epoch": 0.2523, + "grad_norm": 1.132642263631579, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 25230 + }, + { + "epoch": 0.25231, + "grad_norm": 0.858993168241749, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 25231 + }, + { + "epoch": 0.25232, + "grad_norm": 0.8545914715526058, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 25232 + }, + { + "epoch": 0.25233, + "grad_norm": 0.9071076500241047, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 25233 + }, + { + "epoch": 0.25234, + "grad_norm": 0.9597384260018077, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 25234 + }, + { + "epoch": 0.25235, + "grad_norm": 1.0181780271907945, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 25235 + }, + { + "epoch": 0.25236, + "grad_norm": 1.029214817350396, + "learning_rate": 0.003, + "loss": 4.05, + "step": 25236 + }, + { + "epoch": 0.25237, + "grad_norm": 0.8697735414944162, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 25237 + }, + { + "epoch": 0.25238, + "grad_norm": 0.6680952513846449, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 25238 + }, + { + "epoch": 0.25239, + "grad_norm": 0.7636531788845267, + "learning_rate": 0.003, + "loss": 3.999, + "step": 25239 + }, + { + "epoch": 0.2524, + "grad_norm": 0.9350175866905628, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 25240 + }, + { + "epoch": 0.25241, + "grad_norm": 1.0878594655695313, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 25241 + }, + { + "epoch": 0.25242, + "grad_norm": 0.8456444775147571, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 25242 + }, + { + "epoch": 0.25243, + "grad_norm": 0.8017390806244562, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 25243 + }, + { + "epoch": 0.25244, + "grad_norm": 0.8515905375919618, + "learning_rate": 0.003, + "loss": 4.008, + "step": 25244 + }, + { + "epoch": 0.25245, + "grad_norm": 0.6985969117247613, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 25245 + }, + { + "epoch": 0.25246, + "grad_norm": 0.5437843380769579, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 25246 + }, + { + "epoch": 0.25247, + "grad_norm": 0.5641215237084071, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 25247 + }, + { + "epoch": 0.25248, + "grad_norm": 0.49755750420689726, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 25248 + }, + { + "epoch": 0.25249, + "grad_norm": 0.45293347001877643, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 25249 + }, + { + "epoch": 0.2525, + "grad_norm": 0.42206769897343843, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 25250 + }, + { + "epoch": 0.25251, + "grad_norm": 0.5082603661058938, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 25251 + }, + { + "epoch": 0.25252, + "grad_norm": 0.5617489726469287, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 25252 + }, + { + "epoch": 0.25253, + "grad_norm": 0.5971396984053922, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 25253 + }, + { + "epoch": 0.25254, + "grad_norm": 0.5559537834699254, + "learning_rate": 0.003, + "loss": 3.967, + "step": 25254 + }, + { + "epoch": 0.25255, + "grad_norm": 0.5547389434125428, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 25255 + }, + { + "epoch": 0.25256, + "grad_norm": 0.5579141403941402, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 25256 + }, + { + "epoch": 0.25257, + "grad_norm": 0.7091991243471873, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 25257 + }, + { + "epoch": 0.25258, + "grad_norm": 0.9798413965582925, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 25258 + }, + { + "epoch": 0.25259, + "grad_norm": 1.0207806467982317, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 25259 + }, + { + "epoch": 0.2526, + "grad_norm": 0.646707441813099, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 25260 + }, + { + "epoch": 0.25261, + "grad_norm": 0.7172948573708702, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 25261 + }, + { + "epoch": 0.25262, + "grad_norm": 0.9275699834762066, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 25262 + }, + { + "epoch": 0.25263, + "grad_norm": 0.8846040715374622, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 25263 + }, + { + "epoch": 0.25264, + "grad_norm": 0.6654294061768087, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 25264 + }, + { + "epoch": 0.25265, + "grad_norm": 0.6535369036556136, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 25265 + }, + { + "epoch": 0.25266, + "grad_norm": 0.6774917311122981, + "learning_rate": 0.003, + "loss": 3.996, + "step": 25266 + }, + { + "epoch": 0.25267, + "grad_norm": 0.6903506074770176, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 25267 + }, + { + "epoch": 0.25268, + "grad_norm": 0.8060157413203931, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 25268 + }, + { + "epoch": 0.25269, + "grad_norm": 0.9245316702393009, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 25269 + }, + { + "epoch": 0.2527, + "grad_norm": 0.9414651722829217, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 25270 + }, + { + "epoch": 0.25271, + "grad_norm": 0.8123232248485971, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 25271 + }, + { + "epoch": 0.25272, + "grad_norm": 0.699483569575051, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 25272 + }, + { + "epoch": 0.25273, + "grad_norm": 0.6859505330682303, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 25273 + }, + { + "epoch": 0.25274, + "grad_norm": 0.7004237479311264, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 25274 + }, + { + "epoch": 0.25275, + "grad_norm": 0.7042216069804997, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 25275 + }, + { + "epoch": 0.25276, + "grad_norm": 0.6176523389292038, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 25276 + }, + { + "epoch": 0.25277, + "grad_norm": 0.562136589127288, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 25277 + }, + { + "epoch": 0.25278, + "grad_norm": 0.5684012546478096, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 25278 + }, + { + "epoch": 0.25279, + "grad_norm": 0.6255285837225243, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 25279 + }, + { + "epoch": 0.2528, + "grad_norm": 0.6773332850465046, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 25280 + }, + { + "epoch": 0.25281, + "grad_norm": 0.8082400509658089, + "learning_rate": 0.003, + "loss": 4.012, + "step": 25281 + }, + { + "epoch": 0.25282, + "grad_norm": 0.7966530880620241, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 25282 + }, + { + "epoch": 0.25283, + "grad_norm": 0.8733708830241373, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 25283 + }, + { + "epoch": 0.25284, + "grad_norm": 0.8251677848915099, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 25284 + }, + { + "epoch": 0.25285, + "grad_norm": 0.8629376909372586, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 25285 + }, + { + "epoch": 0.25286, + "grad_norm": 0.8553532129640854, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 25286 + }, + { + "epoch": 0.25287, + "grad_norm": 0.9057808828572803, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 25287 + }, + { + "epoch": 0.25288, + "grad_norm": 0.8963894895680895, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 25288 + }, + { + "epoch": 0.25289, + "grad_norm": 0.8458271412081274, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 25289 + }, + { + "epoch": 0.2529, + "grad_norm": 0.700425695828584, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 25290 + }, + { + "epoch": 0.25291, + "grad_norm": 0.8119432034979246, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 25291 + }, + { + "epoch": 0.25292, + "grad_norm": 0.8732507996537779, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 25292 + }, + { + "epoch": 0.25293, + "grad_norm": 0.9893826772839385, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 25293 + }, + { + "epoch": 0.25294, + "grad_norm": 1.0065984129505579, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 25294 + }, + { + "epoch": 0.25295, + "grad_norm": 0.9473659633561528, + "learning_rate": 0.003, + "loss": 3.981, + "step": 25295 + }, + { + "epoch": 0.25296, + "grad_norm": 0.9055327000933188, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 25296 + }, + { + "epoch": 0.25297, + "grad_norm": 0.7450611410875354, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 25297 + }, + { + "epoch": 0.25298, + "grad_norm": 0.6358796353401356, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 25298 + }, + { + "epoch": 0.25299, + "grad_norm": 0.6292501024552208, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 25299 + }, + { + "epoch": 0.253, + "grad_norm": 0.6888370638635901, + "learning_rate": 0.003, + "loss": 4.005, + "step": 25300 + }, + { + "epoch": 0.25301, + "grad_norm": 0.7726977472061521, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 25301 + }, + { + "epoch": 0.25302, + "grad_norm": 0.878288231243068, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 25302 + }, + { + "epoch": 0.25303, + "grad_norm": 0.8532255555554215, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 25303 + }, + { + "epoch": 0.25304, + "grad_norm": 0.8438612382858182, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 25304 + }, + { + "epoch": 0.25305, + "grad_norm": 0.7162722834568511, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 25305 + }, + { + "epoch": 0.25306, + "grad_norm": 0.6813924851174841, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 25306 + }, + { + "epoch": 0.25307, + "grad_norm": 0.6497735644108534, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 25307 + }, + { + "epoch": 0.25308, + "grad_norm": 0.6410835264347201, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 25308 + }, + { + "epoch": 0.25309, + "grad_norm": 0.7735360514639792, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 25309 + }, + { + "epoch": 0.2531, + "grad_norm": 0.8995561534450598, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 25310 + }, + { + "epoch": 0.25311, + "grad_norm": 0.9527481873232487, + "learning_rate": 0.003, + "loss": 4.011, + "step": 25311 + }, + { + "epoch": 0.25312, + "grad_norm": 0.8875153445502881, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 25312 + }, + { + "epoch": 0.25313, + "grad_norm": 0.7346393004989178, + "learning_rate": 0.003, + "loss": 4.02, + "step": 25313 + }, + { + "epoch": 0.25314, + "grad_norm": 0.842380285204154, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 25314 + }, + { + "epoch": 0.25315, + "grad_norm": 0.8512092370159562, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 25315 + }, + { + "epoch": 0.25316, + "grad_norm": 0.8009864342976616, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 25316 + }, + { + "epoch": 0.25317, + "grad_norm": 0.7946971356147469, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 25317 + }, + { + "epoch": 0.25318, + "grad_norm": 0.7945501021132003, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 25318 + }, + { + "epoch": 0.25319, + "grad_norm": 0.7015811579215454, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 25319 + }, + { + "epoch": 0.2532, + "grad_norm": 0.6132145877343083, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 25320 + }, + { + "epoch": 0.25321, + "grad_norm": 0.6331164652561513, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 25321 + }, + { + "epoch": 0.25322, + "grad_norm": 0.6184013136005434, + "learning_rate": 0.003, + "loss": 4.01, + "step": 25322 + }, + { + "epoch": 0.25323, + "grad_norm": 0.5959546731514845, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 25323 + }, + { + "epoch": 0.25324, + "grad_norm": 0.607284248416114, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 25324 + }, + { + "epoch": 0.25325, + "grad_norm": 0.6480387004401679, + "learning_rate": 0.003, + "loss": 3.983, + "step": 25325 + }, + { + "epoch": 0.25326, + "grad_norm": 0.7330648215491412, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 25326 + }, + { + "epoch": 0.25327, + "grad_norm": 0.7989235944060188, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 25327 + }, + { + "epoch": 0.25328, + "grad_norm": 0.816186442622755, + "learning_rate": 0.003, + "loss": 4.001, + "step": 25328 + }, + { + "epoch": 0.25329, + "grad_norm": 0.7874245143181086, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 25329 + }, + { + "epoch": 0.2533, + "grad_norm": 0.7453935035939373, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 25330 + }, + { + "epoch": 0.25331, + "grad_norm": 0.6839246179945869, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 25331 + }, + { + "epoch": 0.25332, + "grad_norm": 0.6967238715455218, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 25332 + }, + { + "epoch": 0.25333, + "grad_norm": 0.724163946072045, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 25333 + }, + { + "epoch": 0.25334, + "grad_norm": 0.6277495048234754, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 25334 + }, + { + "epoch": 0.25335, + "grad_norm": 0.6921434811703087, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 25335 + }, + { + "epoch": 0.25336, + "grad_norm": 0.7112724631757505, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 25336 + }, + { + "epoch": 0.25337, + "grad_norm": 0.7805415805857253, + "learning_rate": 0.003, + "loss": 4.011, + "step": 25337 + }, + { + "epoch": 0.25338, + "grad_norm": 0.8856070195815144, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 25338 + }, + { + "epoch": 0.25339, + "grad_norm": 1.0267223891373098, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 25339 + }, + { + "epoch": 0.2534, + "grad_norm": 0.8600261988717706, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 25340 + }, + { + "epoch": 0.25341, + "grad_norm": 0.7682525326603936, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 25341 + }, + { + "epoch": 0.25342, + "grad_norm": 0.7353144245458141, + "learning_rate": 0.003, + "loss": 3.983, + "step": 25342 + }, + { + "epoch": 0.25343, + "grad_norm": 0.6962000739196281, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 25343 + }, + { + "epoch": 0.25344, + "grad_norm": 0.720962317600038, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 25344 + }, + { + "epoch": 0.25345, + "grad_norm": 0.7674135311430074, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 25345 + }, + { + "epoch": 0.25346, + "grad_norm": 0.8697654597695212, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 25346 + }, + { + "epoch": 0.25347, + "grad_norm": 1.0482735155523815, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 25347 + }, + { + "epoch": 0.25348, + "grad_norm": 1.0506186950476684, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 25348 + }, + { + "epoch": 0.25349, + "grad_norm": 0.7395006444791304, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 25349 + }, + { + "epoch": 0.2535, + "grad_norm": 0.7666810244336393, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 25350 + }, + { + "epoch": 0.25351, + "grad_norm": 0.7462297876931889, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 25351 + }, + { + "epoch": 0.25352, + "grad_norm": 0.7053593629723057, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 25352 + }, + { + "epoch": 0.25353, + "grad_norm": 0.6915568979132151, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 25353 + }, + { + "epoch": 0.25354, + "grad_norm": 0.6877005695529735, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 25354 + }, + { + "epoch": 0.25355, + "grad_norm": 0.6630703447549963, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 25355 + }, + { + "epoch": 0.25356, + "grad_norm": 0.6330092572356901, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 25356 + }, + { + "epoch": 0.25357, + "grad_norm": 0.6255540475784858, + "learning_rate": 0.003, + "loss": 3.981, + "step": 25357 + }, + { + "epoch": 0.25358, + "grad_norm": 0.6572895135166179, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 25358 + }, + { + "epoch": 0.25359, + "grad_norm": 0.5649320150276217, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 25359 + }, + { + "epoch": 0.2536, + "grad_norm": 0.5610247303514114, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 25360 + }, + { + "epoch": 0.25361, + "grad_norm": 0.6249619953272241, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 25361 + }, + { + "epoch": 0.25362, + "grad_norm": 0.7924133261837428, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 25362 + }, + { + "epoch": 0.25363, + "grad_norm": 0.9354280412675385, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 25363 + }, + { + "epoch": 0.25364, + "grad_norm": 0.9484066835267222, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 25364 + }, + { + "epoch": 0.25365, + "grad_norm": 0.795122133112041, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 25365 + }, + { + "epoch": 0.25366, + "grad_norm": 0.6011980509162627, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 25366 + }, + { + "epoch": 0.25367, + "grad_norm": 0.68005711447305, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 25367 + }, + { + "epoch": 0.25368, + "grad_norm": 0.8291733640311645, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 25368 + }, + { + "epoch": 0.25369, + "grad_norm": 0.8809454454872463, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 25369 + }, + { + "epoch": 0.2537, + "grad_norm": 0.6900875519842026, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 25370 + }, + { + "epoch": 0.25371, + "grad_norm": 0.6202220023782614, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 25371 + }, + { + "epoch": 0.25372, + "grad_norm": 0.6540304055386289, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 25372 + }, + { + "epoch": 0.25373, + "grad_norm": 0.8007273050517948, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 25373 + }, + { + "epoch": 0.25374, + "grad_norm": 0.8660368599346974, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 25374 + }, + { + "epoch": 0.25375, + "grad_norm": 0.8280854735930749, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 25375 + }, + { + "epoch": 0.25376, + "grad_norm": 0.797279825534588, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 25376 + }, + { + "epoch": 0.25377, + "grad_norm": 0.7967232340517939, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 25377 + }, + { + "epoch": 0.25378, + "grad_norm": 0.7284700578500022, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 25378 + }, + { + "epoch": 0.25379, + "grad_norm": 0.7588754970133204, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 25379 + }, + { + "epoch": 0.2538, + "grad_norm": 0.7252887840680627, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 25380 + }, + { + "epoch": 0.25381, + "grad_norm": 0.762259282110901, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 25381 + }, + { + "epoch": 0.25382, + "grad_norm": 0.7801283293022062, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 25382 + }, + { + "epoch": 0.25383, + "grad_norm": 0.8188309859533837, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 25383 + }, + { + "epoch": 0.25384, + "grad_norm": 0.9264673956599471, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 25384 + }, + { + "epoch": 0.25385, + "grad_norm": 1.0004439950100397, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 25385 + }, + { + "epoch": 0.25386, + "grad_norm": 0.9540214051655357, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 25386 + }, + { + "epoch": 0.25387, + "grad_norm": 0.8939143384308285, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 25387 + }, + { + "epoch": 0.25388, + "grad_norm": 0.8745174540516394, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 25388 + }, + { + "epoch": 0.25389, + "grad_norm": 0.8211626932380731, + "learning_rate": 0.003, + "loss": 4.002, + "step": 25389 + }, + { + "epoch": 0.2539, + "grad_norm": 0.7228890948175118, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 25390 + }, + { + "epoch": 0.25391, + "grad_norm": 0.8071488278132771, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 25391 + }, + { + "epoch": 0.25392, + "grad_norm": 0.8297584948888754, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 25392 + }, + { + "epoch": 0.25393, + "grad_norm": 0.7557703686572788, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 25393 + }, + { + "epoch": 0.25394, + "grad_norm": 0.7497556987882084, + "learning_rate": 0.003, + "loss": 4.0522, + "step": 25394 + }, + { + "epoch": 0.25395, + "grad_norm": 0.825984902163014, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 25395 + }, + { + "epoch": 0.25396, + "grad_norm": 0.702598175493362, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 25396 + }, + { + "epoch": 0.25397, + "grad_norm": 0.6682439511923435, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 25397 + }, + { + "epoch": 0.25398, + "grad_norm": 0.5379474766154301, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 25398 + }, + { + "epoch": 0.25399, + "grad_norm": 0.5659380637203866, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 25399 + }, + { + "epoch": 0.254, + "grad_norm": 0.5632960849094134, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 25400 + }, + { + "epoch": 0.25401, + "grad_norm": 0.6478730807030587, + "learning_rate": 0.003, + "loss": 3.978, + "step": 25401 + }, + { + "epoch": 0.25402, + "grad_norm": 0.7939120382494839, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 25402 + }, + { + "epoch": 0.25403, + "grad_norm": 0.9251508470074107, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 25403 + }, + { + "epoch": 0.25404, + "grad_norm": 0.9481670305603259, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 25404 + }, + { + "epoch": 0.25405, + "grad_norm": 0.683233380186595, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 25405 + }, + { + "epoch": 0.25406, + "grad_norm": 0.5778817964940394, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 25406 + }, + { + "epoch": 0.25407, + "grad_norm": 0.6800666193640319, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 25407 + }, + { + "epoch": 0.25408, + "grad_norm": 0.7538229210406681, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 25408 + }, + { + "epoch": 0.25409, + "grad_norm": 0.6918692786915607, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 25409 + }, + { + "epoch": 0.2541, + "grad_norm": 0.6678708250106287, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 25410 + }, + { + "epoch": 0.25411, + "grad_norm": 0.631036464899064, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 25411 + }, + { + "epoch": 0.25412, + "grad_norm": 0.6140756203177679, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 25412 + }, + { + "epoch": 0.25413, + "grad_norm": 0.6314836422885071, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 25413 + }, + { + "epoch": 0.25414, + "grad_norm": 0.6221996626909133, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 25414 + }, + { + "epoch": 0.25415, + "grad_norm": 0.6064627264497836, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 25415 + }, + { + "epoch": 0.25416, + "grad_norm": 0.5901923248442471, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 25416 + }, + { + "epoch": 0.25417, + "grad_norm": 0.6355930062275439, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 25417 + }, + { + "epoch": 0.25418, + "grad_norm": 0.6877923596317786, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 25418 + }, + { + "epoch": 0.25419, + "grad_norm": 0.7582318410498923, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 25419 + }, + { + "epoch": 0.2542, + "grad_norm": 0.8551451213950809, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 25420 + }, + { + "epoch": 0.25421, + "grad_norm": 1.1084175405153194, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 25421 + }, + { + "epoch": 0.25422, + "grad_norm": 0.8803293529947149, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 25422 + }, + { + "epoch": 0.25423, + "grad_norm": 0.8486593296291137, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 25423 + }, + { + "epoch": 0.25424, + "grad_norm": 0.8823865696099047, + "learning_rate": 0.003, + "loss": 4.022, + "step": 25424 + }, + { + "epoch": 0.25425, + "grad_norm": 0.9059286954690068, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 25425 + }, + { + "epoch": 0.25426, + "grad_norm": 0.8435187315431898, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 25426 + }, + { + "epoch": 0.25427, + "grad_norm": 0.7518367627989422, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 25427 + }, + { + "epoch": 0.25428, + "grad_norm": 0.9520457764649537, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 25428 + }, + { + "epoch": 0.25429, + "grad_norm": 1.2727804265651053, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 25429 + }, + { + "epoch": 0.2543, + "grad_norm": 0.8014596497207541, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 25430 + }, + { + "epoch": 0.25431, + "grad_norm": 0.8571694054192753, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 25431 + }, + { + "epoch": 0.25432, + "grad_norm": 0.8550619350700238, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 25432 + }, + { + "epoch": 0.25433, + "grad_norm": 0.9253148797259186, + "learning_rate": 0.003, + "loss": 3.988, + "step": 25433 + }, + { + "epoch": 0.25434, + "grad_norm": 0.7946492806548523, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 25434 + }, + { + "epoch": 0.25435, + "grad_norm": 0.7045461458452709, + "learning_rate": 0.003, + "loss": 4.01, + "step": 25435 + }, + { + "epoch": 0.25436, + "grad_norm": 0.7113876237402328, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 25436 + }, + { + "epoch": 0.25437, + "grad_norm": 0.7300058154392313, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 25437 + }, + { + "epoch": 0.25438, + "grad_norm": 0.7205684297112328, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 25438 + }, + { + "epoch": 0.25439, + "grad_norm": 0.64180388972255, + "learning_rate": 0.003, + "loss": 3.996, + "step": 25439 + }, + { + "epoch": 0.2544, + "grad_norm": 0.6122011608998068, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 25440 + }, + { + "epoch": 0.25441, + "grad_norm": 0.653036481328176, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 25441 + }, + { + "epoch": 0.25442, + "grad_norm": 0.6949409385006339, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 25442 + }, + { + "epoch": 0.25443, + "grad_norm": 0.7355613731773846, + "learning_rate": 0.003, + "loss": 3.981, + "step": 25443 + }, + { + "epoch": 0.25444, + "grad_norm": 0.7196274771514639, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 25444 + }, + { + "epoch": 0.25445, + "grad_norm": 0.8239078643263902, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 25445 + }, + { + "epoch": 0.25446, + "grad_norm": 0.8992745884515145, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 25446 + }, + { + "epoch": 0.25447, + "grad_norm": 0.8647498622308059, + "learning_rate": 0.003, + "loss": 4.038, + "step": 25447 + }, + { + "epoch": 0.25448, + "grad_norm": 0.8399069841573386, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 25448 + }, + { + "epoch": 0.25449, + "grad_norm": 0.8262573820591249, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 25449 + }, + { + "epoch": 0.2545, + "grad_norm": 0.7820672624500212, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 25450 + }, + { + "epoch": 0.25451, + "grad_norm": 0.7828642992099566, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 25451 + }, + { + "epoch": 0.25452, + "grad_norm": 0.7868427425728068, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 25452 + }, + { + "epoch": 0.25453, + "grad_norm": 0.8127766439408832, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 25453 + }, + { + "epoch": 0.25454, + "grad_norm": 0.8589398441548248, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 25454 + }, + { + "epoch": 0.25455, + "grad_norm": 1.0262336057170587, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 25455 + }, + { + "epoch": 0.25456, + "grad_norm": 1.0909536381016542, + "learning_rate": 0.003, + "loss": 3.997, + "step": 25456 + }, + { + "epoch": 0.25457, + "grad_norm": 0.7383604524573226, + "learning_rate": 0.003, + "loss": 3.988, + "step": 25457 + }, + { + "epoch": 0.25458, + "grad_norm": 0.679684898845004, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 25458 + }, + { + "epoch": 0.25459, + "grad_norm": 0.639993879646903, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 25459 + }, + { + "epoch": 0.2546, + "grad_norm": 0.6311526581285544, + "learning_rate": 0.003, + "loss": 4.004, + "step": 25460 + }, + { + "epoch": 0.25461, + "grad_norm": 0.6900554199161002, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 25461 + }, + { + "epoch": 0.25462, + "grad_norm": 0.6792693337795486, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 25462 + }, + { + "epoch": 0.25463, + "grad_norm": 0.6415789056247584, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 25463 + }, + { + "epoch": 0.25464, + "grad_norm": 0.5882673988388264, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 25464 + }, + { + "epoch": 0.25465, + "grad_norm": 0.581292291247017, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 25465 + }, + { + "epoch": 0.25466, + "grad_norm": 0.5559904697210722, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 25466 + }, + { + "epoch": 0.25467, + "grad_norm": 0.5887623484245418, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 25467 + }, + { + "epoch": 0.25468, + "grad_norm": 0.686131070275882, + "learning_rate": 0.003, + "loss": 3.978, + "step": 25468 + }, + { + "epoch": 0.25469, + "grad_norm": 0.7926311194667497, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 25469 + }, + { + "epoch": 0.2547, + "grad_norm": 0.9047237248488333, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 25470 + }, + { + "epoch": 0.25471, + "grad_norm": 0.8182794518684563, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 25471 + }, + { + "epoch": 0.25472, + "grad_norm": 0.7607802129048049, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 25472 + }, + { + "epoch": 0.25473, + "grad_norm": 0.6869341056984961, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 25473 + }, + { + "epoch": 0.25474, + "grad_norm": 0.6531393073899648, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 25474 + }, + { + "epoch": 0.25475, + "grad_norm": 0.6726353699236546, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 25475 + }, + { + "epoch": 0.25476, + "grad_norm": 0.725520630983986, + "learning_rate": 0.003, + "loss": 3.992, + "step": 25476 + }, + { + "epoch": 0.25477, + "grad_norm": 0.8612155034653864, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 25477 + }, + { + "epoch": 0.25478, + "grad_norm": 0.7891473250090933, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 25478 + }, + { + "epoch": 0.25479, + "grad_norm": 0.81586993888372, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 25479 + }, + { + "epoch": 0.2548, + "grad_norm": 0.8245360847362493, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 25480 + }, + { + "epoch": 0.25481, + "grad_norm": 0.9820164545172112, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 25481 + }, + { + "epoch": 0.25482, + "grad_norm": 0.8596778025387962, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 25482 + }, + { + "epoch": 0.25483, + "grad_norm": 0.9283014978835801, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 25483 + }, + { + "epoch": 0.25484, + "grad_norm": 0.960816464282305, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 25484 + }, + { + "epoch": 0.25485, + "grad_norm": 0.9574305422239849, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 25485 + }, + { + "epoch": 0.25486, + "grad_norm": 1.0181855054484386, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 25486 + }, + { + "epoch": 0.25487, + "grad_norm": 0.9268995614292281, + "learning_rate": 0.003, + "loss": 4.02, + "step": 25487 + }, + { + "epoch": 0.25488, + "grad_norm": 0.8384611720953153, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 25488 + }, + { + "epoch": 0.25489, + "grad_norm": 0.8690472734533232, + "learning_rate": 0.003, + "loss": 4.019, + "step": 25489 + }, + { + "epoch": 0.2549, + "grad_norm": 0.6510916130046333, + "learning_rate": 0.003, + "loss": 4.042, + "step": 25490 + }, + { + "epoch": 0.25491, + "grad_norm": 0.608674357883734, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 25491 + }, + { + "epoch": 0.25492, + "grad_norm": 0.6221709775960125, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 25492 + }, + { + "epoch": 0.25493, + "grad_norm": 0.6858281641494856, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 25493 + }, + { + "epoch": 0.25494, + "grad_norm": 0.816472247173102, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 25494 + }, + { + "epoch": 0.25495, + "grad_norm": 0.9962459920316992, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 25495 + }, + { + "epoch": 0.25496, + "grad_norm": 1.0145574708863896, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 25496 + }, + { + "epoch": 0.25497, + "grad_norm": 0.8046488506185994, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 25497 + }, + { + "epoch": 0.25498, + "grad_norm": 0.7816045931497415, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 25498 + }, + { + "epoch": 0.25499, + "grad_norm": 0.850502161672614, + "learning_rate": 0.003, + "loss": 3.998, + "step": 25499 + }, + { + "epoch": 0.255, + "grad_norm": 0.9177939895306166, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 25500 + }, + { + "epoch": 0.25501, + "grad_norm": 0.9400804580523009, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 25501 + }, + { + "epoch": 0.25502, + "grad_norm": 0.850266268895049, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 25502 + }, + { + "epoch": 0.25503, + "grad_norm": 0.7009848440104431, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 25503 + }, + { + "epoch": 0.25504, + "grad_norm": 0.6487089854547475, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 25504 + }, + { + "epoch": 0.25505, + "grad_norm": 0.6059595750393046, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 25505 + }, + { + "epoch": 0.25506, + "grad_norm": 0.6750065457786308, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 25506 + }, + { + "epoch": 0.25507, + "grad_norm": 0.7563864501325127, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 25507 + }, + { + "epoch": 0.25508, + "grad_norm": 0.8942304244639986, + "learning_rate": 0.003, + "loss": 4.0515, + "step": 25508 + }, + { + "epoch": 0.25509, + "grad_norm": 0.8512316250747243, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 25509 + }, + { + "epoch": 0.2551, + "grad_norm": 1.0140762782144968, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 25510 + }, + { + "epoch": 0.25511, + "grad_norm": 0.8741934019828556, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 25511 + }, + { + "epoch": 0.25512, + "grad_norm": 0.7254176452030355, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 25512 + }, + { + "epoch": 0.25513, + "grad_norm": 0.6626025218473678, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 25513 + }, + { + "epoch": 0.25514, + "grad_norm": 0.5512204285454552, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 25514 + }, + { + "epoch": 0.25515, + "grad_norm": 0.5578171860381422, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 25515 + }, + { + "epoch": 0.25516, + "grad_norm": 0.6045827019013809, + "learning_rate": 0.003, + "loss": 4.002, + "step": 25516 + }, + { + "epoch": 0.25517, + "grad_norm": 0.6002958041733738, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 25517 + }, + { + "epoch": 0.25518, + "grad_norm": 0.6889589382498947, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 25518 + }, + { + "epoch": 0.25519, + "grad_norm": 0.7112281667974483, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 25519 + }, + { + "epoch": 0.2552, + "grad_norm": 0.7437067931458179, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 25520 + }, + { + "epoch": 0.25521, + "grad_norm": 0.8735846873353997, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 25521 + }, + { + "epoch": 0.25522, + "grad_norm": 0.9551977802938961, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 25522 + }, + { + "epoch": 0.25523, + "grad_norm": 0.8208708151476047, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 25523 + }, + { + "epoch": 0.25524, + "grad_norm": 0.6367796658153898, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 25524 + }, + { + "epoch": 0.25525, + "grad_norm": 0.6241505590054908, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 25525 + }, + { + "epoch": 0.25526, + "grad_norm": 0.710597383749558, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 25526 + }, + { + "epoch": 0.25527, + "grad_norm": 0.7503062332205629, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 25527 + }, + { + "epoch": 0.25528, + "grad_norm": 0.7369822250222329, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 25528 + }, + { + "epoch": 0.25529, + "grad_norm": 0.6778415913173695, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 25529 + }, + { + "epoch": 0.2553, + "grad_norm": 0.6192959321630241, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 25530 + }, + { + "epoch": 0.25531, + "grad_norm": 0.5434113924250463, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 25531 + }, + { + "epoch": 0.25532, + "grad_norm": 0.4944039025256062, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 25532 + }, + { + "epoch": 0.25533, + "grad_norm": 0.5876296561681366, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 25533 + }, + { + "epoch": 0.25534, + "grad_norm": 0.5985119131318761, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 25534 + }, + { + "epoch": 0.25535, + "grad_norm": 0.5866652948868764, + "learning_rate": 0.003, + "loss": 3.96, + "step": 25535 + }, + { + "epoch": 0.25536, + "grad_norm": 0.5694498330555057, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 25536 + }, + { + "epoch": 0.25537, + "grad_norm": 0.6850278369708546, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 25537 + }, + { + "epoch": 0.25538, + "grad_norm": 0.7325277254907513, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 25538 + }, + { + "epoch": 0.25539, + "grad_norm": 0.727113516196007, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 25539 + }, + { + "epoch": 0.2554, + "grad_norm": 0.7149228003920488, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 25540 + }, + { + "epoch": 0.25541, + "grad_norm": 0.7375516332286035, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 25541 + }, + { + "epoch": 0.25542, + "grad_norm": 0.8260539903368033, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 25542 + }, + { + "epoch": 0.25543, + "grad_norm": 0.7407986153580236, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 25543 + }, + { + "epoch": 0.25544, + "grad_norm": 0.5494781017819345, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 25544 + }, + { + "epoch": 0.25545, + "grad_norm": 0.6157053447253684, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 25545 + }, + { + "epoch": 0.25546, + "grad_norm": 0.601146076171225, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 25546 + }, + { + "epoch": 0.25547, + "grad_norm": 0.5978646452585052, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 25547 + }, + { + "epoch": 0.25548, + "grad_norm": 0.5921810367988009, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 25548 + }, + { + "epoch": 0.25549, + "grad_norm": 0.6504251264412593, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 25549 + }, + { + "epoch": 0.2555, + "grad_norm": 0.7149852758180801, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 25550 + }, + { + "epoch": 0.25551, + "grad_norm": 0.7329710126802479, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 25551 + }, + { + "epoch": 0.25552, + "grad_norm": 0.8664551308024498, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 25552 + }, + { + "epoch": 0.25553, + "grad_norm": 1.1170184671593595, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 25553 + }, + { + "epoch": 0.25554, + "grad_norm": 0.8864175783140392, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 25554 + }, + { + "epoch": 0.25555, + "grad_norm": 0.9488996900616827, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 25555 + }, + { + "epoch": 0.25556, + "grad_norm": 1.1122646085807328, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 25556 + }, + { + "epoch": 0.25557, + "grad_norm": 0.9759352302037101, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 25557 + }, + { + "epoch": 0.25558, + "grad_norm": 0.8614054274418461, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 25558 + }, + { + "epoch": 0.25559, + "grad_norm": 0.7555091926293641, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 25559 + }, + { + "epoch": 0.2556, + "grad_norm": 0.7185455620467491, + "learning_rate": 0.003, + "loss": 3.978, + "step": 25560 + }, + { + "epoch": 0.25561, + "grad_norm": 0.6612831346640872, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 25561 + }, + { + "epoch": 0.25562, + "grad_norm": 0.6233467047977633, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 25562 + }, + { + "epoch": 0.25563, + "grad_norm": 0.5925955716633008, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 25563 + }, + { + "epoch": 0.25564, + "grad_norm": 0.5932790201356751, + "learning_rate": 0.003, + "loss": 3.995, + "step": 25564 + }, + { + "epoch": 0.25565, + "grad_norm": 0.6347760602746424, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 25565 + }, + { + "epoch": 0.25566, + "grad_norm": 0.7546863958574103, + "learning_rate": 0.003, + "loss": 4.009, + "step": 25566 + }, + { + "epoch": 0.25567, + "grad_norm": 0.8971728228652533, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 25567 + }, + { + "epoch": 0.25568, + "grad_norm": 1.0649328465671695, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 25568 + }, + { + "epoch": 0.25569, + "grad_norm": 1.0116014275301923, + "learning_rate": 0.003, + "loss": 4.0447, + "step": 25569 + }, + { + "epoch": 0.2557, + "grad_norm": 1.0013557249719502, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 25570 + }, + { + "epoch": 0.25571, + "grad_norm": 0.9786564546517564, + "learning_rate": 0.003, + "loss": 4.035, + "step": 25571 + }, + { + "epoch": 0.25572, + "grad_norm": 0.9331727417845722, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 25572 + }, + { + "epoch": 0.25573, + "grad_norm": 0.8860148385712391, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 25573 + }, + { + "epoch": 0.25574, + "grad_norm": 0.9400125428627063, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 25574 + }, + { + "epoch": 0.25575, + "grad_norm": 0.9405442558772052, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 25575 + }, + { + "epoch": 0.25576, + "grad_norm": 0.8264687787900928, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 25576 + }, + { + "epoch": 0.25577, + "grad_norm": 0.8301032621809606, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 25577 + }, + { + "epoch": 0.25578, + "grad_norm": 0.8885937023543972, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 25578 + }, + { + "epoch": 0.25579, + "grad_norm": 0.8861942460408078, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 25579 + }, + { + "epoch": 0.2558, + "grad_norm": 0.8717229483108176, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 25580 + }, + { + "epoch": 0.25581, + "grad_norm": 0.7610206889156189, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 25581 + }, + { + "epoch": 0.25582, + "grad_norm": 0.7463426156614087, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 25582 + }, + { + "epoch": 0.25583, + "grad_norm": 0.7437114372574096, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 25583 + }, + { + "epoch": 0.25584, + "grad_norm": 0.6402546001562329, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 25584 + }, + { + "epoch": 0.25585, + "grad_norm": 0.625800996832183, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 25585 + }, + { + "epoch": 0.25586, + "grad_norm": 0.5651262972683494, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 25586 + }, + { + "epoch": 0.25587, + "grad_norm": 0.6792235187137988, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 25587 + }, + { + "epoch": 0.25588, + "grad_norm": 0.7029509784449727, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 25588 + }, + { + "epoch": 0.25589, + "grad_norm": 0.699652086117887, + "learning_rate": 0.003, + "loss": 3.986, + "step": 25589 + }, + { + "epoch": 0.2559, + "grad_norm": 0.6737034672116039, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 25590 + }, + { + "epoch": 0.25591, + "grad_norm": 0.6879343970598786, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 25591 + }, + { + "epoch": 0.25592, + "grad_norm": 0.6529887611116654, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 25592 + }, + { + "epoch": 0.25593, + "grad_norm": 0.7799969482381576, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 25593 + }, + { + "epoch": 0.25594, + "grad_norm": 0.9401810085751409, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 25594 + }, + { + "epoch": 0.25595, + "grad_norm": 1.1184328583136895, + "learning_rate": 0.003, + "loss": 4.005, + "step": 25595 + }, + { + "epoch": 0.25596, + "grad_norm": 0.8488956207200474, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 25596 + }, + { + "epoch": 0.25597, + "grad_norm": 0.7351101296849926, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 25597 + }, + { + "epoch": 0.25598, + "grad_norm": 0.868674979072741, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 25598 + }, + { + "epoch": 0.25599, + "grad_norm": 0.7097398227369804, + "learning_rate": 0.003, + "loss": 4.01, + "step": 25599 + }, + { + "epoch": 0.256, + "grad_norm": 0.6536306394474336, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 25600 + }, + { + "epoch": 0.25601, + "grad_norm": 0.6030798015840327, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 25601 + }, + { + "epoch": 0.25602, + "grad_norm": 0.5166314861905847, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 25602 + }, + { + "epoch": 0.25603, + "grad_norm": 0.4754017744551712, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 25603 + }, + { + "epoch": 0.25604, + "grad_norm": 0.5078244771442668, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 25604 + }, + { + "epoch": 0.25605, + "grad_norm": 0.5689672298129599, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 25605 + }, + { + "epoch": 0.25606, + "grad_norm": 0.5798913308070845, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 25606 + }, + { + "epoch": 0.25607, + "grad_norm": 0.6351001568808357, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 25607 + }, + { + "epoch": 0.25608, + "grad_norm": 0.7044138686717234, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 25608 + }, + { + "epoch": 0.25609, + "grad_norm": 0.6546682254709966, + "learning_rate": 0.003, + "loss": 3.967, + "step": 25609 + }, + { + "epoch": 0.2561, + "grad_norm": 0.7061175935464133, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 25610 + }, + { + "epoch": 0.25611, + "grad_norm": 0.7179833488052864, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 25611 + }, + { + "epoch": 0.25612, + "grad_norm": 0.6893316050857081, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 25612 + }, + { + "epoch": 0.25613, + "grad_norm": 0.7076753490837825, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 25613 + }, + { + "epoch": 0.25614, + "grad_norm": 0.7643874015567181, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 25614 + }, + { + "epoch": 0.25615, + "grad_norm": 0.7580148918984284, + "learning_rate": 0.003, + "loss": 4.023, + "step": 25615 + }, + { + "epoch": 0.25616, + "grad_norm": 0.7957566449869538, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 25616 + }, + { + "epoch": 0.25617, + "grad_norm": 0.9229218324914377, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 25617 + }, + { + "epoch": 0.25618, + "grad_norm": 1.0028397364601365, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 25618 + }, + { + "epoch": 0.25619, + "grad_norm": 0.9312467548237616, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 25619 + }, + { + "epoch": 0.2562, + "grad_norm": 0.9672760882645325, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 25620 + }, + { + "epoch": 0.25621, + "grad_norm": 0.919082611726874, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 25621 + }, + { + "epoch": 0.25622, + "grad_norm": 0.882180450113223, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 25622 + }, + { + "epoch": 0.25623, + "grad_norm": 0.9273494353624908, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 25623 + }, + { + "epoch": 0.25624, + "grad_norm": 0.8311333086518223, + "learning_rate": 0.003, + "loss": 3.997, + "step": 25624 + }, + { + "epoch": 0.25625, + "grad_norm": 0.8082634897348785, + "learning_rate": 0.003, + "loss": 4.0356, + "step": 25625 + }, + { + "epoch": 0.25626, + "grad_norm": 0.8243360734545833, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 25626 + }, + { + "epoch": 0.25627, + "grad_norm": 0.9377123119477923, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 25627 + }, + { + "epoch": 0.25628, + "grad_norm": 1.0558678772678192, + "learning_rate": 0.003, + "loss": 4.013, + "step": 25628 + }, + { + "epoch": 0.25629, + "grad_norm": 0.971226847300283, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 25629 + }, + { + "epoch": 0.2563, + "grad_norm": 0.7718960187391117, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 25630 + }, + { + "epoch": 0.25631, + "grad_norm": 0.7021455244366933, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 25631 + }, + { + "epoch": 0.25632, + "grad_norm": 0.754452106199192, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 25632 + }, + { + "epoch": 0.25633, + "grad_norm": 0.7668616097884348, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 25633 + }, + { + "epoch": 0.25634, + "grad_norm": 0.6981061596968788, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 25634 + }, + { + "epoch": 0.25635, + "grad_norm": 0.6974878661664287, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 25635 + }, + { + "epoch": 0.25636, + "grad_norm": 0.6788219544514772, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 25636 + }, + { + "epoch": 0.25637, + "grad_norm": 0.6015027298560717, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 25637 + }, + { + "epoch": 0.25638, + "grad_norm": 0.546599054297129, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 25638 + }, + { + "epoch": 0.25639, + "grad_norm": 0.6521115454656986, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 25639 + }, + { + "epoch": 0.2564, + "grad_norm": 0.6452082901862529, + "learning_rate": 0.003, + "loss": 4.007, + "step": 25640 + }, + { + "epoch": 0.25641, + "grad_norm": 0.6743334987317481, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 25641 + }, + { + "epoch": 0.25642, + "grad_norm": 0.7945511827753887, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 25642 + }, + { + "epoch": 0.25643, + "grad_norm": 0.9181761955278108, + "learning_rate": 0.003, + "loss": 3.999, + "step": 25643 + }, + { + "epoch": 0.25644, + "grad_norm": 1.018555434461602, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 25644 + }, + { + "epoch": 0.25645, + "grad_norm": 0.9271649662521755, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 25645 + }, + { + "epoch": 0.25646, + "grad_norm": 0.7536195903290381, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 25646 + }, + { + "epoch": 0.25647, + "grad_norm": 0.819594010781068, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 25647 + }, + { + "epoch": 0.25648, + "grad_norm": 0.9233146300535235, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 25648 + }, + { + "epoch": 0.25649, + "grad_norm": 0.7928307382639301, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 25649 + }, + { + "epoch": 0.2565, + "grad_norm": 0.6740062097408116, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 25650 + }, + { + "epoch": 0.25651, + "grad_norm": 0.6055737512573666, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 25651 + }, + { + "epoch": 0.25652, + "grad_norm": 0.6147768794924507, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 25652 + }, + { + "epoch": 0.25653, + "grad_norm": 0.6367293878208827, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 25653 + }, + { + "epoch": 0.25654, + "grad_norm": 0.6822481969639564, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 25654 + }, + { + "epoch": 0.25655, + "grad_norm": 0.7655303022891675, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 25655 + }, + { + "epoch": 0.25656, + "grad_norm": 0.7625151714877938, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 25656 + }, + { + "epoch": 0.25657, + "grad_norm": 0.7538796962536223, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 25657 + }, + { + "epoch": 0.25658, + "grad_norm": 0.7958796589656196, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 25658 + }, + { + "epoch": 0.25659, + "grad_norm": 0.7384590831144624, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 25659 + }, + { + "epoch": 0.2566, + "grad_norm": 0.7197818502372787, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 25660 + }, + { + "epoch": 0.25661, + "grad_norm": 0.8429665047255122, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 25661 + }, + { + "epoch": 0.25662, + "grad_norm": 0.7983454003910998, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 25662 + }, + { + "epoch": 0.25663, + "grad_norm": 0.6846179544041878, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 25663 + }, + { + "epoch": 0.25664, + "grad_norm": 0.5885246264493451, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 25664 + }, + { + "epoch": 0.25665, + "grad_norm": 0.6646171925285543, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 25665 + }, + { + "epoch": 0.25666, + "grad_norm": 0.5861955133134406, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 25666 + }, + { + "epoch": 0.25667, + "grad_norm": 0.5988792535591564, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 25667 + }, + { + "epoch": 0.25668, + "grad_norm": 0.5592830786640824, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 25668 + }, + { + "epoch": 0.25669, + "grad_norm": 0.6466724399074221, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 25669 + }, + { + "epoch": 0.2567, + "grad_norm": 0.6664049404426042, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 25670 + }, + { + "epoch": 0.25671, + "grad_norm": 0.6078690929831227, + "learning_rate": 0.003, + "loss": 3.99, + "step": 25671 + }, + { + "epoch": 0.25672, + "grad_norm": 0.6240752454140357, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 25672 + }, + { + "epoch": 0.25673, + "grad_norm": 0.6452431961398054, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 25673 + }, + { + "epoch": 0.25674, + "grad_norm": 0.7843396757781105, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 25674 + }, + { + "epoch": 0.25675, + "grad_norm": 1.0488910541033787, + "learning_rate": 0.003, + "loss": 3.993, + "step": 25675 + }, + { + "epoch": 0.25676, + "grad_norm": 1.0476893157336367, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 25676 + }, + { + "epoch": 0.25677, + "grad_norm": 0.8487058146825386, + "learning_rate": 0.003, + "loss": 4.038, + "step": 25677 + }, + { + "epoch": 0.25678, + "grad_norm": 0.8492239343607961, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 25678 + }, + { + "epoch": 0.25679, + "grad_norm": 0.9167031394919601, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 25679 + }, + { + "epoch": 0.2568, + "grad_norm": 1.0245002092638682, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 25680 + }, + { + "epoch": 0.25681, + "grad_norm": 1.0320847816071217, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 25681 + }, + { + "epoch": 0.25682, + "grad_norm": 0.8640152348124235, + "learning_rate": 0.003, + "loss": 4.0355, + "step": 25682 + }, + { + "epoch": 0.25683, + "grad_norm": 0.8057015643354172, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 25683 + }, + { + "epoch": 0.25684, + "grad_norm": 0.8283500446808987, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 25684 + }, + { + "epoch": 0.25685, + "grad_norm": 0.858289932797647, + "learning_rate": 0.003, + "loss": 3.992, + "step": 25685 + }, + { + "epoch": 0.25686, + "grad_norm": 0.8055094775854319, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 25686 + }, + { + "epoch": 0.25687, + "grad_norm": 0.736713392235155, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 25687 + }, + { + "epoch": 0.25688, + "grad_norm": 0.7135731809971687, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 25688 + }, + { + "epoch": 0.25689, + "grad_norm": 0.6491836224323598, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 25689 + }, + { + "epoch": 0.2569, + "grad_norm": 0.6437073612998326, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 25690 + }, + { + "epoch": 0.25691, + "grad_norm": 0.5723652057240622, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 25691 + }, + { + "epoch": 0.25692, + "grad_norm": 0.5713785118639422, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 25692 + }, + { + "epoch": 0.25693, + "grad_norm": 0.6183933051613799, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 25693 + }, + { + "epoch": 0.25694, + "grad_norm": 0.6617836935110404, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 25694 + }, + { + "epoch": 0.25695, + "grad_norm": 0.6966715342519467, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 25695 + }, + { + "epoch": 0.25696, + "grad_norm": 0.7128858825143706, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 25696 + }, + { + "epoch": 0.25697, + "grad_norm": 0.73165956892888, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 25697 + }, + { + "epoch": 0.25698, + "grad_norm": 0.7296722570364994, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 25698 + }, + { + "epoch": 0.25699, + "grad_norm": 0.7322993028018979, + "learning_rate": 0.003, + "loss": 4.0, + "step": 25699 + }, + { + "epoch": 0.257, + "grad_norm": 0.841868833069187, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 25700 + }, + { + "epoch": 0.25701, + "grad_norm": 0.9519950508509126, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 25701 + }, + { + "epoch": 0.25702, + "grad_norm": 0.9214987287634264, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 25702 + }, + { + "epoch": 0.25703, + "grad_norm": 0.805136415178069, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 25703 + }, + { + "epoch": 0.25704, + "grad_norm": 0.7501715643761309, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 25704 + }, + { + "epoch": 0.25705, + "grad_norm": 0.7151176296912337, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 25705 + }, + { + "epoch": 0.25706, + "grad_norm": 0.6133927094004685, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 25706 + }, + { + "epoch": 0.25707, + "grad_norm": 0.5768130028353978, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 25707 + }, + { + "epoch": 0.25708, + "grad_norm": 0.6050281407590715, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 25708 + }, + { + "epoch": 0.25709, + "grad_norm": 0.6860310505716491, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 25709 + }, + { + "epoch": 0.2571, + "grad_norm": 0.810340348526892, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 25710 + }, + { + "epoch": 0.25711, + "grad_norm": 0.9264515951559826, + "learning_rate": 0.003, + "loss": 3.997, + "step": 25711 + }, + { + "epoch": 0.25712, + "grad_norm": 0.8957607726362142, + "learning_rate": 0.003, + "loss": 4.001, + "step": 25712 + }, + { + "epoch": 0.25713, + "grad_norm": 0.7473259267532758, + "learning_rate": 0.003, + "loss": 4.0, + "step": 25713 + }, + { + "epoch": 0.25714, + "grad_norm": 0.8052763658305429, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 25714 + }, + { + "epoch": 0.25715, + "grad_norm": 0.8182730523028852, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 25715 + }, + { + "epoch": 0.25716, + "grad_norm": 0.7486531783744813, + "learning_rate": 0.003, + "loss": 3.993, + "step": 25716 + }, + { + "epoch": 0.25717, + "grad_norm": 0.7437805404921433, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 25717 + }, + { + "epoch": 0.25718, + "grad_norm": 0.8266083296837775, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 25718 + }, + { + "epoch": 0.25719, + "grad_norm": 0.6829753138976218, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 25719 + }, + { + "epoch": 0.2572, + "grad_norm": 0.5819764926107772, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 25720 + }, + { + "epoch": 0.25721, + "grad_norm": 0.6141319476853941, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 25721 + }, + { + "epoch": 0.25722, + "grad_norm": 0.6650296028711956, + "learning_rate": 0.003, + "loss": 4.009, + "step": 25722 + }, + { + "epoch": 0.25723, + "grad_norm": 0.7718796613898917, + "learning_rate": 0.003, + "loss": 3.998, + "step": 25723 + }, + { + "epoch": 0.25724, + "grad_norm": 0.8776406577421502, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 25724 + }, + { + "epoch": 0.25725, + "grad_norm": 0.9630383634335493, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 25725 + }, + { + "epoch": 0.25726, + "grad_norm": 0.9062903282659963, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 25726 + }, + { + "epoch": 0.25727, + "grad_norm": 0.6754756393424014, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 25727 + }, + { + "epoch": 0.25728, + "grad_norm": 0.5841452759950624, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 25728 + }, + { + "epoch": 0.25729, + "grad_norm": 0.5599954318645919, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 25729 + }, + { + "epoch": 0.2573, + "grad_norm": 0.624792976381153, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 25730 + }, + { + "epoch": 0.25731, + "grad_norm": 0.7322703575327189, + "learning_rate": 0.003, + "loss": 4.003, + "step": 25731 + }, + { + "epoch": 0.25732, + "grad_norm": 0.8447653958918117, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 25732 + }, + { + "epoch": 0.25733, + "grad_norm": 0.9815027529187014, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 25733 + }, + { + "epoch": 0.25734, + "grad_norm": 1.0690301965936904, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 25734 + }, + { + "epoch": 0.25735, + "grad_norm": 0.9300837030516816, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 25735 + }, + { + "epoch": 0.25736, + "grad_norm": 0.9213072725504964, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 25736 + }, + { + "epoch": 0.25737, + "grad_norm": 0.7554870034713885, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 25737 + }, + { + "epoch": 0.25738, + "grad_norm": 0.7176188300570437, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 25738 + }, + { + "epoch": 0.25739, + "grad_norm": 0.7238754218462933, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 25739 + }, + { + "epoch": 0.2574, + "grad_norm": 0.8581722958109705, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 25740 + }, + { + "epoch": 0.25741, + "grad_norm": 0.9964869607798694, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 25741 + }, + { + "epoch": 0.25742, + "grad_norm": 0.9615384353267187, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 25742 + }, + { + "epoch": 0.25743, + "grad_norm": 0.9727730855508675, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 25743 + }, + { + "epoch": 0.25744, + "grad_norm": 1.0575801056107998, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 25744 + }, + { + "epoch": 0.25745, + "grad_norm": 0.8803018048719221, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 25745 + }, + { + "epoch": 0.25746, + "grad_norm": 0.9287695864754041, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 25746 + }, + { + "epoch": 0.25747, + "grad_norm": 0.9975459886625878, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 25747 + }, + { + "epoch": 0.25748, + "grad_norm": 1.0740377832254446, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 25748 + }, + { + "epoch": 0.25749, + "grad_norm": 0.9891785018815772, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 25749 + }, + { + "epoch": 0.2575, + "grad_norm": 0.8779080282774572, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 25750 + }, + { + "epoch": 0.25751, + "grad_norm": 0.7579920334839702, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 25751 + }, + { + "epoch": 0.25752, + "grad_norm": 0.7727165222994488, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 25752 + }, + { + "epoch": 0.25753, + "grad_norm": 0.7487788333298403, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 25753 + }, + { + "epoch": 0.25754, + "grad_norm": 0.7309652803083906, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 25754 + }, + { + "epoch": 0.25755, + "grad_norm": 0.8159148739405169, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 25755 + }, + { + "epoch": 0.25756, + "grad_norm": 0.7994312910586372, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 25756 + }, + { + "epoch": 0.25757, + "grad_norm": 0.7696149386166918, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 25757 + }, + { + "epoch": 0.25758, + "grad_norm": 0.6630247288099946, + "learning_rate": 0.003, + "loss": 3.986, + "step": 25758 + }, + { + "epoch": 0.25759, + "grad_norm": 0.6002475405167886, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 25759 + }, + { + "epoch": 0.2576, + "grad_norm": 0.5424294552312766, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 25760 + }, + { + "epoch": 0.25761, + "grad_norm": 0.5202405325856521, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 25761 + }, + { + "epoch": 0.25762, + "grad_norm": 0.5995112877021826, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 25762 + }, + { + "epoch": 0.25763, + "grad_norm": 0.6722439355160511, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 25763 + }, + { + "epoch": 0.25764, + "grad_norm": 0.7257973161764948, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 25764 + }, + { + "epoch": 0.25765, + "grad_norm": 0.7684242780625132, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 25765 + }, + { + "epoch": 0.25766, + "grad_norm": 0.7306611104766396, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 25766 + }, + { + "epoch": 0.25767, + "grad_norm": 0.6141077378693675, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 25767 + }, + { + "epoch": 0.25768, + "grad_norm": 0.5816923300911143, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 25768 + }, + { + "epoch": 0.25769, + "grad_norm": 0.786536513322659, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 25769 + }, + { + "epoch": 0.2577, + "grad_norm": 0.9206747604720813, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 25770 + }, + { + "epoch": 0.25771, + "grad_norm": 0.7789589789815513, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 25771 + }, + { + "epoch": 0.25772, + "grad_norm": 0.6561515059717363, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 25772 + }, + { + "epoch": 0.25773, + "grad_norm": 0.6041003893375905, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 25773 + }, + { + "epoch": 0.25774, + "grad_norm": 0.6853973476655179, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 25774 + }, + { + "epoch": 0.25775, + "grad_norm": 0.6761893046101025, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 25775 + }, + { + "epoch": 0.25776, + "grad_norm": 0.6513831736592494, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 25776 + }, + { + "epoch": 0.25777, + "grad_norm": 0.7261545683204972, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 25777 + }, + { + "epoch": 0.25778, + "grad_norm": 0.7237865827958263, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 25778 + }, + { + "epoch": 0.25779, + "grad_norm": 0.7529229501731367, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 25779 + }, + { + "epoch": 0.2578, + "grad_norm": 0.7105197639548669, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 25780 + }, + { + "epoch": 0.25781, + "grad_norm": 0.6840606060316551, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 25781 + }, + { + "epoch": 0.25782, + "grad_norm": 0.7116114046924501, + "learning_rate": 0.003, + "loss": 3.987, + "step": 25782 + }, + { + "epoch": 0.25783, + "grad_norm": 0.6579752292720844, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 25783 + }, + { + "epoch": 0.25784, + "grad_norm": 0.6063457429793268, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 25784 + }, + { + "epoch": 0.25785, + "grad_norm": 0.6169119631547623, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 25785 + }, + { + "epoch": 0.25786, + "grad_norm": 0.6495558222694768, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 25786 + }, + { + "epoch": 0.25787, + "grad_norm": 0.7409993227951851, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 25787 + }, + { + "epoch": 0.25788, + "grad_norm": 0.8756068465268173, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 25788 + }, + { + "epoch": 0.25789, + "grad_norm": 0.9779544613888419, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 25789 + }, + { + "epoch": 0.2579, + "grad_norm": 0.8511247764594082, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 25790 + }, + { + "epoch": 0.25791, + "grad_norm": 0.7024032086809704, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 25791 + }, + { + "epoch": 0.25792, + "grad_norm": 0.7394326779274484, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 25792 + }, + { + "epoch": 0.25793, + "grad_norm": 0.7577865493144107, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 25793 + }, + { + "epoch": 0.25794, + "grad_norm": 0.8242213452850051, + "learning_rate": 0.003, + "loss": 4.022, + "step": 25794 + }, + { + "epoch": 0.25795, + "grad_norm": 0.8064404820257959, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 25795 + }, + { + "epoch": 0.25796, + "grad_norm": 0.8881363293493777, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 25796 + }, + { + "epoch": 0.25797, + "grad_norm": 0.9576139770930719, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 25797 + }, + { + "epoch": 0.25798, + "grad_norm": 1.0208088795341497, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 25798 + }, + { + "epoch": 0.25799, + "grad_norm": 0.9417362395356639, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 25799 + }, + { + "epoch": 0.258, + "grad_norm": 0.8670348788346847, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 25800 + }, + { + "epoch": 0.25801, + "grad_norm": 0.9357867490894624, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 25801 + }, + { + "epoch": 0.25802, + "grad_norm": 0.8443497116913447, + "learning_rate": 0.003, + "loss": 4.019, + "step": 25802 + }, + { + "epoch": 0.25803, + "grad_norm": 0.7178590593695229, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 25803 + }, + { + "epoch": 0.25804, + "grad_norm": 0.6323150207508803, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 25804 + }, + { + "epoch": 0.25805, + "grad_norm": 0.6558362082995537, + "learning_rate": 0.003, + "loss": 4.021, + "step": 25805 + }, + { + "epoch": 0.25806, + "grad_norm": 0.6971687512285489, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 25806 + }, + { + "epoch": 0.25807, + "grad_norm": 0.6890973809409023, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 25807 + }, + { + "epoch": 0.25808, + "grad_norm": 0.7537307218177992, + "learning_rate": 0.003, + "loss": 3.986, + "step": 25808 + }, + { + "epoch": 0.25809, + "grad_norm": 0.7451757641292533, + "learning_rate": 0.003, + "loss": 3.987, + "step": 25809 + }, + { + "epoch": 0.2581, + "grad_norm": 0.6413254255040753, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 25810 + }, + { + "epoch": 0.25811, + "grad_norm": 0.5636975883178494, + "learning_rate": 0.003, + "loss": 3.999, + "step": 25811 + }, + { + "epoch": 0.25812, + "grad_norm": 0.6209993071806241, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 25812 + }, + { + "epoch": 0.25813, + "grad_norm": 0.6087543624838568, + "learning_rate": 0.003, + "loss": 3.987, + "step": 25813 + }, + { + "epoch": 0.25814, + "grad_norm": 0.5787562100095643, + "learning_rate": 0.003, + "loss": 4.022, + "step": 25814 + }, + { + "epoch": 0.25815, + "grad_norm": 0.5766486833956456, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 25815 + }, + { + "epoch": 0.25816, + "grad_norm": 0.5250814697268411, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 25816 + }, + { + "epoch": 0.25817, + "grad_norm": 0.6123314244980667, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 25817 + }, + { + "epoch": 0.25818, + "grad_norm": 0.6112333891372669, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 25818 + }, + { + "epoch": 0.25819, + "grad_norm": 0.5649905190854119, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 25819 + }, + { + "epoch": 0.2582, + "grad_norm": 0.660919807758348, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 25820 + }, + { + "epoch": 0.25821, + "grad_norm": 0.8024685884682884, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 25821 + }, + { + "epoch": 0.25822, + "grad_norm": 0.9256996321504599, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 25822 + }, + { + "epoch": 0.25823, + "grad_norm": 1.0290329015159783, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 25823 + }, + { + "epoch": 0.25824, + "grad_norm": 0.9496848872106158, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 25824 + }, + { + "epoch": 0.25825, + "grad_norm": 1.009926930474451, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 25825 + }, + { + "epoch": 0.25826, + "grad_norm": 0.9695147861231678, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 25826 + }, + { + "epoch": 0.25827, + "grad_norm": 0.866049966663142, + "learning_rate": 0.003, + "loss": 4.012, + "step": 25827 + }, + { + "epoch": 0.25828, + "grad_norm": 0.8355726193453581, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 25828 + }, + { + "epoch": 0.25829, + "grad_norm": 0.7880030971049653, + "learning_rate": 0.003, + "loss": 4.028, + "step": 25829 + }, + { + "epoch": 0.2583, + "grad_norm": 0.7837319877156004, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 25830 + }, + { + "epoch": 0.25831, + "grad_norm": 0.7888597837230424, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 25831 + }, + { + "epoch": 0.25832, + "grad_norm": 0.8786783410402059, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 25832 + }, + { + "epoch": 0.25833, + "grad_norm": 0.8225649170634657, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 25833 + }, + { + "epoch": 0.25834, + "grad_norm": 0.6662304873061651, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 25834 + }, + { + "epoch": 0.25835, + "grad_norm": 0.796724834396577, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 25835 + }, + { + "epoch": 0.25836, + "grad_norm": 0.803846837966274, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 25836 + }, + { + "epoch": 0.25837, + "grad_norm": 0.8642026286966626, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 25837 + }, + { + "epoch": 0.25838, + "grad_norm": 0.8024252063638414, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 25838 + }, + { + "epoch": 0.25839, + "grad_norm": 0.7468107032657433, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 25839 + }, + { + "epoch": 0.2584, + "grad_norm": 0.7374915314269607, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 25840 + }, + { + "epoch": 0.25841, + "grad_norm": 0.8728348243212926, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 25841 + }, + { + "epoch": 0.25842, + "grad_norm": 0.7288884691304257, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 25842 + }, + { + "epoch": 0.25843, + "grad_norm": 0.6265552281791, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 25843 + }, + { + "epoch": 0.25844, + "grad_norm": 0.6365448891902122, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 25844 + }, + { + "epoch": 0.25845, + "grad_norm": 0.6603833243612346, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 25845 + }, + { + "epoch": 0.25846, + "grad_norm": 0.6015820305864525, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 25846 + }, + { + "epoch": 0.25847, + "grad_norm": 0.6766145095317606, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 25847 + }, + { + "epoch": 0.25848, + "grad_norm": 0.6824915357813415, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 25848 + }, + { + "epoch": 0.25849, + "grad_norm": 0.6525280878690334, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 25849 + }, + { + "epoch": 0.2585, + "grad_norm": 0.6781463556164732, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 25850 + }, + { + "epoch": 0.25851, + "grad_norm": 0.6173626374144492, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 25851 + }, + { + "epoch": 0.25852, + "grad_norm": 0.536308835524837, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 25852 + }, + { + "epoch": 0.25853, + "grad_norm": 0.5562984851907227, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 25853 + }, + { + "epoch": 0.25854, + "grad_norm": 0.6906922384930473, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 25854 + }, + { + "epoch": 0.25855, + "grad_norm": 0.8553276801895474, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 25855 + }, + { + "epoch": 0.25856, + "grad_norm": 1.0250262798586793, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 25856 + }, + { + "epoch": 0.25857, + "grad_norm": 1.1410895024320051, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 25857 + }, + { + "epoch": 0.25858, + "grad_norm": 0.8350437802720613, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 25858 + }, + { + "epoch": 0.25859, + "grad_norm": 0.9072819600019499, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 25859 + }, + { + "epoch": 0.2586, + "grad_norm": 0.8229352875487207, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 25860 + }, + { + "epoch": 0.25861, + "grad_norm": 0.8946440518324706, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 25861 + }, + { + "epoch": 0.25862, + "grad_norm": 0.989633615287119, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 25862 + }, + { + "epoch": 0.25863, + "grad_norm": 1.0619891801890922, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 25863 + }, + { + "epoch": 0.25864, + "grad_norm": 0.9113172676147648, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 25864 + }, + { + "epoch": 0.25865, + "grad_norm": 0.8535573177519261, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 25865 + }, + { + "epoch": 0.25866, + "grad_norm": 0.8492996966192603, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 25866 + }, + { + "epoch": 0.25867, + "grad_norm": 0.7825194032648987, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 25867 + }, + { + "epoch": 0.25868, + "grad_norm": 0.8338944576973274, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 25868 + }, + { + "epoch": 0.25869, + "grad_norm": 0.9114433923637915, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 25869 + }, + { + "epoch": 0.2587, + "grad_norm": 1.0634640620059637, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 25870 + }, + { + "epoch": 0.25871, + "grad_norm": 1.0926304136916354, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 25871 + }, + { + "epoch": 0.25872, + "grad_norm": 0.7968977946567644, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 25872 + }, + { + "epoch": 0.25873, + "grad_norm": 0.7333697437160344, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 25873 + }, + { + "epoch": 0.25874, + "grad_norm": 0.7230804737182784, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 25874 + }, + { + "epoch": 0.25875, + "grad_norm": 0.6958651075971463, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 25875 + }, + { + "epoch": 0.25876, + "grad_norm": 0.7449420703306316, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 25876 + }, + { + "epoch": 0.25877, + "grad_norm": 0.7803178337945204, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 25877 + }, + { + "epoch": 0.25878, + "grad_norm": 0.7659190284734136, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 25878 + }, + { + "epoch": 0.25879, + "grad_norm": 0.6709559265141024, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 25879 + }, + { + "epoch": 0.2588, + "grad_norm": 0.5501552749972288, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 25880 + }, + { + "epoch": 0.25881, + "grad_norm": 0.5268671695804499, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 25881 + }, + { + "epoch": 0.25882, + "grad_norm": 0.5187956951505505, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 25882 + }, + { + "epoch": 0.25883, + "grad_norm": 0.5083516501136742, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 25883 + }, + { + "epoch": 0.25884, + "grad_norm": 0.5253397426475062, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 25884 + }, + { + "epoch": 0.25885, + "grad_norm": 0.5771935166492537, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 25885 + }, + { + "epoch": 0.25886, + "grad_norm": 0.5963056439086044, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 25886 + }, + { + "epoch": 0.25887, + "grad_norm": 0.5842085492584715, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 25887 + }, + { + "epoch": 0.25888, + "grad_norm": 0.6030339902099127, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 25888 + }, + { + "epoch": 0.25889, + "grad_norm": 0.5702455186903419, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 25889 + }, + { + "epoch": 0.2589, + "grad_norm": 0.5518483809249196, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 25890 + }, + { + "epoch": 0.25891, + "grad_norm": 0.579137287376548, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 25891 + }, + { + "epoch": 0.25892, + "grad_norm": 0.6402065470554613, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 25892 + }, + { + "epoch": 0.25893, + "grad_norm": 0.7000893080591981, + "learning_rate": 0.003, + "loss": 4.002, + "step": 25893 + }, + { + "epoch": 0.25894, + "grad_norm": 0.8114724482015901, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 25894 + }, + { + "epoch": 0.25895, + "grad_norm": 0.9133634823515683, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 25895 + }, + { + "epoch": 0.25896, + "grad_norm": 1.0014889001472882, + "learning_rate": 0.003, + "loss": 3.976, + "step": 25896 + }, + { + "epoch": 0.25897, + "grad_norm": 1.0034042576153435, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 25897 + }, + { + "epoch": 0.25898, + "grad_norm": 0.8571112440686903, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 25898 + }, + { + "epoch": 0.25899, + "grad_norm": 0.741356367932479, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 25899 + }, + { + "epoch": 0.259, + "grad_norm": 0.6285633704761491, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 25900 + }, + { + "epoch": 0.25901, + "grad_norm": 0.638368547966785, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 25901 + }, + { + "epoch": 0.25902, + "grad_norm": 0.6543557203449528, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 25902 + }, + { + "epoch": 0.25903, + "grad_norm": 0.6429517165544505, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 25903 + }, + { + "epoch": 0.25904, + "grad_norm": 0.6059470823405309, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 25904 + }, + { + "epoch": 0.25905, + "grad_norm": 0.6887924290622045, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 25905 + }, + { + "epoch": 0.25906, + "grad_norm": 0.81984458451692, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 25906 + }, + { + "epoch": 0.25907, + "grad_norm": 0.783881538088212, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 25907 + }, + { + "epoch": 0.25908, + "grad_norm": 0.6433407422274032, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 25908 + }, + { + "epoch": 0.25909, + "grad_norm": 0.6244018203306202, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 25909 + }, + { + "epoch": 0.2591, + "grad_norm": 0.5985269156593956, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 25910 + }, + { + "epoch": 0.25911, + "grad_norm": 0.6639761254449599, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 25911 + }, + { + "epoch": 0.25912, + "grad_norm": 0.7190635021411698, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 25912 + }, + { + "epoch": 0.25913, + "grad_norm": 0.8716740111007982, + "learning_rate": 0.003, + "loss": 3.978, + "step": 25913 + }, + { + "epoch": 0.25914, + "grad_norm": 0.9984336102759979, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 25914 + }, + { + "epoch": 0.25915, + "grad_norm": 1.0230175952699183, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 25915 + }, + { + "epoch": 0.25916, + "grad_norm": 0.9754543519439587, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 25916 + }, + { + "epoch": 0.25917, + "grad_norm": 1.0030866414722643, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 25917 + }, + { + "epoch": 0.25918, + "grad_norm": 1.0405726525151329, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 25918 + }, + { + "epoch": 0.25919, + "grad_norm": 0.90786514088945, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 25919 + }, + { + "epoch": 0.2592, + "grad_norm": 0.978471087065511, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 25920 + }, + { + "epoch": 0.25921, + "grad_norm": 1.094345341349745, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 25921 + }, + { + "epoch": 0.25922, + "grad_norm": 0.9170771171563766, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 25922 + }, + { + "epoch": 0.25923, + "grad_norm": 0.8998641189883237, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 25923 + }, + { + "epoch": 0.25924, + "grad_norm": 0.9043278428464304, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 25924 + }, + { + "epoch": 0.25925, + "grad_norm": 0.8289233754303955, + "learning_rate": 0.003, + "loss": 3.997, + "step": 25925 + }, + { + "epoch": 0.25926, + "grad_norm": 0.7352227861531934, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 25926 + }, + { + "epoch": 0.25927, + "grad_norm": 0.7222367268724076, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 25927 + }, + { + "epoch": 0.25928, + "grad_norm": 0.71604314726032, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 25928 + }, + { + "epoch": 0.25929, + "grad_norm": 0.7286565228365426, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 25929 + }, + { + "epoch": 0.2593, + "grad_norm": 0.8071138151263517, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 25930 + }, + { + "epoch": 0.25931, + "grad_norm": 0.7862317271326709, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 25931 + }, + { + "epoch": 0.25932, + "grad_norm": 0.7927672264537482, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 25932 + }, + { + "epoch": 0.25933, + "grad_norm": 0.8429998301347975, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 25933 + }, + { + "epoch": 0.25934, + "grad_norm": 0.7554209118014498, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 25934 + }, + { + "epoch": 0.25935, + "grad_norm": 0.727345401104779, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 25935 + }, + { + "epoch": 0.25936, + "grad_norm": 0.6497345775922052, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 25936 + }, + { + "epoch": 0.25937, + "grad_norm": 0.654881861889486, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 25937 + }, + { + "epoch": 0.25938, + "grad_norm": 0.6544081519475444, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 25938 + }, + { + "epoch": 0.25939, + "grad_norm": 0.6242498805028595, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 25939 + }, + { + "epoch": 0.2594, + "grad_norm": 0.6628125600760323, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 25940 + }, + { + "epoch": 0.25941, + "grad_norm": 0.7918498154777424, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 25941 + }, + { + "epoch": 0.25942, + "grad_norm": 0.910095801940323, + "learning_rate": 0.003, + "loss": 4.01, + "step": 25942 + }, + { + "epoch": 0.25943, + "grad_norm": 1.0695165767491903, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 25943 + }, + { + "epoch": 0.25944, + "grad_norm": 0.8208242788823407, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 25944 + }, + { + "epoch": 0.25945, + "grad_norm": 0.6985931289138766, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 25945 + }, + { + "epoch": 0.25946, + "grad_norm": 0.7052858332483457, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 25946 + }, + { + "epoch": 0.25947, + "grad_norm": 0.7407373280759094, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 25947 + }, + { + "epoch": 0.25948, + "grad_norm": 0.6548281982887982, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 25948 + }, + { + "epoch": 0.25949, + "grad_norm": 0.6610869516755906, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 25949 + }, + { + "epoch": 0.2595, + "grad_norm": 0.6533941370630999, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 25950 + }, + { + "epoch": 0.25951, + "grad_norm": 0.626467057007967, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 25951 + }, + { + "epoch": 0.25952, + "grad_norm": 0.6390765210412209, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 25952 + }, + { + "epoch": 0.25953, + "grad_norm": 0.6596761545163181, + "learning_rate": 0.003, + "loss": 4.01, + "step": 25953 + }, + { + "epoch": 0.25954, + "grad_norm": 0.7411841097800066, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 25954 + }, + { + "epoch": 0.25955, + "grad_norm": 0.8807892660345327, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 25955 + }, + { + "epoch": 0.25956, + "grad_norm": 0.8900989087045995, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 25956 + }, + { + "epoch": 0.25957, + "grad_norm": 0.7842057752187997, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 25957 + }, + { + "epoch": 0.25958, + "grad_norm": 0.785640736342632, + "learning_rate": 0.003, + "loss": 4.02, + "step": 25958 + }, + { + "epoch": 0.25959, + "grad_norm": 0.6817052207485199, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 25959 + }, + { + "epoch": 0.2596, + "grad_norm": 0.7936306706110211, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 25960 + }, + { + "epoch": 0.25961, + "grad_norm": 0.8994814891714366, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 25961 + }, + { + "epoch": 0.25962, + "grad_norm": 0.946976498144478, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 25962 + }, + { + "epoch": 0.25963, + "grad_norm": 0.7573570695153992, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 25963 + }, + { + "epoch": 0.25964, + "grad_norm": 0.7068692918150853, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 25964 + }, + { + "epoch": 0.25965, + "grad_norm": 0.7075032385982376, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 25965 + }, + { + "epoch": 0.25966, + "grad_norm": 0.6816685712758698, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 25966 + }, + { + "epoch": 0.25967, + "grad_norm": 0.6896350180122806, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 25967 + }, + { + "epoch": 0.25968, + "grad_norm": 0.6830367964571379, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 25968 + }, + { + "epoch": 0.25969, + "grad_norm": 0.6367133498502637, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 25969 + }, + { + "epoch": 0.2597, + "grad_norm": 0.6786766619274487, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 25970 + }, + { + "epoch": 0.25971, + "grad_norm": 0.6826357394898753, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 25971 + }, + { + "epoch": 0.25972, + "grad_norm": 0.7776151051206429, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 25972 + }, + { + "epoch": 0.25973, + "grad_norm": 0.7871364212521182, + "learning_rate": 0.003, + "loss": 3.981, + "step": 25973 + }, + { + "epoch": 0.25974, + "grad_norm": 0.7792519418788296, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 25974 + }, + { + "epoch": 0.25975, + "grad_norm": 0.6537469187713923, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 25975 + }, + { + "epoch": 0.25976, + "grad_norm": 0.5943326027417286, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 25976 + }, + { + "epoch": 0.25977, + "grad_norm": 0.6441384877190028, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 25977 + }, + { + "epoch": 0.25978, + "grad_norm": 0.6738672444782089, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 25978 + }, + { + "epoch": 0.25979, + "grad_norm": 0.6711160955786699, + "learning_rate": 0.003, + "loss": 3.984, + "step": 25979 + }, + { + "epoch": 0.2598, + "grad_norm": 0.7125555333590661, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 25980 + }, + { + "epoch": 0.25981, + "grad_norm": 0.786454200193559, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 25981 + }, + { + "epoch": 0.25982, + "grad_norm": 0.7413175791109545, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 25982 + }, + { + "epoch": 0.25983, + "grad_norm": 0.6864646207701878, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 25983 + }, + { + "epoch": 0.25984, + "grad_norm": 0.8154944750349111, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 25984 + }, + { + "epoch": 0.25985, + "grad_norm": 0.734001158616293, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 25985 + }, + { + "epoch": 0.25986, + "grad_norm": 0.8161329040155119, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 25986 + }, + { + "epoch": 0.25987, + "grad_norm": 0.9826281515856761, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 25987 + }, + { + "epoch": 0.25988, + "grad_norm": 0.9932699626001956, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 25988 + }, + { + "epoch": 0.25989, + "grad_norm": 0.8634270266811019, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 25989 + }, + { + "epoch": 0.2599, + "grad_norm": 0.8716362810165114, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 25990 + }, + { + "epoch": 0.25991, + "grad_norm": 0.7938552217105829, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 25991 + }, + { + "epoch": 0.25992, + "grad_norm": 0.840988398291006, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 25992 + }, + { + "epoch": 0.25993, + "grad_norm": 0.719539811859336, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 25993 + }, + { + "epoch": 0.25994, + "grad_norm": 0.7409394321722911, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 25994 + }, + { + "epoch": 0.25995, + "grad_norm": 0.6637196043533848, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 25995 + }, + { + "epoch": 0.25996, + "grad_norm": 0.5852566376776857, + "learning_rate": 0.003, + "loss": 3.969, + "step": 25996 + }, + { + "epoch": 0.25997, + "grad_norm": 0.6862270797626431, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 25997 + }, + { + "epoch": 0.25998, + "grad_norm": 0.7673534399819706, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 25998 + }, + { + "epoch": 0.25999, + "grad_norm": 0.8025680177636404, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 25999 + }, + { + "epoch": 0.26, + "grad_norm": 0.8116394010362521, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 26000 + }, + { + "epoch": 0.26001, + "grad_norm": 0.8822751739490251, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 26001 + }, + { + "epoch": 0.26002, + "grad_norm": 0.9830082993068349, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 26002 + }, + { + "epoch": 0.26003, + "grad_norm": 0.9659066362638261, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 26003 + }, + { + "epoch": 0.26004, + "grad_norm": 0.823004004709134, + "learning_rate": 0.003, + "loss": 4.031, + "step": 26004 + }, + { + "epoch": 0.26005, + "grad_norm": 0.8184006404000064, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 26005 + }, + { + "epoch": 0.26006, + "grad_norm": 0.921414383560556, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 26006 + }, + { + "epoch": 0.26007, + "grad_norm": 1.014259596594752, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 26007 + }, + { + "epoch": 0.26008, + "grad_norm": 1.069109935044181, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 26008 + }, + { + "epoch": 0.26009, + "grad_norm": 0.8704330151431847, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 26009 + }, + { + "epoch": 0.2601, + "grad_norm": 0.826119270848297, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 26010 + }, + { + "epoch": 0.26011, + "grad_norm": 0.7704523175909997, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 26011 + }, + { + "epoch": 0.26012, + "grad_norm": 0.8038520625066321, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 26012 + }, + { + "epoch": 0.26013, + "grad_norm": 0.7277619626579584, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 26013 + }, + { + "epoch": 0.26014, + "grad_norm": 0.6634909625550028, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 26014 + }, + { + "epoch": 0.26015, + "grad_norm": 0.6196171732449021, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 26015 + }, + { + "epoch": 0.26016, + "grad_norm": 0.6361044289954033, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 26016 + }, + { + "epoch": 0.26017, + "grad_norm": 0.7004765663283469, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 26017 + }, + { + "epoch": 0.26018, + "grad_norm": 0.6466800094354244, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 26018 + }, + { + "epoch": 0.26019, + "grad_norm": 0.6863385477219729, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 26019 + }, + { + "epoch": 0.2602, + "grad_norm": 0.7110950404927185, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 26020 + }, + { + "epoch": 0.26021, + "grad_norm": 0.673349294993344, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 26021 + }, + { + "epoch": 0.26022, + "grad_norm": 0.6424167867284339, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 26022 + }, + { + "epoch": 0.26023, + "grad_norm": 0.6592475130904633, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 26023 + }, + { + "epoch": 0.26024, + "grad_norm": 0.6394890358451556, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 26024 + }, + { + "epoch": 0.26025, + "grad_norm": 0.6809225374281279, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 26025 + }, + { + "epoch": 0.26026, + "grad_norm": 0.8790673436019829, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 26026 + }, + { + "epoch": 0.26027, + "grad_norm": 0.9551670546266879, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 26027 + }, + { + "epoch": 0.26028, + "grad_norm": 0.790954806427979, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 26028 + }, + { + "epoch": 0.26029, + "grad_norm": 0.7677118020659153, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 26029 + }, + { + "epoch": 0.2603, + "grad_norm": 0.8829638031561572, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 26030 + }, + { + "epoch": 0.26031, + "grad_norm": 0.860576887573745, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 26031 + }, + { + "epoch": 0.26032, + "grad_norm": 0.7248003057960902, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 26032 + }, + { + "epoch": 0.26033, + "grad_norm": 0.7535191025573703, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 26033 + }, + { + "epoch": 0.26034, + "grad_norm": 0.7959862292720805, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 26034 + }, + { + "epoch": 0.26035, + "grad_norm": 0.717717868714385, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 26035 + }, + { + "epoch": 0.26036, + "grad_norm": 0.715610643146189, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 26036 + }, + { + "epoch": 0.26037, + "grad_norm": 0.7434578627195779, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 26037 + }, + { + "epoch": 0.26038, + "grad_norm": 0.6649719676952001, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 26038 + }, + { + "epoch": 0.26039, + "grad_norm": 0.6233142967479163, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 26039 + }, + { + "epoch": 0.2604, + "grad_norm": 0.7122306035517032, + "learning_rate": 0.003, + "loss": 4.015, + "step": 26040 + }, + { + "epoch": 0.26041, + "grad_norm": 0.7395196119652168, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 26041 + }, + { + "epoch": 0.26042, + "grad_norm": 0.7471400068416033, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 26042 + }, + { + "epoch": 0.26043, + "grad_norm": 0.6596524056930249, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 26043 + }, + { + "epoch": 0.26044, + "grad_norm": 0.6539178582168308, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 26044 + }, + { + "epoch": 0.26045, + "grad_norm": 0.6571529604805604, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 26045 + }, + { + "epoch": 0.26046, + "grad_norm": 0.7676075230647341, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 26046 + }, + { + "epoch": 0.26047, + "grad_norm": 0.792554937994675, + "learning_rate": 0.003, + "loss": 3.985, + "step": 26047 + }, + { + "epoch": 0.26048, + "grad_norm": 0.9279911071972702, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 26048 + }, + { + "epoch": 0.26049, + "grad_norm": 1.0670641045379186, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 26049 + }, + { + "epoch": 0.2605, + "grad_norm": 1.0737416321482858, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 26050 + }, + { + "epoch": 0.26051, + "grad_norm": 0.8069685428417478, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 26051 + }, + { + "epoch": 0.26052, + "grad_norm": 0.7307079430435364, + "learning_rate": 0.003, + "loss": 3.974, + "step": 26052 + }, + { + "epoch": 0.26053, + "grad_norm": 0.6691255892738766, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 26053 + }, + { + "epoch": 0.26054, + "grad_norm": 0.6687867120654216, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 26054 + }, + { + "epoch": 0.26055, + "grad_norm": 0.5678160456693774, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 26055 + }, + { + "epoch": 0.26056, + "grad_norm": 0.6496936904361467, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 26056 + }, + { + "epoch": 0.26057, + "grad_norm": 0.6710734764496329, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26057 + }, + { + "epoch": 0.26058, + "grad_norm": 0.7433835916138102, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 26058 + }, + { + "epoch": 0.26059, + "grad_norm": 0.8243020210017044, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 26059 + }, + { + "epoch": 0.2606, + "grad_norm": 0.6931598190026206, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 26060 + }, + { + "epoch": 0.26061, + "grad_norm": 0.675707125530016, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 26061 + }, + { + "epoch": 0.26062, + "grad_norm": 0.8134062869907965, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 26062 + }, + { + "epoch": 0.26063, + "grad_norm": 0.7785705523395529, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 26063 + }, + { + "epoch": 0.26064, + "grad_norm": 0.6429225562718363, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 26064 + }, + { + "epoch": 0.26065, + "grad_norm": 0.6640521618351013, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 26065 + }, + { + "epoch": 0.26066, + "grad_norm": 0.659333468666555, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 26066 + }, + { + "epoch": 0.26067, + "grad_norm": 0.7586399715526804, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26067 + }, + { + "epoch": 0.26068, + "grad_norm": 0.8119124777894898, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 26068 + }, + { + "epoch": 0.26069, + "grad_norm": 0.8522691395269644, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 26069 + }, + { + "epoch": 0.2607, + "grad_norm": 0.7866273973656637, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 26070 + }, + { + "epoch": 0.26071, + "grad_norm": 0.7478352712307655, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 26071 + }, + { + "epoch": 0.26072, + "grad_norm": 0.7143481493300967, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 26072 + }, + { + "epoch": 0.26073, + "grad_norm": 0.8202527513403985, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 26073 + }, + { + "epoch": 0.26074, + "grad_norm": 0.993769823052839, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 26074 + }, + { + "epoch": 0.26075, + "grad_norm": 1.0177548480657497, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 26075 + }, + { + "epoch": 0.26076, + "grad_norm": 0.9922606350560447, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 26076 + }, + { + "epoch": 0.26077, + "grad_norm": 1.10750926141, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 26077 + }, + { + "epoch": 0.26078, + "grad_norm": 1.0304497648587116, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 26078 + }, + { + "epoch": 0.26079, + "grad_norm": 0.8559337493583932, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26079 + }, + { + "epoch": 0.2608, + "grad_norm": 0.8416454514786842, + "learning_rate": 0.003, + "loss": 4.0596, + "step": 26080 + }, + { + "epoch": 0.26081, + "grad_norm": 0.7330982648846999, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 26081 + }, + { + "epoch": 0.26082, + "grad_norm": 0.6761588584000612, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 26082 + }, + { + "epoch": 0.26083, + "grad_norm": 0.6874275489689876, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 26083 + }, + { + "epoch": 0.26084, + "grad_norm": 0.722389723912535, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 26084 + }, + { + "epoch": 0.26085, + "grad_norm": 0.7283265356724368, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 26085 + }, + { + "epoch": 0.26086, + "grad_norm": 0.7096047724582384, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 26086 + }, + { + "epoch": 0.26087, + "grad_norm": 0.6845194904217047, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 26087 + }, + { + "epoch": 0.26088, + "grad_norm": 0.6385751164705501, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 26088 + }, + { + "epoch": 0.26089, + "grad_norm": 0.5961053690973993, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 26089 + }, + { + "epoch": 0.2609, + "grad_norm": 0.6203946124200477, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 26090 + }, + { + "epoch": 0.26091, + "grad_norm": 0.529737593248414, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 26091 + }, + { + "epoch": 0.26092, + "grad_norm": 0.557023162706738, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 26092 + }, + { + "epoch": 0.26093, + "grad_norm": 0.6789865779383332, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 26093 + }, + { + "epoch": 0.26094, + "grad_norm": 0.7345193598398556, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 26094 + }, + { + "epoch": 0.26095, + "grad_norm": 0.6653971261781838, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 26095 + }, + { + "epoch": 0.26096, + "grad_norm": 0.5994198795333124, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 26096 + }, + { + "epoch": 0.26097, + "grad_norm": 0.6293696398356655, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26097 + }, + { + "epoch": 0.26098, + "grad_norm": 0.582270086516519, + "learning_rate": 0.003, + "loss": 3.987, + "step": 26098 + }, + { + "epoch": 0.26099, + "grad_norm": 0.5493201189918787, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 26099 + }, + { + "epoch": 0.261, + "grad_norm": 0.5094384813825007, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 26100 + }, + { + "epoch": 0.26101, + "grad_norm": 0.5563578178497137, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 26101 + }, + { + "epoch": 0.26102, + "grad_norm": 0.5964168039601315, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 26102 + }, + { + "epoch": 0.26103, + "grad_norm": 0.7644318485678426, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 26103 + }, + { + "epoch": 0.26104, + "grad_norm": 0.9095760060490704, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 26104 + }, + { + "epoch": 0.26105, + "grad_norm": 0.9175919688352919, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 26105 + }, + { + "epoch": 0.26106, + "grad_norm": 0.8461657955165931, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 26106 + }, + { + "epoch": 0.26107, + "grad_norm": 0.9374657794562733, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 26107 + }, + { + "epoch": 0.26108, + "grad_norm": 1.1117285297775497, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 26108 + }, + { + "epoch": 0.26109, + "grad_norm": 0.9797303526285226, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 26109 + }, + { + "epoch": 0.2611, + "grad_norm": 0.7945588343397423, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 26110 + }, + { + "epoch": 0.26111, + "grad_norm": 0.8111726755079948, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 26111 + }, + { + "epoch": 0.26112, + "grad_norm": 0.8776448126122436, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 26112 + }, + { + "epoch": 0.26113, + "grad_norm": 0.8383392396317223, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 26113 + }, + { + "epoch": 0.26114, + "grad_norm": 0.677021133799328, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 26114 + }, + { + "epoch": 0.26115, + "grad_norm": 0.6610868701683519, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 26115 + }, + { + "epoch": 0.26116, + "grad_norm": 0.5797462092374662, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 26116 + }, + { + "epoch": 0.26117, + "grad_norm": 0.5905981723078754, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 26117 + }, + { + "epoch": 0.26118, + "grad_norm": 0.6132215556758122, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 26118 + }, + { + "epoch": 0.26119, + "grad_norm": 0.735387192090587, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 26119 + }, + { + "epoch": 0.2612, + "grad_norm": 0.769630235375842, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 26120 + }, + { + "epoch": 0.26121, + "grad_norm": 0.7532910678861705, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 26121 + }, + { + "epoch": 0.26122, + "grad_norm": 0.8835102577321698, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 26122 + }, + { + "epoch": 0.26123, + "grad_norm": 0.9073633868947663, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 26123 + }, + { + "epoch": 0.26124, + "grad_norm": 0.7842887150390577, + "learning_rate": 0.003, + "loss": 3.993, + "step": 26124 + }, + { + "epoch": 0.26125, + "grad_norm": 0.8232996121775272, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 26125 + }, + { + "epoch": 0.26126, + "grad_norm": 0.8098494640947403, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 26126 + }, + { + "epoch": 0.26127, + "grad_norm": 0.9140064879258748, + "learning_rate": 0.003, + "loss": 3.981, + "step": 26127 + }, + { + "epoch": 0.26128, + "grad_norm": 1.0653241057598908, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 26128 + }, + { + "epoch": 0.26129, + "grad_norm": 1.0058837477500961, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 26129 + }, + { + "epoch": 0.2613, + "grad_norm": 0.9746627877492858, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 26130 + }, + { + "epoch": 0.26131, + "grad_norm": 0.8231004167438208, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 26131 + }, + { + "epoch": 0.26132, + "grad_norm": 0.7608830426100928, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 26132 + }, + { + "epoch": 0.26133, + "grad_norm": 0.6570830510539312, + "learning_rate": 0.003, + "loss": 3.998, + "step": 26133 + }, + { + "epoch": 0.26134, + "grad_norm": 0.6789247111899815, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 26134 + }, + { + "epoch": 0.26135, + "grad_norm": 0.6896652417866405, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 26135 + }, + { + "epoch": 0.26136, + "grad_norm": 0.7724563139412732, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 26136 + }, + { + "epoch": 0.26137, + "grad_norm": 0.7921459711697307, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 26137 + }, + { + "epoch": 0.26138, + "grad_norm": 0.8178202347327709, + "learning_rate": 0.003, + "loss": 4.024, + "step": 26138 + }, + { + "epoch": 0.26139, + "grad_norm": 0.7708039724883085, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 26139 + }, + { + "epoch": 0.2614, + "grad_norm": 0.7460708301284401, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 26140 + }, + { + "epoch": 0.26141, + "grad_norm": 0.7112996272887191, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 26141 + }, + { + "epoch": 0.26142, + "grad_norm": 0.6543843253989641, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 26142 + }, + { + "epoch": 0.26143, + "grad_norm": 0.6045253423220172, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 26143 + }, + { + "epoch": 0.26144, + "grad_norm": 0.5597015758596728, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 26144 + }, + { + "epoch": 0.26145, + "grad_norm": 0.6154376309628511, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 26145 + }, + { + "epoch": 0.26146, + "grad_norm": 0.6508882445767316, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 26146 + }, + { + "epoch": 0.26147, + "grad_norm": 0.7652077135620806, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 26147 + }, + { + "epoch": 0.26148, + "grad_norm": 0.8496304641520322, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 26148 + }, + { + "epoch": 0.26149, + "grad_norm": 0.8332112824143975, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 26149 + }, + { + "epoch": 0.2615, + "grad_norm": 0.7300216149799403, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 26150 + }, + { + "epoch": 0.26151, + "grad_norm": 0.6260719286806871, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 26151 + }, + { + "epoch": 0.26152, + "grad_norm": 0.6391254952257498, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 26152 + }, + { + "epoch": 0.26153, + "grad_norm": 0.6245135201459748, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 26153 + }, + { + "epoch": 0.26154, + "grad_norm": 0.6881907050902347, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 26154 + }, + { + "epoch": 0.26155, + "grad_norm": 0.744381343236013, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 26155 + }, + { + "epoch": 0.26156, + "grad_norm": 0.7969569165937368, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 26156 + }, + { + "epoch": 0.26157, + "grad_norm": 0.7804072641400521, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 26157 + }, + { + "epoch": 0.26158, + "grad_norm": 0.8393698897650171, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 26158 + }, + { + "epoch": 0.26159, + "grad_norm": 0.8322519982755444, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 26159 + }, + { + "epoch": 0.2616, + "grad_norm": 0.7256456694841533, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 26160 + }, + { + "epoch": 0.26161, + "grad_norm": 0.599322727734571, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 26161 + }, + { + "epoch": 0.26162, + "grad_norm": 0.5902853305746428, + "learning_rate": 0.003, + "loss": 4.007, + "step": 26162 + }, + { + "epoch": 0.26163, + "grad_norm": 0.7152073022993888, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 26163 + }, + { + "epoch": 0.26164, + "grad_norm": 0.7797109425573814, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 26164 + }, + { + "epoch": 0.26165, + "grad_norm": 0.7389012277784791, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26165 + }, + { + "epoch": 0.26166, + "grad_norm": 0.7937827042006101, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 26166 + }, + { + "epoch": 0.26167, + "grad_norm": 0.768409910705224, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 26167 + }, + { + "epoch": 0.26168, + "grad_norm": 0.7164565521745153, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 26168 + }, + { + "epoch": 0.26169, + "grad_norm": 0.6884413354849003, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 26169 + }, + { + "epoch": 0.2617, + "grad_norm": 0.6506468779678474, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 26170 + }, + { + "epoch": 0.26171, + "grad_norm": 0.7022309880181435, + "learning_rate": 0.003, + "loss": 3.984, + "step": 26171 + }, + { + "epoch": 0.26172, + "grad_norm": 0.7395518618272883, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 26172 + }, + { + "epoch": 0.26173, + "grad_norm": 0.8292726135557593, + "learning_rate": 0.003, + "loss": 4.005, + "step": 26173 + }, + { + "epoch": 0.26174, + "grad_norm": 1.0320383507625723, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 26174 + }, + { + "epoch": 0.26175, + "grad_norm": 1.1798772337579444, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 26175 + }, + { + "epoch": 0.26176, + "grad_norm": 0.725000277974863, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 26176 + }, + { + "epoch": 0.26177, + "grad_norm": 0.7401059380633696, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 26177 + }, + { + "epoch": 0.26178, + "grad_norm": 0.8797222463240028, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 26178 + }, + { + "epoch": 0.26179, + "grad_norm": 0.9726815534757691, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 26179 + }, + { + "epoch": 0.2618, + "grad_norm": 1.0829890787012912, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 26180 + }, + { + "epoch": 0.26181, + "grad_norm": 0.9108213468826298, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 26181 + }, + { + "epoch": 0.26182, + "grad_norm": 0.9471245256045061, + "learning_rate": 0.003, + "loss": 3.994, + "step": 26182 + }, + { + "epoch": 0.26183, + "grad_norm": 1.0569263949322745, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 26183 + }, + { + "epoch": 0.26184, + "grad_norm": 0.8049260050041577, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 26184 + }, + { + "epoch": 0.26185, + "grad_norm": 0.8553396783955233, + "learning_rate": 0.003, + "loss": 3.989, + "step": 26185 + }, + { + "epoch": 0.26186, + "grad_norm": 0.8291801440283996, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 26186 + }, + { + "epoch": 0.26187, + "grad_norm": 0.7987633945544224, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 26187 + }, + { + "epoch": 0.26188, + "grad_norm": 0.8790242962015699, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 26188 + }, + { + "epoch": 0.26189, + "grad_norm": 0.8579693665233601, + "learning_rate": 0.003, + "loss": 3.974, + "step": 26189 + }, + { + "epoch": 0.2619, + "grad_norm": 0.8166629640466326, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 26190 + }, + { + "epoch": 0.26191, + "grad_norm": 0.7966126603038657, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 26191 + }, + { + "epoch": 0.26192, + "grad_norm": 0.6243296748853999, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 26192 + }, + { + "epoch": 0.26193, + "grad_norm": 0.5778218898617955, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 26193 + }, + { + "epoch": 0.26194, + "grad_norm": 0.5924462033435215, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 26194 + }, + { + "epoch": 0.26195, + "grad_norm": 0.668257076993702, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 26195 + }, + { + "epoch": 0.26196, + "grad_norm": 0.6829420850354883, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 26196 + }, + { + "epoch": 0.26197, + "grad_norm": 0.7007100003161213, + "learning_rate": 0.003, + "loss": 3.994, + "step": 26197 + }, + { + "epoch": 0.26198, + "grad_norm": 0.7046826198081203, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 26198 + }, + { + "epoch": 0.26199, + "grad_norm": 0.8092639970739356, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 26199 + }, + { + "epoch": 0.262, + "grad_norm": 1.0470631682927052, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 26200 + }, + { + "epoch": 0.26201, + "grad_norm": 1.0110897585679768, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 26201 + }, + { + "epoch": 0.26202, + "grad_norm": 0.8745929918770774, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 26202 + }, + { + "epoch": 0.26203, + "grad_norm": 0.6604714221242695, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 26203 + }, + { + "epoch": 0.26204, + "grad_norm": 0.7427948307616672, + "learning_rate": 0.003, + "loss": 3.986, + "step": 26204 + }, + { + "epoch": 0.26205, + "grad_norm": 0.8489850819680498, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 26205 + }, + { + "epoch": 0.26206, + "grad_norm": 0.8320494829168119, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 26206 + }, + { + "epoch": 0.26207, + "grad_norm": 0.6689013978095567, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 26207 + }, + { + "epoch": 0.26208, + "grad_norm": 0.6649212479052008, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 26208 + }, + { + "epoch": 0.26209, + "grad_norm": 0.6612715082228301, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26209 + }, + { + "epoch": 0.2621, + "grad_norm": 0.6098941851635603, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 26210 + }, + { + "epoch": 0.26211, + "grad_norm": 0.6155973131801863, + "learning_rate": 0.003, + "loss": 3.9474, + "step": 26211 + }, + { + "epoch": 0.26212, + "grad_norm": 0.6587314207785143, + "learning_rate": 0.003, + "loss": 4.031, + "step": 26212 + }, + { + "epoch": 0.26213, + "grad_norm": 0.7095264073179519, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26213 + }, + { + "epoch": 0.26214, + "grad_norm": 0.5859500360717373, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 26214 + }, + { + "epoch": 0.26215, + "grad_norm": 0.5641473990883844, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 26215 + }, + { + "epoch": 0.26216, + "grad_norm": 0.5470455177967265, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 26216 + }, + { + "epoch": 0.26217, + "grad_norm": 0.6369239019732527, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 26217 + }, + { + "epoch": 0.26218, + "grad_norm": 0.643424414174699, + "learning_rate": 0.003, + "loss": 3.979, + "step": 26218 + }, + { + "epoch": 0.26219, + "grad_norm": 0.6675734624682966, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26219 + }, + { + "epoch": 0.2622, + "grad_norm": 0.8309223805624838, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 26220 + }, + { + "epoch": 0.26221, + "grad_norm": 0.8755972287052307, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 26221 + }, + { + "epoch": 0.26222, + "grad_norm": 1.014566099929869, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 26222 + }, + { + "epoch": 0.26223, + "grad_norm": 0.8973774443440645, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 26223 + }, + { + "epoch": 0.26224, + "grad_norm": 0.775166520568272, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 26224 + }, + { + "epoch": 0.26225, + "grad_norm": 0.8759025922088268, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 26225 + }, + { + "epoch": 0.26226, + "grad_norm": 0.6775260802465609, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 26226 + }, + { + "epoch": 0.26227, + "grad_norm": 0.5525123809377475, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 26227 + }, + { + "epoch": 0.26228, + "grad_norm": 0.5857638011000433, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 26228 + }, + { + "epoch": 0.26229, + "grad_norm": 0.6495691122974436, + "learning_rate": 0.003, + "loss": 3.981, + "step": 26229 + }, + { + "epoch": 0.2623, + "grad_norm": 0.5875695471999056, + "learning_rate": 0.003, + "loss": 3.991, + "step": 26230 + }, + { + "epoch": 0.26231, + "grad_norm": 0.6087440041110799, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 26231 + }, + { + "epoch": 0.26232, + "grad_norm": 0.6380012808727954, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 26232 + }, + { + "epoch": 0.26233, + "grad_norm": 0.6466048992345179, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 26233 + }, + { + "epoch": 0.26234, + "grad_norm": 0.6397404857341289, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 26234 + }, + { + "epoch": 0.26235, + "grad_norm": 0.7383820580720589, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 26235 + }, + { + "epoch": 0.26236, + "grad_norm": 0.8917673266132623, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 26236 + }, + { + "epoch": 0.26237, + "grad_norm": 1.046833047087125, + "learning_rate": 0.003, + "loss": 4.008, + "step": 26237 + }, + { + "epoch": 0.26238, + "grad_norm": 1.148049435493214, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 26238 + }, + { + "epoch": 0.26239, + "grad_norm": 0.7003032695658858, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 26239 + }, + { + "epoch": 0.2624, + "grad_norm": 0.7745526148184764, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 26240 + }, + { + "epoch": 0.26241, + "grad_norm": 0.824572503455801, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 26241 + }, + { + "epoch": 0.26242, + "grad_norm": 0.8104024637273716, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 26242 + }, + { + "epoch": 0.26243, + "grad_norm": 0.8000180806173033, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 26243 + }, + { + "epoch": 0.26244, + "grad_norm": 0.8325510615979714, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 26244 + }, + { + "epoch": 0.26245, + "grad_norm": 0.7425622703951866, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 26245 + }, + { + "epoch": 0.26246, + "grad_norm": 0.6542429968086126, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 26246 + }, + { + "epoch": 0.26247, + "grad_norm": 0.6267838509860705, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 26247 + }, + { + "epoch": 0.26248, + "grad_norm": 0.6722418554903627, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 26248 + }, + { + "epoch": 0.26249, + "grad_norm": 0.6830021025898023, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 26249 + }, + { + "epoch": 0.2625, + "grad_norm": 0.8020614927640074, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 26250 + }, + { + "epoch": 0.26251, + "grad_norm": 0.7849537593892247, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 26251 + }, + { + "epoch": 0.26252, + "grad_norm": 0.7829070750203108, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 26252 + }, + { + "epoch": 0.26253, + "grad_norm": 0.816837724337339, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 26253 + }, + { + "epoch": 0.26254, + "grad_norm": 0.7918166374449456, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 26254 + }, + { + "epoch": 0.26255, + "grad_norm": 0.8227284291263228, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 26255 + }, + { + "epoch": 0.26256, + "grad_norm": 0.885648298612498, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 26256 + }, + { + "epoch": 0.26257, + "grad_norm": 0.9584280427890032, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 26257 + }, + { + "epoch": 0.26258, + "grad_norm": 0.8893919479862507, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 26258 + }, + { + "epoch": 0.26259, + "grad_norm": 0.891883932950665, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 26259 + }, + { + "epoch": 0.2626, + "grad_norm": 0.9108285649672216, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 26260 + }, + { + "epoch": 0.26261, + "grad_norm": 0.9506961838498585, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 26261 + }, + { + "epoch": 0.26262, + "grad_norm": 1.0041202060558028, + "learning_rate": 0.003, + "loss": 3.999, + "step": 26262 + }, + { + "epoch": 0.26263, + "grad_norm": 0.9779472359920998, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 26263 + }, + { + "epoch": 0.26264, + "grad_norm": 0.9591674604730607, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 26264 + }, + { + "epoch": 0.26265, + "grad_norm": 0.8120307755811267, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 26265 + }, + { + "epoch": 0.26266, + "grad_norm": 0.8487861450953242, + "learning_rate": 0.003, + "loss": 3.997, + "step": 26266 + }, + { + "epoch": 0.26267, + "grad_norm": 0.947709523753156, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 26267 + }, + { + "epoch": 0.26268, + "grad_norm": 0.9303157297238327, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 26268 + }, + { + "epoch": 0.26269, + "grad_norm": 0.8850525797944616, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 26269 + }, + { + "epoch": 0.2627, + "grad_norm": 0.9413519105095204, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 26270 + }, + { + "epoch": 0.26271, + "grad_norm": 0.9188442532655978, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 26271 + }, + { + "epoch": 0.26272, + "grad_norm": 0.8609247137315225, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 26272 + }, + { + "epoch": 0.26273, + "grad_norm": 0.8279446243950823, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 26273 + }, + { + "epoch": 0.26274, + "grad_norm": 0.7573194527476875, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 26274 + }, + { + "epoch": 0.26275, + "grad_norm": 0.7678455056340567, + "learning_rate": 0.003, + "loss": 3.971, + "step": 26275 + }, + { + "epoch": 0.26276, + "grad_norm": 0.6458043876636899, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 26276 + }, + { + "epoch": 0.26277, + "grad_norm": 0.6584955739819889, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 26277 + }, + { + "epoch": 0.26278, + "grad_norm": 0.666547639520703, + "learning_rate": 0.003, + "loss": 3.96, + "step": 26278 + }, + { + "epoch": 0.26279, + "grad_norm": 0.5728389379262726, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 26279 + }, + { + "epoch": 0.2628, + "grad_norm": 0.5612444692584716, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 26280 + }, + { + "epoch": 0.26281, + "grad_norm": 0.4254473163068477, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 26281 + }, + { + "epoch": 0.26282, + "grad_norm": 0.4840543952845364, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 26282 + }, + { + "epoch": 0.26283, + "grad_norm": 0.48520555618875144, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 26283 + }, + { + "epoch": 0.26284, + "grad_norm": 0.5703227747761753, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 26284 + }, + { + "epoch": 0.26285, + "grad_norm": 0.5973343806446286, + "learning_rate": 0.003, + "loss": 3.972, + "step": 26285 + }, + { + "epoch": 0.26286, + "grad_norm": 0.5970397000352272, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 26286 + }, + { + "epoch": 0.26287, + "grad_norm": 0.5894621907365729, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 26287 + }, + { + "epoch": 0.26288, + "grad_norm": 0.6212238552885954, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 26288 + }, + { + "epoch": 0.26289, + "grad_norm": 0.635731809348941, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 26289 + }, + { + "epoch": 0.2629, + "grad_norm": 0.7455946244426015, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 26290 + }, + { + "epoch": 0.26291, + "grad_norm": 0.886669139321867, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 26291 + }, + { + "epoch": 0.26292, + "grad_norm": 0.895967610860179, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 26292 + }, + { + "epoch": 0.26293, + "grad_norm": 0.8384576781274317, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 26293 + }, + { + "epoch": 0.26294, + "grad_norm": 0.7718217292488896, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 26294 + }, + { + "epoch": 0.26295, + "grad_norm": 0.7878856675167177, + "learning_rate": 0.003, + "loss": 3.991, + "step": 26295 + }, + { + "epoch": 0.26296, + "grad_norm": 0.711665441963161, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 26296 + }, + { + "epoch": 0.26297, + "grad_norm": 0.6917900202260475, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 26297 + }, + { + "epoch": 0.26298, + "grad_norm": 0.6355889660202451, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 26298 + }, + { + "epoch": 0.26299, + "grad_norm": 0.6891416576272283, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 26299 + }, + { + "epoch": 0.263, + "grad_norm": 0.7228057362905981, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 26300 + }, + { + "epoch": 0.26301, + "grad_norm": 0.7907841525113859, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 26301 + }, + { + "epoch": 0.26302, + "grad_norm": 0.7172430469511146, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 26302 + }, + { + "epoch": 0.26303, + "grad_norm": 0.652410596638339, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 26303 + }, + { + "epoch": 0.26304, + "grad_norm": 0.6207964885805554, + "learning_rate": 0.003, + "loss": 3.994, + "step": 26304 + }, + { + "epoch": 0.26305, + "grad_norm": 0.6712598647881084, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 26305 + }, + { + "epoch": 0.26306, + "grad_norm": 0.7613465952906502, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 26306 + }, + { + "epoch": 0.26307, + "grad_norm": 0.8541293904057283, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 26307 + }, + { + "epoch": 0.26308, + "grad_norm": 0.9592966492739377, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 26308 + }, + { + "epoch": 0.26309, + "grad_norm": 0.9592180336628212, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 26309 + }, + { + "epoch": 0.2631, + "grad_norm": 0.95506273908706, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 26310 + }, + { + "epoch": 0.26311, + "grad_norm": 1.1119421279308885, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 26311 + }, + { + "epoch": 0.26312, + "grad_norm": 1.0198823175369107, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 26312 + }, + { + "epoch": 0.26313, + "grad_norm": 1.0436725654902035, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 26313 + }, + { + "epoch": 0.26314, + "grad_norm": 1.1287859273543115, + "learning_rate": 0.003, + "loss": 4.038, + "step": 26314 + }, + { + "epoch": 0.26315, + "grad_norm": 0.8027321243020565, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 26315 + }, + { + "epoch": 0.26316, + "grad_norm": 0.8194057647724963, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 26316 + }, + { + "epoch": 0.26317, + "grad_norm": 0.7594625375756923, + "learning_rate": 0.003, + "loss": 4.013, + "step": 26317 + }, + { + "epoch": 0.26318, + "grad_norm": 0.8821783569929856, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 26318 + }, + { + "epoch": 0.26319, + "grad_norm": 1.029738785563517, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 26319 + }, + { + "epoch": 0.2632, + "grad_norm": 1.099286059999389, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 26320 + }, + { + "epoch": 0.26321, + "grad_norm": 0.9756159170900421, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 26321 + }, + { + "epoch": 0.26322, + "grad_norm": 0.9248846079266425, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 26322 + }, + { + "epoch": 0.26323, + "grad_norm": 1.0855085239451194, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 26323 + }, + { + "epoch": 0.26324, + "grad_norm": 1.1214049938873627, + "learning_rate": 0.003, + "loss": 4.053, + "step": 26324 + }, + { + "epoch": 0.26325, + "grad_norm": 0.7845980285801205, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 26325 + }, + { + "epoch": 0.26326, + "grad_norm": 0.7286339918760594, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 26326 + }, + { + "epoch": 0.26327, + "grad_norm": 0.6845118351242365, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 26327 + }, + { + "epoch": 0.26328, + "grad_norm": 0.6203377730460515, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 26328 + }, + { + "epoch": 0.26329, + "grad_norm": 0.6762034441593396, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 26329 + }, + { + "epoch": 0.2633, + "grad_norm": 0.7198284990679666, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 26330 + }, + { + "epoch": 0.26331, + "grad_norm": 0.7094139849395142, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 26331 + }, + { + "epoch": 0.26332, + "grad_norm": 0.6319003294774005, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 26332 + }, + { + "epoch": 0.26333, + "grad_norm": 0.6355603623594617, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 26333 + }, + { + "epoch": 0.26334, + "grad_norm": 0.544334925920631, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 26334 + }, + { + "epoch": 0.26335, + "grad_norm": 0.41195520377233175, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 26335 + }, + { + "epoch": 0.26336, + "grad_norm": 0.43011580628511853, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 26336 + }, + { + "epoch": 0.26337, + "grad_norm": 0.3903727252660624, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 26337 + }, + { + "epoch": 0.26338, + "grad_norm": 0.4297539104230795, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 26338 + }, + { + "epoch": 0.26339, + "grad_norm": 0.43722562692304034, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 26339 + }, + { + "epoch": 0.2634, + "grad_norm": 0.4314807059814629, + "learning_rate": 0.003, + "loss": 4.005, + "step": 26340 + }, + { + "epoch": 0.26341, + "grad_norm": 0.55687597669832, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 26341 + }, + { + "epoch": 0.26342, + "grad_norm": 0.6805357911044635, + "learning_rate": 0.003, + "loss": 3.987, + "step": 26342 + }, + { + "epoch": 0.26343, + "grad_norm": 0.7983856349447098, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 26343 + }, + { + "epoch": 0.26344, + "grad_norm": 0.8960646896455012, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 26344 + }, + { + "epoch": 0.26345, + "grad_norm": 0.9328716568414556, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 26345 + }, + { + "epoch": 0.26346, + "grad_norm": 0.8174356855529514, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 26346 + }, + { + "epoch": 0.26347, + "grad_norm": 0.7330862988721213, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 26347 + }, + { + "epoch": 0.26348, + "grad_norm": 0.7329359989855151, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 26348 + }, + { + "epoch": 0.26349, + "grad_norm": 0.7097107082688937, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 26349 + }, + { + "epoch": 0.2635, + "grad_norm": 0.7224814929908094, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 26350 + }, + { + "epoch": 0.26351, + "grad_norm": 0.756612815624217, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 26351 + }, + { + "epoch": 0.26352, + "grad_norm": 0.8949183849404176, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 26352 + }, + { + "epoch": 0.26353, + "grad_norm": 0.883300508017289, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 26353 + }, + { + "epoch": 0.26354, + "grad_norm": 0.8761568566424887, + "learning_rate": 0.003, + "loss": 3.974, + "step": 26354 + }, + { + "epoch": 0.26355, + "grad_norm": 0.7042436126060613, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 26355 + }, + { + "epoch": 0.26356, + "grad_norm": 0.6620749324766956, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 26356 + }, + { + "epoch": 0.26357, + "grad_norm": 0.59783594698038, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 26357 + }, + { + "epoch": 0.26358, + "grad_norm": 0.606206249324627, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 26358 + }, + { + "epoch": 0.26359, + "grad_norm": 0.6036094526783191, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 26359 + }, + { + "epoch": 0.2636, + "grad_norm": 0.6870718697287079, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 26360 + }, + { + "epoch": 0.26361, + "grad_norm": 0.7961307662748317, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 26361 + }, + { + "epoch": 0.26362, + "grad_norm": 0.8862305512935246, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 26362 + }, + { + "epoch": 0.26363, + "grad_norm": 0.7920153121348741, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 26363 + }, + { + "epoch": 0.26364, + "grad_norm": 0.809737219490461, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 26364 + }, + { + "epoch": 0.26365, + "grad_norm": 0.8520893855206655, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 26365 + }, + { + "epoch": 0.26366, + "grad_norm": 0.8333670169874201, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 26366 + }, + { + "epoch": 0.26367, + "grad_norm": 0.7409629927189934, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 26367 + }, + { + "epoch": 0.26368, + "grad_norm": 0.7447805104220392, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 26368 + }, + { + "epoch": 0.26369, + "grad_norm": 0.7608464513598222, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 26369 + }, + { + "epoch": 0.2637, + "grad_norm": 0.8334381994863077, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 26370 + }, + { + "epoch": 0.26371, + "grad_norm": 0.7851823843839212, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 26371 + }, + { + "epoch": 0.26372, + "grad_norm": 0.6869230423726493, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 26372 + }, + { + "epoch": 0.26373, + "grad_norm": 0.7774541486529938, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 26373 + }, + { + "epoch": 0.26374, + "grad_norm": 0.8624430385802809, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 26374 + }, + { + "epoch": 0.26375, + "grad_norm": 0.8437587020961863, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 26375 + }, + { + "epoch": 0.26376, + "grad_norm": 0.9011804107129507, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 26376 + }, + { + "epoch": 0.26377, + "grad_norm": 0.940183810135435, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 26377 + }, + { + "epoch": 0.26378, + "grad_norm": 0.9343148474182764, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 26378 + }, + { + "epoch": 0.26379, + "grad_norm": 1.0489452308122011, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 26379 + }, + { + "epoch": 0.2638, + "grad_norm": 1.0069675411765124, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 26380 + }, + { + "epoch": 0.26381, + "grad_norm": 0.9519338391477984, + "learning_rate": 0.003, + "loss": 4.005, + "step": 26381 + }, + { + "epoch": 0.26382, + "grad_norm": 0.9066618631948455, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 26382 + }, + { + "epoch": 0.26383, + "grad_norm": 0.7913270992303794, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 26383 + }, + { + "epoch": 0.26384, + "grad_norm": 0.8785263701005728, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 26384 + }, + { + "epoch": 0.26385, + "grad_norm": 0.9137986986076582, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 26385 + }, + { + "epoch": 0.26386, + "grad_norm": 0.9166430805062935, + "learning_rate": 0.003, + "loss": 4.0469, + "step": 26386 + }, + { + "epoch": 0.26387, + "grad_norm": 0.9776918979755093, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 26387 + }, + { + "epoch": 0.26388, + "grad_norm": 0.8333669383278723, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 26388 + }, + { + "epoch": 0.26389, + "grad_norm": 0.8601068157262132, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 26389 + }, + { + "epoch": 0.2639, + "grad_norm": 0.8687156950314937, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 26390 + }, + { + "epoch": 0.26391, + "grad_norm": 0.8316036217914919, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 26391 + }, + { + "epoch": 0.26392, + "grad_norm": 0.7945321291985832, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 26392 + }, + { + "epoch": 0.26393, + "grad_norm": 0.7626670685968702, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 26393 + }, + { + "epoch": 0.26394, + "grad_norm": 0.6648700443966071, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 26394 + }, + { + "epoch": 0.26395, + "grad_norm": 0.6568061036322543, + "learning_rate": 0.003, + "loss": 3.991, + "step": 26395 + }, + { + "epoch": 0.26396, + "grad_norm": 0.7273435137890154, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 26396 + }, + { + "epoch": 0.26397, + "grad_norm": 0.7866637453734419, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 26397 + }, + { + "epoch": 0.26398, + "grad_norm": 0.8753968165237239, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 26398 + }, + { + "epoch": 0.26399, + "grad_norm": 0.8885314177097811, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 26399 + }, + { + "epoch": 0.264, + "grad_norm": 0.796688078269642, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 26400 + }, + { + "epoch": 0.26401, + "grad_norm": 0.7088025427515653, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 26401 + }, + { + "epoch": 0.26402, + "grad_norm": 0.6612870221318008, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 26402 + }, + { + "epoch": 0.26403, + "grad_norm": 0.5799935305641143, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 26403 + }, + { + "epoch": 0.26404, + "grad_norm": 0.5593147630574922, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 26404 + }, + { + "epoch": 0.26405, + "grad_norm": 0.5806028313113434, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 26405 + }, + { + "epoch": 0.26406, + "grad_norm": 0.5862491532632181, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 26406 + }, + { + "epoch": 0.26407, + "grad_norm": 0.5564663277928468, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 26407 + }, + { + "epoch": 0.26408, + "grad_norm": 0.5858735272248832, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 26408 + }, + { + "epoch": 0.26409, + "grad_norm": 0.7194863234891057, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 26409 + }, + { + "epoch": 0.2641, + "grad_norm": 0.7441312034956169, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 26410 + }, + { + "epoch": 0.26411, + "grad_norm": 0.8073612566178119, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 26411 + }, + { + "epoch": 0.26412, + "grad_norm": 0.7196750200858563, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 26412 + }, + { + "epoch": 0.26413, + "grad_norm": 0.5785245082517381, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 26413 + }, + { + "epoch": 0.26414, + "grad_norm": 0.5596176586044977, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 26414 + }, + { + "epoch": 0.26415, + "grad_norm": 0.5254245345504698, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 26415 + }, + { + "epoch": 0.26416, + "grad_norm": 0.6475905299405218, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 26416 + }, + { + "epoch": 0.26417, + "grad_norm": 0.7329946203530521, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 26417 + }, + { + "epoch": 0.26418, + "grad_norm": 0.7425966659814502, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 26418 + }, + { + "epoch": 0.26419, + "grad_norm": 0.7537601163787155, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 26419 + }, + { + "epoch": 0.2642, + "grad_norm": 0.7499369184617038, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 26420 + }, + { + "epoch": 0.26421, + "grad_norm": 0.9529277124533485, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 26421 + }, + { + "epoch": 0.26422, + "grad_norm": 1.0377226495928036, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 26422 + }, + { + "epoch": 0.26423, + "grad_norm": 0.9468730857064199, + "learning_rate": 0.003, + "loss": 4.033, + "step": 26423 + }, + { + "epoch": 0.26424, + "grad_norm": 0.851252416224338, + "learning_rate": 0.003, + "loss": 3.999, + "step": 26424 + }, + { + "epoch": 0.26425, + "grad_norm": 0.9088966197039849, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26425 + }, + { + "epoch": 0.26426, + "grad_norm": 1.0066275669101565, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 26426 + }, + { + "epoch": 0.26427, + "grad_norm": 1.033289353314282, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 26427 + }, + { + "epoch": 0.26428, + "grad_norm": 0.7744200118590315, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 26428 + }, + { + "epoch": 0.26429, + "grad_norm": 0.8571581065150157, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 26429 + }, + { + "epoch": 0.2643, + "grad_norm": 0.8215252709110393, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 26430 + }, + { + "epoch": 0.26431, + "grad_norm": 0.9015653384138416, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 26431 + }, + { + "epoch": 0.26432, + "grad_norm": 0.8897090373549321, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 26432 + }, + { + "epoch": 0.26433, + "grad_norm": 0.778138054378948, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 26433 + }, + { + "epoch": 0.26434, + "grad_norm": 0.7047087598060302, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 26434 + }, + { + "epoch": 0.26435, + "grad_norm": 0.731131861350555, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 26435 + }, + { + "epoch": 0.26436, + "grad_norm": 0.7104659056110243, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 26436 + }, + { + "epoch": 0.26437, + "grad_norm": 0.6573462242550815, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 26437 + }, + { + "epoch": 0.26438, + "grad_norm": 0.7387713928590051, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 26438 + }, + { + "epoch": 0.26439, + "grad_norm": 0.744436981562007, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 26439 + }, + { + "epoch": 0.2644, + "grad_norm": 0.7757607114314731, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 26440 + }, + { + "epoch": 0.26441, + "grad_norm": 0.7845351124402219, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 26441 + }, + { + "epoch": 0.26442, + "grad_norm": 0.723075060949612, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 26442 + }, + { + "epoch": 0.26443, + "grad_norm": 0.5970464296348136, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 26443 + }, + { + "epoch": 0.26444, + "grad_norm": 0.6286322241023924, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 26444 + }, + { + "epoch": 0.26445, + "grad_norm": 0.605088629668354, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 26445 + }, + { + "epoch": 0.26446, + "grad_norm": 0.6070148413904642, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 26446 + }, + { + "epoch": 0.26447, + "grad_norm": 0.6546811966469883, + "learning_rate": 0.003, + "loss": 4.024, + "step": 26447 + }, + { + "epoch": 0.26448, + "grad_norm": 0.7353827599312232, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 26448 + }, + { + "epoch": 0.26449, + "grad_norm": 0.7801882403044758, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 26449 + }, + { + "epoch": 0.2645, + "grad_norm": 0.7347266022411059, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 26450 + }, + { + "epoch": 0.26451, + "grad_norm": 0.7011410843640933, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 26451 + }, + { + "epoch": 0.26452, + "grad_norm": 0.7336222921860851, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 26452 + }, + { + "epoch": 0.26453, + "grad_norm": 0.7289791945437285, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 26453 + }, + { + "epoch": 0.26454, + "grad_norm": 0.7592427829971796, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 26454 + }, + { + "epoch": 0.26455, + "grad_norm": 0.8061128579964962, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 26455 + }, + { + "epoch": 0.26456, + "grad_norm": 0.8059057324447464, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 26456 + }, + { + "epoch": 0.26457, + "grad_norm": 0.7288574150274718, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 26457 + }, + { + "epoch": 0.26458, + "grad_norm": 0.7351728066805783, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 26458 + }, + { + "epoch": 0.26459, + "grad_norm": 0.755457019496774, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 26459 + }, + { + "epoch": 0.2646, + "grad_norm": 0.7704787834033859, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 26460 + }, + { + "epoch": 0.26461, + "grad_norm": 0.7708170721637932, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 26461 + }, + { + "epoch": 0.26462, + "grad_norm": 0.8263529257414622, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 26462 + }, + { + "epoch": 0.26463, + "grad_norm": 0.9052030146263234, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 26463 + }, + { + "epoch": 0.26464, + "grad_norm": 0.8918843417519189, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 26464 + }, + { + "epoch": 0.26465, + "grad_norm": 0.6940367804086317, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 26465 + }, + { + "epoch": 0.26466, + "grad_norm": 0.6170044218517776, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 26466 + }, + { + "epoch": 0.26467, + "grad_norm": 0.6908857160658133, + "learning_rate": 0.003, + "loss": 3.989, + "step": 26467 + }, + { + "epoch": 0.26468, + "grad_norm": 0.71798115380739, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 26468 + }, + { + "epoch": 0.26469, + "grad_norm": 0.685899425288107, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 26469 + }, + { + "epoch": 0.2647, + "grad_norm": 0.7469052055202383, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 26470 + }, + { + "epoch": 0.26471, + "grad_norm": 0.7897159629185788, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 26471 + }, + { + "epoch": 0.26472, + "grad_norm": 0.8233983834781099, + "learning_rate": 0.003, + "loss": 4.03, + "step": 26472 + }, + { + "epoch": 0.26473, + "grad_norm": 0.846740595518943, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 26473 + }, + { + "epoch": 0.26474, + "grad_norm": 0.9348487175850875, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 26474 + }, + { + "epoch": 0.26475, + "grad_norm": 0.8862239781594405, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 26475 + }, + { + "epoch": 0.26476, + "grad_norm": 0.9261310256723885, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 26476 + }, + { + "epoch": 0.26477, + "grad_norm": 0.977753498955085, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 26477 + }, + { + "epoch": 0.26478, + "grad_norm": 1.0073853673448017, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 26478 + }, + { + "epoch": 0.26479, + "grad_norm": 1.3983754162307647, + "learning_rate": 0.003, + "loss": 4.0597, + "step": 26479 + }, + { + "epoch": 0.2648, + "grad_norm": 0.7106713265548592, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 26480 + }, + { + "epoch": 0.26481, + "grad_norm": 0.8991037626907845, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 26481 + }, + { + "epoch": 0.26482, + "grad_norm": 0.9212227950752446, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 26482 + }, + { + "epoch": 0.26483, + "grad_norm": 1.0619941421275554, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 26483 + }, + { + "epoch": 0.26484, + "grad_norm": 0.9001349605776924, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 26484 + }, + { + "epoch": 0.26485, + "grad_norm": 0.8953221745229455, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 26485 + }, + { + "epoch": 0.26486, + "grad_norm": 0.9452473673923274, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 26486 + }, + { + "epoch": 0.26487, + "grad_norm": 1.0206071636145908, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 26487 + }, + { + "epoch": 0.26488, + "grad_norm": 0.9152603776073721, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 26488 + }, + { + "epoch": 0.26489, + "grad_norm": 1.0972588473327556, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 26489 + }, + { + "epoch": 0.2649, + "grad_norm": 1.1272879079968, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 26490 + }, + { + "epoch": 0.26491, + "grad_norm": 1.2375224695960199, + "learning_rate": 0.003, + "loss": 4.052, + "step": 26491 + }, + { + "epoch": 0.26492, + "grad_norm": 1.1875917188927974, + "learning_rate": 0.003, + "loss": 4.0383, + "step": 26492 + }, + { + "epoch": 0.26493, + "grad_norm": 1.120689610229112, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 26493 + }, + { + "epoch": 0.26494, + "grad_norm": 1.0193672124631388, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 26494 + }, + { + "epoch": 0.26495, + "grad_norm": 0.9479607455008817, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 26495 + }, + { + "epoch": 0.26496, + "grad_norm": 1.0590478247937392, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 26496 + }, + { + "epoch": 0.26497, + "grad_norm": 1.1194834045478064, + "learning_rate": 0.003, + "loss": 4.0426, + "step": 26497 + }, + { + "epoch": 0.26498, + "grad_norm": 1.39188139030375, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 26498 + }, + { + "epoch": 0.26499, + "grad_norm": 1.0954887135183187, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 26499 + }, + { + "epoch": 0.265, + "grad_norm": 0.9852146420567137, + "learning_rate": 0.003, + "loss": 4.0557, + "step": 26500 + }, + { + "epoch": 0.26501, + "grad_norm": 1.097556711276283, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 26501 + }, + { + "epoch": 0.26502, + "grad_norm": 0.7846674562051491, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 26502 + }, + { + "epoch": 0.26503, + "grad_norm": 0.746163659413626, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 26503 + }, + { + "epoch": 0.26504, + "grad_norm": 0.7739427072601085, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 26504 + }, + { + "epoch": 0.26505, + "grad_norm": 0.7434436270680155, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 26505 + }, + { + "epoch": 0.26506, + "grad_norm": 0.606148583168174, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 26506 + }, + { + "epoch": 0.26507, + "grad_norm": 0.6132190654951946, + "learning_rate": 0.003, + "loss": 4.007, + "step": 26507 + }, + { + "epoch": 0.26508, + "grad_norm": 0.717720028763801, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 26508 + }, + { + "epoch": 0.26509, + "grad_norm": 0.8389157831323801, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 26509 + }, + { + "epoch": 0.2651, + "grad_norm": 1.0885277607017927, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 26510 + }, + { + "epoch": 0.26511, + "grad_norm": 0.9220952114711308, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 26511 + }, + { + "epoch": 0.26512, + "grad_norm": 0.9037648862550921, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 26512 + }, + { + "epoch": 0.26513, + "grad_norm": 0.8184524226947304, + "learning_rate": 0.003, + "loss": 3.999, + "step": 26513 + }, + { + "epoch": 0.26514, + "grad_norm": 0.7065495644517304, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 26514 + }, + { + "epoch": 0.26515, + "grad_norm": 0.5635824967308926, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 26515 + }, + { + "epoch": 0.26516, + "grad_norm": 0.5451007133171146, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 26516 + }, + { + "epoch": 0.26517, + "grad_norm": 0.49979392194083827, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 26517 + }, + { + "epoch": 0.26518, + "grad_norm": 0.5058797910642798, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 26518 + }, + { + "epoch": 0.26519, + "grad_norm": 0.5152131935460921, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 26519 + }, + { + "epoch": 0.2652, + "grad_norm": 0.498919235523642, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 26520 + }, + { + "epoch": 0.26521, + "grad_norm": 0.4981155390888175, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 26521 + }, + { + "epoch": 0.26522, + "grad_norm": 0.5222008569522782, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 26522 + }, + { + "epoch": 0.26523, + "grad_norm": 0.655968215161332, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 26523 + }, + { + "epoch": 0.26524, + "grad_norm": 0.9511220336392247, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 26524 + }, + { + "epoch": 0.26525, + "grad_norm": 1.199850895006818, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 26525 + }, + { + "epoch": 0.26526, + "grad_norm": 0.7551666237806448, + "learning_rate": 0.003, + "loss": 4.0, + "step": 26526 + }, + { + "epoch": 0.26527, + "grad_norm": 0.7221392750006405, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 26527 + }, + { + "epoch": 0.26528, + "grad_norm": 0.7606041997243578, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 26528 + }, + { + "epoch": 0.26529, + "grad_norm": 0.7723741158404143, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 26529 + }, + { + "epoch": 0.2653, + "grad_norm": 0.7527697681222097, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 26530 + }, + { + "epoch": 0.26531, + "grad_norm": 0.7109812113103402, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 26531 + }, + { + "epoch": 0.26532, + "grad_norm": 0.7366349328689182, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 26532 + }, + { + "epoch": 0.26533, + "grad_norm": 0.7806277689360517, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 26533 + }, + { + "epoch": 0.26534, + "grad_norm": 0.8523583935866801, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 26534 + }, + { + "epoch": 0.26535, + "grad_norm": 0.97276416495003, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 26535 + }, + { + "epoch": 0.26536, + "grad_norm": 0.8501794649066898, + "learning_rate": 0.003, + "loss": 4.001, + "step": 26536 + }, + { + "epoch": 0.26537, + "grad_norm": 0.7116545340526359, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 26537 + }, + { + "epoch": 0.26538, + "grad_norm": 0.6222260745457721, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 26538 + }, + { + "epoch": 0.26539, + "grad_norm": 0.5291218797209851, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 26539 + }, + { + "epoch": 0.2654, + "grad_norm": 0.566846547513202, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 26540 + }, + { + "epoch": 0.26541, + "grad_norm": 0.437553772191313, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 26541 + }, + { + "epoch": 0.26542, + "grad_norm": 0.5074620530232908, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 26542 + }, + { + "epoch": 0.26543, + "grad_norm": 0.4781838247891866, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 26543 + }, + { + "epoch": 0.26544, + "grad_norm": 0.4775150866196757, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 26544 + }, + { + "epoch": 0.26545, + "grad_norm": 0.47349565477660865, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 26545 + }, + { + "epoch": 0.26546, + "grad_norm": 0.5171409759874409, + "learning_rate": 0.003, + "loss": 3.98, + "step": 26546 + }, + { + "epoch": 0.26547, + "grad_norm": 0.5643738936470258, + "learning_rate": 0.003, + "loss": 3.99, + "step": 26547 + }, + { + "epoch": 0.26548, + "grad_norm": 0.5350696426318361, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 26548 + }, + { + "epoch": 0.26549, + "grad_norm": 0.5601435469650469, + "learning_rate": 0.003, + "loss": 3.965, + "step": 26549 + }, + { + "epoch": 0.2655, + "grad_norm": 0.7870179184031978, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 26550 + }, + { + "epoch": 0.26551, + "grad_norm": 1.0013976383646708, + "learning_rate": 0.003, + "loss": 4.014, + "step": 26551 + }, + { + "epoch": 0.26552, + "grad_norm": 0.9801493814264162, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 26552 + }, + { + "epoch": 0.26553, + "grad_norm": 0.9945496196389119, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 26553 + }, + { + "epoch": 0.26554, + "grad_norm": 1.0598231566032357, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 26554 + }, + { + "epoch": 0.26555, + "grad_norm": 0.8288722206899993, + "learning_rate": 0.003, + "loss": 4.01, + "step": 26555 + }, + { + "epoch": 0.26556, + "grad_norm": 0.7547137407479807, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 26556 + }, + { + "epoch": 0.26557, + "grad_norm": 0.8765608680815681, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 26557 + }, + { + "epoch": 0.26558, + "grad_norm": 0.9797511252736297, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 26558 + }, + { + "epoch": 0.26559, + "grad_norm": 1.0566643728427731, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 26559 + }, + { + "epoch": 0.2656, + "grad_norm": 0.9680648221821827, + "learning_rate": 0.003, + "loss": 3.994, + "step": 26560 + }, + { + "epoch": 0.26561, + "grad_norm": 0.8125206540386436, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 26561 + }, + { + "epoch": 0.26562, + "grad_norm": 0.9085611676956429, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 26562 + }, + { + "epoch": 0.26563, + "grad_norm": 0.9415988668076137, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 26563 + }, + { + "epoch": 0.26564, + "grad_norm": 0.7603359378851267, + "learning_rate": 0.003, + "loss": 4.001, + "step": 26564 + }, + { + "epoch": 0.26565, + "grad_norm": 0.7766648186304322, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 26565 + }, + { + "epoch": 0.26566, + "grad_norm": 0.6974966295001201, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 26566 + }, + { + "epoch": 0.26567, + "grad_norm": 0.7227134582754149, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 26567 + }, + { + "epoch": 0.26568, + "grad_norm": 0.5144830429968877, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 26568 + }, + { + "epoch": 0.26569, + "grad_norm": 0.5528314177218105, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 26569 + }, + { + "epoch": 0.2657, + "grad_norm": 0.5349940893306553, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 26570 + }, + { + "epoch": 0.26571, + "grad_norm": 0.4956703587652507, + "learning_rate": 0.003, + "loss": 3.987, + "step": 26571 + }, + { + "epoch": 0.26572, + "grad_norm": 0.466312180376092, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 26572 + }, + { + "epoch": 0.26573, + "grad_norm": 0.388686678396847, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 26573 + }, + { + "epoch": 0.26574, + "grad_norm": 0.39866836744118606, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 26574 + }, + { + "epoch": 0.26575, + "grad_norm": 0.42367480733671226, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 26575 + }, + { + "epoch": 0.26576, + "grad_norm": 0.504762321882293, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 26576 + }, + { + "epoch": 0.26577, + "grad_norm": 0.6270000145573932, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 26577 + }, + { + "epoch": 0.26578, + "grad_norm": 0.7144040238606794, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 26578 + }, + { + "epoch": 0.26579, + "grad_norm": 0.7125276690014694, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 26579 + }, + { + "epoch": 0.2658, + "grad_norm": 0.7272748022275127, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 26580 + }, + { + "epoch": 0.26581, + "grad_norm": 0.8221769330594352, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 26581 + }, + { + "epoch": 0.26582, + "grad_norm": 0.8972517394738592, + "learning_rate": 0.003, + "loss": 3.986, + "step": 26582 + }, + { + "epoch": 0.26583, + "grad_norm": 0.9528923332806126, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 26583 + }, + { + "epoch": 0.26584, + "grad_norm": 1.1184349991794909, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 26584 + }, + { + "epoch": 0.26585, + "grad_norm": 0.8155092965179608, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 26585 + }, + { + "epoch": 0.26586, + "grad_norm": 0.6862335529146869, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 26586 + }, + { + "epoch": 0.26587, + "grad_norm": 0.6484790871544491, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 26587 + }, + { + "epoch": 0.26588, + "grad_norm": 0.647012602554932, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 26588 + }, + { + "epoch": 0.26589, + "grad_norm": 0.5695916238080658, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 26589 + }, + { + "epoch": 0.2659, + "grad_norm": 0.5740537775255561, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 26590 + }, + { + "epoch": 0.26591, + "grad_norm": 0.6240585905834684, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 26591 + }, + { + "epoch": 0.26592, + "grad_norm": 0.7164866690203429, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 26592 + }, + { + "epoch": 0.26593, + "grad_norm": 0.7429748183608245, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 26593 + }, + { + "epoch": 0.26594, + "grad_norm": 0.7128230131042327, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 26594 + }, + { + "epoch": 0.26595, + "grad_norm": 0.8249856496387634, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 26595 + }, + { + "epoch": 0.26596, + "grad_norm": 0.9451287455046002, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 26596 + }, + { + "epoch": 0.26597, + "grad_norm": 1.0190518590027031, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 26597 + }, + { + "epoch": 0.26598, + "grad_norm": 0.9921424252028198, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 26598 + }, + { + "epoch": 0.26599, + "grad_norm": 0.9262200099177875, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 26599 + }, + { + "epoch": 0.266, + "grad_norm": 0.8199654959504346, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 26600 + }, + { + "epoch": 0.26601, + "grad_norm": 0.7154610332787682, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 26601 + }, + { + "epoch": 0.26602, + "grad_norm": 0.7094043421628909, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 26602 + }, + { + "epoch": 0.26603, + "grad_norm": 0.9369297865665762, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 26603 + }, + { + "epoch": 0.26604, + "grad_norm": 1.1962079636080063, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 26604 + }, + { + "epoch": 0.26605, + "grad_norm": 0.6566578332902203, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 26605 + }, + { + "epoch": 0.26606, + "grad_norm": 0.7357644395094001, + "learning_rate": 0.003, + "loss": 4.031, + "step": 26606 + }, + { + "epoch": 0.26607, + "grad_norm": 0.8132776579503773, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 26607 + }, + { + "epoch": 0.26608, + "grad_norm": 0.9135869936751536, + "learning_rate": 0.003, + "loss": 4.0558, + "step": 26608 + }, + { + "epoch": 0.26609, + "grad_norm": 0.8898792033376062, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 26609 + }, + { + "epoch": 0.2661, + "grad_norm": 0.829834606529551, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 26610 + }, + { + "epoch": 0.26611, + "grad_norm": 0.6908285421366434, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 26611 + }, + { + "epoch": 0.26612, + "grad_norm": 0.740591801823334, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 26612 + }, + { + "epoch": 0.26613, + "grad_norm": 0.7548988573115912, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 26613 + }, + { + "epoch": 0.26614, + "grad_norm": 0.6677253327226647, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 26614 + }, + { + "epoch": 0.26615, + "grad_norm": 0.6611530043606524, + "learning_rate": 0.003, + "loss": 4.008, + "step": 26615 + }, + { + "epoch": 0.26616, + "grad_norm": 0.6558850461223857, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 26616 + }, + { + "epoch": 0.26617, + "grad_norm": 0.613226620060827, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 26617 + }, + { + "epoch": 0.26618, + "grad_norm": 0.5825014150698941, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 26618 + }, + { + "epoch": 0.26619, + "grad_norm": 0.5904791754956599, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 26619 + }, + { + "epoch": 0.2662, + "grad_norm": 0.6065584012148701, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 26620 + }, + { + "epoch": 0.26621, + "grad_norm": 0.5977734607764048, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 26621 + }, + { + "epoch": 0.26622, + "grad_norm": 0.5881055194878632, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 26622 + }, + { + "epoch": 0.26623, + "grad_norm": 0.6512616326616586, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 26623 + }, + { + "epoch": 0.26624, + "grad_norm": 0.7022491499399223, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 26624 + }, + { + "epoch": 0.26625, + "grad_norm": 0.824396078072785, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 26625 + }, + { + "epoch": 0.26626, + "grad_norm": 0.7722817310016116, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 26626 + }, + { + "epoch": 0.26627, + "grad_norm": 0.8137388060962558, + "learning_rate": 0.003, + "loss": 4.0353, + "step": 26627 + }, + { + "epoch": 0.26628, + "grad_norm": 0.8709569265826255, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 26628 + }, + { + "epoch": 0.26629, + "grad_norm": 0.7957382233405534, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 26629 + }, + { + "epoch": 0.2663, + "grad_norm": 0.7183759443187185, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 26630 + }, + { + "epoch": 0.26631, + "grad_norm": 0.6765230708685671, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 26631 + }, + { + "epoch": 0.26632, + "grad_norm": 0.6512594981200249, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 26632 + }, + { + "epoch": 0.26633, + "grad_norm": 0.7301726180603532, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 26633 + }, + { + "epoch": 0.26634, + "grad_norm": 0.7357614368646186, + "learning_rate": 0.003, + "loss": 3.979, + "step": 26634 + }, + { + "epoch": 0.26635, + "grad_norm": 0.7899319536521294, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 26635 + }, + { + "epoch": 0.26636, + "grad_norm": 0.8288863063713523, + "learning_rate": 0.003, + "loss": 3.972, + "step": 26636 + }, + { + "epoch": 0.26637, + "grad_norm": 0.9141668800495653, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 26637 + }, + { + "epoch": 0.26638, + "grad_norm": 0.9328117079351478, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 26638 + }, + { + "epoch": 0.26639, + "grad_norm": 0.998984599155532, + "learning_rate": 0.003, + "loss": 4.046, + "step": 26639 + }, + { + "epoch": 0.2664, + "grad_norm": 0.9048991323589213, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 26640 + }, + { + "epoch": 0.26641, + "grad_norm": 0.8376926964302074, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 26641 + }, + { + "epoch": 0.26642, + "grad_norm": 0.8366380660778405, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 26642 + }, + { + "epoch": 0.26643, + "grad_norm": 0.8225458487431931, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 26643 + }, + { + "epoch": 0.26644, + "grad_norm": 0.8802294815200705, + "learning_rate": 0.003, + "loss": 4.022, + "step": 26644 + }, + { + "epoch": 0.26645, + "grad_norm": 0.7835826031295835, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26645 + }, + { + "epoch": 0.26646, + "grad_norm": 0.8010733028810327, + "learning_rate": 0.003, + "loss": 4.0, + "step": 26646 + }, + { + "epoch": 0.26647, + "grad_norm": 0.8223658624956993, + "learning_rate": 0.003, + "loss": 4.013, + "step": 26647 + }, + { + "epoch": 0.26648, + "grad_norm": 0.8116140832871068, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 26648 + }, + { + "epoch": 0.26649, + "grad_norm": 0.7610962726061673, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 26649 + }, + { + "epoch": 0.2665, + "grad_norm": 0.7156786938845108, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 26650 + }, + { + "epoch": 0.26651, + "grad_norm": 0.6644098702339114, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 26651 + }, + { + "epoch": 0.26652, + "grad_norm": 0.7354384816290965, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 26652 + }, + { + "epoch": 0.26653, + "grad_norm": 0.7247251808609535, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 26653 + }, + { + "epoch": 0.26654, + "grad_norm": 0.7002140359711168, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 26654 + }, + { + "epoch": 0.26655, + "grad_norm": 0.9683235658269378, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 26655 + }, + { + "epoch": 0.26656, + "grad_norm": 1.161960680087703, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 26656 + }, + { + "epoch": 0.26657, + "grad_norm": 0.7282976664343572, + "learning_rate": 0.003, + "loss": 3.979, + "step": 26657 + }, + { + "epoch": 0.26658, + "grad_norm": 0.7448389732343799, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 26658 + }, + { + "epoch": 0.26659, + "grad_norm": 0.7816609963178238, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 26659 + }, + { + "epoch": 0.2666, + "grad_norm": 0.8358615707717857, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 26660 + }, + { + "epoch": 0.26661, + "grad_norm": 0.7757620513561578, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 26661 + }, + { + "epoch": 0.26662, + "grad_norm": 0.8007427123581756, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 26662 + }, + { + "epoch": 0.26663, + "grad_norm": 0.793250748391146, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 26663 + }, + { + "epoch": 0.26664, + "grad_norm": 0.8429349981527812, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 26664 + }, + { + "epoch": 0.26665, + "grad_norm": 0.8222830127761905, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 26665 + }, + { + "epoch": 0.26666, + "grad_norm": 0.8021922785232349, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 26666 + }, + { + "epoch": 0.26667, + "grad_norm": 0.8520347948217646, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 26667 + }, + { + "epoch": 0.26668, + "grad_norm": 0.8727475999433185, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 26668 + }, + { + "epoch": 0.26669, + "grad_norm": 0.8519073769852545, + "learning_rate": 0.003, + "loss": 4.002, + "step": 26669 + }, + { + "epoch": 0.2667, + "grad_norm": 0.8171809484768101, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 26670 + }, + { + "epoch": 0.26671, + "grad_norm": 0.7767380086519637, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 26671 + }, + { + "epoch": 0.26672, + "grad_norm": 0.7737884065375757, + "learning_rate": 0.003, + "loss": 4.012, + "step": 26672 + }, + { + "epoch": 0.26673, + "grad_norm": 0.7178831344378444, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 26673 + }, + { + "epoch": 0.26674, + "grad_norm": 0.7423139327930309, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 26674 + }, + { + "epoch": 0.26675, + "grad_norm": 0.794408389342941, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 26675 + }, + { + "epoch": 0.26676, + "grad_norm": 0.8629437578895319, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 26676 + }, + { + "epoch": 0.26677, + "grad_norm": 0.911604813066286, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 26677 + }, + { + "epoch": 0.26678, + "grad_norm": 0.8413749326298993, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 26678 + }, + { + "epoch": 0.26679, + "grad_norm": 0.6491735973825912, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 26679 + }, + { + "epoch": 0.2668, + "grad_norm": 0.5576332913858884, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 26680 + }, + { + "epoch": 0.26681, + "grad_norm": 0.5342161701883981, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 26681 + }, + { + "epoch": 0.26682, + "grad_norm": 0.5419298984491387, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 26682 + }, + { + "epoch": 0.26683, + "grad_norm": 0.49360628188935873, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 26683 + }, + { + "epoch": 0.26684, + "grad_norm": 0.519696581054315, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 26684 + }, + { + "epoch": 0.26685, + "grad_norm": 0.49036312700406653, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 26685 + }, + { + "epoch": 0.26686, + "grad_norm": 0.5150463889385473, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 26686 + }, + { + "epoch": 0.26687, + "grad_norm": 0.4876564598307529, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 26687 + }, + { + "epoch": 0.26688, + "grad_norm": 0.5048359769027045, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 26688 + }, + { + "epoch": 0.26689, + "grad_norm": 0.5597538218627947, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 26689 + }, + { + "epoch": 0.2669, + "grad_norm": 0.6578818680124969, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 26690 + }, + { + "epoch": 0.26691, + "grad_norm": 0.7115063523987585, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 26691 + }, + { + "epoch": 0.26692, + "grad_norm": 0.8431722839635093, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 26692 + }, + { + "epoch": 0.26693, + "grad_norm": 0.9455240260225357, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 26693 + }, + { + "epoch": 0.26694, + "grad_norm": 0.7565925451972626, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 26694 + }, + { + "epoch": 0.26695, + "grad_norm": 0.5286705178162189, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 26695 + }, + { + "epoch": 0.26696, + "grad_norm": 0.626399397519397, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 26696 + }, + { + "epoch": 0.26697, + "grad_norm": 0.6216200012004461, + "learning_rate": 0.003, + "loss": 4.016, + "step": 26697 + }, + { + "epoch": 0.26698, + "grad_norm": 0.5860333314587821, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 26698 + }, + { + "epoch": 0.26699, + "grad_norm": 0.6253160211383125, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 26699 + }, + { + "epoch": 0.267, + "grad_norm": 0.7217880522186882, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 26700 + }, + { + "epoch": 0.26701, + "grad_norm": 0.7689182320101317, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 26701 + }, + { + "epoch": 0.26702, + "grad_norm": 0.831525241084346, + "learning_rate": 0.003, + "loss": 3.99, + "step": 26702 + }, + { + "epoch": 0.26703, + "grad_norm": 0.781839958290703, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 26703 + }, + { + "epoch": 0.26704, + "grad_norm": 1.0080464086993666, + "learning_rate": 0.003, + "loss": 4.009, + "step": 26704 + }, + { + "epoch": 0.26705, + "grad_norm": 1.0625988737160026, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 26705 + }, + { + "epoch": 0.26706, + "grad_norm": 0.970634280681878, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 26706 + }, + { + "epoch": 0.26707, + "grad_norm": 0.9697071576704877, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 26707 + }, + { + "epoch": 0.26708, + "grad_norm": 0.928716188759671, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 26708 + }, + { + "epoch": 0.26709, + "grad_norm": 0.7503142860792942, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 26709 + }, + { + "epoch": 0.2671, + "grad_norm": 0.7657415143899828, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 26710 + }, + { + "epoch": 0.26711, + "grad_norm": 0.8755614109422263, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 26711 + }, + { + "epoch": 0.26712, + "grad_norm": 1.041079981428904, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 26712 + }, + { + "epoch": 0.26713, + "grad_norm": 0.9851826632893524, + "learning_rate": 0.003, + "loss": 4.008, + "step": 26713 + }, + { + "epoch": 0.26714, + "grad_norm": 0.8879613557764677, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 26714 + }, + { + "epoch": 0.26715, + "grad_norm": 0.8179304508658486, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 26715 + }, + { + "epoch": 0.26716, + "grad_norm": 0.878223375280105, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 26716 + }, + { + "epoch": 0.26717, + "grad_norm": 0.8382307352694376, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 26717 + }, + { + "epoch": 0.26718, + "grad_norm": 0.8920593677334917, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 26718 + }, + { + "epoch": 0.26719, + "grad_norm": 0.9418880762080557, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 26719 + }, + { + "epoch": 0.2672, + "grad_norm": 0.8862976565274936, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 26720 + }, + { + "epoch": 0.26721, + "grad_norm": 0.822311883639645, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 26721 + }, + { + "epoch": 0.26722, + "grad_norm": 0.8228521164753427, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 26722 + }, + { + "epoch": 0.26723, + "grad_norm": 0.8565078983338296, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 26723 + }, + { + "epoch": 0.26724, + "grad_norm": 0.8286861977813731, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 26724 + }, + { + "epoch": 0.26725, + "grad_norm": 0.9366948107267272, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 26725 + }, + { + "epoch": 0.26726, + "grad_norm": 1.0090002607515896, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 26726 + }, + { + "epoch": 0.26727, + "grad_norm": 0.9831788803058118, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 26727 + }, + { + "epoch": 0.26728, + "grad_norm": 0.8908536156121155, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 26728 + }, + { + "epoch": 0.26729, + "grad_norm": 0.8214148449463843, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 26729 + }, + { + "epoch": 0.2673, + "grad_norm": 0.6654152710759001, + "learning_rate": 0.003, + "loss": 3.984, + "step": 26730 + }, + { + "epoch": 0.26731, + "grad_norm": 0.6255886180397238, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 26731 + }, + { + "epoch": 0.26732, + "grad_norm": 0.6318231043402484, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 26732 + }, + { + "epoch": 0.26733, + "grad_norm": 0.6320711412380939, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 26733 + }, + { + "epoch": 0.26734, + "grad_norm": 0.5553732788184841, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 26734 + }, + { + "epoch": 0.26735, + "grad_norm": 0.5253487383983, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 26735 + }, + { + "epoch": 0.26736, + "grad_norm": 0.4981437870405723, + "learning_rate": 0.003, + "loss": 4.004, + "step": 26736 + }, + { + "epoch": 0.26737, + "grad_norm": 0.4955187901682576, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 26737 + }, + { + "epoch": 0.26738, + "grad_norm": 0.3702655202217686, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 26738 + }, + { + "epoch": 0.26739, + "grad_norm": 0.43042024918128846, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 26739 + }, + { + "epoch": 0.2674, + "grad_norm": 0.42396326795853034, + "learning_rate": 0.003, + "loss": 3.987, + "step": 26740 + }, + { + "epoch": 0.26741, + "grad_norm": 0.40047592538169713, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 26741 + }, + { + "epoch": 0.26742, + "grad_norm": 0.4338005627735062, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 26742 + }, + { + "epoch": 0.26743, + "grad_norm": 0.5186661739058808, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 26743 + }, + { + "epoch": 0.26744, + "grad_norm": 0.6308521467792944, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 26744 + }, + { + "epoch": 0.26745, + "grad_norm": 0.7071672180771745, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 26745 + }, + { + "epoch": 0.26746, + "grad_norm": 0.6405745861939225, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 26746 + }, + { + "epoch": 0.26747, + "grad_norm": 0.56116617299239, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 26747 + }, + { + "epoch": 0.26748, + "grad_norm": 0.613813869295296, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 26748 + }, + { + "epoch": 0.26749, + "grad_norm": 0.6475028313115089, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 26749 + }, + { + "epoch": 0.2675, + "grad_norm": 0.7095477938797542, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 26750 + }, + { + "epoch": 0.26751, + "grad_norm": 0.6903687420993456, + "learning_rate": 0.003, + "loss": 4.002, + "step": 26751 + }, + { + "epoch": 0.26752, + "grad_norm": 0.6742541680658156, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 26752 + }, + { + "epoch": 0.26753, + "grad_norm": 0.8814794192607578, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 26753 + }, + { + "epoch": 0.26754, + "grad_norm": 1.2066360829610898, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 26754 + }, + { + "epoch": 0.26755, + "grad_norm": 1.1067475977058625, + "learning_rate": 0.003, + "loss": 3.981, + "step": 26755 + }, + { + "epoch": 0.26756, + "grad_norm": 0.6630819503966724, + "learning_rate": 0.003, + "loss": 3.96, + "step": 26756 + }, + { + "epoch": 0.26757, + "grad_norm": 0.5985213338634652, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 26757 + }, + { + "epoch": 0.26758, + "grad_norm": 0.686736397725092, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 26758 + }, + { + "epoch": 0.26759, + "grad_norm": 0.7714911343758655, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 26759 + }, + { + "epoch": 0.2676, + "grad_norm": 0.8263256802669792, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 26760 + }, + { + "epoch": 0.26761, + "grad_norm": 0.7337354297326808, + "learning_rate": 0.003, + "loss": 3.987, + "step": 26761 + }, + { + "epoch": 0.26762, + "grad_norm": 0.6265023486866033, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 26762 + }, + { + "epoch": 0.26763, + "grad_norm": 0.7492830301924914, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 26763 + }, + { + "epoch": 0.26764, + "grad_norm": 0.8748325392505174, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 26764 + }, + { + "epoch": 0.26765, + "grad_norm": 0.8806560713852518, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 26765 + }, + { + "epoch": 0.26766, + "grad_norm": 0.9591756220104237, + "learning_rate": 0.003, + "loss": 4.017, + "step": 26766 + }, + { + "epoch": 0.26767, + "grad_norm": 0.8426352855288031, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 26767 + }, + { + "epoch": 0.26768, + "grad_norm": 0.8153655838071058, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 26768 + }, + { + "epoch": 0.26769, + "grad_norm": 0.8329675355865075, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 26769 + }, + { + "epoch": 0.2677, + "grad_norm": 0.8259980887291519, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 26770 + }, + { + "epoch": 0.26771, + "grad_norm": 0.7938176393240504, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 26771 + }, + { + "epoch": 0.26772, + "grad_norm": 0.8182890884593896, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 26772 + }, + { + "epoch": 0.26773, + "grad_norm": 0.8356754350391928, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 26773 + }, + { + "epoch": 0.26774, + "grad_norm": 0.8832102149151304, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 26774 + }, + { + "epoch": 0.26775, + "grad_norm": 0.9610251198800766, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 26775 + }, + { + "epoch": 0.26776, + "grad_norm": 0.9053077828453328, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 26776 + }, + { + "epoch": 0.26777, + "grad_norm": 0.8977991071930411, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 26777 + }, + { + "epoch": 0.26778, + "grad_norm": 1.0733339484214834, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 26778 + }, + { + "epoch": 0.26779, + "grad_norm": 0.9290014203701109, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 26779 + }, + { + "epoch": 0.2678, + "grad_norm": 0.8260654180863968, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 26780 + }, + { + "epoch": 0.26781, + "grad_norm": 0.74245492977528, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 26781 + }, + { + "epoch": 0.26782, + "grad_norm": 0.5538960636919399, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 26782 + }, + { + "epoch": 0.26783, + "grad_norm": 0.5021060170625399, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 26783 + }, + { + "epoch": 0.26784, + "grad_norm": 0.46176493793479684, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 26784 + }, + { + "epoch": 0.26785, + "grad_norm": 0.594444637360958, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 26785 + }, + { + "epoch": 0.26786, + "grad_norm": 0.6150356571773834, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 26786 + }, + { + "epoch": 0.26787, + "grad_norm": 0.7019378382358723, + "learning_rate": 0.003, + "loss": 3.995, + "step": 26787 + }, + { + "epoch": 0.26788, + "grad_norm": 0.6732243768647092, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 26788 + }, + { + "epoch": 0.26789, + "grad_norm": 0.6318149149526184, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 26789 + }, + { + "epoch": 0.2679, + "grad_norm": 0.6404533499872277, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 26790 + }, + { + "epoch": 0.26791, + "grad_norm": 0.6650663323427309, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 26791 + }, + { + "epoch": 0.26792, + "grad_norm": 0.7648654646526483, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 26792 + }, + { + "epoch": 0.26793, + "grad_norm": 0.824486795010908, + "learning_rate": 0.003, + "loss": 3.984, + "step": 26793 + }, + { + "epoch": 0.26794, + "grad_norm": 0.8715525699129938, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 26794 + }, + { + "epoch": 0.26795, + "grad_norm": 0.8543343219475428, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 26795 + }, + { + "epoch": 0.26796, + "grad_norm": 0.9003929708407905, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 26796 + }, + { + "epoch": 0.26797, + "grad_norm": 0.9905587819616245, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 26797 + }, + { + "epoch": 0.26798, + "grad_norm": 1.010599157079638, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 26798 + }, + { + "epoch": 0.26799, + "grad_norm": 0.8834150795492143, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 26799 + }, + { + "epoch": 0.268, + "grad_norm": 0.7296562082109985, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 26800 + }, + { + "epoch": 0.26801, + "grad_norm": 0.5735724502254219, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 26801 + }, + { + "epoch": 0.26802, + "grad_norm": 0.6868710058497945, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 26802 + }, + { + "epoch": 0.26803, + "grad_norm": 0.6611120859530559, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26803 + }, + { + "epoch": 0.26804, + "grad_norm": 0.7475649239622205, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 26804 + }, + { + "epoch": 0.26805, + "grad_norm": 0.8567425176926436, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 26805 + }, + { + "epoch": 0.26806, + "grad_norm": 0.8469540544672114, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 26806 + }, + { + "epoch": 0.26807, + "grad_norm": 0.7075887128736791, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 26807 + }, + { + "epoch": 0.26808, + "grad_norm": 0.7108248409606734, + "learning_rate": 0.003, + "loss": 3.975, + "step": 26808 + }, + { + "epoch": 0.26809, + "grad_norm": 0.6846572722673401, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 26809 + }, + { + "epoch": 0.2681, + "grad_norm": 0.7689561453720137, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 26810 + }, + { + "epoch": 0.26811, + "grad_norm": 0.8778187476924374, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 26811 + }, + { + "epoch": 0.26812, + "grad_norm": 0.91743705366749, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 26812 + }, + { + "epoch": 0.26813, + "grad_norm": 0.9019082298658463, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 26813 + }, + { + "epoch": 0.26814, + "grad_norm": 0.8243941659371254, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 26814 + }, + { + "epoch": 0.26815, + "grad_norm": 0.8200534197656225, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 26815 + }, + { + "epoch": 0.26816, + "grad_norm": 0.7922527202726813, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 26816 + }, + { + "epoch": 0.26817, + "grad_norm": 0.7706709089532483, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 26817 + }, + { + "epoch": 0.26818, + "grad_norm": 0.7940374978261461, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 26818 + }, + { + "epoch": 0.26819, + "grad_norm": 0.7683367739473489, + "learning_rate": 0.003, + "loss": 4.016, + "step": 26819 + }, + { + "epoch": 0.2682, + "grad_norm": 0.6877369975274051, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 26820 + }, + { + "epoch": 0.26821, + "grad_norm": 0.7738108029258398, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 26821 + }, + { + "epoch": 0.26822, + "grad_norm": 0.7875186483416055, + "learning_rate": 0.003, + "loss": 4.005, + "step": 26822 + }, + { + "epoch": 0.26823, + "grad_norm": 0.8775891696359782, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 26823 + }, + { + "epoch": 0.26824, + "grad_norm": 1.018250455838107, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 26824 + }, + { + "epoch": 0.26825, + "grad_norm": 0.9278867717625211, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 26825 + }, + { + "epoch": 0.26826, + "grad_norm": 0.7360604860389891, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 26826 + }, + { + "epoch": 0.26827, + "grad_norm": 0.6612529338265013, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 26827 + }, + { + "epoch": 0.26828, + "grad_norm": 0.793710617141989, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 26828 + }, + { + "epoch": 0.26829, + "grad_norm": 0.9552034657513951, + "learning_rate": 0.003, + "loss": 3.997, + "step": 26829 + }, + { + "epoch": 0.2683, + "grad_norm": 0.8366142562979986, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 26830 + }, + { + "epoch": 0.26831, + "grad_norm": 0.6880131731679345, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 26831 + }, + { + "epoch": 0.26832, + "grad_norm": 0.655980970274443, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 26832 + }, + { + "epoch": 0.26833, + "grad_norm": 0.637359286403938, + "learning_rate": 0.003, + "loss": 3.972, + "step": 26833 + }, + { + "epoch": 0.26834, + "grad_norm": 0.6123763025962758, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 26834 + }, + { + "epoch": 0.26835, + "grad_norm": 0.6173097455485234, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 26835 + }, + { + "epoch": 0.26836, + "grad_norm": 0.5700372887967199, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 26836 + }, + { + "epoch": 0.26837, + "grad_norm": 0.5743902479886173, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 26837 + }, + { + "epoch": 0.26838, + "grad_norm": 0.5223482788842044, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 26838 + }, + { + "epoch": 0.26839, + "grad_norm": 0.4701274629618058, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 26839 + }, + { + "epoch": 0.2684, + "grad_norm": 0.5094419649921065, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 26840 + }, + { + "epoch": 0.26841, + "grad_norm": 0.5325829569404611, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 26841 + }, + { + "epoch": 0.26842, + "grad_norm": 0.5736492231521606, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 26842 + }, + { + "epoch": 0.26843, + "grad_norm": 0.5841731772992081, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 26843 + }, + { + "epoch": 0.26844, + "grad_norm": 0.6249893764632605, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 26844 + }, + { + "epoch": 0.26845, + "grad_norm": 0.665031989685072, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 26845 + }, + { + "epoch": 0.26846, + "grad_norm": 0.7524094407470656, + "learning_rate": 0.003, + "loss": 3.999, + "step": 26846 + }, + { + "epoch": 0.26847, + "grad_norm": 0.8269013661917162, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 26847 + }, + { + "epoch": 0.26848, + "grad_norm": 0.9498835787954625, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 26848 + }, + { + "epoch": 0.26849, + "grad_norm": 0.8902234387256162, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 26849 + }, + { + "epoch": 0.2685, + "grad_norm": 0.6938869905205454, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 26850 + }, + { + "epoch": 0.26851, + "grad_norm": 0.697976097697444, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 26851 + }, + { + "epoch": 0.26852, + "grad_norm": 0.701425351581024, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 26852 + }, + { + "epoch": 0.26853, + "grad_norm": 0.6415163613509228, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 26853 + }, + { + "epoch": 0.26854, + "grad_norm": 0.59565691679594, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 26854 + }, + { + "epoch": 0.26855, + "grad_norm": 0.6341473665974464, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 26855 + }, + { + "epoch": 0.26856, + "grad_norm": 0.7339649035481328, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 26856 + }, + { + "epoch": 0.26857, + "grad_norm": 0.8539249018270546, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 26857 + }, + { + "epoch": 0.26858, + "grad_norm": 0.8890324695621952, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 26858 + }, + { + "epoch": 0.26859, + "grad_norm": 0.8602242688764091, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 26859 + }, + { + "epoch": 0.2686, + "grad_norm": 0.7104783084811207, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 26860 + }, + { + "epoch": 0.26861, + "grad_norm": 0.5995019635759061, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 26861 + }, + { + "epoch": 0.26862, + "grad_norm": 0.6502255802860192, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 26862 + }, + { + "epoch": 0.26863, + "grad_norm": 0.6368792642745256, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 26863 + }, + { + "epoch": 0.26864, + "grad_norm": 0.7232256869778158, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 26864 + }, + { + "epoch": 0.26865, + "grad_norm": 0.8166973819619315, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 26865 + }, + { + "epoch": 0.26866, + "grad_norm": 0.883655595997718, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 26866 + }, + { + "epoch": 0.26867, + "grad_norm": 0.9135791181276222, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 26867 + }, + { + "epoch": 0.26868, + "grad_norm": 0.9406262591279699, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 26868 + }, + { + "epoch": 0.26869, + "grad_norm": 0.9630434281799954, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 26869 + }, + { + "epoch": 0.2687, + "grad_norm": 1.102744372151899, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 26870 + }, + { + "epoch": 0.26871, + "grad_norm": 0.9583572483979265, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 26871 + }, + { + "epoch": 0.26872, + "grad_norm": 0.9747086072277106, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 26872 + }, + { + "epoch": 0.26873, + "grad_norm": 0.9751495696219857, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 26873 + }, + { + "epoch": 0.26874, + "grad_norm": 0.9579392425339892, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 26874 + }, + { + "epoch": 0.26875, + "grad_norm": 0.9886060478151821, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 26875 + }, + { + "epoch": 0.26876, + "grad_norm": 1.0776422069274036, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 26876 + }, + { + "epoch": 0.26877, + "grad_norm": 0.9580650676683676, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 26877 + }, + { + "epoch": 0.26878, + "grad_norm": 0.8260533141986884, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 26878 + }, + { + "epoch": 0.26879, + "grad_norm": 0.77503461512026, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 26879 + }, + { + "epoch": 0.2688, + "grad_norm": 0.714568369174464, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 26880 + }, + { + "epoch": 0.26881, + "grad_norm": 0.8982374847491509, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 26881 + }, + { + "epoch": 0.26882, + "grad_norm": 1.0043141158899886, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 26882 + }, + { + "epoch": 0.26883, + "grad_norm": 0.9632574408698719, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 26883 + }, + { + "epoch": 0.26884, + "grad_norm": 0.7354772739475415, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 26884 + }, + { + "epoch": 0.26885, + "grad_norm": 0.7000349615920475, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 26885 + }, + { + "epoch": 0.26886, + "grad_norm": 0.803429045422577, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 26886 + }, + { + "epoch": 0.26887, + "grad_norm": 0.8562468548814842, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 26887 + }, + { + "epoch": 0.26888, + "grad_norm": 0.828344788799538, + "learning_rate": 0.003, + "loss": 4.0, + "step": 26888 + }, + { + "epoch": 0.26889, + "grad_norm": 0.804879862207181, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 26889 + }, + { + "epoch": 0.2689, + "grad_norm": 0.7650538761830946, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 26890 + }, + { + "epoch": 0.26891, + "grad_norm": 0.6754890174482868, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 26891 + }, + { + "epoch": 0.26892, + "grad_norm": 0.6969424025451075, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 26892 + }, + { + "epoch": 0.26893, + "grad_norm": 0.6773574282291159, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 26893 + }, + { + "epoch": 0.26894, + "grad_norm": 0.6533936217326113, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 26894 + }, + { + "epoch": 0.26895, + "grad_norm": 0.6430766667177791, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 26895 + }, + { + "epoch": 0.26896, + "grad_norm": 0.6187055638167063, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 26896 + }, + { + "epoch": 0.26897, + "grad_norm": 0.52850126620231, + "learning_rate": 0.003, + "loss": 4.006, + "step": 26897 + }, + { + "epoch": 0.26898, + "grad_norm": 0.6110062892873914, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 26898 + }, + { + "epoch": 0.26899, + "grad_norm": 0.6715625015787674, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 26899 + }, + { + "epoch": 0.269, + "grad_norm": 0.6937089661897113, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 26900 + }, + { + "epoch": 0.26901, + "grad_norm": 0.6725400912077131, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 26901 + }, + { + "epoch": 0.26902, + "grad_norm": 0.7543869558970819, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 26902 + }, + { + "epoch": 0.26903, + "grad_norm": 0.8533282238050293, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 26903 + }, + { + "epoch": 0.26904, + "grad_norm": 0.8920278788070509, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 26904 + }, + { + "epoch": 0.26905, + "grad_norm": 0.8738059410964506, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 26905 + }, + { + "epoch": 0.26906, + "grad_norm": 0.7889509187928847, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 26906 + }, + { + "epoch": 0.26907, + "grad_norm": 0.7778883808711788, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 26907 + }, + { + "epoch": 0.26908, + "grad_norm": 0.8785525998240638, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 26908 + }, + { + "epoch": 0.26909, + "grad_norm": 0.9719905050713835, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 26909 + }, + { + "epoch": 0.2691, + "grad_norm": 0.881619390554545, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 26910 + }, + { + "epoch": 0.26911, + "grad_norm": 0.8600048627944982, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 26911 + }, + { + "epoch": 0.26912, + "grad_norm": 0.8715861296261459, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 26912 + }, + { + "epoch": 0.26913, + "grad_norm": 0.7385801640792763, + "learning_rate": 0.003, + "loss": 3.998, + "step": 26913 + }, + { + "epoch": 0.26914, + "grad_norm": 0.6807153075866756, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 26914 + }, + { + "epoch": 0.26915, + "grad_norm": 0.7074271544869039, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 26915 + }, + { + "epoch": 0.26916, + "grad_norm": 0.6469487462320639, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 26916 + }, + { + "epoch": 0.26917, + "grad_norm": 0.656636434653867, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 26917 + }, + { + "epoch": 0.26918, + "grad_norm": 0.6626060375840429, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 26918 + }, + { + "epoch": 0.26919, + "grad_norm": 0.642989889025569, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 26919 + }, + { + "epoch": 0.2692, + "grad_norm": 0.695185221158723, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 26920 + }, + { + "epoch": 0.26921, + "grad_norm": 0.6922238545865369, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 26921 + }, + { + "epoch": 0.26922, + "grad_norm": 0.7175666216525578, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 26922 + }, + { + "epoch": 0.26923, + "grad_norm": 0.6757423893690475, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 26923 + }, + { + "epoch": 0.26924, + "grad_norm": 0.6633752855191468, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 26924 + }, + { + "epoch": 0.26925, + "grad_norm": 0.6965220126322785, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 26925 + }, + { + "epoch": 0.26926, + "grad_norm": 0.7288690419181945, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 26926 + }, + { + "epoch": 0.26927, + "grad_norm": 0.7906587281484343, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 26927 + }, + { + "epoch": 0.26928, + "grad_norm": 0.9585781990796858, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 26928 + }, + { + "epoch": 0.26929, + "grad_norm": 0.9093164426768965, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 26929 + }, + { + "epoch": 0.2693, + "grad_norm": 0.7830226331542635, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 26930 + }, + { + "epoch": 0.26931, + "grad_norm": 0.7625494602916146, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 26931 + }, + { + "epoch": 0.26932, + "grad_norm": 0.6939976898304225, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 26932 + }, + { + "epoch": 0.26933, + "grad_norm": 0.6313241114298549, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 26933 + }, + { + "epoch": 0.26934, + "grad_norm": 0.5750969622824721, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 26934 + }, + { + "epoch": 0.26935, + "grad_norm": 0.7120432740628125, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 26935 + }, + { + "epoch": 0.26936, + "grad_norm": 0.789035128472306, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 26936 + }, + { + "epoch": 0.26937, + "grad_norm": 0.8715926405943051, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 26937 + }, + { + "epoch": 0.26938, + "grad_norm": 0.8531180281757417, + "learning_rate": 0.003, + "loss": 3.984, + "step": 26938 + }, + { + "epoch": 0.26939, + "grad_norm": 0.7162759428262979, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 26939 + }, + { + "epoch": 0.2694, + "grad_norm": 0.7342184435370578, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 26940 + }, + { + "epoch": 0.26941, + "grad_norm": 0.734814202336762, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 26941 + }, + { + "epoch": 0.26942, + "grad_norm": 0.7497641790187017, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 26942 + }, + { + "epoch": 0.26943, + "grad_norm": 0.8024205870978998, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 26943 + }, + { + "epoch": 0.26944, + "grad_norm": 0.9506178699864138, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 26944 + }, + { + "epoch": 0.26945, + "grad_norm": 1.074261541736628, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 26945 + }, + { + "epoch": 0.26946, + "grad_norm": 0.946194345619566, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 26946 + }, + { + "epoch": 0.26947, + "grad_norm": 0.9169032120675376, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 26947 + }, + { + "epoch": 0.26948, + "grad_norm": 0.9162423861720758, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 26948 + }, + { + "epoch": 0.26949, + "grad_norm": 0.9600443310809811, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 26949 + }, + { + "epoch": 0.2695, + "grad_norm": 0.8931866114267042, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 26950 + }, + { + "epoch": 0.26951, + "grad_norm": 0.7825213831744771, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 26951 + }, + { + "epoch": 0.26952, + "grad_norm": 0.7085193024293235, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 26952 + }, + { + "epoch": 0.26953, + "grad_norm": 0.7056001878334555, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 26953 + }, + { + "epoch": 0.26954, + "grad_norm": 0.7435871659964588, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 26954 + }, + { + "epoch": 0.26955, + "grad_norm": 0.9623630947530908, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 26955 + }, + { + "epoch": 0.26956, + "grad_norm": 1.0064368362851568, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 26956 + }, + { + "epoch": 0.26957, + "grad_norm": 0.9266680810275008, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 26957 + }, + { + "epoch": 0.26958, + "grad_norm": 0.7723781541800014, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 26958 + }, + { + "epoch": 0.26959, + "grad_norm": 0.7510150587643106, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 26959 + }, + { + "epoch": 0.2696, + "grad_norm": 0.7152310482309302, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 26960 + }, + { + "epoch": 0.26961, + "grad_norm": 0.7817414318108916, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 26961 + }, + { + "epoch": 0.26962, + "grad_norm": 0.7719205488953711, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 26962 + }, + { + "epoch": 0.26963, + "grad_norm": 0.7486620726838898, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 26963 + }, + { + "epoch": 0.26964, + "grad_norm": 0.8175781063879625, + "learning_rate": 0.003, + "loss": 4.0403, + "step": 26964 + }, + { + "epoch": 0.26965, + "grad_norm": 0.8539672095273315, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 26965 + }, + { + "epoch": 0.26966, + "grad_norm": 0.8268055956260941, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 26966 + }, + { + "epoch": 0.26967, + "grad_norm": 0.6721844036372968, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 26967 + }, + { + "epoch": 0.26968, + "grad_norm": 0.6571244030954432, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 26968 + }, + { + "epoch": 0.26969, + "grad_norm": 0.620070557912129, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 26969 + }, + { + "epoch": 0.2697, + "grad_norm": 0.5914965774646024, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 26970 + }, + { + "epoch": 0.26971, + "grad_norm": 0.596897105557298, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 26971 + }, + { + "epoch": 0.26972, + "grad_norm": 0.5667522524937814, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 26972 + }, + { + "epoch": 0.26973, + "grad_norm": 0.49435358393456436, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 26973 + }, + { + "epoch": 0.26974, + "grad_norm": 0.5373118861814745, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 26974 + }, + { + "epoch": 0.26975, + "grad_norm": 0.5090746824871833, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 26975 + }, + { + "epoch": 0.26976, + "grad_norm": 0.49180924124594866, + "learning_rate": 0.003, + "loss": 3.947, + "step": 26976 + }, + { + "epoch": 0.26977, + "grad_norm": 0.5775431732514932, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 26977 + }, + { + "epoch": 0.26978, + "grad_norm": 0.7573703213691377, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 26978 + }, + { + "epoch": 0.26979, + "grad_norm": 0.9089467664517573, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 26979 + }, + { + "epoch": 0.2698, + "grad_norm": 0.9758067092063639, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 26980 + }, + { + "epoch": 0.26981, + "grad_norm": 0.7690658964860904, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 26981 + }, + { + "epoch": 0.26982, + "grad_norm": 0.7486559326634226, + "learning_rate": 0.003, + "loss": 4.014, + "step": 26982 + }, + { + "epoch": 0.26983, + "grad_norm": 0.8827504389320715, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 26983 + }, + { + "epoch": 0.26984, + "grad_norm": 0.8470238820291596, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 26984 + }, + { + "epoch": 0.26985, + "grad_norm": 0.7758158567352583, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 26985 + }, + { + "epoch": 0.26986, + "grad_norm": 0.8540674570474297, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 26986 + }, + { + "epoch": 0.26987, + "grad_norm": 0.8086489104754736, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 26987 + }, + { + "epoch": 0.26988, + "grad_norm": 0.8710482133878473, + "learning_rate": 0.003, + "loss": 3.986, + "step": 26988 + }, + { + "epoch": 0.26989, + "grad_norm": 0.9534819081748404, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 26989 + }, + { + "epoch": 0.2699, + "grad_norm": 0.939856509158281, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 26990 + }, + { + "epoch": 0.26991, + "grad_norm": 1.0475465465168885, + "learning_rate": 0.003, + "loss": 4.0399, + "step": 26991 + }, + { + "epoch": 0.26992, + "grad_norm": 0.8859796853881172, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 26992 + }, + { + "epoch": 0.26993, + "grad_norm": 0.6816437478848607, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 26993 + }, + { + "epoch": 0.26994, + "grad_norm": 0.7035496034885548, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 26994 + }, + { + "epoch": 0.26995, + "grad_norm": 0.6854081654982226, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 26995 + }, + { + "epoch": 0.26996, + "grad_norm": 0.6087383523502591, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 26996 + }, + { + "epoch": 0.26997, + "grad_norm": 0.6594913334181419, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 26997 + }, + { + "epoch": 0.26998, + "grad_norm": 0.6776397243279054, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 26998 + }, + { + "epoch": 0.26999, + "grad_norm": 0.755112147959338, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 26999 + }, + { + "epoch": 0.27, + "grad_norm": 0.8910806988646222, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 27000 + }, + { + "epoch": 0.27001, + "grad_norm": 0.8292158182211544, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 27001 + }, + { + "epoch": 0.27002, + "grad_norm": 0.6410932053152472, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 27002 + }, + { + "epoch": 0.27003, + "grad_norm": 0.6195735430376514, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 27003 + }, + { + "epoch": 0.27004, + "grad_norm": 0.6097118718230534, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 27004 + }, + { + "epoch": 0.27005, + "grad_norm": 0.6707961962113449, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 27005 + }, + { + "epoch": 0.27006, + "grad_norm": 0.7061150010414898, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 27006 + }, + { + "epoch": 0.27007, + "grad_norm": 0.6582862695253152, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 27007 + }, + { + "epoch": 0.27008, + "grad_norm": 0.5970057053314086, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 27008 + }, + { + "epoch": 0.27009, + "grad_norm": 0.5690220577754044, + "learning_rate": 0.003, + "loss": 4.005, + "step": 27009 + }, + { + "epoch": 0.2701, + "grad_norm": 0.5014183047772461, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 27010 + }, + { + "epoch": 0.27011, + "grad_norm": 0.6011162734444813, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 27011 + }, + { + "epoch": 0.27012, + "grad_norm": 0.6102827059102458, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 27012 + }, + { + "epoch": 0.27013, + "grad_norm": 0.7456290156266115, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 27013 + }, + { + "epoch": 0.27014, + "grad_norm": 0.8071169560295329, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 27014 + }, + { + "epoch": 0.27015, + "grad_norm": 0.8915226259982906, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 27015 + }, + { + "epoch": 0.27016, + "grad_norm": 1.0708462239782817, + "learning_rate": 0.003, + "loss": 3.991, + "step": 27016 + }, + { + "epoch": 0.27017, + "grad_norm": 0.9067132564252707, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 27017 + }, + { + "epoch": 0.27018, + "grad_norm": 0.8536141477218243, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 27018 + }, + { + "epoch": 0.27019, + "grad_norm": 0.9086127320629686, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 27019 + }, + { + "epoch": 0.2702, + "grad_norm": 0.8914650500205816, + "learning_rate": 0.003, + "loss": 3.994, + "step": 27020 + }, + { + "epoch": 0.27021, + "grad_norm": 0.6805061644515676, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 27021 + }, + { + "epoch": 0.27022, + "grad_norm": 0.5926508670957916, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 27022 + }, + { + "epoch": 0.27023, + "grad_norm": 0.6312149116773342, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 27023 + }, + { + "epoch": 0.27024, + "grad_norm": 0.626136432412893, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 27024 + }, + { + "epoch": 0.27025, + "grad_norm": 0.6363199863633326, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 27025 + }, + { + "epoch": 0.27026, + "grad_norm": 0.6332200683561012, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 27026 + }, + { + "epoch": 0.27027, + "grad_norm": 0.5901202645700524, + "learning_rate": 0.003, + "loss": 3.986, + "step": 27027 + }, + { + "epoch": 0.27028, + "grad_norm": 0.5571139561899239, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 27028 + }, + { + "epoch": 0.27029, + "grad_norm": 0.5628768440134875, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 27029 + }, + { + "epoch": 0.2703, + "grad_norm": 0.5894423980692308, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 27030 + }, + { + "epoch": 0.27031, + "grad_norm": 0.5888931038135342, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 27031 + }, + { + "epoch": 0.27032, + "grad_norm": 0.666135859151724, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 27032 + }, + { + "epoch": 0.27033, + "grad_norm": 0.8736260864361507, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 27033 + }, + { + "epoch": 0.27034, + "grad_norm": 1.1393876629484707, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 27034 + }, + { + "epoch": 0.27035, + "grad_norm": 0.929175332424337, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 27035 + }, + { + "epoch": 0.27036, + "grad_norm": 0.7700872528752475, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 27036 + }, + { + "epoch": 0.27037, + "grad_norm": 0.8147553985869325, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 27037 + }, + { + "epoch": 0.27038, + "grad_norm": 0.8824709328985694, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 27038 + }, + { + "epoch": 0.27039, + "grad_norm": 0.7958013635141079, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 27039 + }, + { + "epoch": 0.2704, + "grad_norm": 0.9100744754335824, + "learning_rate": 0.003, + "loss": 4.019, + "step": 27040 + }, + { + "epoch": 0.27041, + "grad_norm": 0.9640794914023914, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 27041 + }, + { + "epoch": 0.27042, + "grad_norm": 0.8793343698350267, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 27042 + }, + { + "epoch": 0.27043, + "grad_norm": 0.9208142359437098, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 27043 + }, + { + "epoch": 0.27044, + "grad_norm": 0.8691007494209759, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 27044 + }, + { + "epoch": 0.27045, + "grad_norm": 0.8812723160014985, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 27045 + }, + { + "epoch": 0.27046, + "grad_norm": 0.8618567049581137, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 27046 + }, + { + "epoch": 0.27047, + "grad_norm": 0.8099530445436263, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 27047 + }, + { + "epoch": 0.27048, + "grad_norm": 0.7124278476399976, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 27048 + }, + { + "epoch": 0.27049, + "grad_norm": 0.6471041489706842, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 27049 + }, + { + "epoch": 0.2705, + "grad_norm": 0.549355071324247, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 27050 + }, + { + "epoch": 0.27051, + "grad_norm": 0.5447661522530529, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 27051 + }, + { + "epoch": 0.27052, + "grad_norm": 0.5517907329492222, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 27052 + }, + { + "epoch": 0.27053, + "grad_norm": 0.5797926356024237, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 27053 + }, + { + "epoch": 0.27054, + "grad_norm": 0.7047889904481995, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 27054 + }, + { + "epoch": 0.27055, + "grad_norm": 0.9289278871613974, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 27055 + }, + { + "epoch": 0.27056, + "grad_norm": 0.987141649354768, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 27056 + }, + { + "epoch": 0.27057, + "grad_norm": 0.8721878067712929, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 27057 + }, + { + "epoch": 0.27058, + "grad_norm": 0.8486043127754173, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 27058 + }, + { + "epoch": 0.27059, + "grad_norm": 0.828270699725832, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 27059 + }, + { + "epoch": 0.2706, + "grad_norm": 0.764626658873039, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 27060 + }, + { + "epoch": 0.27061, + "grad_norm": 0.6747076089587668, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 27061 + }, + { + "epoch": 0.27062, + "grad_norm": 0.5620335395966793, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 27062 + }, + { + "epoch": 0.27063, + "grad_norm": 0.6507573139757009, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 27063 + }, + { + "epoch": 0.27064, + "grad_norm": 0.682613599647797, + "learning_rate": 0.003, + "loss": 4.004, + "step": 27064 + }, + { + "epoch": 0.27065, + "grad_norm": 0.713133972311372, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 27065 + }, + { + "epoch": 0.27066, + "grad_norm": 0.8223262014678526, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 27066 + }, + { + "epoch": 0.27067, + "grad_norm": 0.9427962107758378, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 27067 + }, + { + "epoch": 0.27068, + "grad_norm": 1.002410449307041, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 27068 + }, + { + "epoch": 0.27069, + "grad_norm": 0.8233763327842776, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 27069 + }, + { + "epoch": 0.2707, + "grad_norm": 0.7717345411485065, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 27070 + }, + { + "epoch": 0.27071, + "grad_norm": 0.7189538754136773, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 27071 + }, + { + "epoch": 0.27072, + "grad_norm": 0.7973303967243424, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 27072 + }, + { + "epoch": 0.27073, + "grad_norm": 0.8998738135661104, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 27073 + }, + { + "epoch": 0.27074, + "grad_norm": 0.9277897081114539, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27074 + }, + { + "epoch": 0.27075, + "grad_norm": 0.8146236372178668, + "learning_rate": 0.003, + "loss": 4.0, + "step": 27075 + }, + { + "epoch": 0.27076, + "grad_norm": 0.7207266707442345, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 27076 + }, + { + "epoch": 0.27077, + "grad_norm": 0.5936450897523778, + "learning_rate": 0.003, + "loss": 3.983, + "step": 27077 + }, + { + "epoch": 0.27078, + "grad_norm": 0.7274581370545685, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 27078 + }, + { + "epoch": 0.27079, + "grad_norm": 0.8380719284064274, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 27079 + }, + { + "epoch": 0.2708, + "grad_norm": 0.9063616213575739, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 27080 + }, + { + "epoch": 0.27081, + "grad_norm": 0.7863216016398216, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 27081 + }, + { + "epoch": 0.27082, + "grad_norm": 0.6822542685465165, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 27082 + }, + { + "epoch": 0.27083, + "grad_norm": 0.6003254841803793, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 27083 + }, + { + "epoch": 0.27084, + "grad_norm": 0.7458860228055906, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 27084 + }, + { + "epoch": 0.27085, + "grad_norm": 0.7442421306406815, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 27085 + }, + { + "epoch": 0.27086, + "grad_norm": 0.7804641203900975, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 27086 + }, + { + "epoch": 0.27087, + "grad_norm": 0.7851180030888387, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 27087 + }, + { + "epoch": 0.27088, + "grad_norm": 0.7443687605132965, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 27088 + }, + { + "epoch": 0.27089, + "grad_norm": 0.752462604478653, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 27089 + }, + { + "epoch": 0.2709, + "grad_norm": 0.7498157502054276, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 27090 + }, + { + "epoch": 0.27091, + "grad_norm": 0.6738435620676975, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 27091 + }, + { + "epoch": 0.27092, + "grad_norm": 0.7838477254515817, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 27092 + }, + { + "epoch": 0.27093, + "grad_norm": 0.8186805405844124, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 27093 + }, + { + "epoch": 0.27094, + "grad_norm": 0.840373151806823, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 27094 + }, + { + "epoch": 0.27095, + "grad_norm": 0.7705320785562492, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 27095 + }, + { + "epoch": 0.27096, + "grad_norm": 0.6691769667960278, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 27096 + }, + { + "epoch": 0.27097, + "grad_norm": 0.5848749895462273, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 27097 + }, + { + "epoch": 0.27098, + "grad_norm": 0.5833128628234706, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 27098 + }, + { + "epoch": 0.27099, + "grad_norm": 0.6800876089296903, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 27099 + }, + { + "epoch": 0.271, + "grad_norm": 0.7318476595850353, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 27100 + }, + { + "epoch": 0.27101, + "grad_norm": 0.7160473596062032, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 27101 + }, + { + "epoch": 0.27102, + "grad_norm": 0.7112754395565009, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 27102 + }, + { + "epoch": 0.27103, + "grad_norm": 0.7839170352446276, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 27103 + }, + { + "epoch": 0.27104, + "grad_norm": 0.8924268286596833, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 27104 + }, + { + "epoch": 0.27105, + "grad_norm": 1.0646741148098857, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 27105 + }, + { + "epoch": 0.27106, + "grad_norm": 1.0449289464138218, + "learning_rate": 0.003, + "loss": 4.01, + "step": 27106 + }, + { + "epoch": 0.27107, + "grad_norm": 0.8045223963794355, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 27107 + }, + { + "epoch": 0.27108, + "grad_norm": 0.7902699797412109, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 27108 + }, + { + "epoch": 0.27109, + "grad_norm": 0.9227265423020021, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 27109 + }, + { + "epoch": 0.2711, + "grad_norm": 0.9908630064452209, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 27110 + }, + { + "epoch": 0.27111, + "grad_norm": 0.9881333590451387, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 27111 + }, + { + "epoch": 0.27112, + "grad_norm": 0.9313910043834995, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 27112 + }, + { + "epoch": 0.27113, + "grad_norm": 0.824568138937265, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 27113 + }, + { + "epoch": 0.27114, + "grad_norm": 0.7855966133442771, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 27114 + }, + { + "epoch": 0.27115, + "grad_norm": 0.8696422879850004, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 27115 + }, + { + "epoch": 0.27116, + "grad_norm": 0.8638068806804345, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 27116 + }, + { + "epoch": 0.27117, + "grad_norm": 0.8302649422436914, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 27117 + }, + { + "epoch": 0.27118, + "grad_norm": 0.7394527058015873, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 27118 + }, + { + "epoch": 0.27119, + "grad_norm": 0.5809614910954115, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 27119 + }, + { + "epoch": 0.2712, + "grad_norm": 0.5330193555728627, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 27120 + }, + { + "epoch": 0.27121, + "grad_norm": 0.5476860240317996, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 27121 + }, + { + "epoch": 0.27122, + "grad_norm": 0.611723099420332, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 27122 + }, + { + "epoch": 0.27123, + "grad_norm": 0.6263734825362824, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 27123 + }, + { + "epoch": 0.27124, + "grad_norm": 0.6162651794574622, + "learning_rate": 0.003, + "loss": 3.978, + "step": 27124 + }, + { + "epoch": 0.27125, + "grad_norm": 0.6127676418977117, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 27125 + }, + { + "epoch": 0.27126, + "grad_norm": 0.56589719572701, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 27126 + }, + { + "epoch": 0.27127, + "grad_norm": 0.6189438292646732, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 27127 + }, + { + "epoch": 0.27128, + "grad_norm": 0.6744962288365848, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 27128 + }, + { + "epoch": 0.27129, + "grad_norm": 0.7028491202568842, + "learning_rate": 0.003, + "loss": 3.993, + "step": 27129 + }, + { + "epoch": 0.2713, + "grad_norm": 0.7357977537331468, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 27130 + }, + { + "epoch": 0.27131, + "grad_norm": 0.7805280458397165, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 27131 + }, + { + "epoch": 0.27132, + "grad_norm": 0.7915431426166667, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 27132 + }, + { + "epoch": 0.27133, + "grad_norm": 0.7577922676091604, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 27133 + }, + { + "epoch": 0.27134, + "grad_norm": 0.7903518745209365, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 27134 + }, + { + "epoch": 0.27135, + "grad_norm": 0.8735802992814484, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 27135 + }, + { + "epoch": 0.27136, + "grad_norm": 0.8158333135246872, + "learning_rate": 0.003, + "loss": 4.025, + "step": 27136 + }, + { + "epoch": 0.27137, + "grad_norm": 0.9022144060692975, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 27137 + }, + { + "epoch": 0.27138, + "grad_norm": 0.9144661138922101, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 27138 + }, + { + "epoch": 0.27139, + "grad_norm": 1.008644896082268, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 27139 + }, + { + "epoch": 0.2714, + "grad_norm": 1.020994281008986, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 27140 + }, + { + "epoch": 0.27141, + "grad_norm": 1.0002941762815618, + "learning_rate": 0.003, + "loss": 4.002, + "step": 27141 + }, + { + "epoch": 0.27142, + "grad_norm": 0.8744471693195325, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 27142 + }, + { + "epoch": 0.27143, + "grad_norm": 0.8643222080306882, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 27143 + }, + { + "epoch": 0.27144, + "grad_norm": 0.8648430049781212, + "learning_rate": 0.003, + "loss": 4.0461, + "step": 27144 + }, + { + "epoch": 0.27145, + "grad_norm": 0.9161575688026637, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 27145 + }, + { + "epoch": 0.27146, + "grad_norm": 0.8190511530363191, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 27146 + }, + { + "epoch": 0.27147, + "grad_norm": 0.9078127220575233, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 27147 + }, + { + "epoch": 0.27148, + "grad_norm": 0.8301733905057173, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 27148 + }, + { + "epoch": 0.27149, + "grad_norm": 0.798752803989491, + "learning_rate": 0.003, + "loss": 4.011, + "step": 27149 + }, + { + "epoch": 0.2715, + "grad_norm": 0.8255912807241405, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 27150 + }, + { + "epoch": 0.27151, + "grad_norm": 0.8239063362406366, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 27151 + }, + { + "epoch": 0.27152, + "grad_norm": 0.6973219466942645, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 27152 + }, + { + "epoch": 0.27153, + "grad_norm": 0.5914780047032197, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 27153 + }, + { + "epoch": 0.27154, + "grad_norm": 0.6093185689377123, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 27154 + }, + { + "epoch": 0.27155, + "grad_norm": 0.6757061363598604, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 27155 + }, + { + "epoch": 0.27156, + "grad_norm": 0.7623004371584942, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 27156 + }, + { + "epoch": 0.27157, + "grad_norm": 0.7477350884538699, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 27157 + }, + { + "epoch": 0.27158, + "grad_norm": 0.7251423302258162, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 27158 + }, + { + "epoch": 0.27159, + "grad_norm": 0.7050663897915034, + "learning_rate": 0.003, + "loss": 3.979, + "step": 27159 + }, + { + "epoch": 0.2716, + "grad_norm": 0.6821695451049262, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 27160 + }, + { + "epoch": 0.27161, + "grad_norm": 0.6751279506133865, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 27161 + }, + { + "epoch": 0.27162, + "grad_norm": 0.7395754205850383, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 27162 + }, + { + "epoch": 0.27163, + "grad_norm": 0.7667766122187029, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 27163 + }, + { + "epoch": 0.27164, + "grad_norm": 0.7860735535385046, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 27164 + }, + { + "epoch": 0.27165, + "grad_norm": 0.7502541690153248, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 27165 + }, + { + "epoch": 0.27166, + "grad_norm": 0.7493884207385132, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 27166 + }, + { + "epoch": 0.27167, + "grad_norm": 0.6837367309416509, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 27167 + }, + { + "epoch": 0.27168, + "grad_norm": 0.6588116946778555, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 27168 + }, + { + "epoch": 0.27169, + "grad_norm": 0.6412323209296953, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 27169 + }, + { + "epoch": 0.2717, + "grad_norm": 0.6093561777282891, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 27170 + }, + { + "epoch": 0.27171, + "grad_norm": 0.7155804571551677, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 27171 + }, + { + "epoch": 0.27172, + "grad_norm": 0.753788043685158, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 27172 + }, + { + "epoch": 0.27173, + "grad_norm": 0.805225398110181, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 27173 + }, + { + "epoch": 0.27174, + "grad_norm": 0.9418917124329196, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 27174 + }, + { + "epoch": 0.27175, + "grad_norm": 0.9027609139056031, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27175 + }, + { + "epoch": 0.27176, + "grad_norm": 0.8027612349873585, + "learning_rate": 0.003, + "loss": 3.9404, + "step": 27176 + }, + { + "epoch": 0.27177, + "grad_norm": 0.877798688372876, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 27177 + }, + { + "epoch": 0.27178, + "grad_norm": 0.8640155874916797, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 27178 + }, + { + "epoch": 0.27179, + "grad_norm": 0.748970498606992, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 27179 + }, + { + "epoch": 0.2718, + "grad_norm": 0.7830819039454465, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 27180 + }, + { + "epoch": 0.27181, + "grad_norm": 0.8104412571839836, + "learning_rate": 0.003, + "loss": 4.019, + "step": 27181 + }, + { + "epoch": 0.27182, + "grad_norm": 0.8104338925074704, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 27182 + }, + { + "epoch": 0.27183, + "grad_norm": 0.7329224750302011, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 27183 + }, + { + "epoch": 0.27184, + "grad_norm": 0.778042737805402, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 27184 + }, + { + "epoch": 0.27185, + "grad_norm": 0.931690137572689, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 27185 + }, + { + "epoch": 0.27186, + "grad_norm": 0.8923445632922178, + "learning_rate": 0.003, + "loss": 4.034, + "step": 27186 + }, + { + "epoch": 0.27187, + "grad_norm": 0.9754978781362048, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 27187 + }, + { + "epoch": 0.27188, + "grad_norm": 1.0030225131485686, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 27188 + }, + { + "epoch": 0.27189, + "grad_norm": 1.0277577342916675, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 27189 + }, + { + "epoch": 0.2719, + "grad_norm": 1.0937021251478494, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 27190 + }, + { + "epoch": 0.27191, + "grad_norm": 1.0073075841663248, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 27191 + }, + { + "epoch": 0.27192, + "grad_norm": 0.7448418690064501, + "learning_rate": 0.003, + "loss": 4.025, + "step": 27192 + }, + { + "epoch": 0.27193, + "grad_norm": 0.5970779247178024, + "learning_rate": 0.003, + "loss": 3.984, + "step": 27193 + }, + { + "epoch": 0.27194, + "grad_norm": 0.6185181441380783, + "learning_rate": 0.003, + "loss": 3.997, + "step": 27194 + }, + { + "epoch": 0.27195, + "grad_norm": 0.6724186444791291, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 27195 + }, + { + "epoch": 0.27196, + "grad_norm": 0.659685922434594, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 27196 + }, + { + "epoch": 0.27197, + "grad_norm": 0.783761737282814, + "learning_rate": 0.003, + "loss": 4.027, + "step": 27197 + }, + { + "epoch": 0.27198, + "grad_norm": 0.7260461066221339, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 27198 + }, + { + "epoch": 0.27199, + "grad_norm": 0.7694787252553674, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 27199 + }, + { + "epoch": 0.272, + "grad_norm": 0.6864303053391334, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 27200 + }, + { + "epoch": 0.27201, + "grad_norm": 0.5909330667493499, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 27201 + }, + { + "epoch": 0.27202, + "grad_norm": 0.5506209553169783, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 27202 + }, + { + "epoch": 0.27203, + "grad_norm": 0.5410244644652917, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 27203 + }, + { + "epoch": 0.27204, + "grad_norm": 0.5313088803343629, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 27204 + }, + { + "epoch": 0.27205, + "grad_norm": 0.5547496361310756, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 27205 + }, + { + "epoch": 0.27206, + "grad_norm": 0.6549060847034016, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 27206 + }, + { + "epoch": 0.27207, + "grad_norm": 0.6157071245442556, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 27207 + }, + { + "epoch": 0.27208, + "grad_norm": 0.4777003066560462, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 27208 + }, + { + "epoch": 0.27209, + "grad_norm": 0.4710713760514212, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 27209 + }, + { + "epoch": 0.2721, + "grad_norm": 0.529314482820489, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 27210 + }, + { + "epoch": 0.27211, + "grad_norm": 0.5090344319813178, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 27211 + }, + { + "epoch": 0.27212, + "grad_norm": 0.4924117438552204, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 27212 + }, + { + "epoch": 0.27213, + "grad_norm": 0.5632446124906718, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 27213 + }, + { + "epoch": 0.27214, + "grad_norm": 0.6545410657603415, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 27214 + }, + { + "epoch": 0.27215, + "grad_norm": 0.7581622078501002, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 27215 + }, + { + "epoch": 0.27216, + "grad_norm": 0.7973207042121634, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 27216 + }, + { + "epoch": 0.27217, + "grad_norm": 0.8372015087918199, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 27217 + }, + { + "epoch": 0.27218, + "grad_norm": 0.8930086239101943, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 27218 + }, + { + "epoch": 0.27219, + "grad_norm": 0.826199490350316, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 27219 + }, + { + "epoch": 0.2722, + "grad_norm": 0.7680743021609577, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 27220 + }, + { + "epoch": 0.27221, + "grad_norm": 0.7933317302860522, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 27221 + }, + { + "epoch": 0.27222, + "grad_norm": 0.8355082374196331, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 27222 + }, + { + "epoch": 0.27223, + "grad_norm": 0.9144268420610914, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 27223 + }, + { + "epoch": 0.27224, + "grad_norm": 0.9400281849849725, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 27224 + }, + { + "epoch": 0.27225, + "grad_norm": 1.0103183001506941, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 27225 + }, + { + "epoch": 0.27226, + "grad_norm": 0.9995473633219354, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 27226 + }, + { + "epoch": 0.27227, + "grad_norm": 0.8253277473268018, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 27227 + }, + { + "epoch": 0.27228, + "grad_norm": 0.721828244709108, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 27228 + }, + { + "epoch": 0.27229, + "grad_norm": 0.6820815498813747, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 27229 + }, + { + "epoch": 0.2723, + "grad_norm": 0.6968221200441083, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 27230 + }, + { + "epoch": 0.27231, + "grad_norm": 0.6994705684621843, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 27231 + }, + { + "epoch": 0.27232, + "grad_norm": 0.6886788737312122, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 27232 + }, + { + "epoch": 0.27233, + "grad_norm": 0.7045615006919975, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 27233 + }, + { + "epoch": 0.27234, + "grad_norm": 0.6226018033183799, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 27234 + }, + { + "epoch": 0.27235, + "grad_norm": 0.6399481280933077, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 27235 + }, + { + "epoch": 0.27236, + "grad_norm": 0.6682455160920875, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 27236 + }, + { + "epoch": 0.27237, + "grad_norm": 0.7172707260697674, + "learning_rate": 0.003, + "loss": 3.994, + "step": 27237 + }, + { + "epoch": 0.27238, + "grad_norm": 0.8001215652462242, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 27238 + }, + { + "epoch": 0.27239, + "grad_norm": 0.909265750725235, + "learning_rate": 0.003, + "loss": 3.979, + "step": 27239 + }, + { + "epoch": 0.2724, + "grad_norm": 0.9968297067295245, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 27240 + }, + { + "epoch": 0.27241, + "grad_norm": 1.1213318181646175, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 27241 + }, + { + "epoch": 0.27242, + "grad_norm": 0.8968769380823017, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 27242 + }, + { + "epoch": 0.27243, + "grad_norm": 0.9802616240855897, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 27243 + }, + { + "epoch": 0.27244, + "grad_norm": 1.0487405752042398, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 27244 + }, + { + "epoch": 0.27245, + "grad_norm": 0.9197576052801589, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 27245 + }, + { + "epoch": 0.27246, + "grad_norm": 0.84278713760124, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 27246 + }, + { + "epoch": 0.27247, + "grad_norm": 0.878676926528553, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 27247 + }, + { + "epoch": 0.27248, + "grad_norm": 0.8788629727192826, + "learning_rate": 0.003, + "loss": 4.009, + "step": 27248 + }, + { + "epoch": 0.27249, + "grad_norm": 0.8834724062831476, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 27249 + }, + { + "epoch": 0.2725, + "grad_norm": 0.8818261267178296, + "learning_rate": 0.003, + "loss": 4.004, + "step": 27250 + }, + { + "epoch": 0.27251, + "grad_norm": 0.9659696288977246, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 27251 + }, + { + "epoch": 0.27252, + "grad_norm": 0.9505976278683897, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 27252 + }, + { + "epoch": 0.27253, + "grad_norm": 0.9717823764605178, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 27253 + }, + { + "epoch": 0.27254, + "grad_norm": 1.0106153116512624, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 27254 + }, + { + "epoch": 0.27255, + "grad_norm": 0.9008113714710319, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 27255 + }, + { + "epoch": 0.27256, + "grad_norm": 0.8213440368635283, + "learning_rate": 0.003, + "loss": 4.013, + "step": 27256 + }, + { + "epoch": 0.27257, + "grad_norm": 0.7281635720735562, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 27257 + }, + { + "epoch": 0.27258, + "grad_norm": 0.5902175452056194, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 27258 + }, + { + "epoch": 0.27259, + "grad_norm": 0.6130894553795876, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 27259 + }, + { + "epoch": 0.2726, + "grad_norm": 0.6498130028899739, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27260 + }, + { + "epoch": 0.27261, + "grad_norm": 0.6679219695835058, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 27261 + }, + { + "epoch": 0.27262, + "grad_norm": 0.7287529278661643, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 27262 + }, + { + "epoch": 0.27263, + "grad_norm": 0.6789383566375243, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 27263 + }, + { + "epoch": 0.27264, + "grad_norm": 0.5733038097270817, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 27264 + }, + { + "epoch": 0.27265, + "grad_norm": 0.511651091589473, + "learning_rate": 0.003, + "loss": 4.003, + "step": 27265 + }, + { + "epoch": 0.27266, + "grad_norm": 0.5133655278631492, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 27266 + }, + { + "epoch": 0.27267, + "grad_norm": 0.5563160127552483, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 27267 + }, + { + "epoch": 0.27268, + "grad_norm": 0.6459320756695918, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 27268 + }, + { + "epoch": 0.27269, + "grad_norm": 0.8151084341706494, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 27269 + }, + { + "epoch": 0.2727, + "grad_norm": 0.9857892181521545, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 27270 + }, + { + "epoch": 0.27271, + "grad_norm": 1.0729884707044748, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 27271 + }, + { + "epoch": 0.27272, + "grad_norm": 0.7934903654606186, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 27272 + }, + { + "epoch": 0.27273, + "grad_norm": 0.6551060551992505, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 27273 + }, + { + "epoch": 0.27274, + "grad_norm": 0.755275699968913, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 27274 + }, + { + "epoch": 0.27275, + "grad_norm": 0.8929168795175761, + "learning_rate": 0.003, + "loss": 4.024, + "step": 27275 + }, + { + "epoch": 0.27276, + "grad_norm": 0.9332089841057836, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 27276 + }, + { + "epoch": 0.27277, + "grad_norm": 0.880093302226534, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 27277 + }, + { + "epoch": 0.27278, + "grad_norm": 0.8590811915266037, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 27278 + }, + { + "epoch": 0.27279, + "grad_norm": 0.7728997205514954, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 27279 + }, + { + "epoch": 0.2728, + "grad_norm": 0.7564652366629878, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 27280 + }, + { + "epoch": 0.27281, + "grad_norm": 0.6863808647243063, + "learning_rate": 0.003, + "loss": 3.999, + "step": 27281 + }, + { + "epoch": 0.27282, + "grad_norm": 0.6698618537242151, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 27282 + }, + { + "epoch": 0.27283, + "grad_norm": 0.7503502897518572, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 27283 + }, + { + "epoch": 0.27284, + "grad_norm": 0.8166341110369237, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 27284 + }, + { + "epoch": 0.27285, + "grad_norm": 0.8380924233374251, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 27285 + }, + { + "epoch": 0.27286, + "grad_norm": 0.827416620642889, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 27286 + }, + { + "epoch": 0.27287, + "grad_norm": 0.8997382992093753, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 27287 + }, + { + "epoch": 0.27288, + "grad_norm": 0.8222277377067113, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 27288 + }, + { + "epoch": 0.27289, + "grad_norm": 0.7262106536645934, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 27289 + }, + { + "epoch": 0.2729, + "grad_norm": 0.769036380219617, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 27290 + }, + { + "epoch": 0.27291, + "grad_norm": 0.8771252150741305, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 27291 + }, + { + "epoch": 0.27292, + "grad_norm": 0.9163214352354011, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 27292 + }, + { + "epoch": 0.27293, + "grad_norm": 0.7484134503288268, + "learning_rate": 0.003, + "loss": 3.982, + "step": 27293 + }, + { + "epoch": 0.27294, + "grad_norm": 0.5642006740398628, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 27294 + }, + { + "epoch": 0.27295, + "grad_norm": 0.5072314292728279, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 27295 + }, + { + "epoch": 0.27296, + "grad_norm": 0.5234146489233675, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 27296 + }, + { + "epoch": 0.27297, + "grad_norm": 0.5196432740153647, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 27297 + }, + { + "epoch": 0.27298, + "grad_norm": 0.4538329596191541, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 27298 + }, + { + "epoch": 0.27299, + "grad_norm": 0.4327911306070125, + "learning_rate": 0.003, + "loss": 3.985, + "step": 27299 + }, + { + "epoch": 0.273, + "grad_norm": 0.4823701939566797, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 27300 + }, + { + "epoch": 0.27301, + "grad_norm": 0.524778597211109, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 27301 + }, + { + "epoch": 0.27302, + "grad_norm": 0.6133654904551438, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 27302 + }, + { + "epoch": 0.27303, + "grad_norm": 0.6578560298041923, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 27303 + }, + { + "epoch": 0.27304, + "grad_norm": 0.6182804441798923, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 27304 + }, + { + "epoch": 0.27305, + "grad_norm": 0.6282040367208515, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 27305 + }, + { + "epoch": 0.27306, + "grad_norm": 0.603997060653713, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 27306 + }, + { + "epoch": 0.27307, + "grad_norm": 0.6963071953260551, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 27307 + }, + { + "epoch": 0.27308, + "grad_norm": 0.7343069131399572, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 27308 + }, + { + "epoch": 0.27309, + "grad_norm": 0.6436583181246939, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 27309 + }, + { + "epoch": 0.2731, + "grad_norm": 0.6766585978450302, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 27310 + }, + { + "epoch": 0.27311, + "grad_norm": 0.6980548829801871, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 27311 + }, + { + "epoch": 0.27312, + "grad_norm": 0.7929983672045327, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 27312 + }, + { + "epoch": 0.27313, + "grad_norm": 0.9055763095413364, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 27313 + }, + { + "epoch": 0.27314, + "grad_norm": 1.0453716039003869, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 27314 + }, + { + "epoch": 0.27315, + "grad_norm": 0.8359214234493059, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 27315 + }, + { + "epoch": 0.27316, + "grad_norm": 0.7841696975632837, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 27316 + }, + { + "epoch": 0.27317, + "grad_norm": 0.8871178364354252, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 27317 + }, + { + "epoch": 0.27318, + "grad_norm": 0.8241664698868331, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 27318 + }, + { + "epoch": 0.27319, + "grad_norm": 0.7936572245715043, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 27319 + }, + { + "epoch": 0.2732, + "grad_norm": 0.6493362440808795, + "learning_rate": 0.003, + "loss": 3.982, + "step": 27320 + }, + { + "epoch": 0.27321, + "grad_norm": 0.626016613395403, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 27321 + }, + { + "epoch": 0.27322, + "grad_norm": 0.73149472811308, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 27322 + }, + { + "epoch": 0.27323, + "grad_norm": 0.7136109417705138, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 27323 + }, + { + "epoch": 0.27324, + "grad_norm": 0.7517379881892611, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 27324 + }, + { + "epoch": 0.27325, + "grad_norm": 0.8226003350756242, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 27325 + }, + { + "epoch": 0.27326, + "grad_norm": 0.7918100265834269, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 27326 + }, + { + "epoch": 0.27327, + "grad_norm": 0.6855386635463571, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 27327 + }, + { + "epoch": 0.27328, + "grad_norm": 0.7519501175189863, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 27328 + }, + { + "epoch": 0.27329, + "grad_norm": 0.8398643177567885, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 27329 + }, + { + "epoch": 0.2733, + "grad_norm": 0.9490231497829494, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 27330 + }, + { + "epoch": 0.27331, + "grad_norm": 0.9504779299870811, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 27331 + }, + { + "epoch": 0.27332, + "grad_norm": 0.8091083825741897, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 27332 + }, + { + "epoch": 0.27333, + "grad_norm": 0.747209473662579, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 27333 + }, + { + "epoch": 0.27334, + "grad_norm": 0.7089973794825576, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 27334 + }, + { + "epoch": 0.27335, + "grad_norm": 0.6856761165845606, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 27335 + }, + { + "epoch": 0.27336, + "grad_norm": 0.6855728085189408, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 27336 + }, + { + "epoch": 0.27337, + "grad_norm": 0.6873362623843932, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 27337 + }, + { + "epoch": 0.27338, + "grad_norm": 0.6119687936430895, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 27338 + }, + { + "epoch": 0.27339, + "grad_norm": 0.7126292013201697, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 27339 + }, + { + "epoch": 0.2734, + "grad_norm": 0.8069514451994955, + "learning_rate": 0.003, + "loss": 3.986, + "step": 27340 + }, + { + "epoch": 0.27341, + "grad_norm": 1.0147330067245532, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 27341 + }, + { + "epoch": 0.27342, + "grad_norm": 1.0886395163145577, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 27342 + }, + { + "epoch": 0.27343, + "grad_norm": 0.817806900757776, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 27343 + }, + { + "epoch": 0.27344, + "grad_norm": 0.6807825553232755, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 27344 + }, + { + "epoch": 0.27345, + "grad_norm": 0.6534763461717803, + "learning_rate": 0.003, + "loss": 3.999, + "step": 27345 + }, + { + "epoch": 0.27346, + "grad_norm": 0.690000617489994, + "learning_rate": 0.003, + "loss": 3.968, + "step": 27346 + }, + { + "epoch": 0.27347, + "grad_norm": 0.7816541418656591, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 27347 + }, + { + "epoch": 0.27348, + "grad_norm": 0.8168964234650746, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 27348 + }, + { + "epoch": 0.27349, + "grad_norm": 0.822725049724362, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 27349 + }, + { + "epoch": 0.2735, + "grad_norm": 0.7432415284770858, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 27350 + }, + { + "epoch": 0.27351, + "grad_norm": 0.693917552109886, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 27351 + }, + { + "epoch": 0.27352, + "grad_norm": 0.6795673346600425, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 27352 + }, + { + "epoch": 0.27353, + "grad_norm": 0.6188726098383244, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 27353 + }, + { + "epoch": 0.27354, + "grad_norm": 0.6039316901279284, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 27354 + }, + { + "epoch": 0.27355, + "grad_norm": 0.5535894709520285, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 27355 + }, + { + "epoch": 0.27356, + "grad_norm": 0.5759682273271701, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 27356 + }, + { + "epoch": 0.27357, + "grad_norm": 0.5904886492416502, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 27357 + }, + { + "epoch": 0.27358, + "grad_norm": 0.5359867806917052, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 27358 + }, + { + "epoch": 0.27359, + "grad_norm": 0.6281814854007486, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 27359 + }, + { + "epoch": 0.2736, + "grad_norm": 0.6841784709461111, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 27360 + }, + { + "epoch": 0.27361, + "grad_norm": 0.7991706745245093, + "learning_rate": 0.003, + "loss": 3.985, + "step": 27361 + }, + { + "epoch": 0.27362, + "grad_norm": 0.8547516595271809, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 27362 + }, + { + "epoch": 0.27363, + "grad_norm": 0.8716177932151272, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 27363 + }, + { + "epoch": 0.27364, + "grad_norm": 0.97710326273114, + "learning_rate": 0.003, + "loss": 3.999, + "step": 27364 + }, + { + "epoch": 0.27365, + "grad_norm": 1.0141663511554566, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 27365 + }, + { + "epoch": 0.27366, + "grad_norm": 1.0593380342628946, + "learning_rate": 0.003, + "loss": 3.99, + "step": 27366 + }, + { + "epoch": 0.27367, + "grad_norm": 0.9622181782753357, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 27367 + }, + { + "epoch": 0.27368, + "grad_norm": 0.8891019661362763, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 27368 + }, + { + "epoch": 0.27369, + "grad_norm": 0.8205308596103951, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 27369 + }, + { + "epoch": 0.2737, + "grad_norm": 0.8061369467196615, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 27370 + }, + { + "epoch": 0.27371, + "grad_norm": 0.8979324011287657, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 27371 + }, + { + "epoch": 0.27372, + "grad_norm": 0.855880164582388, + "learning_rate": 0.003, + "loss": 4.001, + "step": 27372 + }, + { + "epoch": 0.27373, + "grad_norm": 0.8539588248139159, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 27373 + }, + { + "epoch": 0.27374, + "grad_norm": 0.9541097256765466, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 27374 + }, + { + "epoch": 0.27375, + "grad_norm": 0.9594509625597786, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 27375 + }, + { + "epoch": 0.27376, + "grad_norm": 1.1006836913678988, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 27376 + }, + { + "epoch": 0.27377, + "grad_norm": 0.8215789331139838, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 27377 + }, + { + "epoch": 0.27378, + "grad_norm": 0.733985126995747, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 27378 + }, + { + "epoch": 0.27379, + "grad_norm": 0.6642016511282698, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 27379 + }, + { + "epoch": 0.2738, + "grad_norm": 0.5876558274201218, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 27380 + }, + { + "epoch": 0.27381, + "grad_norm": 0.6257670795703766, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 27381 + }, + { + "epoch": 0.27382, + "grad_norm": 0.738944746323079, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 27382 + }, + { + "epoch": 0.27383, + "grad_norm": 0.8058308169687158, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 27383 + }, + { + "epoch": 0.27384, + "grad_norm": 0.7532683204797291, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27384 + }, + { + "epoch": 0.27385, + "grad_norm": 0.6375918126845437, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 27385 + }, + { + "epoch": 0.27386, + "grad_norm": 0.6947278928209589, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 27386 + }, + { + "epoch": 0.27387, + "grad_norm": 0.74633840372219, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 27387 + }, + { + "epoch": 0.27388, + "grad_norm": 0.7477228759970019, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 27388 + }, + { + "epoch": 0.27389, + "grad_norm": 0.7779350523523124, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 27389 + }, + { + "epoch": 0.2739, + "grad_norm": 0.7899335316079754, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 27390 + }, + { + "epoch": 0.27391, + "grad_norm": 0.7146613520159817, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 27391 + }, + { + "epoch": 0.27392, + "grad_norm": 0.6901773014935291, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 27392 + }, + { + "epoch": 0.27393, + "grad_norm": 0.6125739098416996, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 27393 + }, + { + "epoch": 0.27394, + "grad_norm": 0.5907411096500715, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 27394 + }, + { + "epoch": 0.27395, + "grad_norm": 0.593820215822552, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 27395 + }, + { + "epoch": 0.27396, + "grad_norm": 0.6005307511510178, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 27396 + }, + { + "epoch": 0.27397, + "grad_norm": 0.479662329911353, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 27397 + }, + { + "epoch": 0.27398, + "grad_norm": 0.6059930851958173, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 27398 + }, + { + "epoch": 0.27399, + "grad_norm": 0.6244908470194416, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 27399 + }, + { + "epoch": 0.274, + "grad_norm": 0.7241567795993303, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 27400 + }, + { + "epoch": 0.27401, + "grad_norm": 0.7809781743241709, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 27401 + }, + { + "epoch": 0.27402, + "grad_norm": 0.7879066533384952, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 27402 + }, + { + "epoch": 0.27403, + "grad_norm": 0.8104892957235827, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 27403 + }, + { + "epoch": 0.27404, + "grad_norm": 0.9415889305970484, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 27404 + }, + { + "epoch": 0.27405, + "grad_norm": 0.9640196053346303, + "learning_rate": 0.003, + "loss": 3.981, + "step": 27405 + }, + { + "epoch": 0.27406, + "grad_norm": 0.8690455096171272, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 27406 + }, + { + "epoch": 0.27407, + "grad_norm": 0.8290569912735802, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 27407 + }, + { + "epoch": 0.27408, + "grad_norm": 0.8383170336533754, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 27408 + }, + { + "epoch": 0.27409, + "grad_norm": 0.9003885767242851, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 27409 + }, + { + "epoch": 0.2741, + "grad_norm": 0.7089563369541907, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 27410 + }, + { + "epoch": 0.27411, + "grad_norm": 0.7434237176337065, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 27411 + }, + { + "epoch": 0.27412, + "grad_norm": 0.7337323662301016, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 27412 + }, + { + "epoch": 0.27413, + "grad_norm": 0.7063170908527584, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 27413 + }, + { + "epoch": 0.27414, + "grad_norm": 0.7040558505482633, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 27414 + }, + { + "epoch": 0.27415, + "grad_norm": 0.6886301034828517, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 27415 + }, + { + "epoch": 0.27416, + "grad_norm": 0.6955136403590492, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 27416 + }, + { + "epoch": 0.27417, + "grad_norm": 0.6515905952264904, + "learning_rate": 0.003, + "loss": 3.979, + "step": 27417 + }, + { + "epoch": 0.27418, + "grad_norm": 0.5989265330004819, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 27418 + }, + { + "epoch": 0.27419, + "grad_norm": 0.6562244019085367, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 27419 + }, + { + "epoch": 0.2742, + "grad_norm": 0.7035389256073604, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 27420 + }, + { + "epoch": 0.27421, + "grad_norm": 0.7789557532008093, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 27421 + }, + { + "epoch": 0.27422, + "grad_norm": 0.9593205562933252, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 27422 + }, + { + "epoch": 0.27423, + "grad_norm": 1.1115099671740498, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 27423 + }, + { + "epoch": 0.27424, + "grad_norm": 0.9106572342794219, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 27424 + }, + { + "epoch": 0.27425, + "grad_norm": 0.9907996703449462, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 27425 + }, + { + "epoch": 0.27426, + "grad_norm": 1.0562329783779114, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 27426 + }, + { + "epoch": 0.27427, + "grad_norm": 0.9921058809789303, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 27427 + }, + { + "epoch": 0.27428, + "grad_norm": 1.0257282168632889, + "learning_rate": 0.003, + "loss": 4.002, + "step": 27428 + }, + { + "epoch": 0.27429, + "grad_norm": 1.2361510432440381, + "learning_rate": 0.003, + "loss": 4.039, + "step": 27429 + }, + { + "epoch": 0.2743, + "grad_norm": 0.8145778893385199, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 27430 + }, + { + "epoch": 0.27431, + "grad_norm": 0.7196908690465379, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 27431 + }, + { + "epoch": 0.27432, + "grad_norm": 0.6439930804260944, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 27432 + }, + { + "epoch": 0.27433, + "grad_norm": 0.6760792761710389, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 27433 + }, + { + "epoch": 0.27434, + "grad_norm": 0.6632120468108798, + "learning_rate": 0.003, + "loss": 3.973, + "step": 27434 + }, + { + "epoch": 0.27435, + "grad_norm": 0.594320854685027, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 27435 + }, + { + "epoch": 0.27436, + "grad_norm": 0.6335937420556015, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 27436 + }, + { + "epoch": 0.27437, + "grad_norm": 0.6872400715176129, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 27437 + }, + { + "epoch": 0.27438, + "grad_norm": 0.7504963328780516, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 27438 + }, + { + "epoch": 0.27439, + "grad_norm": 0.8831018888292786, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 27439 + }, + { + "epoch": 0.2744, + "grad_norm": 0.9258936525364058, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 27440 + }, + { + "epoch": 0.27441, + "grad_norm": 0.9388462024342203, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 27441 + }, + { + "epoch": 0.27442, + "grad_norm": 0.7342221942578473, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 27442 + }, + { + "epoch": 0.27443, + "grad_norm": 0.7558947409761582, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 27443 + }, + { + "epoch": 0.27444, + "grad_norm": 0.8431481100856398, + "learning_rate": 0.003, + "loss": 4.003, + "step": 27444 + }, + { + "epoch": 0.27445, + "grad_norm": 0.9373981200889032, + "learning_rate": 0.003, + "loss": 3.997, + "step": 27445 + }, + { + "epoch": 0.27446, + "grad_norm": 0.8508070937303446, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 27446 + }, + { + "epoch": 0.27447, + "grad_norm": 0.6669638882599368, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 27447 + }, + { + "epoch": 0.27448, + "grad_norm": 0.6547270049731772, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 27448 + }, + { + "epoch": 0.27449, + "grad_norm": 0.6690418918023043, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 27449 + }, + { + "epoch": 0.2745, + "grad_norm": 0.6637846246988327, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 27450 + }, + { + "epoch": 0.27451, + "grad_norm": 0.624548399309038, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 27451 + }, + { + "epoch": 0.27452, + "grad_norm": 0.5714731946584503, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 27452 + }, + { + "epoch": 0.27453, + "grad_norm": 0.6004831911503996, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 27453 + }, + { + "epoch": 0.27454, + "grad_norm": 0.6561512698580193, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 27454 + }, + { + "epoch": 0.27455, + "grad_norm": 0.6829397056531255, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 27455 + }, + { + "epoch": 0.27456, + "grad_norm": 0.635983673378462, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 27456 + }, + { + "epoch": 0.27457, + "grad_norm": 0.6045558190831911, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 27457 + }, + { + "epoch": 0.27458, + "grad_norm": 0.5658738779189912, + "learning_rate": 0.003, + "loss": 3.975, + "step": 27458 + }, + { + "epoch": 0.27459, + "grad_norm": 0.49272163401595503, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 27459 + }, + { + "epoch": 0.2746, + "grad_norm": 0.5083649996660085, + "learning_rate": 0.003, + "loss": 3.972, + "step": 27460 + }, + { + "epoch": 0.27461, + "grad_norm": 0.5554077759123707, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 27461 + }, + { + "epoch": 0.27462, + "grad_norm": 0.6327418503741816, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 27462 + }, + { + "epoch": 0.27463, + "grad_norm": 0.7914181724132145, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27463 + }, + { + "epoch": 0.27464, + "grad_norm": 0.9458593327883079, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 27464 + }, + { + "epoch": 0.27465, + "grad_norm": 0.935325721479939, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 27465 + }, + { + "epoch": 0.27466, + "grad_norm": 0.7617981130140338, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 27466 + }, + { + "epoch": 0.27467, + "grad_norm": 0.6462302639807164, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 27467 + }, + { + "epoch": 0.27468, + "grad_norm": 0.8355356312810984, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 27468 + }, + { + "epoch": 0.27469, + "grad_norm": 0.9614904248317762, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 27469 + }, + { + "epoch": 0.2747, + "grad_norm": 0.8372699376447595, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 27470 + }, + { + "epoch": 0.27471, + "grad_norm": 0.7095809350322374, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 27471 + }, + { + "epoch": 0.27472, + "grad_norm": 0.6679935520522007, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 27472 + }, + { + "epoch": 0.27473, + "grad_norm": 0.6659828152725513, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 27473 + }, + { + "epoch": 0.27474, + "grad_norm": 0.6998837906719945, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 27474 + }, + { + "epoch": 0.27475, + "grad_norm": 0.6556183186146022, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27475 + }, + { + "epoch": 0.27476, + "grad_norm": 0.7367114679358231, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 27476 + }, + { + "epoch": 0.27477, + "grad_norm": 0.8906382338240273, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 27477 + }, + { + "epoch": 0.27478, + "grad_norm": 0.9203786090519548, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 27478 + }, + { + "epoch": 0.27479, + "grad_norm": 0.919174459306262, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 27479 + }, + { + "epoch": 0.2748, + "grad_norm": 0.9396984486316355, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 27480 + }, + { + "epoch": 0.27481, + "grad_norm": 0.810374528458854, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 27481 + }, + { + "epoch": 0.27482, + "grad_norm": 0.8499271747675768, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 27482 + }, + { + "epoch": 0.27483, + "grad_norm": 0.9113924972240574, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 27483 + }, + { + "epoch": 0.27484, + "grad_norm": 1.0277983350207294, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 27484 + }, + { + "epoch": 0.27485, + "grad_norm": 0.9689018336568809, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 27485 + }, + { + "epoch": 0.27486, + "grad_norm": 1.0101450372748988, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 27486 + }, + { + "epoch": 0.27487, + "grad_norm": 0.9351295767622417, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 27487 + }, + { + "epoch": 0.27488, + "grad_norm": 0.8293076058907924, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 27488 + }, + { + "epoch": 0.27489, + "grad_norm": 0.8425247628923354, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 27489 + }, + { + "epoch": 0.2749, + "grad_norm": 0.8688011953481125, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 27490 + }, + { + "epoch": 0.27491, + "grad_norm": 0.903806341475528, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 27491 + }, + { + "epoch": 0.27492, + "grad_norm": 0.8283219186437213, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 27492 + }, + { + "epoch": 0.27493, + "grad_norm": 0.8080890859675905, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 27493 + }, + { + "epoch": 0.27494, + "grad_norm": 0.8181186630262883, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 27494 + }, + { + "epoch": 0.27495, + "grad_norm": 0.7822051916380566, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 27495 + }, + { + "epoch": 0.27496, + "grad_norm": 0.823628303933078, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 27496 + }, + { + "epoch": 0.27497, + "grad_norm": 0.8305902837971004, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 27497 + }, + { + "epoch": 0.27498, + "grad_norm": 0.8921734226161337, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 27498 + }, + { + "epoch": 0.27499, + "grad_norm": 0.878086429330449, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 27499 + }, + { + "epoch": 0.275, + "grad_norm": 0.8521314883941742, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 27500 + }, + { + "epoch": 0.27501, + "grad_norm": 0.6723138062030379, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 27501 + }, + { + "epoch": 0.27502, + "grad_norm": 0.6531568487722126, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 27502 + }, + { + "epoch": 0.27503, + "grad_norm": 0.6444000925825306, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 27503 + }, + { + "epoch": 0.27504, + "grad_norm": 0.6683902869889419, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 27504 + }, + { + "epoch": 0.27505, + "grad_norm": 0.642176899988935, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 27505 + }, + { + "epoch": 0.27506, + "grad_norm": 0.5381260874129726, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 27506 + }, + { + "epoch": 0.27507, + "grad_norm": 0.5848132813283372, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 27507 + }, + { + "epoch": 0.27508, + "grad_norm": 0.6471260503298695, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 27508 + }, + { + "epoch": 0.27509, + "grad_norm": 0.6668276676150034, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 27509 + }, + { + "epoch": 0.2751, + "grad_norm": 0.6406822321251413, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 27510 + }, + { + "epoch": 0.27511, + "grad_norm": 0.5504851611221279, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 27511 + }, + { + "epoch": 0.27512, + "grad_norm": 0.6481548846115451, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 27512 + }, + { + "epoch": 0.27513, + "grad_norm": 0.7995398091577213, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 27513 + }, + { + "epoch": 0.27514, + "grad_norm": 0.923520002929306, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 27514 + }, + { + "epoch": 0.27515, + "grad_norm": 0.9916667321113767, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 27515 + }, + { + "epoch": 0.27516, + "grad_norm": 0.8390512234554638, + "learning_rate": 0.003, + "loss": 3.993, + "step": 27516 + }, + { + "epoch": 0.27517, + "grad_norm": 0.731642580220124, + "learning_rate": 0.003, + "loss": 3.994, + "step": 27517 + }, + { + "epoch": 0.27518, + "grad_norm": 0.6570862322995327, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 27518 + }, + { + "epoch": 0.27519, + "grad_norm": 0.6752480432773217, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 27519 + }, + { + "epoch": 0.2752, + "grad_norm": 0.6783977355586, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 27520 + }, + { + "epoch": 0.27521, + "grad_norm": 0.7262301540130852, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 27521 + }, + { + "epoch": 0.27522, + "grad_norm": 0.7644246333612531, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 27522 + }, + { + "epoch": 0.27523, + "grad_norm": 0.7573336855440234, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 27523 + }, + { + "epoch": 0.27524, + "grad_norm": 0.7718915728721562, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 27524 + }, + { + "epoch": 0.27525, + "grad_norm": 0.7601986682817271, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 27525 + }, + { + "epoch": 0.27526, + "grad_norm": 0.7487487597887328, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 27526 + }, + { + "epoch": 0.27527, + "grad_norm": 0.7416749444317515, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 27527 + }, + { + "epoch": 0.27528, + "grad_norm": 0.7387232606744256, + "learning_rate": 0.003, + "loss": 3.993, + "step": 27528 + }, + { + "epoch": 0.27529, + "grad_norm": 0.7913306772563667, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 27529 + }, + { + "epoch": 0.2753, + "grad_norm": 0.8699288603274848, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 27530 + }, + { + "epoch": 0.27531, + "grad_norm": 0.8836207946906436, + "learning_rate": 0.003, + "loss": 4.016, + "step": 27531 + }, + { + "epoch": 0.27532, + "grad_norm": 0.9649490884976414, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 27532 + }, + { + "epoch": 0.27533, + "grad_norm": 1.0089019313459613, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 27533 + }, + { + "epoch": 0.27534, + "grad_norm": 1.029274155202031, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 27534 + }, + { + "epoch": 0.27535, + "grad_norm": 0.8129706030808552, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 27535 + }, + { + "epoch": 0.27536, + "grad_norm": 0.8783377203936273, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 27536 + }, + { + "epoch": 0.27537, + "grad_norm": 0.9035340925752766, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 27537 + }, + { + "epoch": 0.27538, + "grad_norm": 0.8393924439165747, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 27538 + }, + { + "epoch": 0.27539, + "grad_norm": 0.736469124788594, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 27539 + }, + { + "epoch": 0.2754, + "grad_norm": 0.7628384607941366, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 27540 + }, + { + "epoch": 0.27541, + "grad_norm": 0.7947348065415869, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 27541 + }, + { + "epoch": 0.27542, + "grad_norm": 0.7777865778779078, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 27542 + }, + { + "epoch": 0.27543, + "grad_norm": 0.6813461962050159, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 27543 + }, + { + "epoch": 0.27544, + "grad_norm": 0.6672498038801292, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 27544 + }, + { + "epoch": 0.27545, + "grad_norm": 0.5832216867055693, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 27545 + }, + { + "epoch": 0.27546, + "grad_norm": 0.6215819093360376, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 27546 + }, + { + "epoch": 0.27547, + "grad_norm": 0.7632470984282598, + "learning_rate": 0.003, + "loss": 4.005, + "step": 27547 + }, + { + "epoch": 0.27548, + "grad_norm": 0.8206443084810752, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 27548 + }, + { + "epoch": 0.27549, + "grad_norm": 0.8442564822904237, + "learning_rate": 0.003, + "loss": 3.994, + "step": 27549 + }, + { + "epoch": 0.2755, + "grad_norm": 0.9138145305321193, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 27550 + }, + { + "epoch": 0.27551, + "grad_norm": 1.0622509096463812, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 27551 + }, + { + "epoch": 0.27552, + "grad_norm": 0.8713262707192515, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 27552 + }, + { + "epoch": 0.27553, + "grad_norm": 0.7027580411603492, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 27553 + }, + { + "epoch": 0.27554, + "grad_norm": 0.6865197602735642, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 27554 + }, + { + "epoch": 0.27555, + "grad_norm": 0.6784521292153591, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 27555 + }, + { + "epoch": 0.27556, + "grad_norm": 0.8388568795935056, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 27556 + }, + { + "epoch": 0.27557, + "grad_norm": 0.9448427676038776, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 27557 + }, + { + "epoch": 0.27558, + "grad_norm": 1.0758467762120167, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 27558 + }, + { + "epoch": 0.27559, + "grad_norm": 0.7825674066921499, + "learning_rate": 0.003, + "loss": 3.973, + "step": 27559 + }, + { + "epoch": 0.2756, + "grad_norm": 0.6714697127122128, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 27560 + }, + { + "epoch": 0.27561, + "grad_norm": 0.6456204916985722, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 27561 + }, + { + "epoch": 0.27562, + "grad_norm": 0.6227196272269658, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 27562 + }, + { + "epoch": 0.27563, + "grad_norm": 0.6344643455885347, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 27563 + }, + { + "epoch": 0.27564, + "grad_norm": 0.6188582809268938, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 27564 + }, + { + "epoch": 0.27565, + "grad_norm": 0.5526978613913931, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 27565 + }, + { + "epoch": 0.27566, + "grad_norm": 0.5205662909072358, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 27566 + }, + { + "epoch": 0.27567, + "grad_norm": 0.5221541978766028, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 27567 + }, + { + "epoch": 0.27568, + "grad_norm": 0.5914032051406121, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 27568 + }, + { + "epoch": 0.27569, + "grad_norm": 0.7417916795478837, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 27569 + }, + { + "epoch": 0.2757, + "grad_norm": 0.8079729237153062, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 27570 + }, + { + "epoch": 0.27571, + "grad_norm": 0.823719276353226, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 27571 + }, + { + "epoch": 0.27572, + "grad_norm": 0.7412877593544407, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 27572 + }, + { + "epoch": 0.27573, + "grad_norm": 0.7667867819057317, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 27573 + }, + { + "epoch": 0.27574, + "grad_norm": 0.7202919422653935, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 27574 + }, + { + "epoch": 0.27575, + "grad_norm": 0.7434751180870758, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 27575 + }, + { + "epoch": 0.27576, + "grad_norm": 0.7275199115348503, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 27576 + }, + { + "epoch": 0.27577, + "grad_norm": 0.6871569871490335, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 27577 + }, + { + "epoch": 0.27578, + "grad_norm": 0.6110469038990823, + "learning_rate": 0.003, + "loss": 3.981, + "step": 27578 + }, + { + "epoch": 0.27579, + "grad_norm": 0.6036345833626791, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 27579 + }, + { + "epoch": 0.2758, + "grad_norm": 0.6665802714441758, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 27580 + }, + { + "epoch": 0.27581, + "grad_norm": 0.8417997304511834, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 27581 + }, + { + "epoch": 0.27582, + "grad_norm": 0.9239623013136103, + "learning_rate": 0.003, + "loss": 3.978, + "step": 27582 + }, + { + "epoch": 0.27583, + "grad_norm": 0.8908137639633806, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 27583 + }, + { + "epoch": 0.27584, + "grad_norm": 0.7200813901403839, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 27584 + }, + { + "epoch": 0.27585, + "grad_norm": 0.8103924711276943, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 27585 + }, + { + "epoch": 0.27586, + "grad_norm": 0.8954560833325947, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 27586 + }, + { + "epoch": 0.27587, + "grad_norm": 0.9662763930304512, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 27587 + }, + { + "epoch": 0.27588, + "grad_norm": 1.0031360174685702, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 27588 + }, + { + "epoch": 0.27589, + "grad_norm": 0.9113541364345298, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 27589 + }, + { + "epoch": 0.2759, + "grad_norm": 0.8473792586257566, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 27590 + }, + { + "epoch": 0.27591, + "grad_norm": 0.9842073752567351, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 27591 + }, + { + "epoch": 0.27592, + "grad_norm": 1.0426795452649968, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 27592 + }, + { + "epoch": 0.27593, + "grad_norm": 0.9114134012762175, + "learning_rate": 0.003, + "loss": 3.973, + "step": 27593 + }, + { + "epoch": 0.27594, + "grad_norm": 0.8097691701652995, + "learning_rate": 0.003, + "loss": 4.027, + "step": 27594 + }, + { + "epoch": 0.27595, + "grad_norm": 0.7881594808333235, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 27595 + }, + { + "epoch": 0.27596, + "grad_norm": 0.7658335827383506, + "learning_rate": 0.003, + "loss": 4.009, + "step": 27596 + }, + { + "epoch": 0.27597, + "grad_norm": 0.7357219516258428, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 27597 + }, + { + "epoch": 0.27598, + "grad_norm": 0.6786414927948052, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 27598 + }, + { + "epoch": 0.27599, + "grad_norm": 0.7157051108866471, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 27599 + }, + { + "epoch": 0.276, + "grad_norm": 0.7957143143423834, + "learning_rate": 0.003, + "loss": 3.983, + "step": 27600 + }, + { + "epoch": 0.27601, + "grad_norm": 0.7784672604907453, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 27601 + }, + { + "epoch": 0.27602, + "grad_norm": 0.6584121908098993, + "learning_rate": 0.003, + "loss": 4.029, + "step": 27602 + }, + { + "epoch": 0.27603, + "grad_norm": 0.5751317778928002, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 27603 + }, + { + "epoch": 0.27604, + "grad_norm": 0.6598895133908711, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 27604 + }, + { + "epoch": 0.27605, + "grad_norm": 0.7248461608746108, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 27605 + }, + { + "epoch": 0.27606, + "grad_norm": 0.7361840514845751, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 27606 + }, + { + "epoch": 0.27607, + "grad_norm": 0.7861647138744697, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 27607 + }, + { + "epoch": 0.27608, + "grad_norm": 0.8151850231193966, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 27608 + }, + { + "epoch": 0.27609, + "grad_norm": 0.8466396345141237, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 27609 + }, + { + "epoch": 0.2761, + "grad_norm": 0.8358044398611868, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 27610 + }, + { + "epoch": 0.27611, + "grad_norm": 0.8252144156682629, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 27611 + }, + { + "epoch": 0.27612, + "grad_norm": 0.7836096562507018, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 27612 + }, + { + "epoch": 0.27613, + "grad_norm": 0.6310416339691713, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 27613 + }, + { + "epoch": 0.27614, + "grad_norm": 0.5241468942619337, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 27614 + }, + { + "epoch": 0.27615, + "grad_norm": 0.4742128489844521, + "learning_rate": 0.003, + "loss": 3.979, + "step": 27615 + }, + { + "epoch": 0.27616, + "grad_norm": 0.625252165261094, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 27616 + }, + { + "epoch": 0.27617, + "grad_norm": 0.6412631243935891, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 27617 + }, + { + "epoch": 0.27618, + "grad_norm": 0.6406089700712683, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 27618 + }, + { + "epoch": 0.27619, + "grad_norm": 0.7037171877529763, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 27619 + }, + { + "epoch": 0.2762, + "grad_norm": 0.8457227775836256, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 27620 + }, + { + "epoch": 0.27621, + "grad_norm": 0.9010402337046045, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 27621 + }, + { + "epoch": 0.27622, + "grad_norm": 0.8593653865833153, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 27622 + }, + { + "epoch": 0.27623, + "grad_norm": 0.801189691101168, + "learning_rate": 0.003, + "loss": 3.962, + "step": 27623 + }, + { + "epoch": 0.27624, + "grad_norm": 0.6511801560781395, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 27624 + }, + { + "epoch": 0.27625, + "grad_norm": 0.6439500128092362, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 27625 + }, + { + "epoch": 0.27626, + "grad_norm": 0.631365516036763, + "learning_rate": 0.003, + "loss": 3.983, + "step": 27626 + }, + { + "epoch": 0.27627, + "grad_norm": 0.70574554782682, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 27627 + }, + { + "epoch": 0.27628, + "grad_norm": 0.7276842370873595, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 27628 + }, + { + "epoch": 0.27629, + "grad_norm": 0.722156220876055, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 27629 + }, + { + "epoch": 0.2763, + "grad_norm": 0.8226203838764751, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 27630 + }, + { + "epoch": 0.27631, + "grad_norm": 1.0260038120206039, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 27631 + }, + { + "epoch": 0.27632, + "grad_norm": 1.0118300728810863, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 27632 + }, + { + "epoch": 0.27633, + "grad_norm": 0.8218901502406133, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 27633 + }, + { + "epoch": 0.27634, + "grad_norm": 0.7599500469599454, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 27634 + }, + { + "epoch": 0.27635, + "grad_norm": 0.6869429571221999, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 27635 + }, + { + "epoch": 0.27636, + "grad_norm": 0.5858730641804928, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 27636 + }, + { + "epoch": 0.27637, + "grad_norm": 0.6949272884715589, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 27637 + }, + { + "epoch": 0.27638, + "grad_norm": 0.6701329137044186, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 27638 + }, + { + "epoch": 0.27639, + "grad_norm": 0.7108457261439343, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 27639 + }, + { + "epoch": 0.2764, + "grad_norm": 0.7579006929337919, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 27640 + }, + { + "epoch": 0.27641, + "grad_norm": 0.8040500357340891, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 27641 + }, + { + "epoch": 0.27642, + "grad_norm": 0.8151899706906818, + "learning_rate": 0.003, + "loss": 4.016, + "step": 27642 + }, + { + "epoch": 0.27643, + "grad_norm": 0.757362237807463, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 27643 + }, + { + "epoch": 0.27644, + "grad_norm": 0.7393043542845629, + "learning_rate": 0.003, + "loss": 3.994, + "step": 27644 + }, + { + "epoch": 0.27645, + "grad_norm": 0.7056632334754847, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 27645 + }, + { + "epoch": 0.27646, + "grad_norm": 0.7825890412561322, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 27646 + }, + { + "epoch": 0.27647, + "grad_norm": 0.7608598935543437, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 27647 + }, + { + "epoch": 0.27648, + "grad_norm": 0.6749569453049421, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 27648 + }, + { + "epoch": 0.27649, + "grad_norm": 0.7148929052972756, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 27649 + }, + { + "epoch": 0.2765, + "grad_norm": 0.7923230437013273, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 27650 + }, + { + "epoch": 0.27651, + "grad_norm": 0.9531939843792301, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 27651 + }, + { + "epoch": 0.27652, + "grad_norm": 1.0386505663013372, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 27652 + }, + { + "epoch": 0.27653, + "grad_norm": 0.9644580088586797, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 27653 + }, + { + "epoch": 0.27654, + "grad_norm": 1.0213164275924642, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 27654 + }, + { + "epoch": 0.27655, + "grad_norm": 0.959216492842491, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 27655 + }, + { + "epoch": 0.27656, + "grad_norm": 0.8451900086049983, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 27656 + }, + { + "epoch": 0.27657, + "grad_norm": 0.7815879089013716, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 27657 + }, + { + "epoch": 0.27658, + "grad_norm": 0.7826251066935491, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 27658 + }, + { + "epoch": 0.27659, + "grad_norm": 0.78207638458673, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 27659 + }, + { + "epoch": 0.2766, + "grad_norm": 0.9638448662026342, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 27660 + }, + { + "epoch": 0.27661, + "grad_norm": 0.9125434330418678, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 27661 + }, + { + "epoch": 0.27662, + "grad_norm": 0.6949355641029011, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 27662 + }, + { + "epoch": 0.27663, + "grad_norm": 0.645632717522372, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 27663 + }, + { + "epoch": 0.27664, + "grad_norm": 0.7007773528768063, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 27664 + }, + { + "epoch": 0.27665, + "grad_norm": 0.731865747770787, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27665 + }, + { + "epoch": 0.27666, + "grad_norm": 0.701911827865855, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 27666 + }, + { + "epoch": 0.27667, + "grad_norm": 0.5975161835178896, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 27667 + }, + { + "epoch": 0.27668, + "grad_norm": 0.6154980434367272, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 27668 + }, + { + "epoch": 0.27669, + "grad_norm": 0.5724289780878303, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 27669 + }, + { + "epoch": 0.2767, + "grad_norm": 0.5215480654516284, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 27670 + }, + { + "epoch": 0.27671, + "grad_norm": 0.5224429268797122, + "learning_rate": 0.003, + "loss": 3.964, + "step": 27671 + }, + { + "epoch": 0.27672, + "grad_norm": 0.469973559426477, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 27672 + }, + { + "epoch": 0.27673, + "grad_norm": 0.5449871736774428, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 27673 + }, + { + "epoch": 0.27674, + "grad_norm": 0.5622967852190902, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 27674 + }, + { + "epoch": 0.27675, + "grad_norm": 0.5717063260084012, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 27675 + }, + { + "epoch": 0.27676, + "grad_norm": 0.5735794356110594, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 27676 + }, + { + "epoch": 0.27677, + "grad_norm": 0.644250897543418, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 27677 + }, + { + "epoch": 0.27678, + "grad_norm": 0.7261677903504073, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 27678 + }, + { + "epoch": 0.27679, + "grad_norm": 0.9075764376419272, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 27679 + }, + { + "epoch": 0.2768, + "grad_norm": 0.9749430181597617, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 27680 + }, + { + "epoch": 0.27681, + "grad_norm": 0.9211727087459279, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 27681 + }, + { + "epoch": 0.27682, + "grad_norm": 0.8229213858418101, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 27682 + }, + { + "epoch": 0.27683, + "grad_norm": 0.8567310210892333, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 27683 + }, + { + "epoch": 0.27684, + "grad_norm": 0.944183262932085, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 27684 + }, + { + "epoch": 0.27685, + "grad_norm": 0.8743720177163036, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 27685 + }, + { + "epoch": 0.27686, + "grad_norm": 0.9170761876273953, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 27686 + }, + { + "epoch": 0.27687, + "grad_norm": 0.9882915914715316, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 27687 + }, + { + "epoch": 0.27688, + "grad_norm": 1.1539882100853611, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27688 + }, + { + "epoch": 0.27689, + "grad_norm": 0.7832794159576858, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 27689 + }, + { + "epoch": 0.2769, + "grad_norm": 0.7139199129731166, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 27690 + }, + { + "epoch": 0.27691, + "grad_norm": 0.6142703412304795, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 27691 + }, + { + "epoch": 0.27692, + "grad_norm": 0.6808533248402067, + "learning_rate": 0.003, + "loss": 3.987, + "step": 27692 + }, + { + "epoch": 0.27693, + "grad_norm": 0.6121450694217342, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 27693 + }, + { + "epoch": 0.27694, + "grad_norm": 0.5539207283974074, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 27694 + }, + { + "epoch": 0.27695, + "grad_norm": 0.5240083579866068, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 27695 + }, + { + "epoch": 0.27696, + "grad_norm": 0.5789394077180299, + "learning_rate": 0.003, + "loss": 4.007, + "step": 27696 + }, + { + "epoch": 0.27697, + "grad_norm": 0.660786157588859, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 27697 + }, + { + "epoch": 0.27698, + "grad_norm": 0.7242954945639034, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 27698 + }, + { + "epoch": 0.27699, + "grad_norm": 0.7472132876542437, + "learning_rate": 0.003, + "loss": 3.98, + "step": 27699 + }, + { + "epoch": 0.277, + "grad_norm": 0.7518575125085929, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 27700 + }, + { + "epoch": 0.27701, + "grad_norm": 0.7618452113865272, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 27701 + }, + { + "epoch": 0.27702, + "grad_norm": 0.7211460911096701, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 27702 + }, + { + "epoch": 0.27703, + "grad_norm": 0.686169463597501, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 27703 + }, + { + "epoch": 0.27704, + "grad_norm": 0.7256947088895593, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 27704 + }, + { + "epoch": 0.27705, + "grad_norm": 0.8024742025624635, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 27705 + }, + { + "epoch": 0.27706, + "grad_norm": 0.9232298732266953, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 27706 + }, + { + "epoch": 0.27707, + "grad_norm": 1.039295208312718, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 27707 + }, + { + "epoch": 0.27708, + "grad_norm": 1.1453692996866716, + "learning_rate": 0.003, + "loss": 3.995, + "step": 27708 + }, + { + "epoch": 0.27709, + "grad_norm": 0.9462551860403817, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 27709 + }, + { + "epoch": 0.2771, + "grad_norm": 0.8235357753414881, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 27710 + }, + { + "epoch": 0.27711, + "grad_norm": 0.8219738766581622, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 27711 + }, + { + "epoch": 0.27712, + "grad_norm": 0.8458840852358583, + "learning_rate": 0.003, + "loss": 3.971, + "step": 27712 + }, + { + "epoch": 0.27713, + "grad_norm": 0.8444082150001073, + "learning_rate": 0.003, + "loss": 4.003, + "step": 27713 + }, + { + "epoch": 0.27714, + "grad_norm": 0.8054107205017791, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 27714 + }, + { + "epoch": 0.27715, + "grad_norm": 0.739920643830956, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 27715 + }, + { + "epoch": 0.27716, + "grad_norm": 0.6252244912573307, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 27716 + }, + { + "epoch": 0.27717, + "grad_norm": 0.6169070234469615, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 27717 + }, + { + "epoch": 0.27718, + "grad_norm": 0.700148661457577, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 27718 + }, + { + "epoch": 0.27719, + "grad_norm": 0.8036333227764582, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 27719 + }, + { + "epoch": 0.2772, + "grad_norm": 0.6722883931565484, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 27720 + }, + { + "epoch": 0.27721, + "grad_norm": 0.5946736362210918, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 27721 + }, + { + "epoch": 0.27722, + "grad_norm": 0.629720372782294, + "learning_rate": 0.003, + "loss": 4.004, + "step": 27722 + }, + { + "epoch": 0.27723, + "grad_norm": 0.7720409416801831, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 27723 + }, + { + "epoch": 0.27724, + "grad_norm": 0.9600868543073822, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 27724 + }, + { + "epoch": 0.27725, + "grad_norm": 1.1211854066940785, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 27725 + }, + { + "epoch": 0.27726, + "grad_norm": 0.8200206951149853, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 27726 + }, + { + "epoch": 0.27727, + "grad_norm": 0.7844564881956752, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 27727 + }, + { + "epoch": 0.27728, + "grad_norm": 0.7382082843422226, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 27728 + }, + { + "epoch": 0.27729, + "grad_norm": 0.8282710544634673, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 27729 + }, + { + "epoch": 0.2773, + "grad_norm": 0.7597695224039251, + "learning_rate": 0.003, + "loss": 3.976, + "step": 27730 + }, + { + "epoch": 0.27731, + "grad_norm": 0.7476252094537339, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 27731 + }, + { + "epoch": 0.27732, + "grad_norm": 0.7625531606546558, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 27732 + }, + { + "epoch": 0.27733, + "grad_norm": 0.8138227039689894, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 27733 + }, + { + "epoch": 0.27734, + "grad_norm": 0.8223637854297942, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 27734 + }, + { + "epoch": 0.27735, + "grad_norm": 0.7995391312678078, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 27735 + }, + { + "epoch": 0.27736, + "grad_norm": 0.7538534876176207, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 27736 + }, + { + "epoch": 0.27737, + "grad_norm": 0.7323486630995175, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 27737 + }, + { + "epoch": 0.27738, + "grad_norm": 0.7020064957976612, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 27738 + }, + { + "epoch": 0.27739, + "grad_norm": 0.773933699120892, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 27739 + }, + { + "epoch": 0.2774, + "grad_norm": 0.6567677133571167, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 27740 + }, + { + "epoch": 0.27741, + "grad_norm": 0.550802179660512, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 27741 + }, + { + "epoch": 0.27742, + "grad_norm": 0.4293131401495554, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 27742 + }, + { + "epoch": 0.27743, + "grad_norm": 0.5060979579648514, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 27743 + }, + { + "epoch": 0.27744, + "grad_norm": 0.4969406590877531, + "learning_rate": 0.003, + "loss": 3.982, + "step": 27744 + }, + { + "epoch": 0.27745, + "grad_norm": 0.5417372719079292, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 27745 + }, + { + "epoch": 0.27746, + "grad_norm": 0.5836274597267015, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 27746 + }, + { + "epoch": 0.27747, + "grad_norm": 0.6626434163205588, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 27747 + }, + { + "epoch": 0.27748, + "grad_norm": 0.8136754001048411, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 27748 + }, + { + "epoch": 0.27749, + "grad_norm": 0.911189038260259, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 27749 + }, + { + "epoch": 0.2775, + "grad_norm": 0.9805622566524477, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 27750 + }, + { + "epoch": 0.27751, + "grad_norm": 0.9726121129325052, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 27751 + }, + { + "epoch": 0.27752, + "grad_norm": 0.9724927697223574, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 27752 + }, + { + "epoch": 0.27753, + "grad_norm": 1.0308072191619035, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 27753 + }, + { + "epoch": 0.27754, + "grad_norm": 0.9821653929822171, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 27754 + }, + { + "epoch": 0.27755, + "grad_norm": 0.82602389425107, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 27755 + }, + { + "epoch": 0.27756, + "grad_norm": 0.7502878583260826, + "learning_rate": 0.003, + "loss": 4.006, + "step": 27756 + }, + { + "epoch": 0.27757, + "grad_norm": 0.7284087361535233, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 27757 + }, + { + "epoch": 0.27758, + "grad_norm": 0.7634109953176816, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 27758 + }, + { + "epoch": 0.27759, + "grad_norm": 0.8547628339935236, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 27759 + }, + { + "epoch": 0.2776, + "grad_norm": 0.8754545323443959, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 27760 + }, + { + "epoch": 0.27761, + "grad_norm": 0.8626345079335979, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 27761 + }, + { + "epoch": 0.27762, + "grad_norm": 0.8286966756119551, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 27762 + }, + { + "epoch": 0.27763, + "grad_norm": 0.8064080871554284, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 27763 + }, + { + "epoch": 0.27764, + "grad_norm": 0.8392543549049669, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 27764 + }, + { + "epoch": 0.27765, + "grad_norm": 0.8813769530774089, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 27765 + }, + { + "epoch": 0.27766, + "grad_norm": 0.9709375807932362, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 27766 + }, + { + "epoch": 0.27767, + "grad_norm": 0.8539677147220908, + "learning_rate": 0.003, + "loss": 3.994, + "step": 27767 + }, + { + "epoch": 0.27768, + "grad_norm": 0.8995479530432354, + "learning_rate": 0.003, + "loss": 4.005, + "step": 27768 + }, + { + "epoch": 0.27769, + "grad_norm": 0.8741895862894656, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 27769 + }, + { + "epoch": 0.2777, + "grad_norm": 0.8100762909570717, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 27770 + }, + { + "epoch": 0.27771, + "grad_norm": 0.7170806792358297, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 27771 + }, + { + "epoch": 0.27772, + "grad_norm": 0.7803219952599622, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 27772 + }, + { + "epoch": 0.27773, + "grad_norm": 0.7839135208586959, + "learning_rate": 0.003, + "loss": 4.004, + "step": 27773 + }, + { + "epoch": 0.27774, + "grad_norm": 0.754803138124787, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 27774 + }, + { + "epoch": 0.27775, + "grad_norm": 0.7311843510615395, + "learning_rate": 0.003, + "loss": 4.017, + "step": 27775 + }, + { + "epoch": 0.27776, + "grad_norm": 0.673867184481753, + "learning_rate": 0.003, + "loss": 4.013, + "step": 27776 + }, + { + "epoch": 0.27777, + "grad_norm": 0.6410310112439187, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 27777 + }, + { + "epoch": 0.27778, + "grad_norm": 0.7114958767823502, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 27778 + }, + { + "epoch": 0.27779, + "grad_norm": 0.8346441107447827, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 27779 + }, + { + "epoch": 0.2778, + "grad_norm": 0.8595450077818125, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 27780 + }, + { + "epoch": 0.27781, + "grad_norm": 1.0442972169623612, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 27781 + }, + { + "epoch": 0.27782, + "grad_norm": 1.045702093178626, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 27782 + }, + { + "epoch": 0.27783, + "grad_norm": 0.793105292249228, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 27783 + }, + { + "epoch": 0.27784, + "grad_norm": 0.5958669476710579, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 27784 + }, + { + "epoch": 0.27785, + "grad_norm": 0.6891622554260887, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 27785 + }, + { + "epoch": 0.27786, + "grad_norm": 0.7087659497608495, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 27786 + }, + { + "epoch": 0.27787, + "grad_norm": 0.6792706303336151, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 27787 + }, + { + "epoch": 0.27788, + "grad_norm": 0.5671800279548984, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 27788 + }, + { + "epoch": 0.27789, + "grad_norm": 0.598799567549304, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 27789 + }, + { + "epoch": 0.2779, + "grad_norm": 0.5583261379142739, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 27790 + }, + { + "epoch": 0.27791, + "grad_norm": 0.5783110105421457, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 27791 + }, + { + "epoch": 0.27792, + "grad_norm": 0.5017103227328464, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 27792 + }, + { + "epoch": 0.27793, + "grad_norm": 0.4722454401815684, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 27793 + }, + { + "epoch": 0.27794, + "grad_norm": 0.459928944751263, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 27794 + }, + { + "epoch": 0.27795, + "grad_norm": 0.5620192570273422, + "learning_rate": 0.003, + "loss": 3.9266, + "step": 27795 + }, + { + "epoch": 0.27796, + "grad_norm": 0.6057403344975099, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 27796 + }, + { + "epoch": 0.27797, + "grad_norm": 0.6957474859588345, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 27797 + }, + { + "epoch": 0.27798, + "grad_norm": 0.8656902701069367, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 27798 + }, + { + "epoch": 0.27799, + "grad_norm": 1.0741714875673871, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 27799 + }, + { + "epoch": 0.278, + "grad_norm": 0.8966592130137482, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 27800 + }, + { + "epoch": 0.27801, + "grad_norm": 0.718430547191066, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 27801 + }, + { + "epoch": 0.27802, + "grad_norm": 0.690856795018462, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 27802 + }, + { + "epoch": 0.27803, + "grad_norm": 0.8594376963678187, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 27803 + }, + { + "epoch": 0.27804, + "grad_norm": 0.9660369235631169, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 27804 + }, + { + "epoch": 0.27805, + "grad_norm": 1.0349021173191688, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 27805 + }, + { + "epoch": 0.27806, + "grad_norm": 0.9795687779102337, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 27806 + }, + { + "epoch": 0.27807, + "grad_norm": 0.9593712475418819, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 27807 + }, + { + "epoch": 0.27808, + "grad_norm": 0.92598557065114, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 27808 + }, + { + "epoch": 0.27809, + "grad_norm": 0.8375318020113784, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 27809 + }, + { + "epoch": 0.2781, + "grad_norm": 0.7675100623253974, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 27810 + }, + { + "epoch": 0.27811, + "grad_norm": 0.7779115420101179, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 27811 + }, + { + "epoch": 0.27812, + "grad_norm": 0.7658413592440414, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 27812 + }, + { + "epoch": 0.27813, + "grad_norm": 0.7380702092578432, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 27813 + }, + { + "epoch": 0.27814, + "grad_norm": 0.7603000327976729, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 27814 + }, + { + "epoch": 0.27815, + "grad_norm": 0.731349398202973, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 27815 + }, + { + "epoch": 0.27816, + "grad_norm": 0.647644512714406, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 27816 + }, + { + "epoch": 0.27817, + "grad_norm": 0.6512049613229464, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 27817 + }, + { + "epoch": 0.27818, + "grad_norm": 0.6294056856867439, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 27818 + }, + { + "epoch": 0.27819, + "grad_norm": 0.7055670749236224, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 27819 + }, + { + "epoch": 0.2782, + "grad_norm": 0.7893577521081836, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 27820 + }, + { + "epoch": 0.27821, + "grad_norm": 0.9915427717459308, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 27821 + }, + { + "epoch": 0.27822, + "grad_norm": 1.1189166751080297, + "learning_rate": 0.003, + "loss": 4.011, + "step": 27822 + }, + { + "epoch": 0.27823, + "grad_norm": 0.9286175408508832, + "learning_rate": 0.003, + "loss": 4.018, + "step": 27823 + }, + { + "epoch": 0.27824, + "grad_norm": 0.8227642758222663, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 27824 + }, + { + "epoch": 0.27825, + "grad_norm": 0.610998348710799, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 27825 + }, + { + "epoch": 0.27826, + "grad_norm": 0.582614159552692, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 27826 + }, + { + "epoch": 0.27827, + "grad_norm": 0.5907433989549609, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 27827 + }, + { + "epoch": 0.27828, + "grad_norm": 0.6133149473602598, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 27828 + }, + { + "epoch": 0.27829, + "grad_norm": 0.6179626319741037, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 27829 + }, + { + "epoch": 0.2783, + "grad_norm": 0.6840124173423933, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 27830 + }, + { + "epoch": 0.27831, + "grad_norm": 0.7391900766913824, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 27831 + }, + { + "epoch": 0.27832, + "grad_norm": 0.9169965122295695, + "learning_rate": 0.003, + "loss": 4.024, + "step": 27832 + }, + { + "epoch": 0.27833, + "grad_norm": 1.0012881459595293, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 27833 + }, + { + "epoch": 0.27834, + "grad_norm": 0.9882164888348923, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 27834 + }, + { + "epoch": 0.27835, + "grad_norm": 1.0369604368214156, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 27835 + }, + { + "epoch": 0.27836, + "grad_norm": 0.7935970460223312, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 27836 + }, + { + "epoch": 0.27837, + "grad_norm": 0.7176325726289007, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 27837 + }, + { + "epoch": 0.27838, + "grad_norm": 0.625826574027999, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 27838 + }, + { + "epoch": 0.27839, + "grad_norm": 0.623419987213707, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 27839 + }, + { + "epoch": 0.2784, + "grad_norm": 0.6376672108946563, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 27840 + }, + { + "epoch": 0.27841, + "grad_norm": 0.6357806912872545, + "learning_rate": 0.003, + "loss": 4.007, + "step": 27841 + }, + { + "epoch": 0.27842, + "grad_norm": 0.6132474960610723, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 27842 + }, + { + "epoch": 0.27843, + "grad_norm": 0.5964185796147489, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 27843 + }, + { + "epoch": 0.27844, + "grad_norm": 0.6139408342479437, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 27844 + }, + { + "epoch": 0.27845, + "grad_norm": 0.7388786055740236, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 27845 + }, + { + "epoch": 0.27846, + "grad_norm": 0.9133698737682968, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 27846 + }, + { + "epoch": 0.27847, + "grad_norm": 1.096275689283809, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 27847 + }, + { + "epoch": 0.27848, + "grad_norm": 0.8726521350337709, + "learning_rate": 0.003, + "loss": 3.999, + "step": 27848 + }, + { + "epoch": 0.27849, + "grad_norm": 0.7188665558775329, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 27849 + }, + { + "epoch": 0.2785, + "grad_norm": 0.689568813083638, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 27850 + }, + { + "epoch": 0.27851, + "grad_norm": 0.6150186725003074, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 27851 + }, + { + "epoch": 0.27852, + "grad_norm": 0.6247957503596243, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 27852 + }, + { + "epoch": 0.27853, + "grad_norm": 0.5597852320581574, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 27853 + }, + { + "epoch": 0.27854, + "grad_norm": 0.5729357727739846, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 27854 + }, + { + "epoch": 0.27855, + "grad_norm": 0.6169033773415953, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 27855 + }, + { + "epoch": 0.27856, + "grad_norm": 0.7563565560601458, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 27856 + }, + { + "epoch": 0.27857, + "grad_norm": 0.7480720802353464, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 27857 + }, + { + "epoch": 0.27858, + "grad_norm": 0.752154171416138, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 27858 + }, + { + "epoch": 0.27859, + "grad_norm": 0.7677356841340681, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 27859 + }, + { + "epoch": 0.2786, + "grad_norm": 0.8553645781930288, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 27860 + }, + { + "epoch": 0.27861, + "grad_norm": 1.038584202497652, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 27861 + }, + { + "epoch": 0.27862, + "grad_norm": 0.9643368968102282, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 27862 + }, + { + "epoch": 0.27863, + "grad_norm": 1.010962753078654, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 27863 + }, + { + "epoch": 0.27864, + "grad_norm": 1.02046949804226, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 27864 + }, + { + "epoch": 0.27865, + "grad_norm": 1.017371920336619, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 27865 + }, + { + "epoch": 0.27866, + "grad_norm": 1.0707165066685949, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 27866 + }, + { + "epoch": 0.27867, + "grad_norm": 0.9475619492509266, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 27867 + }, + { + "epoch": 0.27868, + "grad_norm": 0.8922701867828415, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 27868 + }, + { + "epoch": 0.27869, + "grad_norm": 0.9224538024608281, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 27869 + }, + { + "epoch": 0.2787, + "grad_norm": 0.9862034877265979, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 27870 + }, + { + "epoch": 0.27871, + "grad_norm": 1.0357693288399177, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 27871 + }, + { + "epoch": 0.27872, + "grad_norm": 1.0291000567944355, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 27872 + }, + { + "epoch": 0.27873, + "grad_norm": 0.8424588949067068, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 27873 + }, + { + "epoch": 0.27874, + "grad_norm": 0.8055168481520061, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 27874 + }, + { + "epoch": 0.27875, + "grad_norm": 0.7960920775642376, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 27875 + }, + { + "epoch": 0.27876, + "grad_norm": 0.7747473242081399, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 27876 + }, + { + "epoch": 0.27877, + "grad_norm": 0.6263439693168625, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 27877 + }, + { + "epoch": 0.27878, + "grad_norm": 0.6888917813833104, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 27878 + }, + { + "epoch": 0.27879, + "grad_norm": 0.7150660160384423, + "learning_rate": 0.003, + "loss": 3.97, + "step": 27879 + }, + { + "epoch": 0.2788, + "grad_norm": 0.7285440691496937, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 27880 + }, + { + "epoch": 0.27881, + "grad_norm": 0.7845753195296852, + "learning_rate": 0.003, + "loss": 3.982, + "step": 27881 + }, + { + "epoch": 0.27882, + "grad_norm": 0.9062329096405427, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 27882 + }, + { + "epoch": 0.27883, + "grad_norm": 0.9189291901020434, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 27883 + }, + { + "epoch": 0.27884, + "grad_norm": 0.804108484299568, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 27884 + }, + { + "epoch": 0.27885, + "grad_norm": 0.6035045094594205, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 27885 + }, + { + "epoch": 0.27886, + "grad_norm": 0.6120488874316106, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 27886 + }, + { + "epoch": 0.27887, + "grad_norm": 0.5366901496765571, + "learning_rate": 0.003, + "loss": 4.001, + "step": 27887 + }, + { + "epoch": 0.27888, + "grad_norm": 0.5089720578664682, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 27888 + }, + { + "epoch": 0.27889, + "grad_norm": 0.534981683588964, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 27889 + }, + { + "epoch": 0.2789, + "grad_norm": 0.6039399192499156, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 27890 + }, + { + "epoch": 0.27891, + "grad_norm": 0.7631013294691419, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 27891 + }, + { + "epoch": 0.27892, + "grad_norm": 0.8279399620173811, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 27892 + }, + { + "epoch": 0.27893, + "grad_norm": 0.814039574737581, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 27893 + }, + { + "epoch": 0.27894, + "grad_norm": 0.7476094668326242, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 27894 + }, + { + "epoch": 0.27895, + "grad_norm": 0.6628166411642974, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 27895 + }, + { + "epoch": 0.27896, + "grad_norm": 0.4989505305749452, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 27896 + }, + { + "epoch": 0.27897, + "grad_norm": 0.4950536176321295, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 27897 + }, + { + "epoch": 0.27898, + "grad_norm": 0.47269068455972013, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 27898 + }, + { + "epoch": 0.27899, + "grad_norm": 0.3866946844939026, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 27899 + }, + { + "epoch": 0.279, + "grad_norm": 0.4540208718025058, + "learning_rate": 0.003, + "loss": 3.981, + "step": 27900 + }, + { + "epoch": 0.27901, + "grad_norm": 0.4829204030091787, + "learning_rate": 0.003, + "loss": 3.9227, + "step": 27901 + }, + { + "epoch": 0.27902, + "grad_norm": 0.48434516279320067, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 27902 + }, + { + "epoch": 0.27903, + "grad_norm": 0.5779255525285589, + "learning_rate": 0.003, + "loss": 3.98, + "step": 27903 + }, + { + "epoch": 0.27904, + "grad_norm": 0.7341193859250307, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 27904 + }, + { + "epoch": 0.27905, + "grad_norm": 0.787413252756451, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 27905 + }, + { + "epoch": 0.27906, + "grad_norm": 0.7658440851449531, + "learning_rate": 0.003, + "loss": 4.003, + "step": 27906 + }, + { + "epoch": 0.27907, + "grad_norm": 0.702876296494054, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 27907 + }, + { + "epoch": 0.27908, + "grad_norm": 0.6317563725401272, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 27908 + }, + { + "epoch": 0.27909, + "grad_norm": 0.6496310406197852, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 27909 + }, + { + "epoch": 0.2791, + "grad_norm": 0.6732521877354694, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 27910 + }, + { + "epoch": 0.27911, + "grad_norm": 0.7717392321896972, + "learning_rate": 0.003, + "loss": 4.0, + "step": 27911 + }, + { + "epoch": 0.27912, + "grad_norm": 0.9497988904397566, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 27912 + }, + { + "epoch": 0.27913, + "grad_norm": 1.1802882425294996, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 27913 + }, + { + "epoch": 0.27914, + "grad_norm": 0.7871556183802577, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 27914 + }, + { + "epoch": 0.27915, + "grad_norm": 0.6943149279898635, + "learning_rate": 0.003, + "loss": 3.986, + "step": 27915 + }, + { + "epoch": 0.27916, + "grad_norm": 0.7057195044901342, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 27916 + }, + { + "epoch": 0.27917, + "grad_norm": 0.7669179099916582, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 27917 + }, + { + "epoch": 0.27918, + "grad_norm": 0.7358310672331628, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 27918 + }, + { + "epoch": 0.27919, + "grad_norm": 0.7411426162277395, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 27919 + }, + { + "epoch": 0.2792, + "grad_norm": 0.7409059953281965, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 27920 + }, + { + "epoch": 0.27921, + "grad_norm": 0.7350926439739405, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 27921 + }, + { + "epoch": 0.27922, + "grad_norm": 0.711893942693435, + "learning_rate": 0.003, + "loss": 4.0368, + "step": 27922 + }, + { + "epoch": 0.27923, + "grad_norm": 0.7095642069379543, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 27923 + }, + { + "epoch": 0.27924, + "grad_norm": 0.8080379210287206, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 27924 + }, + { + "epoch": 0.27925, + "grad_norm": 0.7304281570399103, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 27925 + }, + { + "epoch": 0.27926, + "grad_norm": 0.7025633128388405, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 27926 + }, + { + "epoch": 0.27927, + "grad_norm": 0.7431016131573671, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 27927 + }, + { + "epoch": 0.27928, + "grad_norm": 0.7921709347510457, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 27928 + }, + { + "epoch": 0.27929, + "grad_norm": 0.8058565318127605, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 27929 + }, + { + "epoch": 0.2793, + "grad_norm": 0.7505189963526987, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 27930 + }, + { + "epoch": 0.27931, + "grad_norm": 0.6169169547708828, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 27931 + }, + { + "epoch": 0.27932, + "grad_norm": 0.6117368634719405, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 27932 + }, + { + "epoch": 0.27933, + "grad_norm": 0.6866907935936758, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 27933 + }, + { + "epoch": 0.27934, + "grad_norm": 0.7630707942669168, + "learning_rate": 0.003, + "loss": 3.978, + "step": 27934 + }, + { + "epoch": 0.27935, + "grad_norm": 0.8269097541942597, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 27935 + }, + { + "epoch": 0.27936, + "grad_norm": 0.9568491022137376, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 27936 + }, + { + "epoch": 0.27937, + "grad_norm": 0.9763457045620367, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 27937 + }, + { + "epoch": 0.27938, + "grad_norm": 0.989224266749453, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 27938 + }, + { + "epoch": 0.27939, + "grad_norm": 1.0712861955927675, + "learning_rate": 0.003, + "loss": 4.0349, + "step": 27939 + }, + { + "epoch": 0.2794, + "grad_norm": 1.1163956085504485, + "learning_rate": 0.003, + "loss": 4.0389, + "step": 27940 + }, + { + "epoch": 0.27941, + "grad_norm": 0.7719528777992227, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 27941 + }, + { + "epoch": 0.27942, + "grad_norm": 0.6969794390171589, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 27942 + }, + { + "epoch": 0.27943, + "grad_norm": 0.8047788509508472, + "learning_rate": 0.003, + "loss": 3.996, + "step": 27943 + }, + { + "epoch": 0.27944, + "grad_norm": 0.9476268671886411, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 27944 + }, + { + "epoch": 0.27945, + "grad_norm": 0.9677068864147189, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 27945 + }, + { + "epoch": 0.27946, + "grad_norm": 1.1229074139206032, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 27946 + }, + { + "epoch": 0.27947, + "grad_norm": 0.9909394148781698, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 27947 + }, + { + "epoch": 0.27948, + "grad_norm": 1.038020297402364, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 27948 + }, + { + "epoch": 0.27949, + "grad_norm": 0.9618864792791566, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 27949 + }, + { + "epoch": 0.2795, + "grad_norm": 0.9869348451564467, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 27950 + }, + { + "epoch": 0.27951, + "grad_norm": 0.8213996080557807, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 27951 + }, + { + "epoch": 0.27952, + "grad_norm": 0.7343300616218813, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 27952 + }, + { + "epoch": 0.27953, + "grad_norm": 0.6440821770239036, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 27953 + }, + { + "epoch": 0.27954, + "grad_norm": 0.6731188492209877, + "learning_rate": 0.003, + "loss": 4.001, + "step": 27954 + }, + { + "epoch": 0.27955, + "grad_norm": 0.6568988216701879, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 27955 + }, + { + "epoch": 0.27956, + "grad_norm": 0.6005872545963578, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 27956 + }, + { + "epoch": 0.27957, + "grad_norm": 0.707754955191795, + "learning_rate": 0.003, + "loss": 4.013, + "step": 27957 + }, + { + "epoch": 0.27958, + "grad_norm": 0.7766756688267565, + "learning_rate": 0.003, + "loss": 3.998, + "step": 27958 + }, + { + "epoch": 0.27959, + "grad_norm": 0.7872905185634355, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 27959 + }, + { + "epoch": 0.2796, + "grad_norm": 0.7237477719310835, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 27960 + }, + { + "epoch": 0.27961, + "grad_norm": 0.7072102045799764, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 27961 + }, + { + "epoch": 0.27962, + "grad_norm": 0.7537474488246975, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 27962 + }, + { + "epoch": 0.27963, + "grad_norm": 0.6758453711727852, + "learning_rate": 0.003, + "loss": 3.984, + "step": 27963 + }, + { + "epoch": 0.27964, + "grad_norm": 0.6333236532249056, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 27964 + }, + { + "epoch": 0.27965, + "grad_norm": 0.6411657318353718, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 27965 + }, + { + "epoch": 0.27966, + "grad_norm": 0.6595754072719496, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 27966 + }, + { + "epoch": 0.27967, + "grad_norm": 0.6100783090610761, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 27967 + }, + { + "epoch": 0.27968, + "grad_norm": 0.6019476617672436, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 27968 + }, + { + "epoch": 0.27969, + "grad_norm": 0.5787272211878659, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 27969 + }, + { + "epoch": 0.2797, + "grad_norm": 0.5556069846909792, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 27970 + }, + { + "epoch": 0.27971, + "grad_norm": 0.5627177269725476, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 27971 + }, + { + "epoch": 0.27972, + "grad_norm": 0.6047232732339411, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 27972 + }, + { + "epoch": 0.27973, + "grad_norm": 0.6252014827085357, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 27973 + }, + { + "epoch": 0.27974, + "grad_norm": 0.6762278479760412, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 27974 + }, + { + "epoch": 0.27975, + "grad_norm": 0.7333154526012695, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 27975 + }, + { + "epoch": 0.27976, + "grad_norm": 0.7890851864009999, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 27976 + }, + { + "epoch": 0.27977, + "grad_norm": 0.7762555805324168, + "learning_rate": 0.003, + "loss": 3.997, + "step": 27977 + }, + { + "epoch": 0.27978, + "grad_norm": 0.7201240372894158, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 27978 + }, + { + "epoch": 0.27979, + "grad_norm": 0.8101432631566168, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 27979 + }, + { + "epoch": 0.2798, + "grad_norm": 0.8144242195537768, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 27980 + }, + { + "epoch": 0.27981, + "grad_norm": 0.8495999534868507, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 27981 + }, + { + "epoch": 0.27982, + "grad_norm": 1.0579627971862187, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 27982 + }, + { + "epoch": 0.27983, + "grad_norm": 1.0706089572008683, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 27983 + }, + { + "epoch": 0.27984, + "grad_norm": 0.8952313872242107, + "learning_rate": 0.003, + "loss": 3.98, + "step": 27984 + }, + { + "epoch": 0.27985, + "grad_norm": 0.9163132686387959, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 27985 + }, + { + "epoch": 0.27986, + "grad_norm": 0.8282930757574292, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 27986 + }, + { + "epoch": 0.27987, + "grad_norm": 0.8163365616319417, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 27987 + }, + { + "epoch": 0.27988, + "grad_norm": 0.8129804381085529, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 27988 + }, + { + "epoch": 0.27989, + "grad_norm": 0.8611229347308631, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 27989 + }, + { + "epoch": 0.2799, + "grad_norm": 0.8060516501867137, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 27990 + }, + { + "epoch": 0.27991, + "grad_norm": 0.7077223529380927, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 27991 + }, + { + "epoch": 0.27992, + "grad_norm": 0.7924007226769582, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 27992 + }, + { + "epoch": 0.27993, + "grad_norm": 0.7683755491549037, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 27993 + }, + { + "epoch": 0.27994, + "grad_norm": 0.6847419999446193, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 27994 + }, + { + "epoch": 0.27995, + "grad_norm": 0.843338472827538, + "learning_rate": 0.003, + "loss": 4.0549, + "step": 27995 + }, + { + "epoch": 0.27996, + "grad_norm": 0.8300407353522945, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 27996 + }, + { + "epoch": 0.27997, + "grad_norm": 0.7568868562236738, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 27997 + }, + { + "epoch": 0.27998, + "grad_norm": 0.7172518708469536, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 27998 + }, + { + "epoch": 0.27999, + "grad_norm": 0.6441483605133784, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 27999 + }, + { + "epoch": 0.28, + "grad_norm": 0.6199767519908963, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 28000 + }, + { + "epoch": 0.28001, + "grad_norm": 0.5197956576976696, + "learning_rate": 0.003, + "loss": 3.965, + "step": 28001 + }, + { + "epoch": 0.28002, + "grad_norm": 0.4979183748505758, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 28002 + }, + { + "epoch": 0.28003, + "grad_norm": 0.5502708020645166, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 28003 + }, + { + "epoch": 0.28004, + "grad_norm": 0.5957195609154033, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28004 + }, + { + "epoch": 0.28005, + "grad_norm": 0.5611797096815452, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 28005 + }, + { + "epoch": 0.28006, + "grad_norm": 0.6018085610135447, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 28006 + }, + { + "epoch": 0.28007, + "grad_norm": 0.7612643612460983, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 28007 + }, + { + "epoch": 0.28008, + "grad_norm": 0.8868189596027714, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 28008 + }, + { + "epoch": 0.28009, + "grad_norm": 0.8885054379008013, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 28009 + }, + { + "epoch": 0.2801, + "grad_norm": 0.8371500605171657, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 28010 + }, + { + "epoch": 0.28011, + "grad_norm": 0.9140023041010049, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 28011 + }, + { + "epoch": 0.28012, + "grad_norm": 0.9821517352513933, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 28012 + }, + { + "epoch": 0.28013, + "grad_norm": 1.103429503282482, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 28013 + }, + { + "epoch": 0.28014, + "grad_norm": 0.7835679258211823, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28014 + }, + { + "epoch": 0.28015, + "grad_norm": 0.8435636870579056, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 28015 + }, + { + "epoch": 0.28016, + "grad_norm": 0.8959590906870629, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 28016 + }, + { + "epoch": 0.28017, + "grad_norm": 1.058701831347882, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 28017 + }, + { + "epoch": 0.28018, + "grad_norm": 1.1288622013810152, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 28018 + }, + { + "epoch": 0.28019, + "grad_norm": 1.012589981327254, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 28019 + }, + { + "epoch": 0.2802, + "grad_norm": 0.9652250685665759, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 28020 + }, + { + "epoch": 0.28021, + "grad_norm": 1.0155823499893226, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 28021 + }, + { + "epoch": 0.28022, + "grad_norm": 0.8652428019009831, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 28022 + }, + { + "epoch": 0.28023, + "grad_norm": 0.971842271585843, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28023 + }, + { + "epoch": 0.28024, + "grad_norm": 1.1255843412828723, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 28024 + }, + { + "epoch": 0.28025, + "grad_norm": 1.0198293027425493, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 28025 + }, + { + "epoch": 0.28026, + "grad_norm": 1.1222720212050699, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 28026 + }, + { + "epoch": 0.28027, + "grad_norm": 1.3387866975038227, + "learning_rate": 0.003, + "loss": 4.0348, + "step": 28027 + }, + { + "epoch": 0.28028, + "grad_norm": 2.2666158475032865, + "learning_rate": 0.003, + "loss": 4.0818, + "step": 28028 + }, + { + "epoch": 0.28029, + "grad_norm": 1.0838481823076804, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 28029 + }, + { + "epoch": 0.2803, + "grad_norm": 1.0386196515530106, + "learning_rate": 0.003, + "loss": 4.019, + "step": 28030 + }, + { + "epoch": 0.28031, + "grad_norm": 1.3056822594366702, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 28031 + }, + { + "epoch": 0.28032, + "grad_norm": 1.1700799198629679, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 28032 + }, + { + "epoch": 0.28033, + "grad_norm": 1.1225220221527388, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 28033 + }, + { + "epoch": 0.28034, + "grad_norm": 0.9890629373530953, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 28034 + }, + { + "epoch": 0.28035, + "grad_norm": 1.1240331213097943, + "learning_rate": 0.003, + "loss": 4.034, + "step": 28035 + }, + { + "epoch": 0.28036, + "grad_norm": 0.9802363666274811, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 28036 + }, + { + "epoch": 0.28037, + "grad_norm": 1.1959205844115197, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 28037 + }, + { + "epoch": 0.28038, + "grad_norm": 1.5794889201572253, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 28038 + }, + { + "epoch": 0.28039, + "grad_norm": 1.1188005990410421, + "learning_rate": 0.003, + "loss": 4.0511, + "step": 28039 + }, + { + "epoch": 0.2804, + "grad_norm": 1.0960748773189777, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 28040 + }, + { + "epoch": 0.28041, + "grad_norm": 0.9495732825892943, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 28041 + }, + { + "epoch": 0.28042, + "grad_norm": 0.8232896287643701, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 28042 + }, + { + "epoch": 0.28043, + "grad_norm": 0.7360603885401349, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 28043 + }, + { + "epoch": 0.28044, + "grad_norm": 0.7839389894251249, + "learning_rate": 0.003, + "loss": 4.0314, + "step": 28044 + }, + { + "epoch": 0.28045, + "grad_norm": 0.78584570981466, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 28045 + }, + { + "epoch": 0.28046, + "grad_norm": 0.756495749316015, + "learning_rate": 0.003, + "loss": 4.0443, + "step": 28046 + }, + { + "epoch": 0.28047, + "grad_norm": 0.717814942687467, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 28047 + }, + { + "epoch": 0.28048, + "grad_norm": 0.6299534397801231, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 28048 + }, + { + "epoch": 0.28049, + "grad_norm": 0.611030816371035, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 28049 + }, + { + "epoch": 0.2805, + "grad_norm": 0.5379310243818175, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 28050 + }, + { + "epoch": 0.28051, + "grad_norm": 0.5173805825882627, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 28051 + }, + { + "epoch": 0.28052, + "grad_norm": 0.5079742187391297, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 28052 + }, + { + "epoch": 0.28053, + "grad_norm": 0.512734214213078, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 28053 + }, + { + "epoch": 0.28054, + "grad_norm": 0.5190998253943722, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 28054 + }, + { + "epoch": 0.28055, + "grad_norm": 0.5624157208324606, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 28055 + }, + { + "epoch": 0.28056, + "grad_norm": 0.5547649544540759, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 28056 + }, + { + "epoch": 0.28057, + "grad_norm": 0.5369845174457818, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 28057 + }, + { + "epoch": 0.28058, + "grad_norm": 0.6006002188324524, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 28058 + }, + { + "epoch": 0.28059, + "grad_norm": 0.7413491918984849, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 28059 + }, + { + "epoch": 0.2806, + "grad_norm": 0.7681000564824877, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28060 + }, + { + "epoch": 0.28061, + "grad_norm": 0.7634082593616937, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 28061 + }, + { + "epoch": 0.28062, + "grad_norm": 0.8807906084992299, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 28062 + }, + { + "epoch": 0.28063, + "grad_norm": 1.066261376977965, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 28063 + }, + { + "epoch": 0.28064, + "grad_norm": 0.8189800081189879, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 28064 + }, + { + "epoch": 0.28065, + "grad_norm": 0.6622596757164085, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 28065 + }, + { + "epoch": 0.28066, + "grad_norm": 0.5717820837347959, + "learning_rate": 0.003, + "loss": 3.997, + "step": 28066 + }, + { + "epoch": 0.28067, + "grad_norm": 0.6582870185455014, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 28067 + }, + { + "epoch": 0.28068, + "grad_norm": 0.6865179408904513, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 28068 + }, + { + "epoch": 0.28069, + "grad_norm": 0.7796236707756719, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 28069 + }, + { + "epoch": 0.2807, + "grad_norm": 0.9542330036708985, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 28070 + }, + { + "epoch": 0.28071, + "grad_norm": 1.2199702367525254, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 28071 + }, + { + "epoch": 0.28072, + "grad_norm": 0.8321064508691784, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 28072 + }, + { + "epoch": 0.28073, + "grad_norm": 0.8441326826825378, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 28073 + }, + { + "epoch": 0.28074, + "grad_norm": 0.6913150708312623, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 28074 + }, + { + "epoch": 0.28075, + "grad_norm": 0.7440213582196883, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 28075 + }, + { + "epoch": 0.28076, + "grad_norm": 0.7743734677847184, + "learning_rate": 0.003, + "loss": 3.972, + "step": 28076 + }, + { + "epoch": 0.28077, + "grad_norm": 0.756347906365762, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 28077 + }, + { + "epoch": 0.28078, + "grad_norm": 0.8086535571818279, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 28078 + }, + { + "epoch": 0.28079, + "grad_norm": 0.7793585315069942, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 28079 + }, + { + "epoch": 0.2808, + "grad_norm": 0.8975381748734678, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 28080 + }, + { + "epoch": 0.28081, + "grad_norm": 0.9166205243301296, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28081 + }, + { + "epoch": 0.28082, + "grad_norm": 0.7606546729349047, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 28082 + }, + { + "epoch": 0.28083, + "grad_norm": 0.6454114925638089, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 28083 + }, + { + "epoch": 0.28084, + "grad_norm": 0.6830938860392994, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 28084 + }, + { + "epoch": 0.28085, + "grad_norm": 0.7681508159953039, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 28085 + }, + { + "epoch": 0.28086, + "grad_norm": 0.9706072036005362, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 28086 + }, + { + "epoch": 0.28087, + "grad_norm": 1.0312943452613206, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 28087 + }, + { + "epoch": 0.28088, + "grad_norm": 0.8164135506741607, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 28088 + }, + { + "epoch": 0.28089, + "grad_norm": 0.8135692646190997, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 28089 + }, + { + "epoch": 0.2809, + "grad_norm": 0.7792388500818118, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 28090 + }, + { + "epoch": 0.28091, + "grad_norm": 0.8073910914689659, + "learning_rate": 0.003, + "loss": 4.001, + "step": 28091 + }, + { + "epoch": 0.28092, + "grad_norm": 0.9291643053014461, + "learning_rate": 0.003, + "loss": 4.0438, + "step": 28092 + }, + { + "epoch": 0.28093, + "grad_norm": 0.8189297573446741, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 28093 + }, + { + "epoch": 0.28094, + "grad_norm": 0.8697320598962878, + "learning_rate": 0.003, + "loss": 3.999, + "step": 28094 + }, + { + "epoch": 0.28095, + "grad_norm": 1.1080735543222093, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 28095 + }, + { + "epoch": 0.28096, + "grad_norm": 0.9193665123746955, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 28096 + }, + { + "epoch": 0.28097, + "grad_norm": 0.7949214641914621, + "learning_rate": 0.003, + "loss": 4.001, + "step": 28097 + }, + { + "epoch": 0.28098, + "grad_norm": 0.7890013965000933, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 28098 + }, + { + "epoch": 0.28099, + "grad_norm": 0.7474282621063514, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 28099 + }, + { + "epoch": 0.281, + "grad_norm": 0.7371240567445875, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 28100 + }, + { + "epoch": 0.28101, + "grad_norm": 0.8403644898057471, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 28101 + }, + { + "epoch": 0.28102, + "grad_norm": 0.925539440528669, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 28102 + }, + { + "epoch": 0.28103, + "grad_norm": 0.8826989140055331, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 28103 + }, + { + "epoch": 0.28104, + "grad_norm": 0.7605253667932622, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 28104 + }, + { + "epoch": 0.28105, + "grad_norm": 0.6461294902033333, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 28105 + }, + { + "epoch": 0.28106, + "grad_norm": 0.5769891731765743, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 28106 + }, + { + "epoch": 0.28107, + "grad_norm": 0.4958782392039952, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 28107 + }, + { + "epoch": 0.28108, + "grad_norm": 0.592643929165787, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 28108 + }, + { + "epoch": 0.28109, + "grad_norm": 0.6183998626832836, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 28109 + }, + { + "epoch": 0.2811, + "grad_norm": 0.6408239068233306, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 28110 + }, + { + "epoch": 0.28111, + "grad_norm": 0.6734846353830174, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 28111 + }, + { + "epoch": 0.28112, + "grad_norm": 0.6209237947807472, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 28112 + }, + { + "epoch": 0.28113, + "grad_norm": 0.614510105658126, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 28113 + }, + { + "epoch": 0.28114, + "grad_norm": 0.5615882835317904, + "learning_rate": 0.003, + "loss": 3.972, + "step": 28114 + }, + { + "epoch": 0.28115, + "grad_norm": 0.5800040516294106, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 28115 + }, + { + "epoch": 0.28116, + "grad_norm": 0.633342281575454, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 28116 + }, + { + "epoch": 0.28117, + "grad_norm": 0.647274857690177, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 28117 + }, + { + "epoch": 0.28118, + "grad_norm": 0.6947186159909409, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 28118 + }, + { + "epoch": 0.28119, + "grad_norm": 0.6726646115740998, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 28119 + }, + { + "epoch": 0.2812, + "grad_norm": 0.5364292599230475, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 28120 + }, + { + "epoch": 0.28121, + "grad_norm": 0.6387531866799078, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 28121 + }, + { + "epoch": 0.28122, + "grad_norm": 0.7560636272809754, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 28122 + }, + { + "epoch": 0.28123, + "grad_norm": 0.8008255447688624, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 28123 + }, + { + "epoch": 0.28124, + "grad_norm": 0.7517257139755842, + "learning_rate": 0.003, + "loss": 4.0, + "step": 28124 + }, + { + "epoch": 0.28125, + "grad_norm": 0.8380149033749134, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 28125 + }, + { + "epoch": 0.28126, + "grad_norm": 0.7493285366909612, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 28126 + }, + { + "epoch": 0.28127, + "grad_norm": 0.6813595377927877, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 28127 + }, + { + "epoch": 0.28128, + "grad_norm": 0.6620020199389005, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 28128 + }, + { + "epoch": 0.28129, + "grad_norm": 0.6667287475411924, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 28129 + }, + { + "epoch": 0.2813, + "grad_norm": 0.645103381326521, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 28130 + }, + { + "epoch": 0.28131, + "grad_norm": 0.6514736998748679, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 28131 + }, + { + "epoch": 0.28132, + "grad_norm": 0.6116170970502841, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 28132 + }, + { + "epoch": 0.28133, + "grad_norm": 0.6259378155576876, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 28133 + }, + { + "epoch": 0.28134, + "grad_norm": 0.68380138539075, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 28134 + }, + { + "epoch": 0.28135, + "grad_norm": 0.9102971154626672, + "learning_rate": 0.003, + "loss": 4.009, + "step": 28135 + }, + { + "epoch": 0.28136, + "grad_norm": 1.277354549473003, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 28136 + }, + { + "epoch": 0.28137, + "grad_norm": 1.007466702430169, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 28137 + }, + { + "epoch": 0.28138, + "grad_norm": 1.0163617799479683, + "learning_rate": 0.003, + "loss": 4.0, + "step": 28138 + }, + { + "epoch": 0.28139, + "grad_norm": 0.9025980307021135, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 28139 + }, + { + "epoch": 0.2814, + "grad_norm": 0.8680710359775037, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 28140 + }, + { + "epoch": 0.28141, + "grad_norm": 0.7041157187442274, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 28141 + }, + { + "epoch": 0.28142, + "grad_norm": 0.7068029291073717, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 28142 + }, + { + "epoch": 0.28143, + "grad_norm": 0.6983602829340811, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 28143 + }, + { + "epoch": 0.28144, + "grad_norm": 0.7024636125780322, + "learning_rate": 0.003, + "loss": 4.0523, + "step": 28144 + }, + { + "epoch": 0.28145, + "grad_norm": 0.9021795689702531, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 28145 + }, + { + "epoch": 0.28146, + "grad_norm": 0.8267895831290366, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 28146 + }, + { + "epoch": 0.28147, + "grad_norm": 0.7480636233411961, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 28147 + }, + { + "epoch": 0.28148, + "grad_norm": 0.694462258851557, + "learning_rate": 0.003, + "loss": 3.974, + "step": 28148 + }, + { + "epoch": 0.28149, + "grad_norm": 0.7527071423281987, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 28149 + }, + { + "epoch": 0.2815, + "grad_norm": 0.7735770284281949, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 28150 + }, + { + "epoch": 0.28151, + "grad_norm": 0.7861341450918164, + "learning_rate": 0.003, + "loss": 4.014, + "step": 28151 + }, + { + "epoch": 0.28152, + "grad_norm": 0.9383477059759189, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 28152 + }, + { + "epoch": 0.28153, + "grad_norm": 1.0572294360362224, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 28153 + }, + { + "epoch": 0.28154, + "grad_norm": 0.9379006156802691, + "learning_rate": 0.003, + "loss": 4.013, + "step": 28154 + }, + { + "epoch": 0.28155, + "grad_norm": 0.8431807017845062, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 28155 + }, + { + "epoch": 0.28156, + "grad_norm": 0.7899999556012107, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 28156 + }, + { + "epoch": 0.28157, + "grad_norm": 0.6544222436141703, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 28157 + }, + { + "epoch": 0.28158, + "grad_norm": 0.6969356947127896, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 28158 + }, + { + "epoch": 0.28159, + "grad_norm": 0.6379520678525118, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 28159 + }, + { + "epoch": 0.2816, + "grad_norm": 0.6627853571886558, + "learning_rate": 0.003, + "loss": 4.021, + "step": 28160 + }, + { + "epoch": 0.28161, + "grad_norm": 0.5600584309384199, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 28161 + }, + { + "epoch": 0.28162, + "grad_norm": 0.6163593942795728, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 28162 + }, + { + "epoch": 0.28163, + "grad_norm": 0.7063225570986313, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 28163 + }, + { + "epoch": 0.28164, + "grad_norm": 0.8912941645063169, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 28164 + }, + { + "epoch": 0.28165, + "grad_norm": 0.9639334966007046, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 28165 + }, + { + "epoch": 0.28166, + "grad_norm": 1.0714656558844617, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 28166 + }, + { + "epoch": 0.28167, + "grad_norm": 0.9916315588007596, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 28167 + }, + { + "epoch": 0.28168, + "grad_norm": 0.8774004269337856, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 28168 + }, + { + "epoch": 0.28169, + "grad_norm": 0.9057875109901088, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 28169 + }, + { + "epoch": 0.2817, + "grad_norm": 0.870856545058607, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 28170 + }, + { + "epoch": 0.28171, + "grad_norm": 0.7827388198460151, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 28171 + }, + { + "epoch": 0.28172, + "grad_norm": 0.7311635337806948, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 28172 + }, + { + "epoch": 0.28173, + "grad_norm": 0.5653994462344406, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 28173 + }, + { + "epoch": 0.28174, + "grad_norm": 0.5924005771926022, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 28174 + }, + { + "epoch": 0.28175, + "grad_norm": 0.5709394421553473, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 28175 + }, + { + "epoch": 0.28176, + "grad_norm": 0.5500702792715431, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 28176 + }, + { + "epoch": 0.28177, + "grad_norm": 0.6547067247905749, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 28177 + }, + { + "epoch": 0.28178, + "grad_norm": 0.7133267344227339, + "learning_rate": 0.003, + "loss": 3.99, + "step": 28178 + }, + { + "epoch": 0.28179, + "grad_norm": 0.6437940691110232, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 28179 + }, + { + "epoch": 0.2818, + "grad_norm": 0.6556641573435348, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 28180 + }, + { + "epoch": 0.28181, + "grad_norm": 0.5410091439414125, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 28181 + }, + { + "epoch": 0.28182, + "grad_norm": 0.5355852134269649, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 28182 + }, + { + "epoch": 0.28183, + "grad_norm": 0.5845009214342385, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 28183 + }, + { + "epoch": 0.28184, + "grad_norm": 0.5504298790735975, + "learning_rate": 0.003, + "loss": 3.9287, + "step": 28184 + }, + { + "epoch": 0.28185, + "grad_norm": 0.6340993504256774, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 28185 + }, + { + "epoch": 0.28186, + "grad_norm": 0.7404353990997765, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 28186 + }, + { + "epoch": 0.28187, + "grad_norm": 0.7870448451170329, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 28187 + }, + { + "epoch": 0.28188, + "grad_norm": 0.8097681222968827, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 28188 + }, + { + "epoch": 0.28189, + "grad_norm": 0.9261350646461081, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 28189 + }, + { + "epoch": 0.2819, + "grad_norm": 1.0268856263205288, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 28190 + }, + { + "epoch": 0.28191, + "grad_norm": 1.0667210551200432, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 28191 + }, + { + "epoch": 0.28192, + "grad_norm": 0.781071643260067, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 28192 + }, + { + "epoch": 0.28193, + "grad_norm": 0.6913178602163792, + "learning_rate": 0.003, + "loss": 3.965, + "step": 28193 + }, + { + "epoch": 0.28194, + "grad_norm": 0.716574431875639, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 28194 + }, + { + "epoch": 0.28195, + "grad_norm": 0.9893540567414196, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 28195 + }, + { + "epoch": 0.28196, + "grad_norm": 1.3184934540390083, + "learning_rate": 0.003, + "loss": 4.026, + "step": 28196 + }, + { + "epoch": 0.28197, + "grad_norm": 0.5497225758481923, + "learning_rate": 0.003, + "loss": 3.992, + "step": 28197 + }, + { + "epoch": 0.28198, + "grad_norm": 0.8915164900413133, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 28198 + }, + { + "epoch": 0.28199, + "grad_norm": 1.0569495504039939, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 28199 + }, + { + "epoch": 0.282, + "grad_norm": 0.7864926748055652, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 28200 + }, + { + "epoch": 0.28201, + "grad_norm": 0.690115611216789, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 28201 + }, + { + "epoch": 0.28202, + "grad_norm": 0.619998339330392, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 28202 + }, + { + "epoch": 0.28203, + "grad_norm": 0.6591337648904464, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 28203 + }, + { + "epoch": 0.28204, + "grad_norm": 0.6865876449836436, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 28204 + }, + { + "epoch": 0.28205, + "grad_norm": 0.7316339136765639, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 28205 + }, + { + "epoch": 0.28206, + "grad_norm": 0.8394149186024958, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 28206 + }, + { + "epoch": 0.28207, + "grad_norm": 0.8517166814187899, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 28207 + }, + { + "epoch": 0.28208, + "grad_norm": 0.8653144110625345, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 28208 + }, + { + "epoch": 0.28209, + "grad_norm": 0.8494398065241906, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 28209 + }, + { + "epoch": 0.2821, + "grad_norm": 0.7619891319749994, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 28210 + }, + { + "epoch": 0.28211, + "grad_norm": 0.7519811340870335, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 28211 + }, + { + "epoch": 0.28212, + "grad_norm": 0.7060282595838953, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28212 + }, + { + "epoch": 0.28213, + "grad_norm": 0.6693090596052362, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 28213 + }, + { + "epoch": 0.28214, + "grad_norm": 0.6624609972331746, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 28214 + }, + { + "epoch": 0.28215, + "grad_norm": 0.5966601219332484, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 28215 + }, + { + "epoch": 0.28216, + "grad_norm": 0.6716613041571552, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 28216 + }, + { + "epoch": 0.28217, + "grad_norm": 0.6442796212638237, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 28217 + }, + { + "epoch": 0.28218, + "grad_norm": 0.5883539097140557, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 28218 + }, + { + "epoch": 0.28219, + "grad_norm": 0.6132517288004553, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 28219 + }, + { + "epoch": 0.2822, + "grad_norm": 0.6160922014461736, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 28220 + }, + { + "epoch": 0.28221, + "grad_norm": 0.7066639688516728, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 28221 + }, + { + "epoch": 0.28222, + "grad_norm": 0.7759564065882493, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 28222 + }, + { + "epoch": 0.28223, + "grad_norm": 0.785285624223619, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 28223 + }, + { + "epoch": 0.28224, + "grad_norm": 0.7486595687086757, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 28224 + }, + { + "epoch": 0.28225, + "grad_norm": 0.7839725224933732, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 28225 + }, + { + "epoch": 0.28226, + "grad_norm": 0.9033777815043428, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 28226 + }, + { + "epoch": 0.28227, + "grad_norm": 0.9940327825831813, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 28227 + }, + { + "epoch": 0.28228, + "grad_norm": 0.8652198110569805, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 28228 + }, + { + "epoch": 0.28229, + "grad_norm": 0.9650149151131087, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 28229 + }, + { + "epoch": 0.2823, + "grad_norm": 1.0505559553017678, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 28230 + }, + { + "epoch": 0.28231, + "grad_norm": 0.9260436738427014, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 28231 + }, + { + "epoch": 0.28232, + "grad_norm": 0.8448959391219902, + "learning_rate": 0.003, + "loss": 3.999, + "step": 28232 + }, + { + "epoch": 0.28233, + "grad_norm": 0.854065541388492, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 28233 + }, + { + "epoch": 0.28234, + "grad_norm": 0.9039084617252149, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 28234 + }, + { + "epoch": 0.28235, + "grad_norm": 0.8056686418464506, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 28235 + }, + { + "epoch": 0.28236, + "grad_norm": 0.7238918377473705, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 28236 + }, + { + "epoch": 0.28237, + "grad_norm": 0.755927466004355, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 28237 + }, + { + "epoch": 0.28238, + "grad_norm": 0.7486326811388554, + "learning_rate": 0.003, + "loss": 3.999, + "step": 28238 + }, + { + "epoch": 0.28239, + "grad_norm": 0.898584183484748, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 28239 + }, + { + "epoch": 0.2824, + "grad_norm": 0.9203174940292991, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 28240 + }, + { + "epoch": 0.28241, + "grad_norm": 0.7916603441866569, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 28241 + }, + { + "epoch": 0.28242, + "grad_norm": 0.7368963099252765, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 28242 + }, + { + "epoch": 0.28243, + "grad_norm": 0.713011063214836, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 28243 + }, + { + "epoch": 0.28244, + "grad_norm": 0.7187787810567695, + "learning_rate": 0.003, + "loss": 3.989, + "step": 28244 + }, + { + "epoch": 0.28245, + "grad_norm": 0.6048856841484191, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 28245 + }, + { + "epoch": 0.28246, + "grad_norm": 0.6276594540255627, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 28246 + }, + { + "epoch": 0.28247, + "grad_norm": 0.7256002092131119, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 28247 + }, + { + "epoch": 0.28248, + "grad_norm": 0.7194112790244956, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 28248 + }, + { + "epoch": 0.28249, + "grad_norm": 0.7617675724248898, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 28249 + }, + { + "epoch": 0.2825, + "grad_norm": 0.7048571206025385, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 28250 + }, + { + "epoch": 0.28251, + "grad_norm": 0.6727870126136609, + "learning_rate": 0.003, + "loss": 3.997, + "step": 28251 + }, + { + "epoch": 0.28252, + "grad_norm": 0.6391664980620589, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 28252 + }, + { + "epoch": 0.28253, + "grad_norm": 0.5595095200598855, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 28253 + }, + { + "epoch": 0.28254, + "grad_norm": 0.5028201130303203, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 28254 + }, + { + "epoch": 0.28255, + "grad_norm": 0.6164608309172973, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 28255 + }, + { + "epoch": 0.28256, + "grad_norm": 0.6227772607836147, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 28256 + }, + { + "epoch": 0.28257, + "grad_norm": 0.71639579355047, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 28257 + }, + { + "epoch": 0.28258, + "grad_norm": 0.6852452305989063, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 28258 + }, + { + "epoch": 0.28259, + "grad_norm": 0.7689273792091248, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 28259 + }, + { + "epoch": 0.2826, + "grad_norm": 0.8486915009544521, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 28260 + }, + { + "epoch": 0.28261, + "grad_norm": 0.8883362068787933, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 28261 + }, + { + "epoch": 0.28262, + "grad_norm": 0.8856128498726811, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 28262 + }, + { + "epoch": 0.28263, + "grad_norm": 0.7673017124785033, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 28263 + }, + { + "epoch": 0.28264, + "grad_norm": 0.774690179075431, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 28264 + }, + { + "epoch": 0.28265, + "grad_norm": 0.7568937170783783, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 28265 + }, + { + "epoch": 0.28266, + "grad_norm": 0.7024783375755262, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 28266 + }, + { + "epoch": 0.28267, + "grad_norm": 0.6882022799303211, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28267 + }, + { + "epoch": 0.28268, + "grad_norm": 0.6722836342686084, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 28268 + }, + { + "epoch": 0.28269, + "grad_norm": 0.677180922510174, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 28269 + }, + { + "epoch": 0.2827, + "grad_norm": 0.7518538907727039, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 28270 + }, + { + "epoch": 0.28271, + "grad_norm": 0.8253518685471088, + "learning_rate": 0.003, + "loss": 4.0416, + "step": 28271 + }, + { + "epoch": 0.28272, + "grad_norm": 0.9648530796142871, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 28272 + }, + { + "epoch": 0.28273, + "grad_norm": 1.1408953092561733, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 28273 + }, + { + "epoch": 0.28274, + "grad_norm": 0.8363920927152508, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 28274 + }, + { + "epoch": 0.28275, + "grad_norm": 0.8131858539602597, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 28275 + }, + { + "epoch": 0.28276, + "grad_norm": 0.8397903936904251, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 28276 + }, + { + "epoch": 0.28277, + "grad_norm": 0.840637094428657, + "learning_rate": 0.003, + "loss": 3.992, + "step": 28277 + }, + { + "epoch": 0.28278, + "grad_norm": 0.8517315325822327, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 28278 + }, + { + "epoch": 0.28279, + "grad_norm": 0.9732413783430856, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 28279 + }, + { + "epoch": 0.2828, + "grad_norm": 1.0144553525981848, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 28280 + }, + { + "epoch": 0.28281, + "grad_norm": 1.0270291943173742, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 28281 + }, + { + "epoch": 0.28282, + "grad_norm": 0.9726743288226722, + "learning_rate": 0.003, + "loss": 4.032, + "step": 28282 + }, + { + "epoch": 0.28283, + "grad_norm": 0.9323884968111951, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 28283 + }, + { + "epoch": 0.28284, + "grad_norm": 0.8156990533045573, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 28284 + }, + { + "epoch": 0.28285, + "grad_norm": 0.7559292031530866, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 28285 + }, + { + "epoch": 0.28286, + "grad_norm": 0.6575825408947644, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 28286 + }, + { + "epoch": 0.28287, + "grad_norm": 0.7335762441480888, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 28287 + }, + { + "epoch": 0.28288, + "grad_norm": 0.6563981409433736, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 28288 + }, + { + "epoch": 0.28289, + "grad_norm": 0.5409188825009004, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 28289 + }, + { + "epoch": 0.2829, + "grad_norm": 0.6652507380529679, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 28290 + }, + { + "epoch": 0.28291, + "grad_norm": 0.7119508416845842, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 28291 + }, + { + "epoch": 0.28292, + "grad_norm": 0.818258364148135, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 28292 + }, + { + "epoch": 0.28293, + "grad_norm": 0.7312030589568758, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 28293 + }, + { + "epoch": 0.28294, + "grad_norm": 0.674345790174293, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 28294 + }, + { + "epoch": 0.28295, + "grad_norm": 0.7357905606820149, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 28295 + }, + { + "epoch": 0.28296, + "grad_norm": 0.6278600325498223, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 28296 + }, + { + "epoch": 0.28297, + "grad_norm": 0.6191733518284912, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 28297 + }, + { + "epoch": 0.28298, + "grad_norm": 0.5760800825782375, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 28298 + }, + { + "epoch": 0.28299, + "grad_norm": 0.6061593056069524, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 28299 + }, + { + "epoch": 0.283, + "grad_norm": 0.5560511069751762, + "learning_rate": 0.003, + "loss": 3.9469, + "step": 28300 + }, + { + "epoch": 0.28301, + "grad_norm": 0.49838448592531015, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 28301 + }, + { + "epoch": 0.28302, + "grad_norm": 0.5327194987882621, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 28302 + }, + { + "epoch": 0.28303, + "grad_norm": 0.5502732205483247, + "learning_rate": 0.003, + "loss": 3.955, + "step": 28303 + }, + { + "epoch": 0.28304, + "grad_norm": 0.6522601608345278, + "learning_rate": 0.003, + "loss": 3.995, + "step": 28304 + }, + { + "epoch": 0.28305, + "grad_norm": 0.8203439309198607, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 28305 + }, + { + "epoch": 0.28306, + "grad_norm": 0.9097752847419966, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 28306 + }, + { + "epoch": 0.28307, + "grad_norm": 0.9134806842370512, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 28307 + }, + { + "epoch": 0.28308, + "grad_norm": 0.797392273630133, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 28308 + }, + { + "epoch": 0.28309, + "grad_norm": 0.7747222560372234, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 28309 + }, + { + "epoch": 0.2831, + "grad_norm": 0.8107527431283518, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 28310 + }, + { + "epoch": 0.28311, + "grad_norm": 0.8382005527916963, + "learning_rate": 0.003, + "loss": 3.977, + "step": 28311 + }, + { + "epoch": 0.28312, + "grad_norm": 0.8003606741311015, + "learning_rate": 0.003, + "loss": 4.018, + "step": 28312 + }, + { + "epoch": 0.28313, + "grad_norm": 0.7852079404111765, + "learning_rate": 0.003, + "loss": 3.999, + "step": 28313 + }, + { + "epoch": 0.28314, + "grad_norm": 0.8006657021156711, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 28314 + }, + { + "epoch": 0.28315, + "grad_norm": 0.7381151504631043, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 28315 + }, + { + "epoch": 0.28316, + "grad_norm": 0.72348615643052, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 28316 + }, + { + "epoch": 0.28317, + "grad_norm": 0.8876186704510156, + "learning_rate": 0.003, + "loss": 4.006, + "step": 28317 + }, + { + "epoch": 0.28318, + "grad_norm": 1.0622363325814512, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 28318 + }, + { + "epoch": 0.28319, + "grad_norm": 0.9384571871484173, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 28319 + }, + { + "epoch": 0.2832, + "grad_norm": 0.8820312833539243, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 28320 + }, + { + "epoch": 0.28321, + "grad_norm": 0.7764217222801517, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 28321 + }, + { + "epoch": 0.28322, + "grad_norm": 0.7128729546060582, + "learning_rate": 0.003, + "loss": 3.988, + "step": 28322 + }, + { + "epoch": 0.28323, + "grad_norm": 0.6399493690542621, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 28323 + }, + { + "epoch": 0.28324, + "grad_norm": 0.615509818148519, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 28324 + }, + { + "epoch": 0.28325, + "grad_norm": 0.6056599284914175, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 28325 + }, + { + "epoch": 0.28326, + "grad_norm": 0.6250297796985825, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 28326 + }, + { + "epoch": 0.28327, + "grad_norm": 0.695358969169569, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 28327 + }, + { + "epoch": 0.28328, + "grad_norm": 0.7862095866799819, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 28328 + }, + { + "epoch": 0.28329, + "grad_norm": 0.7838825740380239, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 28329 + }, + { + "epoch": 0.2833, + "grad_norm": 0.9444742471775321, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 28330 + }, + { + "epoch": 0.28331, + "grad_norm": 0.9603526650300601, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 28331 + }, + { + "epoch": 0.28332, + "grad_norm": 0.9980735188829994, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 28332 + }, + { + "epoch": 0.28333, + "grad_norm": 1.0618628868442437, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 28333 + }, + { + "epoch": 0.28334, + "grad_norm": 0.9908019375835256, + "learning_rate": 0.003, + "loss": 4.0431, + "step": 28334 + }, + { + "epoch": 0.28335, + "grad_norm": 0.8745005630106284, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 28335 + }, + { + "epoch": 0.28336, + "grad_norm": 0.7373055652488252, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 28336 + }, + { + "epoch": 0.28337, + "grad_norm": 0.73346439783573, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 28337 + }, + { + "epoch": 0.28338, + "grad_norm": 0.8077116571828927, + "learning_rate": 0.003, + "loss": 4.003, + "step": 28338 + }, + { + "epoch": 0.28339, + "grad_norm": 0.9398073450859554, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 28339 + }, + { + "epoch": 0.2834, + "grad_norm": 0.7750975389601992, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 28340 + }, + { + "epoch": 0.28341, + "grad_norm": 0.8614588296060142, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 28341 + }, + { + "epoch": 0.28342, + "grad_norm": 0.8983437407131765, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 28342 + }, + { + "epoch": 0.28343, + "grad_norm": 0.8785225151406981, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 28343 + }, + { + "epoch": 0.28344, + "grad_norm": 0.901432825979212, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 28344 + }, + { + "epoch": 0.28345, + "grad_norm": 0.7451317755755678, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 28345 + }, + { + "epoch": 0.28346, + "grad_norm": 0.6525241960677777, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 28346 + }, + { + "epoch": 0.28347, + "grad_norm": 0.68491997838708, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 28347 + }, + { + "epoch": 0.28348, + "grad_norm": 0.6615286767180331, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 28348 + }, + { + "epoch": 0.28349, + "grad_norm": 0.7589607390624518, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 28349 + }, + { + "epoch": 0.2835, + "grad_norm": 0.8262785999619078, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 28350 + }, + { + "epoch": 0.28351, + "grad_norm": 0.827039546134529, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 28351 + }, + { + "epoch": 0.28352, + "grad_norm": 0.8471267368889693, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 28352 + }, + { + "epoch": 0.28353, + "grad_norm": 0.7346938159089187, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 28353 + }, + { + "epoch": 0.28354, + "grad_norm": 0.6634466927361584, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 28354 + }, + { + "epoch": 0.28355, + "grad_norm": 0.6572950609782401, + "learning_rate": 0.003, + "loss": 3.986, + "step": 28355 + }, + { + "epoch": 0.28356, + "grad_norm": 0.7136904869520773, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 28356 + }, + { + "epoch": 0.28357, + "grad_norm": 0.6915092203594485, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 28357 + }, + { + "epoch": 0.28358, + "grad_norm": 0.7020868879672683, + "learning_rate": 0.003, + "loss": 4.006, + "step": 28358 + }, + { + "epoch": 0.28359, + "grad_norm": 0.7874910923543389, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 28359 + }, + { + "epoch": 0.2836, + "grad_norm": 0.7189558520624726, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 28360 + }, + { + "epoch": 0.28361, + "grad_norm": 0.7948782978728129, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 28361 + }, + { + "epoch": 0.28362, + "grad_norm": 1.0461490576181984, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 28362 + }, + { + "epoch": 0.28363, + "grad_norm": 1.034964934282036, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 28363 + }, + { + "epoch": 0.28364, + "grad_norm": 0.9971439143014055, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 28364 + }, + { + "epoch": 0.28365, + "grad_norm": 0.8764850845786479, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 28365 + }, + { + "epoch": 0.28366, + "grad_norm": 0.648824434514544, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 28366 + }, + { + "epoch": 0.28367, + "grad_norm": 0.6448210445397512, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 28367 + }, + { + "epoch": 0.28368, + "grad_norm": 0.647011333615255, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 28368 + }, + { + "epoch": 0.28369, + "grad_norm": 0.6005522255940693, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 28369 + }, + { + "epoch": 0.2837, + "grad_norm": 0.5780467928314005, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 28370 + }, + { + "epoch": 0.28371, + "grad_norm": 0.5502465496608934, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 28371 + }, + { + "epoch": 0.28372, + "grad_norm": 0.5516739038541788, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 28372 + }, + { + "epoch": 0.28373, + "grad_norm": 0.5351655952773391, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 28373 + }, + { + "epoch": 0.28374, + "grad_norm": 0.5275886130745399, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 28374 + }, + { + "epoch": 0.28375, + "grad_norm": 0.5727083665168334, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 28375 + }, + { + "epoch": 0.28376, + "grad_norm": 0.5584764881761762, + "learning_rate": 0.003, + "loss": 3.989, + "step": 28376 + }, + { + "epoch": 0.28377, + "grad_norm": 0.5758539364990977, + "learning_rate": 0.003, + "loss": 3.993, + "step": 28377 + }, + { + "epoch": 0.28378, + "grad_norm": 0.5499514176517846, + "learning_rate": 0.003, + "loss": 3.987, + "step": 28378 + }, + { + "epoch": 0.28379, + "grad_norm": 0.58446357350955, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 28379 + }, + { + "epoch": 0.2838, + "grad_norm": 0.7212084133072366, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 28380 + }, + { + "epoch": 0.28381, + "grad_norm": 0.7741343164313197, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 28381 + }, + { + "epoch": 0.28382, + "grad_norm": 0.8007679668682183, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 28382 + }, + { + "epoch": 0.28383, + "grad_norm": 0.8504089098437465, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 28383 + }, + { + "epoch": 0.28384, + "grad_norm": 0.7930598355159839, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 28384 + }, + { + "epoch": 0.28385, + "grad_norm": 0.8675283712248099, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 28385 + }, + { + "epoch": 0.28386, + "grad_norm": 0.9147211166877625, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 28386 + }, + { + "epoch": 0.28387, + "grad_norm": 1.004611360663758, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 28387 + }, + { + "epoch": 0.28388, + "grad_norm": 0.8985134024578418, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 28388 + }, + { + "epoch": 0.28389, + "grad_norm": 0.8641402369004274, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 28389 + }, + { + "epoch": 0.2839, + "grad_norm": 0.8017950954136089, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 28390 + }, + { + "epoch": 0.28391, + "grad_norm": 0.7391391593402493, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 28391 + }, + { + "epoch": 0.28392, + "grad_norm": 0.8225200107388612, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 28392 + }, + { + "epoch": 0.28393, + "grad_norm": 0.891291145785445, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 28393 + }, + { + "epoch": 0.28394, + "grad_norm": 0.9698397422919041, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 28394 + }, + { + "epoch": 0.28395, + "grad_norm": 0.8673630342746288, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 28395 + }, + { + "epoch": 0.28396, + "grad_norm": 0.9409653919227748, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 28396 + }, + { + "epoch": 0.28397, + "grad_norm": 1.1080268634309693, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 28397 + }, + { + "epoch": 0.28398, + "grad_norm": 0.9829883087178152, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 28398 + }, + { + "epoch": 0.28399, + "grad_norm": 0.800483880222118, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 28399 + }, + { + "epoch": 0.284, + "grad_norm": 0.6968075368068753, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 28400 + }, + { + "epoch": 0.28401, + "grad_norm": 0.7365453641447024, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 28401 + }, + { + "epoch": 0.28402, + "grad_norm": 0.7184161491183083, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 28402 + }, + { + "epoch": 0.28403, + "grad_norm": 0.7164720153949019, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 28403 + }, + { + "epoch": 0.28404, + "grad_norm": 0.7520122193854667, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 28404 + }, + { + "epoch": 0.28405, + "grad_norm": 0.7706018682937796, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 28405 + }, + { + "epoch": 0.28406, + "grad_norm": 0.7662894703459542, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 28406 + }, + { + "epoch": 0.28407, + "grad_norm": 0.6603740345203238, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 28407 + }, + { + "epoch": 0.28408, + "grad_norm": 0.6612986415791804, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 28408 + }, + { + "epoch": 0.28409, + "grad_norm": 0.7455858608940914, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 28409 + }, + { + "epoch": 0.2841, + "grad_norm": 0.8403708372791682, + "learning_rate": 0.003, + "loss": 4.001, + "step": 28410 + }, + { + "epoch": 0.28411, + "grad_norm": 0.8311906832203397, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 28411 + }, + { + "epoch": 0.28412, + "grad_norm": 0.9480756814733876, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 28412 + }, + { + "epoch": 0.28413, + "grad_norm": 0.916692630539645, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 28413 + }, + { + "epoch": 0.28414, + "grad_norm": 0.7326215069834059, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 28414 + }, + { + "epoch": 0.28415, + "grad_norm": 0.6690148905490471, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 28415 + }, + { + "epoch": 0.28416, + "grad_norm": 0.6591886957700172, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 28416 + }, + { + "epoch": 0.28417, + "grad_norm": 0.6224925405557238, + "learning_rate": 0.003, + "loss": 3.978, + "step": 28417 + }, + { + "epoch": 0.28418, + "grad_norm": 0.8084610317947564, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 28418 + }, + { + "epoch": 0.28419, + "grad_norm": 0.9735466930698421, + "learning_rate": 0.003, + "loss": 4.0483, + "step": 28419 + }, + { + "epoch": 0.2842, + "grad_norm": 1.0690744417996756, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 28420 + }, + { + "epoch": 0.28421, + "grad_norm": 0.929923453516141, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 28421 + }, + { + "epoch": 0.28422, + "grad_norm": 0.8017409749321587, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 28422 + }, + { + "epoch": 0.28423, + "grad_norm": 0.7021039178487563, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 28423 + }, + { + "epoch": 0.28424, + "grad_norm": 0.7071981751091035, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 28424 + }, + { + "epoch": 0.28425, + "grad_norm": 0.763562307990982, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 28425 + }, + { + "epoch": 0.28426, + "grad_norm": 0.8007644069118905, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 28426 + }, + { + "epoch": 0.28427, + "grad_norm": 0.6936231514097444, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 28427 + }, + { + "epoch": 0.28428, + "grad_norm": 0.6262158887485195, + "learning_rate": 0.003, + "loss": 3.993, + "step": 28428 + }, + { + "epoch": 0.28429, + "grad_norm": 0.6130530570569147, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 28429 + }, + { + "epoch": 0.2843, + "grad_norm": 0.6492449406601271, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 28430 + }, + { + "epoch": 0.28431, + "grad_norm": 0.7198363249482194, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 28431 + }, + { + "epoch": 0.28432, + "grad_norm": 0.7509850839021031, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 28432 + }, + { + "epoch": 0.28433, + "grad_norm": 0.7564465138816069, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 28433 + }, + { + "epoch": 0.28434, + "grad_norm": 0.7841205443915378, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 28434 + }, + { + "epoch": 0.28435, + "grad_norm": 0.7205793538642267, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 28435 + }, + { + "epoch": 0.28436, + "grad_norm": 0.5927435396866955, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 28436 + }, + { + "epoch": 0.28437, + "grad_norm": 0.643000520467839, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 28437 + }, + { + "epoch": 0.28438, + "grad_norm": 0.580535440851499, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 28438 + }, + { + "epoch": 0.28439, + "grad_norm": 0.6450679574901047, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 28439 + }, + { + "epoch": 0.2844, + "grad_norm": 0.7969155989447262, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 28440 + }, + { + "epoch": 0.28441, + "grad_norm": 0.9471014061589611, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 28441 + }, + { + "epoch": 0.28442, + "grad_norm": 1.0225021655452011, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 28442 + }, + { + "epoch": 0.28443, + "grad_norm": 0.8759326997960274, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 28443 + }, + { + "epoch": 0.28444, + "grad_norm": 0.693438282904106, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 28444 + }, + { + "epoch": 0.28445, + "grad_norm": 0.5634453265803961, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 28445 + }, + { + "epoch": 0.28446, + "grad_norm": 0.6228933661945061, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 28446 + }, + { + "epoch": 0.28447, + "grad_norm": 0.6341181836278057, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 28447 + }, + { + "epoch": 0.28448, + "grad_norm": 0.619761712608331, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 28448 + }, + { + "epoch": 0.28449, + "grad_norm": 0.6669706458259693, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 28449 + }, + { + "epoch": 0.2845, + "grad_norm": 0.8748871390232635, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 28450 + }, + { + "epoch": 0.28451, + "grad_norm": 1.241407050876863, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 28451 + }, + { + "epoch": 0.28452, + "grad_norm": 0.678882754006759, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 28452 + }, + { + "epoch": 0.28453, + "grad_norm": 0.6286598253689947, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 28453 + }, + { + "epoch": 0.28454, + "grad_norm": 0.709646973915133, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 28454 + }, + { + "epoch": 0.28455, + "grad_norm": 0.7569961148587386, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 28455 + }, + { + "epoch": 0.28456, + "grad_norm": 0.7297601581341333, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 28456 + }, + { + "epoch": 0.28457, + "grad_norm": 0.7435477013160595, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 28457 + }, + { + "epoch": 0.28458, + "grad_norm": 0.802623175352527, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 28458 + }, + { + "epoch": 0.28459, + "grad_norm": 0.7244510753569313, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 28459 + }, + { + "epoch": 0.2846, + "grad_norm": 0.7652550555293283, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 28460 + }, + { + "epoch": 0.28461, + "grad_norm": 0.8581372724540486, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 28461 + }, + { + "epoch": 0.28462, + "grad_norm": 0.7605502103274916, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 28462 + }, + { + "epoch": 0.28463, + "grad_norm": 0.6670198829159583, + "learning_rate": 0.003, + "loss": 3.971, + "step": 28463 + }, + { + "epoch": 0.28464, + "grad_norm": 0.8107343809519582, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 28464 + }, + { + "epoch": 0.28465, + "grad_norm": 0.8605071532222128, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 28465 + }, + { + "epoch": 0.28466, + "grad_norm": 0.8570382484304282, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 28466 + }, + { + "epoch": 0.28467, + "grad_norm": 0.8657599087077388, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 28467 + }, + { + "epoch": 0.28468, + "grad_norm": 0.8551904264727741, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 28468 + }, + { + "epoch": 0.28469, + "grad_norm": 0.7853449844418583, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 28469 + }, + { + "epoch": 0.2847, + "grad_norm": 0.7738471080150783, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 28470 + }, + { + "epoch": 0.28471, + "grad_norm": 0.7197017887445892, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 28471 + }, + { + "epoch": 0.28472, + "grad_norm": 0.804582567251711, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 28472 + }, + { + "epoch": 0.28473, + "grad_norm": 0.8555040464270035, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 28473 + }, + { + "epoch": 0.28474, + "grad_norm": 0.8342518218376203, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 28474 + }, + { + "epoch": 0.28475, + "grad_norm": 0.7610845477842737, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 28475 + }, + { + "epoch": 0.28476, + "grad_norm": 0.8018173719089388, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 28476 + }, + { + "epoch": 0.28477, + "grad_norm": 0.8325039945454805, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 28477 + }, + { + "epoch": 0.28478, + "grad_norm": 0.7515181238314781, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 28478 + }, + { + "epoch": 0.28479, + "grad_norm": 0.7070177864384047, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 28479 + }, + { + "epoch": 0.2848, + "grad_norm": 0.7928457218702176, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 28480 + }, + { + "epoch": 0.28481, + "grad_norm": 0.8726020252586085, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 28481 + }, + { + "epoch": 0.28482, + "grad_norm": 1.007482203793029, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 28482 + }, + { + "epoch": 0.28483, + "grad_norm": 1.1672135880351133, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 28483 + }, + { + "epoch": 0.28484, + "grad_norm": 0.9301945498731726, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 28484 + }, + { + "epoch": 0.28485, + "grad_norm": 0.6941606159324162, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 28485 + }, + { + "epoch": 0.28486, + "grad_norm": 0.6796199319068719, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 28486 + }, + { + "epoch": 0.28487, + "grad_norm": 0.7450406670334169, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 28487 + }, + { + "epoch": 0.28488, + "grad_norm": 0.7895913265922636, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 28488 + }, + { + "epoch": 0.28489, + "grad_norm": 0.8402470609283078, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 28489 + }, + { + "epoch": 0.2849, + "grad_norm": 0.8771307536981336, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 28490 + }, + { + "epoch": 0.28491, + "grad_norm": 0.9219273292194755, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 28491 + }, + { + "epoch": 0.28492, + "grad_norm": 0.832921842298198, + "learning_rate": 0.003, + "loss": 4.005, + "step": 28492 + }, + { + "epoch": 0.28493, + "grad_norm": 0.7075178062433362, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 28493 + }, + { + "epoch": 0.28494, + "grad_norm": 0.631439212353034, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 28494 + }, + { + "epoch": 0.28495, + "grad_norm": 0.5926312841749776, + "learning_rate": 0.003, + "loss": 4.006, + "step": 28495 + }, + { + "epoch": 0.28496, + "grad_norm": 0.5430031208006622, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 28496 + }, + { + "epoch": 0.28497, + "grad_norm": 0.5409631126196203, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 28497 + }, + { + "epoch": 0.28498, + "grad_norm": 0.4610355456986479, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 28498 + }, + { + "epoch": 0.28499, + "grad_norm": 0.5742601844645909, + "learning_rate": 0.003, + "loss": 4.027, + "step": 28499 + }, + { + "epoch": 0.285, + "grad_norm": 0.7865466489001767, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 28500 + }, + { + "epoch": 0.28501, + "grad_norm": 0.880884577313961, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 28501 + }, + { + "epoch": 0.28502, + "grad_norm": 0.8021080038797826, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 28502 + }, + { + "epoch": 0.28503, + "grad_norm": 0.7213780229750658, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 28503 + }, + { + "epoch": 0.28504, + "grad_norm": 0.7164679077020295, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 28504 + }, + { + "epoch": 0.28505, + "grad_norm": 0.7755653138818177, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 28505 + }, + { + "epoch": 0.28506, + "grad_norm": 0.7711939071239874, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 28506 + }, + { + "epoch": 0.28507, + "grad_norm": 0.7538486594059517, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 28507 + }, + { + "epoch": 0.28508, + "grad_norm": 0.8577526010991279, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 28508 + }, + { + "epoch": 0.28509, + "grad_norm": 0.928428584957478, + "learning_rate": 0.003, + "loss": 4.016, + "step": 28509 + }, + { + "epoch": 0.2851, + "grad_norm": 0.9378683458876823, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 28510 + }, + { + "epoch": 0.28511, + "grad_norm": 0.9473509757392815, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 28511 + }, + { + "epoch": 0.28512, + "grad_norm": 0.8665992341087055, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 28512 + }, + { + "epoch": 0.28513, + "grad_norm": 0.8327236779165277, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 28513 + }, + { + "epoch": 0.28514, + "grad_norm": 0.7562420540917288, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 28514 + }, + { + "epoch": 0.28515, + "grad_norm": 0.7067597793984113, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 28515 + }, + { + "epoch": 0.28516, + "grad_norm": 0.6710260635422052, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 28516 + }, + { + "epoch": 0.28517, + "grad_norm": 0.5847332563008711, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 28517 + }, + { + "epoch": 0.28518, + "grad_norm": 0.5422744810247769, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 28518 + }, + { + "epoch": 0.28519, + "grad_norm": 0.6386106137481073, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 28519 + }, + { + "epoch": 0.2852, + "grad_norm": 0.7794209463976665, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 28520 + }, + { + "epoch": 0.28521, + "grad_norm": 0.8417234447996743, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 28521 + }, + { + "epoch": 0.28522, + "grad_norm": 0.8237768512060667, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 28522 + }, + { + "epoch": 0.28523, + "grad_norm": 0.8533655631370525, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 28523 + }, + { + "epoch": 0.28524, + "grad_norm": 0.864100321839732, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 28524 + }, + { + "epoch": 0.28525, + "grad_norm": 0.8612955071772045, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 28525 + }, + { + "epoch": 0.28526, + "grad_norm": 0.803824091883269, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 28526 + }, + { + "epoch": 0.28527, + "grad_norm": 0.7021673195713922, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 28527 + }, + { + "epoch": 0.28528, + "grad_norm": 0.6977452062092181, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 28528 + }, + { + "epoch": 0.28529, + "grad_norm": 0.8459863500848802, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 28529 + }, + { + "epoch": 0.2853, + "grad_norm": 0.9287103061249479, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 28530 + }, + { + "epoch": 0.28531, + "grad_norm": 0.8512155148942122, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28531 + }, + { + "epoch": 0.28532, + "grad_norm": 0.85828724318292, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 28532 + }, + { + "epoch": 0.28533, + "grad_norm": 0.9565231039474482, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 28533 + }, + { + "epoch": 0.28534, + "grad_norm": 0.971005130723497, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 28534 + }, + { + "epoch": 0.28535, + "grad_norm": 0.8776206471842124, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 28535 + }, + { + "epoch": 0.28536, + "grad_norm": 0.7608422656877792, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 28536 + }, + { + "epoch": 0.28537, + "grad_norm": 0.7067764178301, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 28537 + }, + { + "epoch": 0.28538, + "grad_norm": 0.7277167527424805, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 28538 + }, + { + "epoch": 0.28539, + "grad_norm": 0.7402288411857731, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 28539 + }, + { + "epoch": 0.2854, + "grad_norm": 0.8270046228784681, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 28540 + }, + { + "epoch": 0.28541, + "grad_norm": 0.7296848380428308, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 28541 + }, + { + "epoch": 0.28542, + "grad_norm": 0.7014103662632768, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 28542 + }, + { + "epoch": 0.28543, + "grad_norm": 0.7052045762480726, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 28543 + }, + { + "epoch": 0.28544, + "grad_norm": 0.6124073149486036, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 28544 + }, + { + "epoch": 0.28545, + "grad_norm": 0.5760414682410314, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 28545 + }, + { + "epoch": 0.28546, + "grad_norm": 0.6173010902530285, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 28546 + }, + { + "epoch": 0.28547, + "grad_norm": 0.708123674413082, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 28547 + }, + { + "epoch": 0.28548, + "grad_norm": 0.7035209240029212, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 28548 + }, + { + "epoch": 0.28549, + "grad_norm": 0.7115377094636446, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 28549 + }, + { + "epoch": 0.2855, + "grad_norm": 0.8736605150763647, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 28550 + }, + { + "epoch": 0.28551, + "grad_norm": 1.1006106689406145, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 28551 + }, + { + "epoch": 0.28552, + "grad_norm": 0.9249531143744473, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 28552 + }, + { + "epoch": 0.28553, + "grad_norm": 0.7593717585061671, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 28553 + }, + { + "epoch": 0.28554, + "grad_norm": 0.7029100352923715, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 28554 + }, + { + "epoch": 0.28555, + "grad_norm": 0.8288636219600447, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 28555 + }, + { + "epoch": 0.28556, + "grad_norm": 0.8585943310905848, + "learning_rate": 0.003, + "loss": 4.004, + "step": 28556 + }, + { + "epoch": 0.28557, + "grad_norm": 0.8734766686178013, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 28557 + }, + { + "epoch": 0.28558, + "grad_norm": 0.7091359703417977, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 28558 + }, + { + "epoch": 0.28559, + "grad_norm": 0.6904961161059832, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 28559 + }, + { + "epoch": 0.2856, + "grad_norm": 0.7533091236272765, + "learning_rate": 0.003, + "loss": 3.99, + "step": 28560 + }, + { + "epoch": 0.28561, + "grad_norm": 0.8169105490265738, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 28561 + }, + { + "epoch": 0.28562, + "grad_norm": 0.7392551684846981, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 28562 + }, + { + "epoch": 0.28563, + "grad_norm": 0.6440552999009729, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 28563 + }, + { + "epoch": 0.28564, + "grad_norm": 0.6973577077856893, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 28564 + }, + { + "epoch": 0.28565, + "grad_norm": 0.7091052192580284, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 28565 + }, + { + "epoch": 0.28566, + "grad_norm": 0.6327805181673413, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 28566 + }, + { + "epoch": 0.28567, + "grad_norm": 0.5639932508400719, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 28567 + }, + { + "epoch": 0.28568, + "grad_norm": 0.5890960640774938, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 28568 + }, + { + "epoch": 0.28569, + "grad_norm": 0.6825688427735956, + "learning_rate": 0.003, + "loss": 3.9446, + "step": 28569 + }, + { + "epoch": 0.2857, + "grad_norm": 0.6826970961998534, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 28570 + }, + { + "epoch": 0.28571, + "grad_norm": 0.6650625282304661, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 28571 + }, + { + "epoch": 0.28572, + "grad_norm": 0.5596536789245763, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 28572 + }, + { + "epoch": 0.28573, + "grad_norm": 0.5550849414745797, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 28573 + }, + { + "epoch": 0.28574, + "grad_norm": 0.6566520507136073, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 28574 + }, + { + "epoch": 0.28575, + "grad_norm": 0.9049752849472558, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 28575 + }, + { + "epoch": 0.28576, + "grad_norm": 1.0572148761823876, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 28576 + }, + { + "epoch": 0.28577, + "grad_norm": 0.8863077843222614, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 28577 + }, + { + "epoch": 0.28578, + "grad_norm": 0.8381405168342643, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 28578 + }, + { + "epoch": 0.28579, + "grad_norm": 0.7920500744305454, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 28579 + }, + { + "epoch": 0.2858, + "grad_norm": 0.9483016890950892, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 28580 + }, + { + "epoch": 0.28581, + "grad_norm": 0.9044977560471354, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 28581 + }, + { + "epoch": 0.28582, + "grad_norm": 0.8573514495527154, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 28582 + }, + { + "epoch": 0.28583, + "grad_norm": 0.8807998544084445, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 28583 + }, + { + "epoch": 0.28584, + "grad_norm": 1.0264934466388176, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 28584 + }, + { + "epoch": 0.28585, + "grad_norm": 1.0899408273100761, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 28585 + }, + { + "epoch": 0.28586, + "grad_norm": 0.7275607744740608, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 28586 + }, + { + "epoch": 0.28587, + "grad_norm": 0.714747692845323, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 28587 + }, + { + "epoch": 0.28588, + "grad_norm": 0.7151232593503533, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 28588 + }, + { + "epoch": 0.28589, + "grad_norm": 0.7053158085348762, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 28589 + }, + { + "epoch": 0.2859, + "grad_norm": 0.6820243370481746, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 28590 + }, + { + "epoch": 0.28591, + "grad_norm": 0.6990324715557826, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 28591 + }, + { + "epoch": 0.28592, + "grad_norm": 0.8446276349383453, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 28592 + }, + { + "epoch": 0.28593, + "grad_norm": 0.9390185433928152, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 28593 + }, + { + "epoch": 0.28594, + "grad_norm": 0.7955407663741889, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 28594 + }, + { + "epoch": 0.28595, + "grad_norm": 0.6749025722309887, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 28595 + }, + { + "epoch": 0.28596, + "grad_norm": 0.7956282593476754, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 28596 + }, + { + "epoch": 0.28597, + "grad_norm": 0.8066533646645022, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 28597 + }, + { + "epoch": 0.28598, + "grad_norm": 0.8465758157065394, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 28598 + }, + { + "epoch": 0.28599, + "grad_norm": 0.8784585537618788, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 28599 + }, + { + "epoch": 0.286, + "grad_norm": 0.9101871340064128, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 28600 + }, + { + "epoch": 0.28601, + "grad_norm": 0.7887217401788844, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 28601 + }, + { + "epoch": 0.28602, + "grad_norm": 0.807356405406069, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 28602 + }, + { + "epoch": 0.28603, + "grad_norm": 0.8750806591055245, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 28603 + }, + { + "epoch": 0.28604, + "grad_norm": 0.8475653778032013, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 28604 + }, + { + "epoch": 0.28605, + "grad_norm": 0.7522181953338535, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 28605 + }, + { + "epoch": 0.28606, + "grad_norm": 0.7774701321707688, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 28606 + }, + { + "epoch": 0.28607, + "grad_norm": 0.7620137587378143, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 28607 + }, + { + "epoch": 0.28608, + "grad_norm": 0.752402552154278, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 28608 + }, + { + "epoch": 0.28609, + "grad_norm": 0.7571821078638671, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 28609 + }, + { + "epoch": 0.2861, + "grad_norm": 0.7394809613890069, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 28610 + }, + { + "epoch": 0.28611, + "grad_norm": 0.7748169915032589, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 28611 + }, + { + "epoch": 0.28612, + "grad_norm": 0.7421770447154334, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 28612 + }, + { + "epoch": 0.28613, + "grad_norm": 0.7327373305734988, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 28613 + }, + { + "epoch": 0.28614, + "grad_norm": 0.7378079469807353, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 28614 + }, + { + "epoch": 0.28615, + "grad_norm": 0.7663249292273762, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 28615 + }, + { + "epoch": 0.28616, + "grad_norm": 0.7668123510041531, + "learning_rate": 0.003, + "loss": 3.9367, + "step": 28616 + }, + { + "epoch": 0.28617, + "grad_norm": 0.7366556240552035, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 28617 + }, + { + "epoch": 0.28618, + "grad_norm": 0.7045790263747104, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 28618 + }, + { + "epoch": 0.28619, + "grad_norm": 0.7417311264735508, + "learning_rate": 0.003, + "loss": 4.008, + "step": 28619 + }, + { + "epoch": 0.2862, + "grad_norm": 0.7474826405335621, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 28620 + }, + { + "epoch": 0.28621, + "grad_norm": 0.7945850114229644, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 28621 + }, + { + "epoch": 0.28622, + "grad_norm": 0.8938843897717419, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 28622 + }, + { + "epoch": 0.28623, + "grad_norm": 0.9861225606195932, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 28623 + }, + { + "epoch": 0.28624, + "grad_norm": 0.8913961391211164, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 28624 + }, + { + "epoch": 0.28625, + "grad_norm": 0.7727246420856531, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 28625 + }, + { + "epoch": 0.28626, + "grad_norm": 0.7419843096083002, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 28626 + }, + { + "epoch": 0.28627, + "grad_norm": 0.7801299500265775, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 28627 + }, + { + "epoch": 0.28628, + "grad_norm": 0.7737211985434601, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 28628 + }, + { + "epoch": 0.28629, + "grad_norm": 0.7292440404750106, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 28629 + }, + { + "epoch": 0.2863, + "grad_norm": 0.7097336130553655, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 28630 + }, + { + "epoch": 0.28631, + "grad_norm": 0.6327587650457639, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 28631 + }, + { + "epoch": 0.28632, + "grad_norm": 0.5963581713769722, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 28632 + }, + { + "epoch": 0.28633, + "grad_norm": 0.5756293544793428, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 28633 + }, + { + "epoch": 0.28634, + "grad_norm": 0.4964957023032938, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 28634 + }, + { + "epoch": 0.28635, + "grad_norm": 0.5144742772130423, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 28635 + }, + { + "epoch": 0.28636, + "grad_norm": 0.5711548588067467, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 28636 + }, + { + "epoch": 0.28637, + "grad_norm": 0.6978970294856269, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 28637 + }, + { + "epoch": 0.28638, + "grad_norm": 0.8341035554315185, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 28638 + }, + { + "epoch": 0.28639, + "grad_norm": 0.9001688214076096, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 28639 + }, + { + "epoch": 0.2864, + "grad_norm": 0.8898691470084044, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 28640 + }, + { + "epoch": 0.28641, + "grad_norm": 0.7939206918603191, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 28641 + }, + { + "epoch": 0.28642, + "grad_norm": 0.7235599555429012, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 28642 + }, + { + "epoch": 0.28643, + "grad_norm": 0.724903100370458, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 28643 + }, + { + "epoch": 0.28644, + "grad_norm": 0.8690377161006495, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 28644 + }, + { + "epoch": 0.28645, + "grad_norm": 1.0122887052242897, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 28645 + }, + { + "epoch": 0.28646, + "grad_norm": 0.9330596605941197, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 28646 + }, + { + "epoch": 0.28647, + "grad_norm": 0.7908768214426771, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 28647 + }, + { + "epoch": 0.28648, + "grad_norm": 0.7460101473301081, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 28648 + }, + { + "epoch": 0.28649, + "grad_norm": 0.7304321225088394, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 28649 + }, + { + "epoch": 0.2865, + "grad_norm": 0.7344107208176298, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 28650 + }, + { + "epoch": 0.28651, + "grad_norm": 0.7163105746744981, + "learning_rate": 0.003, + "loss": 3.983, + "step": 28651 + }, + { + "epoch": 0.28652, + "grad_norm": 0.5699473684016327, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 28652 + }, + { + "epoch": 0.28653, + "grad_norm": 0.5874911225081659, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 28653 + }, + { + "epoch": 0.28654, + "grad_norm": 0.6064182405072168, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 28654 + }, + { + "epoch": 0.28655, + "grad_norm": 0.7197206700256654, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 28655 + }, + { + "epoch": 0.28656, + "grad_norm": 0.8136966379782312, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 28656 + }, + { + "epoch": 0.28657, + "grad_norm": 0.911702463370428, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 28657 + }, + { + "epoch": 0.28658, + "grad_norm": 1.0184032810608903, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 28658 + }, + { + "epoch": 0.28659, + "grad_norm": 0.901880417180161, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 28659 + }, + { + "epoch": 0.2866, + "grad_norm": 0.9526520475143019, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 28660 + }, + { + "epoch": 0.28661, + "grad_norm": 1.0160611418028218, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 28661 + }, + { + "epoch": 0.28662, + "grad_norm": 0.9617920559596149, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 28662 + }, + { + "epoch": 0.28663, + "grad_norm": 0.7996165809836892, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 28663 + }, + { + "epoch": 0.28664, + "grad_norm": 0.6662600644781557, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 28664 + }, + { + "epoch": 0.28665, + "grad_norm": 0.7330719929674251, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 28665 + }, + { + "epoch": 0.28666, + "grad_norm": 0.6864551477086547, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 28666 + }, + { + "epoch": 0.28667, + "grad_norm": 0.7219137241981136, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 28667 + }, + { + "epoch": 0.28668, + "grad_norm": 0.6517366504765542, + "learning_rate": 0.003, + "loss": 3.984, + "step": 28668 + }, + { + "epoch": 0.28669, + "grad_norm": 0.6007586165301317, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 28669 + }, + { + "epoch": 0.2867, + "grad_norm": 0.6339377111512589, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 28670 + }, + { + "epoch": 0.28671, + "grad_norm": 0.6949155192653682, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 28671 + }, + { + "epoch": 0.28672, + "grad_norm": 0.6955017708769978, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 28672 + }, + { + "epoch": 0.28673, + "grad_norm": 0.6599029101095054, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 28673 + }, + { + "epoch": 0.28674, + "grad_norm": 0.7174880401528303, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 28674 + }, + { + "epoch": 0.28675, + "grad_norm": 0.9191223779128278, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 28675 + }, + { + "epoch": 0.28676, + "grad_norm": 1.122778894079271, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 28676 + }, + { + "epoch": 0.28677, + "grad_norm": 0.7823650248114659, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 28677 + }, + { + "epoch": 0.28678, + "grad_norm": 0.728237351355932, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 28678 + }, + { + "epoch": 0.28679, + "grad_norm": 0.746403956772074, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 28679 + }, + { + "epoch": 0.2868, + "grad_norm": 0.8941916133819491, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 28680 + }, + { + "epoch": 0.28681, + "grad_norm": 0.8967280690127301, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 28681 + }, + { + "epoch": 0.28682, + "grad_norm": 0.8707841574641632, + "learning_rate": 0.003, + "loss": 3.993, + "step": 28682 + }, + { + "epoch": 0.28683, + "grad_norm": 0.8594745783839677, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 28683 + }, + { + "epoch": 0.28684, + "grad_norm": 0.7328146922730358, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 28684 + }, + { + "epoch": 0.28685, + "grad_norm": 0.7641889855187659, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 28685 + }, + { + "epoch": 0.28686, + "grad_norm": 0.6977474022430248, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 28686 + }, + { + "epoch": 0.28687, + "grad_norm": 0.6657254017980513, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 28687 + }, + { + "epoch": 0.28688, + "grad_norm": 0.5636779314425178, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 28688 + }, + { + "epoch": 0.28689, + "grad_norm": 0.4815157522196794, + "learning_rate": 0.003, + "loss": 4.007, + "step": 28689 + }, + { + "epoch": 0.2869, + "grad_norm": 0.49231465499925564, + "learning_rate": 0.003, + "loss": 3.991, + "step": 28690 + }, + { + "epoch": 0.28691, + "grad_norm": 0.4506541389897436, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 28691 + }, + { + "epoch": 0.28692, + "grad_norm": 0.5226524757914959, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 28692 + }, + { + "epoch": 0.28693, + "grad_norm": 0.5268538431860278, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 28693 + }, + { + "epoch": 0.28694, + "grad_norm": 0.5496146102493956, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 28694 + }, + { + "epoch": 0.28695, + "grad_norm": 0.5887590995711113, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 28695 + }, + { + "epoch": 0.28696, + "grad_norm": 0.7045951648278592, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 28696 + }, + { + "epoch": 0.28697, + "grad_norm": 0.8288801775097465, + "learning_rate": 0.003, + "loss": 3.974, + "step": 28697 + }, + { + "epoch": 0.28698, + "grad_norm": 0.7886516000164684, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 28698 + }, + { + "epoch": 0.28699, + "grad_norm": 0.740358733352863, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 28699 + }, + { + "epoch": 0.287, + "grad_norm": 0.7726890507874913, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 28700 + }, + { + "epoch": 0.28701, + "grad_norm": 0.7844397241733164, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 28701 + }, + { + "epoch": 0.28702, + "grad_norm": 0.8007129381341971, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 28702 + }, + { + "epoch": 0.28703, + "grad_norm": 0.8807443328057704, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 28703 + }, + { + "epoch": 0.28704, + "grad_norm": 0.9077408970981528, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 28704 + }, + { + "epoch": 0.28705, + "grad_norm": 0.8906997397784541, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 28705 + }, + { + "epoch": 0.28706, + "grad_norm": 0.9518092169414303, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 28706 + }, + { + "epoch": 0.28707, + "grad_norm": 0.9508145905195546, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 28707 + }, + { + "epoch": 0.28708, + "grad_norm": 0.962387072386018, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 28708 + }, + { + "epoch": 0.28709, + "grad_norm": 1.054154713071217, + "learning_rate": 0.003, + "loss": 4.028, + "step": 28709 + }, + { + "epoch": 0.2871, + "grad_norm": 1.1078372377385834, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 28710 + }, + { + "epoch": 0.28711, + "grad_norm": 0.9401542086069987, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 28711 + }, + { + "epoch": 0.28712, + "grad_norm": 0.9085651151879093, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 28712 + }, + { + "epoch": 0.28713, + "grad_norm": 1.0440325487911837, + "learning_rate": 0.003, + "loss": 4.0385, + "step": 28713 + }, + { + "epoch": 0.28714, + "grad_norm": 0.9723440915379193, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 28714 + }, + { + "epoch": 0.28715, + "grad_norm": 0.8648117512631041, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 28715 + }, + { + "epoch": 0.28716, + "grad_norm": 0.8212241329506011, + "learning_rate": 0.003, + "loss": 3.989, + "step": 28716 + }, + { + "epoch": 0.28717, + "grad_norm": 0.8839809949496011, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 28717 + }, + { + "epoch": 0.28718, + "grad_norm": 0.8517469469202673, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 28718 + }, + { + "epoch": 0.28719, + "grad_norm": 0.8183908809326157, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 28719 + }, + { + "epoch": 0.2872, + "grad_norm": 0.8451916277568393, + "learning_rate": 0.003, + "loss": 3.995, + "step": 28720 + }, + { + "epoch": 0.28721, + "grad_norm": 0.7008498555944116, + "learning_rate": 0.003, + "loss": 4.01, + "step": 28721 + }, + { + "epoch": 0.28722, + "grad_norm": 0.5583582511714491, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 28722 + }, + { + "epoch": 0.28723, + "grad_norm": 0.6037566364523945, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 28723 + }, + { + "epoch": 0.28724, + "grad_norm": 0.5214579335403081, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 28724 + }, + { + "epoch": 0.28725, + "grad_norm": 0.5193097207786129, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 28725 + }, + { + "epoch": 0.28726, + "grad_norm": 0.5555030221570131, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 28726 + }, + { + "epoch": 0.28727, + "grad_norm": 0.5885054881019711, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 28727 + }, + { + "epoch": 0.28728, + "grad_norm": 0.655387671408372, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 28728 + }, + { + "epoch": 0.28729, + "grad_norm": 0.7183886393406946, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 28729 + }, + { + "epoch": 0.2873, + "grad_norm": 0.8728298230285202, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 28730 + }, + { + "epoch": 0.28731, + "grad_norm": 1.0245675412754043, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 28731 + }, + { + "epoch": 0.28732, + "grad_norm": 0.8948557238933322, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 28732 + }, + { + "epoch": 0.28733, + "grad_norm": 0.7757602416723336, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 28733 + }, + { + "epoch": 0.28734, + "grad_norm": 0.747465709572551, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 28734 + }, + { + "epoch": 0.28735, + "grad_norm": 0.6424629770979013, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 28735 + }, + { + "epoch": 0.28736, + "grad_norm": 0.5836742396901898, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 28736 + }, + { + "epoch": 0.28737, + "grad_norm": 0.5807755009915383, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 28737 + }, + { + "epoch": 0.28738, + "grad_norm": 0.6643849284762878, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 28738 + }, + { + "epoch": 0.28739, + "grad_norm": 0.655221452162131, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 28739 + }, + { + "epoch": 0.2874, + "grad_norm": 0.6403084124172433, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 28740 + }, + { + "epoch": 0.28741, + "grad_norm": 0.5953020481663295, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 28741 + }, + { + "epoch": 0.28742, + "grad_norm": 0.5944437581865059, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 28742 + }, + { + "epoch": 0.28743, + "grad_norm": 0.5671819815520974, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 28743 + }, + { + "epoch": 0.28744, + "grad_norm": 0.4892877967133075, + "learning_rate": 0.003, + "loss": 3.9334, + "step": 28744 + }, + { + "epoch": 0.28745, + "grad_norm": 0.6023981489358661, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 28745 + }, + { + "epoch": 0.28746, + "grad_norm": 0.5788903958612148, + "learning_rate": 0.003, + "loss": 3.969, + "step": 28746 + }, + { + "epoch": 0.28747, + "grad_norm": 0.5882385754929698, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 28747 + }, + { + "epoch": 0.28748, + "grad_norm": 0.6604635952765031, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 28748 + }, + { + "epoch": 0.28749, + "grad_norm": 0.6907610525711467, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 28749 + }, + { + "epoch": 0.2875, + "grad_norm": 0.6633194736835004, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 28750 + }, + { + "epoch": 0.28751, + "grad_norm": 0.7297141369119897, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 28751 + }, + { + "epoch": 0.28752, + "grad_norm": 0.8821911766269082, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 28752 + }, + { + "epoch": 0.28753, + "grad_norm": 0.9218718071121683, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 28753 + }, + { + "epoch": 0.28754, + "grad_norm": 0.8669294125307733, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 28754 + }, + { + "epoch": 0.28755, + "grad_norm": 0.8006230347372199, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 28755 + }, + { + "epoch": 0.28756, + "grad_norm": 0.8508188620598965, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 28756 + }, + { + "epoch": 0.28757, + "grad_norm": 0.7081693742840817, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 28757 + }, + { + "epoch": 0.28758, + "grad_norm": 0.6760611545899491, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 28758 + }, + { + "epoch": 0.28759, + "grad_norm": 0.6427564452675797, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 28759 + }, + { + "epoch": 0.2876, + "grad_norm": 0.6807158848232217, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 28760 + }, + { + "epoch": 0.28761, + "grad_norm": 0.7886502321676498, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 28761 + }, + { + "epoch": 0.28762, + "grad_norm": 0.8641390129283947, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 28762 + }, + { + "epoch": 0.28763, + "grad_norm": 0.8876554919841523, + "learning_rate": 0.003, + "loss": 3.995, + "step": 28763 + }, + { + "epoch": 0.28764, + "grad_norm": 0.8540561483315106, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 28764 + }, + { + "epoch": 0.28765, + "grad_norm": 1.059017116792444, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 28765 + }, + { + "epoch": 0.28766, + "grad_norm": 1.1183631441761557, + "learning_rate": 0.003, + "loss": 4.0488, + "step": 28766 + }, + { + "epoch": 0.28767, + "grad_norm": 0.8282363152736308, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 28767 + }, + { + "epoch": 0.28768, + "grad_norm": 0.7982700705252146, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 28768 + }, + { + "epoch": 0.28769, + "grad_norm": 0.7367923467160957, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 28769 + }, + { + "epoch": 0.2877, + "grad_norm": 0.8486855298299855, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 28770 + }, + { + "epoch": 0.28771, + "grad_norm": 0.966235309244078, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 28771 + }, + { + "epoch": 0.28772, + "grad_norm": 1.0797547670001626, + "learning_rate": 0.003, + "loss": 4.024, + "step": 28772 + }, + { + "epoch": 0.28773, + "grad_norm": 1.0268324659808883, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 28773 + }, + { + "epoch": 0.28774, + "grad_norm": 0.976983536354853, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 28774 + }, + { + "epoch": 0.28775, + "grad_norm": 0.7806202596389413, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 28775 + }, + { + "epoch": 0.28776, + "grad_norm": 0.7669893736402347, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 28776 + }, + { + "epoch": 0.28777, + "grad_norm": 0.811560804544977, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 28777 + }, + { + "epoch": 0.28778, + "grad_norm": 0.7933160544392452, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 28778 + }, + { + "epoch": 0.28779, + "grad_norm": 0.7762390319824016, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 28779 + }, + { + "epoch": 0.2878, + "grad_norm": 0.7113277974578087, + "learning_rate": 0.003, + "loss": 3.988, + "step": 28780 + }, + { + "epoch": 0.28781, + "grad_norm": 0.7591470845026849, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 28781 + }, + { + "epoch": 0.28782, + "grad_norm": 0.8374751295451974, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 28782 + }, + { + "epoch": 0.28783, + "grad_norm": 0.8899705558459082, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 28783 + }, + { + "epoch": 0.28784, + "grad_norm": 0.9204692647180323, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 28784 + }, + { + "epoch": 0.28785, + "grad_norm": 0.8019215970770237, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 28785 + }, + { + "epoch": 0.28786, + "grad_norm": 0.6867983497382394, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 28786 + }, + { + "epoch": 0.28787, + "grad_norm": 0.7082671383521263, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 28787 + }, + { + "epoch": 0.28788, + "grad_norm": 0.5788615784778828, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 28788 + }, + { + "epoch": 0.28789, + "grad_norm": 0.5997403709589947, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 28789 + }, + { + "epoch": 0.2879, + "grad_norm": 0.6188058677046814, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 28790 + }, + { + "epoch": 0.28791, + "grad_norm": 0.5960225101953891, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 28791 + }, + { + "epoch": 0.28792, + "grad_norm": 0.6078743698663117, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 28792 + }, + { + "epoch": 0.28793, + "grad_norm": 0.6298605289682171, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 28793 + }, + { + "epoch": 0.28794, + "grad_norm": 0.6336737381493822, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 28794 + }, + { + "epoch": 0.28795, + "grad_norm": 0.678733922027492, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 28795 + }, + { + "epoch": 0.28796, + "grad_norm": 0.7033841685134298, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 28796 + }, + { + "epoch": 0.28797, + "grad_norm": 0.7680806461711276, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 28797 + }, + { + "epoch": 0.28798, + "grad_norm": 0.7805474102026234, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 28798 + }, + { + "epoch": 0.28799, + "grad_norm": 0.6980003399421475, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 28799 + }, + { + "epoch": 0.288, + "grad_norm": 0.5847489524602746, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 28800 + }, + { + "epoch": 0.28801, + "grad_norm": 0.6448866167771963, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 28801 + }, + { + "epoch": 0.28802, + "grad_norm": 0.6833364332162286, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 28802 + }, + { + "epoch": 0.28803, + "grad_norm": 0.7086872124543757, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 28803 + }, + { + "epoch": 0.28804, + "grad_norm": 0.7436934004358247, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 28804 + }, + { + "epoch": 0.28805, + "grad_norm": 0.8503644567721197, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 28805 + }, + { + "epoch": 0.28806, + "grad_norm": 0.9757735957333539, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 28806 + }, + { + "epoch": 0.28807, + "grad_norm": 1.1537591977707358, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 28807 + }, + { + "epoch": 0.28808, + "grad_norm": 0.8380767065527954, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 28808 + }, + { + "epoch": 0.28809, + "grad_norm": 0.8053722453442826, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 28809 + }, + { + "epoch": 0.2881, + "grad_norm": 0.8077082361822924, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 28810 + }, + { + "epoch": 0.28811, + "grad_norm": 0.8839525280353095, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 28811 + }, + { + "epoch": 0.28812, + "grad_norm": 0.8818248013044384, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 28812 + }, + { + "epoch": 0.28813, + "grad_norm": 0.8769533836929929, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 28813 + }, + { + "epoch": 0.28814, + "grad_norm": 0.8255260014291517, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 28814 + }, + { + "epoch": 0.28815, + "grad_norm": 0.8462530270686036, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 28815 + }, + { + "epoch": 0.28816, + "grad_norm": 0.7821386832700968, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 28816 + }, + { + "epoch": 0.28817, + "grad_norm": 0.759801687608987, + "learning_rate": 0.003, + "loss": 3.994, + "step": 28817 + }, + { + "epoch": 0.28818, + "grad_norm": 0.6380981828562937, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 28818 + }, + { + "epoch": 0.28819, + "grad_norm": 0.6771864552708032, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 28819 + }, + { + "epoch": 0.2882, + "grad_norm": 0.7494792449193802, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 28820 + }, + { + "epoch": 0.28821, + "grad_norm": 0.7533398150493125, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 28821 + }, + { + "epoch": 0.28822, + "grad_norm": 0.8038308827779231, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 28822 + }, + { + "epoch": 0.28823, + "grad_norm": 0.8656496983728752, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 28823 + }, + { + "epoch": 0.28824, + "grad_norm": 0.9100555272995714, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 28824 + }, + { + "epoch": 0.28825, + "grad_norm": 0.9173526788346051, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 28825 + }, + { + "epoch": 0.28826, + "grad_norm": 0.9249360258760644, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 28826 + }, + { + "epoch": 0.28827, + "grad_norm": 0.9457136776621945, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 28827 + }, + { + "epoch": 0.28828, + "grad_norm": 0.8232223661732548, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 28828 + }, + { + "epoch": 0.28829, + "grad_norm": 0.8178424983707473, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 28829 + }, + { + "epoch": 0.2883, + "grad_norm": 0.7334779313394126, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 28830 + }, + { + "epoch": 0.28831, + "grad_norm": 0.7836083307181675, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 28831 + }, + { + "epoch": 0.28832, + "grad_norm": 0.7995859667180076, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 28832 + }, + { + "epoch": 0.28833, + "grad_norm": 0.8225929649455891, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 28833 + }, + { + "epoch": 0.28834, + "grad_norm": 0.8121641591530454, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 28834 + }, + { + "epoch": 0.28835, + "grad_norm": 0.6364323106678802, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 28835 + }, + { + "epoch": 0.28836, + "grad_norm": 0.6287411994701955, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 28836 + }, + { + "epoch": 0.28837, + "grad_norm": 0.6342144526686979, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 28837 + }, + { + "epoch": 0.28838, + "grad_norm": 0.6182885166418327, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 28838 + }, + { + "epoch": 0.28839, + "grad_norm": 0.6529187019538201, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 28839 + }, + { + "epoch": 0.2884, + "grad_norm": 0.5936580750420638, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 28840 + }, + { + "epoch": 0.28841, + "grad_norm": 0.6341250803856746, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 28841 + }, + { + "epoch": 0.28842, + "grad_norm": 0.6357086755496728, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 28842 + }, + { + "epoch": 0.28843, + "grad_norm": 0.7358312194385965, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 28843 + }, + { + "epoch": 0.28844, + "grad_norm": 0.8576394596378868, + "learning_rate": 0.003, + "loss": 3.993, + "step": 28844 + }, + { + "epoch": 0.28845, + "grad_norm": 0.9302637641597773, + "learning_rate": 0.003, + "loss": 4.006, + "step": 28845 + }, + { + "epoch": 0.28846, + "grad_norm": 0.8630662842594814, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 28846 + }, + { + "epoch": 0.28847, + "grad_norm": 0.8074643862300122, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 28847 + }, + { + "epoch": 0.28848, + "grad_norm": 0.7803662254830535, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 28848 + }, + { + "epoch": 0.28849, + "grad_norm": 0.8776112068783575, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 28849 + }, + { + "epoch": 0.2885, + "grad_norm": 0.8993023179839624, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 28850 + }, + { + "epoch": 0.28851, + "grad_norm": 0.9407780273556812, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 28851 + }, + { + "epoch": 0.28852, + "grad_norm": 1.0052272515396816, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 28852 + }, + { + "epoch": 0.28853, + "grad_norm": 0.9898209791675259, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 28853 + }, + { + "epoch": 0.28854, + "grad_norm": 0.9094585398595642, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 28854 + }, + { + "epoch": 0.28855, + "grad_norm": 0.8493926035402978, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 28855 + }, + { + "epoch": 0.28856, + "grad_norm": 0.7840377721899097, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 28856 + }, + { + "epoch": 0.28857, + "grad_norm": 0.8126229926352431, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 28857 + }, + { + "epoch": 0.28858, + "grad_norm": 0.8384277263831825, + "learning_rate": 0.003, + "loss": 3.996, + "step": 28858 + }, + { + "epoch": 0.28859, + "grad_norm": 0.9482386450637119, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 28859 + }, + { + "epoch": 0.2886, + "grad_norm": 0.8412373601446361, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 28860 + }, + { + "epoch": 0.28861, + "grad_norm": 0.7577812250158196, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 28861 + }, + { + "epoch": 0.28862, + "grad_norm": 0.7308332475828689, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 28862 + }, + { + "epoch": 0.28863, + "grad_norm": 0.6831861942787278, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 28863 + }, + { + "epoch": 0.28864, + "grad_norm": 0.6611956181971473, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 28864 + }, + { + "epoch": 0.28865, + "grad_norm": 0.7033592970540502, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 28865 + }, + { + "epoch": 0.28866, + "grad_norm": 0.8804695309274176, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 28866 + }, + { + "epoch": 0.28867, + "grad_norm": 1.0373692587661065, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 28867 + }, + { + "epoch": 0.28868, + "grad_norm": 0.8242444671016238, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 28868 + }, + { + "epoch": 0.28869, + "grad_norm": 0.6721049619458732, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 28869 + }, + { + "epoch": 0.2887, + "grad_norm": 0.6371955262692464, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 28870 + }, + { + "epoch": 0.28871, + "grad_norm": 0.7208977852823251, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 28871 + }, + { + "epoch": 0.28872, + "grad_norm": 0.6309777122548029, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 28872 + }, + { + "epoch": 0.28873, + "grad_norm": 0.5437800303269015, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 28873 + }, + { + "epoch": 0.28874, + "grad_norm": 0.46233104526513535, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 28874 + }, + { + "epoch": 0.28875, + "grad_norm": 0.5359097840244346, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 28875 + }, + { + "epoch": 0.28876, + "grad_norm": 0.5913758434826768, + "learning_rate": 0.003, + "loss": 3.993, + "step": 28876 + }, + { + "epoch": 0.28877, + "grad_norm": 0.6993082908101579, + "learning_rate": 0.003, + "loss": 3.992, + "step": 28877 + }, + { + "epoch": 0.28878, + "grad_norm": 0.6550317051928308, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 28878 + }, + { + "epoch": 0.28879, + "grad_norm": 0.6592511642348952, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 28879 + }, + { + "epoch": 0.2888, + "grad_norm": 0.6900658011881659, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 28880 + }, + { + "epoch": 0.28881, + "grad_norm": 0.7023349249984456, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 28881 + }, + { + "epoch": 0.28882, + "grad_norm": 0.7415092665856214, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 28882 + }, + { + "epoch": 0.28883, + "grad_norm": 0.5674633796027246, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 28883 + }, + { + "epoch": 0.28884, + "grad_norm": 0.5947523847643436, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 28884 + }, + { + "epoch": 0.28885, + "grad_norm": 0.6953415629374083, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 28885 + }, + { + "epoch": 0.28886, + "grad_norm": 0.7205060656776175, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 28886 + }, + { + "epoch": 0.28887, + "grad_norm": 0.719119206789323, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 28887 + }, + { + "epoch": 0.28888, + "grad_norm": 0.6726117706680739, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 28888 + }, + { + "epoch": 0.28889, + "grad_norm": 0.651137898161609, + "learning_rate": 0.003, + "loss": 3.993, + "step": 28889 + }, + { + "epoch": 0.2889, + "grad_norm": 0.525112252448548, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 28890 + }, + { + "epoch": 0.28891, + "grad_norm": 0.47810847228447434, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 28891 + }, + { + "epoch": 0.28892, + "grad_norm": 0.6071896881900148, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 28892 + }, + { + "epoch": 0.28893, + "grad_norm": 0.8115663618726393, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 28893 + }, + { + "epoch": 0.28894, + "grad_norm": 0.9913529657798992, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 28894 + }, + { + "epoch": 0.28895, + "grad_norm": 1.0844767964481188, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 28895 + }, + { + "epoch": 0.28896, + "grad_norm": 0.7846763676822288, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 28896 + }, + { + "epoch": 0.28897, + "grad_norm": 0.8427637809968884, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 28897 + }, + { + "epoch": 0.28898, + "grad_norm": 0.913134996653207, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 28898 + }, + { + "epoch": 0.28899, + "grad_norm": 0.8572038644885641, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 28899 + }, + { + "epoch": 0.289, + "grad_norm": 0.9195789447180325, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 28900 + }, + { + "epoch": 0.28901, + "grad_norm": 1.096826347535638, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 28901 + }, + { + "epoch": 0.28902, + "grad_norm": 0.922316050697249, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 28902 + }, + { + "epoch": 0.28903, + "grad_norm": 0.9902610528580529, + "learning_rate": 0.003, + "loss": 4.036, + "step": 28903 + }, + { + "epoch": 0.28904, + "grad_norm": 0.9928917735913816, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 28904 + }, + { + "epoch": 0.28905, + "grad_norm": 0.7894015491173564, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 28905 + }, + { + "epoch": 0.28906, + "grad_norm": 0.6729032444759708, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 28906 + }, + { + "epoch": 0.28907, + "grad_norm": 0.7038116826692382, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 28907 + }, + { + "epoch": 0.28908, + "grad_norm": 0.7666599294997288, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 28908 + }, + { + "epoch": 0.28909, + "grad_norm": 0.7667178938997856, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 28909 + }, + { + "epoch": 0.2891, + "grad_norm": 0.7094792963386013, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 28910 + }, + { + "epoch": 0.28911, + "grad_norm": 0.701917180739602, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 28911 + }, + { + "epoch": 0.28912, + "grad_norm": 0.7696762106289583, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 28912 + }, + { + "epoch": 0.28913, + "grad_norm": 0.8103822115317261, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 28913 + }, + { + "epoch": 0.28914, + "grad_norm": 0.7369525052754885, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 28914 + }, + { + "epoch": 0.28915, + "grad_norm": 0.7837605972818484, + "learning_rate": 0.003, + "loss": 4.011, + "step": 28915 + }, + { + "epoch": 0.28916, + "grad_norm": 0.7789977306228713, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 28916 + }, + { + "epoch": 0.28917, + "grad_norm": 0.780525552333569, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 28917 + }, + { + "epoch": 0.28918, + "grad_norm": 0.8257503987581821, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 28918 + }, + { + "epoch": 0.28919, + "grad_norm": 0.8662010768736224, + "learning_rate": 0.003, + "loss": 3.974, + "step": 28919 + }, + { + "epoch": 0.2892, + "grad_norm": 0.8101526347664642, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 28920 + }, + { + "epoch": 0.28921, + "grad_norm": 0.7738672440293899, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 28921 + }, + { + "epoch": 0.28922, + "grad_norm": 0.7655308684407215, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 28922 + }, + { + "epoch": 0.28923, + "grad_norm": 0.7760949960334174, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 28923 + }, + { + "epoch": 0.28924, + "grad_norm": 0.7085421443694349, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 28924 + }, + { + "epoch": 0.28925, + "grad_norm": 0.7910513439528407, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 28925 + }, + { + "epoch": 0.28926, + "grad_norm": 0.8302276002555681, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 28926 + }, + { + "epoch": 0.28927, + "grad_norm": 0.7331051402146737, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 28927 + }, + { + "epoch": 0.28928, + "grad_norm": 0.741538018602652, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 28928 + }, + { + "epoch": 0.28929, + "grad_norm": 0.803249513267659, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 28929 + }, + { + "epoch": 0.2893, + "grad_norm": 0.7127884825297143, + "learning_rate": 0.003, + "loss": 3.967, + "step": 28930 + }, + { + "epoch": 0.28931, + "grad_norm": 0.5816880786260653, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 28931 + }, + { + "epoch": 0.28932, + "grad_norm": 0.5395332452078966, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 28932 + }, + { + "epoch": 0.28933, + "grad_norm": 0.5799149102005418, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 28933 + }, + { + "epoch": 0.28934, + "grad_norm": 0.5516474781519948, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 28934 + }, + { + "epoch": 0.28935, + "grad_norm": 0.5693618513387377, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 28935 + }, + { + "epoch": 0.28936, + "grad_norm": 0.623632025204069, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 28936 + }, + { + "epoch": 0.28937, + "grad_norm": 0.7099621381845325, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 28937 + }, + { + "epoch": 0.28938, + "grad_norm": 0.8019275909567384, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 28938 + }, + { + "epoch": 0.28939, + "grad_norm": 0.9001925796743125, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 28939 + }, + { + "epoch": 0.2894, + "grad_norm": 0.9944186170556674, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 28940 + }, + { + "epoch": 0.28941, + "grad_norm": 0.963734940111436, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 28941 + }, + { + "epoch": 0.28942, + "grad_norm": 0.9866172027005317, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 28942 + }, + { + "epoch": 0.28943, + "grad_norm": 0.8445994458369273, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 28943 + }, + { + "epoch": 0.28944, + "grad_norm": 0.7600183813925717, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 28944 + }, + { + "epoch": 0.28945, + "grad_norm": 0.7376334776607343, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 28945 + }, + { + "epoch": 0.28946, + "grad_norm": 0.7415865889046022, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 28946 + }, + { + "epoch": 0.28947, + "grad_norm": 0.7187055876027814, + "learning_rate": 0.003, + "loss": 3.975, + "step": 28947 + }, + { + "epoch": 0.28948, + "grad_norm": 0.697154031660856, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 28948 + }, + { + "epoch": 0.28949, + "grad_norm": 0.7084675477355067, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 28949 + }, + { + "epoch": 0.2895, + "grad_norm": 0.7030802971954061, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 28950 + }, + { + "epoch": 0.28951, + "grad_norm": 0.6839656662002784, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 28951 + }, + { + "epoch": 0.28952, + "grad_norm": 0.686234475261678, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 28952 + }, + { + "epoch": 0.28953, + "grad_norm": 0.779460210283587, + "learning_rate": 0.003, + "loss": 3.977, + "step": 28953 + }, + { + "epoch": 0.28954, + "grad_norm": 0.6935893648249049, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 28954 + }, + { + "epoch": 0.28955, + "grad_norm": 0.7239906925344279, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 28955 + }, + { + "epoch": 0.28956, + "grad_norm": 0.7655467831485101, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 28956 + }, + { + "epoch": 0.28957, + "grad_norm": 0.7796625549793937, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 28957 + }, + { + "epoch": 0.28958, + "grad_norm": 0.9419422840925985, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 28958 + }, + { + "epoch": 0.28959, + "grad_norm": 1.0746977860796094, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 28959 + }, + { + "epoch": 0.2896, + "grad_norm": 0.9638028190358495, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 28960 + }, + { + "epoch": 0.28961, + "grad_norm": 0.9701128954611202, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 28961 + }, + { + "epoch": 0.28962, + "grad_norm": 0.9306833705826599, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 28962 + }, + { + "epoch": 0.28963, + "grad_norm": 0.8035585175168991, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 28963 + }, + { + "epoch": 0.28964, + "grad_norm": 0.7630686981488338, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 28964 + }, + { + "epoch": 0.28965, + "grad_norm": 0.7370275123885407, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 28965 + }, + { + "epoch": 0.28966, + "grad_norm": 0.7777499758533084, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 28966 + }, + { + "epoch": 0.28967, + "grad_norm": 0.9484382693672839, + "learning_rate": 0.003, + "loss": 4.006, + "step": 28967 + }, + { + "epoch": 0.28968, + "grad_norm": 1.1241390778311844, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 28968 + }, + { + "epoch": 0.28969, + "grad_norm": 0.8684303989409903, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 28969 + }, + { + "epoch": 0.2897, + "grad_norm": 0.7799296255337896, + "learning_rate": 0.003, + "loss": 3.974, + "step": 28970 + }, + { + "epoch": 0.28971, + "grad_norm": 0.6114078481752275, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 28971 + }, + { + "epoch": 0.28972, + "grad_norm": 0.5852222844495506, + "learning_rate": 0.003, + "loss": 3.993, + "step": 28972 + }, + { + "epoch": 0.28973, + "grad_norm": 0.5612615557879295, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 28973 + }, + { + "epoch": 0.28974, + "grad_norm": 0.4956551823190419, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 28974 + }, + { + "epoch": 0.28975, + "grad_norm": 0.6165402704836782, + "learning_rate": 0.003, + "loss": 3.981, + "step": 28975 + }, + { + "epoch": 0.28976, + "grad_norm": 0.6681269584421324, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 28976 + }, + { + "epoch": 0.28977, + "grad_norm": 0.7297245867224068, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 28977 + }, + { + "epoch": 0.28978, + "grad_norm": 0.8421150109680015, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 28978 + }, + { + "epoch": 0.28979, + "grad_norm": 0.9777071805382231, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 28979 + }, + { + "epoch": 0.2898, + "grad_norm": 0.8988663455022907, + "learning_rate": 0.003, + "loss": 3.971, + "step": 28980 + }, + { + "epoch": 0.28981, + "grad_norm": 0.7729295865979282, + "learning_rate": 0.003, + "loss": 3.991, + "step": 28981 + }, + { + "epoch": 0.28982, + "grad_norm": 0.8359029453684749, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 28982 + }, + { + "epoch": 0.28983, + "grad_norm": 0.9325358855772021, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 28983 + }, + { + "epoch": 0.28984, + "grad_norm": 0.7874775530684137, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 28984 + }, + { + "epoch": 0.28985, + "grad_norm": 0.6718369168130991, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 28985 + }, + { + "epoch": 0.28986, + "grad_norm": 0.5334432391093924, + "learning_rate": 0.003, + "loss": 3.995, + "step": 28986 + }, + { + "epoch": 0.28987, + "grad_norm": 0.5797416309659227, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 28987 + }, + { + "epoch": 0.28988, + "grad_norm": 0.6971283735726758, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 28988 + }, + { + "epoch": 0.28989, + "grad_norm": 0.8019453742251933, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 28989 + }, + { + "epoch": 0.2899, + "grad_norm": 0.6664075365988784, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 28990 + }, + { + "epoch": 0.28991, + "grad_norm": 0.5683460389895647, + "learning_rate": 0.003, + "loss": 3.982, + "step": 28991 + }, + { + "epoch": 0.28992, + "grad_norm": 0.5498324261206263, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 28992 + }, + { + "epoch": 0.28993, + "grad_norm": 0.5437725312848026, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 28993 + }, + { + "epoch": 0.28994, + "grad_norm": 0.640631387443895, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 28994 + }, + { + "epoch": 0.28995, + "grad_norm": 0.738208676815241, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 28995 + }, + { + "epoch": 0.28996, + "grad_norm": 0.6906953324961179, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 28996 + }, + { + "epoch": 0.28997, + "grad_norm": 0.6226171890415986, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 28997 + }, + { + "epoch": 0.28998, + "grad_norm": 0.618481198708161, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 28998 + }, + { + "epoch": 0.28999, + "grad_norm": 0.5562480011765798, + "learning_rate": 0.003, + "loss": 4.003, + "step": 28999 + }, + { + "epoch": 0.29, + "grad_norm": 0.6095227136846973, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 29000 + }, + { + "epoch": 0.29001, + "grad_norm": 0.6150262859560228, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 29001 + }, + { + "epoch": 0.29002, + "grad_norm": 0.6827589472566742, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 29002 + }, + { + "epoch": 0.29003, + "grad_norm": 0.7469990910060453, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 29003 + }, + { + "epoch": 0.29004, + "grad_norm": 0.7329801829423754, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 29004 + }, + { + "epoch": 0.29005, + "grad_norm": 0.8098482030436697, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 29005 + }, + { + "epoch": 0.29006, + "grad_norm": 0.8042145811055601, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 29006 + }, + { + "epoch": 0.29007, + "grad_norm": 0.6670889938694611, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 29007 + }, + { + "epoch": 0.29008, + "grad_norm": 0.6478272895272812, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 29008 + }, + { + "epoch": 0.29009, + "grad_norm": 0.6673386565294221, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 29009 + }, + { + "epoch": 0.2901, + "grad_norm": 0.929451396188917, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 29010 + }, + { + "epoch": 0.29011, + "grad_norm": 1.283698997983296, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 29011 + }, + { + "epoch": 0.29012, + "grad_norm": 0.6766213800274731, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 29012 + }, + { + "epoch": 0.29013, + "grad_norm": 0.615473399632724, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 29013 + }, + { + "epoch": 0.29014, + "grad_norm": 0.7321457397361003, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 29014 + }, + { + "epoch": 0.29015, + "grad_norm": 0.7675205972504451, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 29015 + }, + { + "epoch": 0.29016, + "grad_norm": 0.7783648280867138, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 29016 + }, + { + "epoch": 0.29017, + "grad_norm": 0.8724408024732141, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 29017 + }, + { + "epoch": 0.29018, + "grad_norm": 1.1146665849453996, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 29018 + }, + { + "epoch": 0.29019, + "grad_norm": 0.8874101383961757, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 29019 + }, + { + "epoch": 0.2902, + "grad_norm": 0.8965233034725544, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29020 + }, + { + "epoch": 0.29021, + "grad_norm": 0.8969486403515544, + "learning_rate": 0.003, + "loss": 3.971, + "step": 29021 + }, + { + "epoch": 0.29022, + "grad_norm": 0.8509163474164182, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 29022 + }, + { + "epoch": 0.29023, + "grad_norm": 0.9585779142716769, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 29023 + }, + { + "epoch": 0.29024, + "grad_norm": 1.0347615744477119, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 29024 + }, + { + "epoch": 0.29025, + "grad_norm": 0.906475775440978, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 29025 + }, + { + "epoch": 0.29026, + "grad_norm": 0.8877059322855378, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 29026 + }, + { + "epoch": 0.29027, + "grad_norm": 1.0042724757839963, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 29027 + }, + { + "epoch": 0.29028, + "grad_norm": 1.14056987027801, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 29028 + }, + { + "epoch": 0.29029, + "grad_norm": 0.9425245979931934, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 29029 + }, + { + "epoch": 0.2903, + "grad_norm": 0.9310815147494657, + "learning_rate": 0.003, + "loss": 4.006, + "step": 29030 + }, + { + "epoch": 0.29031, + "grad_norm": 1.0235300201692512, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 29031 + }, + { + "epoch": 0.29032, + "grad_norm": 0.9951973786744017, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 29032 + }, + { + "epoch": 0.29033, + "grad_norm": 0.9549520347638851, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 29033 + }, + { + "epoch": 0.29034, + "grad_norm": 0.9750995621321259, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 29034 + }, + { + "epoch": 0.29035, + "grad_norm": 0.9860445265259772, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 29035 + }, + { + "epoch": 0.29036, + "grad_norm": 0.7468348851912704, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 29036 + }, + { + "epoch": 0.29037, + "grad_norm": 0.6398961616891066, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 29037 + }, + { + "epoch": 0.29038, + "grad_norm": 0.703853086848993, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 29038 + }, + { + "epoch": 0.29039, + "grad_norm": 0.8177456912510629, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 29039 + }, + { + "epoch": 0.2904, + "grad_norm": 0.837978859819443, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 29040 + }, + { + "epoch": 0.29041, + "grad_norm": 0.8208961356392038, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 29041 + }, + { + "epoch": 0.29042, + "grad_norm": 0.7228415329176576, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 29042 + }, + { + "epoch": 0.29043, + "grad_norm": 0.6840359375526551, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 29043 + }, + { + "epoch": 0.29044, + "grad_norm": 0.6880301762628598, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 29044 + }, + { + "epoch": 0.29045, + "grad_norm": 0.6234047425164094, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 29045 + }, + { + "epoch": 0.29046, + "grad_norm": 0.54381140931352, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 29046 + }, + { + "epoch": 0.29047, + "grad_norm": 0.5188486794033457, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 29047 + }, + { + "epoch": 0.29048, + "grad_norm": 0.5267693721202221, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 29048 + }, + { + "epoch": 0.29049, + "grad_norm": 0.5661718176937909, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 29049 + }, + { + "epoch": 0.2905, + "grad_norm": 0.5986091625649279, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 29050 + }, + { + "epoch": 0.29051, + "grad_norm": 0.5717841763671405, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 29051 + }, + { + "epoch": 0.29052, + "grad_norm": 0.6348461321654943, + "learning_rate": 0.003, + "loss": 3.995, + "step": 29052 + }, + { + "epoch": 0.29053, + "grad_norm": 0.864964582460035, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 29053 + }, + { + "epoch": 0.29054, + "grad_norm": 1.1197302821472068, + "learning_rate": 0.003, + "loss": 3.973, + "step": 29054 + }, + { + "epoch": 0.29055, + "grad_norm": 0.8451214615068529, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 29055 + }, + { + "epoch": 0.29056, + "grad_norm": 0.6199420747393054, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 29056 + }, + { + "epoch": 0.29057, + "grad_norm": 0.7290144307690649, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 29057 + }, + { + "epoch": 0.29058, + "grad_norm": 0.8180613450642034, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 29058 + }, + { + "epoch": 0.29059, + "grad_norm": 0.8259709183902504, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 29059 + }, + { + "epoch": 0.2906, + "grad_norm": 0.7247999573140556, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 29060 + }, + { + "epoch": 0.29061, + "grad_norm": 0.6649200342317862, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 29061 + }, + { + "epoch": 0.29062, + "grad_norm": 0.6857709470201275, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 29062 + }, + { + "epoch": 0.29063, + "grad_norm": 0.5795757395350442, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 29063 + }, + { + "epoch": 0.29064, + "grad_norm": 0.5398272185828156, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 29064 + }, + { + "epoch": 0.29065, + "grad_norm": 0.5755466246682691, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 29065 + }, + { + "epoch": 0.29066, + "grad_norm": 0.6768774822003555, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 29066 + }, + { + "epoch": 0.29067, + "grad_norm": 0.7476446538330686, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 29067 + }, + { + "epoch": 0.29068, + "grad_norm": 0.8016445434445023, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 29068 + }, + { + "epoch": 0.29069, + "grad_norm": 0.7171230923954156, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 29069 + }, + { + "epoch": 0.2907, + "grad_norm": 0.6757078269838416, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 29070 + }, + { + "epoch": 0.29071, + "grad_norm": 0.6890922908961596, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 29071 + }, + { + "epoch": 0.29072, + "grad_norm": 0.7706041218502686, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 29072 + }, + { + "epoch": 0.29073, + "grad_norm": 0.83131762208579, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 29073 + }, + { + "epoch": 0.29074, + "grad_norm": 0.939202674233422, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 29074 + }, + { + "epoch": 0.29075, + "grad_norm": 1.0258755296072568, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 29075 + }, + { + "epoch": 0.29076, + "grad_norm": 1.0440090071916712, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 29076 + }, + { + "epoch": 0.29077, + "grad_norm": 0.9331941167670125, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 29077 + }, + { + "epoch": 0.29078, + "grad_norm": 0.8086725062898679, + "learning_rate": 0.003, + "loss": 3.975, + "step": 29078 + }, + { + "epoch": 0.29079, + "grad_norm": 0.8962445839934151, + "learning_rate": 0.003, + "loss": 4.015, + "step": 29079 + }, + { + "epoch": 0.2908, + "grad_norm": 0.8831532940861592, + "learning_rate": 0.003, + "loss": 4.032, + "step": 29080 + }, + { + "epoch": 0.29081, + "grad_norm": 0.9229469247687192, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 29081 + }, + { + "epoch": 0.29082, + "grad_norm": 0.9771469288256475, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 29082 + }, + { + "epoch": 0.29083, + "grad_norm": 0.9274521876741013, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 29083 + }, + { + "epoch": 0.29084, + "grad_norm": 0.9243905294203499, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 29084 + }, + { + "epoch": 0.29085, + "grad_norm": 0.9519308530647652, + "learning_rate": 0.003, + "loss": 4.003, + "step": 29085 + }, + { + "epoch": 0.29086, + "grad_norm": 0.9843984233323052, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 29086 + }, + { + "epoch": 0.29087, + "grad_norm": 1.0671165176787387, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 29087 + }, + { + "epoch": 0.29088, + "grad_norm": 0.9906769607621855, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 29088 + }, + { + "epoch": 0.29089, + "grad_norm": 1.0286847989766836, + "learning_rate": 0.003, + "loss": 4.0509, + "step": 29089 + }, + { + "epoch": 0.2909, + "grad_norm": 0.8736085951220042, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 29090 + }, + { + "epoch": 0.29091, + "grad_norm": 0.8436458056237586, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 29091 + }, + { + "epoch": 0.29092, + "grad_norm": 0.8442363505863237, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 29092 + }, + { + "epoch": 0.29093, + "grad_norm": 1.0857893341591494, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 29093 + }, + { + "epoch": 0.29094, + "grad_norm": 1.0322030125712487, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 29094 + }, + { + "epoch": 0.29095, + "grad_norm": 0.9100219144894679, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 29095 + }, + { + "epoch": 0.29096, + "grad_norm": 0.8256429771186227, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 29096 + }, + { + "epoch": 0.29097, + "grad_norm": 0.8277693808001977, + "learning_rate": 0.003, + "loss": 4.0451, + "step": 29097 + }, + { + "epoch": 0.29098, + "grad_norm": 0.6654563370037714, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 29098 + }, + { + "epoch": 0.29099, + "grad_norm": 0.6311168301548642, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 29099 + }, + { + "epoch": 0.291, + "grad_norm": 0.6192334455316543, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 29100 + }, + { + "epoch": 0.29101, + "grad_norm": 0.6124916803210945, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 29101 + }, + { + "epoch": 0.29102, + "grad_norm": 0.5815339779800843, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 29102 + }, + { + "epoch": 0.29103, + "grad_norm": 0.529792722912882, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 29103 + }, + { + "epoch": 0.29104, + "grad_norm": 0.5253041220424238, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 29104 + }, + { + "epoch": 0.29105, + "grad_norm": 0.47978520304939193, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 29105 + }, + { + "epoch": 0.29106, + "grad_norm": 0.44465509729960295, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 29106 + }, + { + "epoch": 0.29107, + "grad_norm": 0.47801384333904207, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 29107 + }, + { + "epoch": 0.29108, + "grad_norm": 0.49019806880022443, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 29108 + }, + { + "epoch": 0.29109, + "grad_norm": 0.49610275169305335, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 29109 + }, + { + "epoch": 0.2911, + "grad_norm": 0.6636204555527928, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 29110 + }, + { + "epoch": 0.29111, + "grad_norm": 0.8474322777748091, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 29111 + }, + { + "epoch": 0.29112, + "grad_norm": 1.0459258926910142, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 29112 + }, + { + "epoch": 0.29113, + "grad_norm": 1.0050975139325078, + "learning_rate": 0.003, + "loss": 3.96, + "step": 29113 + }, + { + "epoch": 0.29114, + "grad_norm": 0.7371200219146781, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 29114 + }, + { + "epoch": 0.29115, + "grad_norm": 0.5953364765321901, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 29115 + }, + { + "epoch": 0.29116, + "grad_norm": 0.6199852727313893, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 29116 + }, + { + "epoch": 0.29117, + "grad_norm": 0.666349798046929, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 29117 + }, + { + "epoch": 0.29118, + "grad_norm": 0.6526055117245178, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 29118 + }, + { + "epoch": 0.29119, + "grad_norm": 0.6914677707683174, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 29119 + }, + { + "epoch": 0.2912, + "grad_norm": 0.7089138070934429, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 29120 + }, + { + "epoch": 0.29121, + "grad_norm": 0.6540976489391765, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 29121 + }, + { + "epoch": 0.29122, + "grad_norm": 0.6945885430348223, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 29122 + }, + { + "epoch": 0.29123, + "grad_norm": 0.66355954555136, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 29123 + }, + { + "epoch": 0.29124, + "grad_norm": 0.637203393735581, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 29124 + }, + { + "epoch": 0.29125, + "grad_norm": 0.583662252982062, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 29125 + }, + { + "epoch": 0.29126, + "grad_norm": 0.5697025121253659, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 29126 + }, + { + "epoch": 0.29127, + "grad_norm": 0.582271123086911, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 29127 + }, + { + "epoch": 0.29128, + "grad_norm": 0.6209953342596568, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 29128 + }, + { + "epoch": 0.29129, + "grad_norm": 0.6019497455976438, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 29129 + }, + { + "epoch": 0.2913, + "grad_norm": 0.6667779541952022, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 29130 + }, + { + "epoch": 0.29131, + "grad_norm": 0.7710649339492551, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 29131 + }, + { + "epoch": 0.29132, + "grad_norm": 0.786711458981371, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 29132 + }, + { + "epoch": 0.29133, + "grad_norm": 0.8922272095749912, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 29133 + }, + { + "epoch": 0.29134, + "grad_norm": 1.0451762673789893, + "learning_rate": 0.003, + "loss": 3.974, + "step": 29134 + }, + { + "epoch": 0.29135, + "grad_norm": 1.070648995262492, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 29135 + }, + { + "epoch": 0.29136, + "grad_norm": 0.9800730196202062, + "learning_rate": 0.003, + "loss": 4.015, + "step": 29136 + }, + { + "epoch": 0.29137, + "grad_norm": 0.8883995421357561, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 29137 + }, + { + "epoch": 0.29138, + "grad_norm": 0.8225485052630563, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 29138 + }, + { + "epoch": 0.29139, + "grad_norm": 0.9388755360012683, + "learning_rate": 0.003, + "loss": 4.009, + "step": 29139 + }, + { + "epoch": 0.2914, + "grad_norm": 1.0475191864632727, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 29140 + }, + { + "epoch": 0.29141, + "grad_norm": 0.9745512395001695, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 29141 + }, + { + "epoch": 0.29142, + "grad_norm": 0.8928383537539698, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 29142 + }, + { + "epoch": 0.29143, + "grad_norm": 0.9246209859535098, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 29143 + }, + { + "epoch": 0.29144, + "grad_norm": 0.7564301705849273, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 29144 + }, + { + "epoch": 0.29145, + "grad_norm": 0.6500726569717896, + "learning_rate": 0.003, + "loss": 4.024, + "step": 29145 + }, + { + "epoch": 0.29146, + "grad_norm": 0.5690930500767499, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 29146 + }, + { + "epoch": 0.29147, + "grad_norm": 0.5641121272967531, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 29147 + }, + { + "epoch": 0.29148, + "grad_norm": 0.48023339229248174, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 29148 + }, + { + "epoch": 0.29149, + "grad_norm": 0.5021143021741842, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 29149 + }, + { + "epoch": 0.2915, + "grad_norm": 0.5108131410952206, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 29150 + }, + { + "epoch": 0.29151, + "grad_norm": 0.6411837304742906, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 29151 + }, + { + "epoch": 0.29152, + "grad_norm": 0.7872175981188828, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 29152 + }, + { + "epoch": 0.29153, + "grad_norm": 0.8072077011064899, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 29153 + }, + { + "epoch": 0.29154, + "grad_norm": 0.8364212147197133, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 29154 + }, + { + "epoch": 0.29155, + "grad_norm": 1.010517050399461, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 29155 + }, + { + "epoch": 0.29156, + "grad_norm": 0.8867830079450383, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 29156 + }, + { + "epoch": 0.29157, + "grad_norm": 0.6598078258610969, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 29157 + }, + { + "epoch": 0.29158, + "grad_norm": 0.5806821656028088, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 29158 + }, + { + "epoch": 0.29159, + "grad_norm": 0.6179097151667364, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 29159 + }, + { + "epoch": 0.2916, + "grad_norm": 0.61896126648342, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 29160 + }, + { + "epoch": 0.29161, + "grad_norm": 0.648686480638205, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 29161 + }, + { + "epoch": 0.29162, + "grad_norm": 0.6902280913657456, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 29162 + }, + { + "epoch": 0.29163, + "grad_norm": 0.744980595423205, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 29163 + }, + { + "epoch": 0.29164, + "grad_norm": 0.9048909014611434, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 29164 + }, + { + "epoch": 0.29165, + "grad_norm": 1.052318643127296, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 29165 + }, + { + "epoch": 0.29166, + "grad_norm": 0.9725736916039918, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 29166 + }, + { + "epoch": 0.29167, + "grad_norm": 1.1329023871245156, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 29167 + }, + { + "epoch": 0.29168, + "grad_norm": 0.8327009293107315, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 29168 + }, + { + "epoch": 0.29169, + "grad_norm": 0.7082360340443907, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 29169 + }, + { + "epoch": 0.2917, + "grad_norm": 0.7088977332207412, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 29170 + }, + { + "epoch": 0.29171, + "grad_norm": 0.738265862080483, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 29171 + }, + { + "epoch": 0.29172, + "grad_norm": 0.7848198398043105, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 29172 + }, + { + "epoch": 0.29173, + "grad_norm": 0.8327795460288404, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 29173 + }, + { + "epoch": 0.29174, + "grad_norm": 0.7278303798113588, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 29174 + }, + { + "epoch": 0.29175, + "grad_norm": 0.6717093487547027, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 29175 + }, + { + "epoch": 0.29176, + "grad_norm": 0.7582726604551611, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 29176 + }, + { + "epoch": 0.29177, + "grad_norm": 0.8547604981037369, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 29177 + }, + { + "epoch": 0.29178, + "grad_norm": 0.7880421841108747, + "learning_rate": 0.003, + "loss": 3.984, + "step": 29178 + }, + { + "epoch": 0.29179, + "grad_norm": 0.704717631254811, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 29179 + }, + { + "epoch": 0.2918, + "grad_norm": 0.6159429159019117, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 29180 + }, + { + "epoch": 0.29181, + "grad_norm": 0.5209220428132018, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 29181 + }, + { + "epoch": 0.29182, + "grad_norm": 0.487861485335546, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 29182 + }, + { + "epoch": 0.29183, + "grad_norm": 0.5590537107068817, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 29183 + }, + { + "epoch": 0.29184, + "grad_norm": 0.5415688025046631, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 29184 + }, + { + "epoch": 0.29185, + "grad_norm": 0.5608950170475457, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 29185 + }, + { + "epoch": 0.29186, + "grad_norm": 0.6405572219858611, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 29186 + }, + { + "epoch": 0.29187, + "grad_norm": 0.7374181620166151, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 29187 + }, + { + "epoch": 0.29188, + "grad_norm": 0.8526043861508683, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 29188 + }, + { + "epoch": 0.29189, + "grad_norm": 0.7756946750573237, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 29189 + }, + { + "epoch": 0.2919, + "grad_norm": 0.7288738863981304, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 29190 + }, + { + "epoch": 0.29191, + "grad_norm": 0.8250845586903196, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 29191 + }, + { + "epoch": 0.29192, + "grad_norm": 0.8895675889141333, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 29192 + }, + { + "epoch": 0.29193, + "grad_norm": 0.9131150020348504, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 29193 + }, + { + "epoch": 0.29194, + "grad_norm": 1.0411154329741796, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 29194 + }, + { + "epoch": 0.29195, + "grad_norm": 0.9677208546494961, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 29195 + }, + { + "epoch": 0.29196, + "grad_norm": 0.9322154246669071, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 29196 + }, + { + "epoch": 0.29197, + "grad_norm": 0.8863108987821383, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 29197 + }, + { + "epoch": 0.29198, + "grad_norm": 0.8188232006363682, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 29198 + }, + { + "epoch": 0.29199, + "grad_norm": 0.8303744384155518, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 29199 + }, + { + "epoch": 0.292, + "grad_norm": 0.7614795291274781, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 29200 + }, + { + "epoch": 0.29201, + "grad_norm": 0.6896152533438243, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 29201 + }, + { + "epoch": 0.29202, + "grad_norm": 0.6135992040562676, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 29202 + }, + { + "epoch": 0.29203, + "grad_norm": 0.6297174894419485, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 29203 + }, + { + "epoch": 0.29204, + "grad_norm": 0.5802419151297199, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 29204 + }, + { + "epoch": 0.29205, + "grad_norm": 0.6202045017577059, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 29205 + }, + { + "epoch": 0.29206, + "grad_norm": 0.6720974397564019, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 29206 + }, + { + "epoch": 0.29207, + "grad_norm": 0.737111326060542, + "learning_rate": 0.003, + "loss": 3.974, + "step": 29207 + }, + { + "epoch": 0.29208, + "grad_norm": 0.6962883660498225, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 29208 + }, + { + "epoch": 0.29209, + "grad_norm": 0.6428893267175624, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 29209 + }, + { + "epoch": 0.2921, + "grad_norm": 0.7002459510610622, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 29210 + }, + { + "epoch": 0.29211, + "grad_norm": 0.6884514646252049, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 29211 + }, + { + "epoch": 0.29212, + "grad_norm": 0.621536376138663, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 29212 + }, + { + "epoch": 0.29213, + "grad_norm": 0.578733195407767, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 29213 + }, + { + "epoch": 0.29214, + "grad_norm": 0.6035152141729493, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 29214 + }, + { + "epoch": 0.29215, + "grad_norm": 0.6517062531060942, + "learning_rate": 0.003, + "loss": 3.98, + "step": 29215 + }, + { + "epoch": 0.29216, + "grad_norm": 0.7199492184621274, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 29216 + }, + { + "epoch": 0.29217, + "grad_norm": 0.8580139293593181, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 29217 + }, + { + "epoch": 0.29218, + "grad_norm": 1.2560089620074695, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 29218 + }, + { + "epoch": 0.29219, + "grad_norm": 0.965119087239774, + "learning_rate": 0.003, + "loss": 4.0476, + "step": 29219 + }, + { + "epoch": 0.2922, + "grad_norm": 0.7874062709909846, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 29220 + }, + { + "epoch": 0.29221, + "grad_norm": 0.7204135200760464, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 29221 + }, + { + "epoch": 0.29222, + "grad_norm": 0.7783479936179355, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 29222 + }, + { + "epoch": 0.29223, + "grad_norm": 0.8340039623144208, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 29223 + }, + { + "epoch": 0.29224, + "grad_norm": 0.9324757810879306, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 29224 + }, + { + "epoch": 0.29225, + "grad_norm": 0.8453193149655341, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 29225 + }, + { + "epoch": 0.29226, + "grad_norm": 0.8849495820444804, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 29226 + }, + { + "epoch": 0.29227, + "grad_norm": 0.9776833386060176, + "learning_rate": 0.003, + "loss": 4.0343, + "step": 29227 + }, + { + "epoch": 0.29228, + "grad_norm": 0.8631717206026415, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 29228 + }, + { + "epoch": 0.29229, + "grad_norm": 0.8767286378253705, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 29229 + }, + { + "epoch": 0.2923, + "grad_norm": 0.858647909063817, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 29230 + }, + { + "epoch": 0.29231, + "grad_norm": 0.8151146625494223, + "learning_rate": 0.003, + "loss": 4.02, + "step": 29231 + }, + { + "epoch": 0.29232, + "grad_norm": 0.7857974728951488, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 29232 + }, + { + "epoch": 0.29233, + "grad_norm": 0.824175750530619, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 29233 + }, + { + "epoch": 0.29234, + "grad_norm": 0.8749969371662478, + "learning_rate": 0.003, + "loss": 4.026, + "step": 29234 + }, + { + "epoch": 0.29235, + "grad_norm": 1.0611501978417877, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 29235 + }, + { + "epoch": 0.29236, + "grad_norm": 0.92274984553015, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 29236 + }, + { + "epoch": 0.29237, + "grad_norm": 0.9411453045765609, + "learning_rate": 0.003, + "loss": 4.018, + "step": 29237 + }, + { + "epoch": 0.29238, + "grad_norm": 1.1074433100604488, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 29238 + }, + { + "epoch": 0.29239, + "grad_norm": 0.8725979607339377, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 29239 + }, + { + "epoch": 0.2924, + "grad_norm": 0.6795002715727138, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 29240 + }, + { + "epoch": 0.29241, + "grad_norm": 0.6824879418413742, + "learning_rate": 0.003, + "loss": 3.982, + "step": 29241 + }, + { + "epoch": 0.29242, + "grad_norm": 0.8155359484461635, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 29242 + }, + { + "epoch": 0.29243, + "grad_norm": 0.8464857489998692, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 29243 + }, + { + "epoch": 0.29244, + "grad_norm": 0.8708695052535179, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 29244 + }, + { + "epoch": 0.29245, + "grad_norm": 0.8002696837849606, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 29245 + }, + { + "epoch": 0.29246, + "grad_norm": 0.845510429682268, + "learning_rate": 0.003, + "loss": 4.0, + "step": 29246 + }, + { + "epoch": 0.29247, + "grad_norm": 0.7831875092037741, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 29247 + }, + { + "epoch": 0.29248, + "grad_norm": 0.769171755421859, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 29248 + }, + { + "epoch": 0.29249, + "grad_norm": 0.841497277220039, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 29249 + }, + { + "epoch": 0.2925, + "grad_norm": 0.7918885079877239, + "learning_rate": 0.003, + "loss": 4.006, + "step": 29250 + }, + { + "epoch": 0.29251, + "grad_norm": 0.8513840546378774, + "learning_rate": 0.003, + "loss": 4.01, + "step": 29251 + }, + { + "epoch": 0.29252, + "grad_norm": 0.9205928618788782, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 29252 + }, + { + "epoch": 0.29253, + "grad_norm": 0.8295206059833581, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 29253 + }, + { + "epoch": 0.29254, + "grad_norm": 0.8492604987736674, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 29254 + }, + { + "epoch": 0.29255, + "grad_norm": 0.9313722313668851, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 29255 + }, + { + "epoch": 0.29256, + "grad_norm": 0.9046883481393639, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 29256 + }, + { + "epoch": 0.29257, + "grad_norm": 0.8168136709390841, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 29257 + }, + { + "epoch": 0.29258, + "grad_norm": 0.7590651290241187, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 29258 + }, + { + "epoch": 0.29259, + "grad_norm": 0.7555764652698012, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 29259 + }, + { + "epoch": 0.2926, + "grad_norm": 0.7516519924238995, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 29260 + }, + { + "epoch": 0.29261, + "grad_norm": 0.7313376765276988, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 29261 + }, + { + "epoch": 0.29262, + "grad_norm": 0.6700826117050457, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 29262 + }, + { + "epoch": 0.29263, + "grad_norm": 0.6813184353185037, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 29263 + }, + { + "epoch": 0.29264, + "grad_norm": 0.6971896014632456, + "learning_rate": 0.003, + "loss": 3.991, + "step": 29264 + }, + { + "epoch": 0.29265, + "grad_norm": 0.74928870854146, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 29265 + }, + { + "epoch": 0.29266, + "grad_norm": 0.8323856083753789, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 29266 + }, + { + "epoch": 0.29267, + "grad_norm": 0.7585970700319604, + "learning_rate": 0.003, + "loss": 4.012, + "step": 29267 + }, + { + "epoch": 0.29268, + "grad_norm": 0.7468433136802444, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 29268 + }, + { + "epoch": 0.29269, + "grad_norm": 0.7895593906566535, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 29269 + }, + { + "epoch": 0.2927, + "grad_norm": 0.770719795099204, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 29270 + }, + { + "epoch": 0.29271, + "grad_norm": 0.7965605689846244, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 29271 + }, + { + "epoch": 0.29272, + "grad_norm": 0.7036472124090232, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 29272 + }, + { + "epoch": 0.29273, + "grad_norm": 0.7289699253011798, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 29273 + }, + { + "epoch": 0.29274, + "grad_norm": 0.763925496855333, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 29274 + }, + { + "epoch": 0.29275, + "grad_norm": 0.7193222533175551, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 29275 + }, + { + "epoch": 0.29276, + "grad_norm": 0.5469648677000987, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 29276 + }, + { + "epoch": 0.29277, + "grad_norm": 0.4256062792272416, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 29277 + }, + { + "epoch": 0.29278, + "grad_norm": 0.4472970638393823, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 29278 + }, + { + "epoch": 0.29279, + "grad_norm": 0.5235285447972209, + "learning_rate": 0.003, + "loss": 3.998, + "step": 29279 + }, + { + "epoch": 0.2928, + "grad_norm": 0.5755859252821207, + "learning_rate": 0.003, + "loss": 3.984, + "step": 29280 + }, + { + "epoch": 0.29281, + "grad_norm": 0.659180966282442, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 29281 + }, + { + "epoch": 0.29282, + "grad_norm": 0.7366928454678977, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 29282 + }, + { + "epoch": 0.29283, + "grad_norm": 0.778135791891431, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 29283 + }, + { + "epoch": 0.29284, + "grad_norm": 0.726682923106269, + "learning_rate": 0.003, + "loss": 3.968, + "step": 29284 + }, + { + "epoch": 0.29285, + "grad_norm": 0.6762737193023285, + "learning_rate": 0.003, + "loss": 3.99, + "step": 29285 + }, + { + "epoch": 0.29286, + "grad_norm": 0.6530575029275636, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 29286 + }, + { + "epoch": 0.29287, + "grad_norm": 0.6224368668288923, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 29287 + }, + { + "epoch": 0.29288, + "grad_norm": 0.6899621882603325, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 29288 + }, + { + "epoch": 0.29289, + "grad_norm": 0.7549397078874776, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 29289 + }, + { + "epoch": 0.2929, + "grad_norm": 0.7990148661065954, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 29290 + }, + { + "epoch": 0.29291, + "grad_norm": 0.8128672644863656, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 29291 + }, + { + "epoch": 0.29292, + "grad_norm": 0.793369904358933, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 29292 + }, + { + "epoch": 0.29293, + "grad_norm": 0.8906487595112834, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 29293 + }, + { + "epoch": 0.29294, + "grad_norm": 0.8285253468336629, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 29294 + }, + { + "epoch": 0.29295, + "grad_norm": 0.7534326041580075, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 29295 + }, + { + "epoch": 0.29296, + "grad_norm": 0.8653135220914354, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 29296 + }, + { + "epoch": 0.29297, + "grad_norm": 0.8921504366185774, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 29297 + }, + { + "epoch": 0.29298, + "grad_norm": 1.0420573316557133, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 29298 + }, + { + "epoch": 0.29299, + "grad_norm": 1.022584231790248, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 29299 + }, + { + "epoch": 0.293, + "grad_norm": 0.8363905613790299, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 29300 + }, + { + "epoch": 0.29301, + "grad_norm": 0.7789740912796472, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 29301 + }, + { + "epoch": 0.29302, + "grad_norm": 0.8639598020049275, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 29302 + }, + { + "epoch": 0.29303, + "grad_norm": 1.004293546373114, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 29303 + }, + { + "epoch": 0.29304, + "grad_norm": 0.9578406213262481, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 29304 + }, + { + "epoch": 0.29305, + "grad_norm": 0.8765628124868031, + "learning_rate": 0.003, + "loss": 3.974, + "step": 29305 + }, + { + "epoch": 0.29306, + "grad_norm": 0.8501867689537596, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 29306 + }, + { + "epoch": 0.29307, + "grad_norm": 0.7847727389890123, + "learning_rate": 0.003, + "loss": 3.996, + "step": 29307 + }, + { + "epoch": 0.29308, + "grad_norm": 0.7035898265374224, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 29308 + }, + { + "epoch": 0.29309, + "grad_norm": 0.7693760205030056, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 29309 + }, + { + "epoch": 0.2931, + "grad_norm": 0.7665621315582554, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 29310 + }, + { + "epoch": 0.29311, + "grad_norm": 0.7010218140873162, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 29311 + }, + { + "epoch": 0.29312, + "grad_norm": 0.6480514411289946, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 29312 + }, + { + "epoch": 0.29313, + "grad_norm": 0.7139492666101543, + "learning_rate": 0.003, + "loss": 3.988, + "step": 29313 + }, + { + "epoch": 0.29314, + "grad_norm": 0.697836653591749, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29314 + }, + { + "epoch": 0.29315, + "grad_norm": 0.6612627061433417, + "learning_rate": 0.003, + "loss": 4.013, + "step": 29315 + }, + { + "epoch": 0.29316, + "grad_norm": 0.6398163789708815, + "learning_rate": 0.003, + "loss": 3.98, + "step": 29316 + }, + { + "epoch": 0.29317, + "grad_norm": 0.5280463454460362, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 29317 + }, + { + "epoch": 0.29318, + "grad_norm": 0.5701704242649501, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 29318 + }, + { + "epoch": 0.29319, + "grad_norm": 0.5880254101989479, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 29319 + }, + { + "epoch": 0.2932, + "grad_norm": 0.5755135864935446, + "learning_rate": 0.003, + "loss": 3.992, + "step": 29320 + }, + { + "epoch": 0.29321, + "grad_norm": 0.6312370402197489, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 29321 + }, + { + "epoch": 0.29322, + "grad_norm": 0.752277284820976, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 29322 + }, + { + "epoch": 0.29323, + "grad_norm": 0.8022067552835913, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 29323 + }, + { + "epoch": 0.29324, + "grad_norm": 0.8763802548061362, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 29324 + }, + { + "epoch": 0.29325, + "grad_norm": 0.9178848237603747, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 29325 + }, + { + "epoch": 0.29326, + "grad_norm": 0.9205847345946144, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 29326 + }, + { + "epoch": 0.29327, + "grad_norm": 0.911836917724606, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 29327 + }, + { + "epoch": 0.29328, + "grad_norm": 0.7313389443084723, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 29328 + }, + { + "epoch": 0.29329, + "grad_norm": 0.7861548760619529, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 29329 + }, + { + "epoch": 0.2933, + "grad_norm": 0.8505688333994301, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 29330 + }, + { + "epoch": 0.29331, + "grad_norm": 0.9131120046149678, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 29331 + }, + { + "epoch": 0.29332, + "grad_norm": 0.96922645021494, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 29332 + }, + { + "epoch": 0.29333, + "grad_norm": 0.8234026094652779, + "learning_rate": 0.003, + "loss": 4.006, + "step": 29333 + }, + { + "epoch": 0.29334, + "grad_norm": 0.7395579444548966, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 29334 + }, + { + "epoch": 0.29335, + "grad_norm": 0.6732685788744871, + "learning_rate": 0.003, + "loss": 4.008, + "step": 29335 + }, + { + "epoch": 0.29336, + "grad_norm": 0.5997832354399194, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 29336 + }, + { + "epoch": 0.29337, + "grad_norm": 0.6086901824372792, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 29337 + }, + { + "epoch": 0.29338, + "grad_norm": 0.6668450791714321, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 29338 + }, + { + "epoch": 0.29339, + "grad_norm": 0.5465977912905894, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 29339 + }, + { + "epoch": 0.2934, + "grad_norm": 0.5882315603691591, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 29340 + }, + { + "epoch": 0.29341, + "grad_norm": 0.6409806614920435, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 29341 + }, + { + "epoch": 0.29342, + "grad_norm": 0.7365497786600791, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 29342 + }, + { + "epoch": 0.29343, + "grad_norm": 0.8944761968011127, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 29343 + }, + { + "epoch": 0.29344, + "grad_norm": 0.9911484387549534, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 29344 + }, + { + "epoch": 0.29345, + "grad_norm": 1.1030503169628496, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 29345 + }, + { + "epoch": 0.29346, + "grad_norm": 0.8597826165260466, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 29346 + }, + { + "epoch": 0.29347, + "grad_norm": 0.7007103802237137, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 29347 + }, + { + "epoch": 0.29348, + "grad_norm": 0.7732668372067562, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 29348 + }, + { + "epoch": 0.29349, + "grad_norm": 0.8284220825576974, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 29349 + }, + { + "epoch": 0.2935, + "grad_norm": 0.7743825818004184, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 29350 + }, + { + "epoch": 0.29351, + "grad_norm": 0.6985997150157517, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 29351 + }, + { + "epoch": 0.29352, + "grad_norm": 0.6774968158471381, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 29352 + }, + { + "epoch": 0.29353, + "grad_norm": 0.7456861656596154, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 29353 + }, + { + "epoch": 0.29354, + "grad_norm": 0.8960550913123437, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 29354 + }, + { + "epoch": 0.29355, + "grad_norm": 0.8784393772475424, + "learning_rate": 0.003, + "loss": 4.008, + "step": 29355 + }, + { + "epoch": 0.29356, + "grad_norm": 0.9191324999855951, + "learning_rate": 0.003, + "loss": 4.017, + "step": 29356 + }, + { + "epoch": 0.29357, + "grad_norm": 1.081547543098129, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 29357 + }, + { + "epoch": 0.29358, + "grad_norm": 0.9924230242719123, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 29358 + }, + { + "epoch": 0.29359, + "grad_norm": 1.1454260544681314, + "learning_rate": 0.003, + "loss": 4.0595, + "step": 29359 + }, + { + "epoch": 0.2936, + "grad_norm": 0.9513992605590219, + "learning_rate": 0.003, + "loss": 4.003, + "step": 29360 + }, + { + "epoch": 0.29361, + "grad_norm": 1.0570883304541334, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 29361 + }, + { + "epoch": 0.29362, + "grad_norm": 0.9445823881915193, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 29362 + }, + { + "epoch": 0.29363, + "grad_norm": 0.9360621650582738, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 29363 + }, + { + "epoch": 0.29364, + "grad_norm": 0.893193262479448, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 29364 + }, + { + "epoch": 0.29365, + "grad_norm": 0.9885417520032967, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 29365 + }, + { + "epoch": 0.29366, + "grad_norm": 0.9179195284162254, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 29366 + }, + { + "epoch": 0.29367, + "grad_norm": 0.7979178131245881, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 29367 + }, + { + "epoch": 0.29368, + "grad_norm": 0.7670640179369295, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 29368 + }, + { + "epoch": 0.29369, + "grad_norm": 0.6240055915540093, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 29369 + }, + { + "epoch": 0.2937, + "grad_norm": 0.5805201787622493, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 29370 + }, + { + "epoch": 0.29371, + "grad_norm": 0.6149534455916423, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29371 + }, + { + "epoch": 0.29372, + "grad_norm": 0.7191651320325667, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 29372 + }, + { + "epoch": 0.29373, + "grad_norm": 0.7844738250229614, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 29373 + }, + { + "epoch": 0.29374, + "grad_norm": 0.8494057556275006, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 29374 + }, + { + "epoch": 0.29375, + "grad_norm": 0.7159063661993588, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 29375 + }, + { + "epoch": 0.29376, + "grad_norm": 0.5478979887332192, + "learning_rate": 0.003, + "loss": 3.9215, + "step": 29376 + }, + { + "epoch": 0.29377, + "grad_norm": 0.5389916048710421, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 29377 + }, + { + "epoch": 0.29378, + "grad_norm": 0.5728262142024918, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 29378 + }, + { + "epoch": 0.29379, + "grad_norm": 0.5977092491725832, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 29379 + }, + { + "epoch": 0.2938, + "grad_norm": 0.5370816271742048, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 29380 + }, + { + "epoch": 0.29381, + "grad_norm": 0.5404996566035587, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 29381 + }, + { + "epoch": 0.29382, + "grad_norm": 0.527543541138819, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 29382 + }, + { + "epoch": 0.29383, + "grad_norm": 0.5631674317948602, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 29383 + }, + { + "epoch": 0.29384, + "grad_norm": 0.7526162411615163, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 29384 + }, + { + "epoch": 0.29385, + "grad_norm": 0.9548477235461489, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 29385 + }, + { + "epoch": 0.29386, + "grad_norm": 0.9272746261071719, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 29386 + }, + { + "epoch": 0.29387, + "grad_norm": 0.8066314468804966, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 29387 + }, + { + "epoch": 0.29388, + "grad_norm": 0.7628099595061114, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 29388 + }, + { + "epoch": 0.29389, + "grad_norm": 0.7382666786932583, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 29389 + }, + { + "epoch": 0.2939, + "grad_norm": 0.8481861583693572, + "learning_rate": 0.003, + "loss": 3.983, + "step": 29390 + }, + { + "epoch": 0.29391, + "grad_norm": 0.8205216341147652, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 29391 + }, + { + "epoch": 0.29392, + "grad_norm": 0.6556158203698397, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 29392 + }, + { + "epoch": 0.29393, + "grad_norm": 0.6324980109215108, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 29393 + }, + { + "epoch": 0.29394, + "grad_norm": 0.8020315084131391, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 29394 + }, + { + "epoch": 0.29395, + "grad_norm": 0.8771929907478794, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 29395 + }, + { + "epoch": 0.29396, + "grad_norm": 0.7619793335339354, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 29396 + }, + { + "epoch": 0.29397, + "grad_norm": 0.7864994980251905, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 29397 + }, + { + "epoch": 0.29398, + "grad_norm": 0.7745457092359624, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 29398 + }, + { + "epoch": 0.29399, + "grad_norm": 0.7304782868036144, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 29399 + }, + { + "epoch": 0.294, + "grad_norm": 0.6425035216392407, + "learning_rate": 0.003, + "loss": 3.992, + "step": 29400 + }, + { + "epoch": 0.29401, + "grad_norm": 0.6291870531140866, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 29401 + }, + { + "epoch": 0.29402, + "grad_norm": 0.7140452951638114, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 29402 + }, + { + "epoch": 0.29403, + "grad_norm": 0.8097464227155743, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 29403 + }, + { + "epoch": 0.29404, + "grad_norm": 0.8777692547030502, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 29404 + }, + { + "epoch": 0.29405, + "grad_norm": 0.864950183018294, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 29405 + }, + { + "epoch": 0.29406, + "grad_norm": 0.8114737971726349, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 29406 + }, + { + "epoch": 0.29407, + "grad_norm": 0.8585947373538686, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 29407 + }, + { + "epoch": 0.29408, + "grad_norm": 0.9124349422524495, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 29408 + }, + { + "epoch": 0.29409, + "grad_norm": 0.8327080222943699, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 29409 + }, + { + "epoch": 0.2941, + "grad_norm": 0.7396392781464761, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 29410 + }, + { + "epoch": 0.29411, + "grad_norm": 0.6968113987686425, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 29411 + }, + { + "epoch": 0.29412, + "grad_norm": 0.6025878733582353, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 29412 + }, + { + "epoch": 0.29413, + "grad_norm": 0.6820837418563388, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 29413 + }, + { + "epoch": 0.29414, + "grad_norm": 0.7678274987568827, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 29414 + }, + { + "epoch": 0.29415, + "grad_norm": 0.9378865226367614, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 29415 + }, + { + "epoch": 0.29416, + "grad_norm": 1.0594512844341786, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 29416 + }, + { + "epoch": 0.29417, + "grad_norm": 0.905515812900681, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 29417 + }, + { + "epoch": 0.29418, + "grad_norm": 0.8165455154806486, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 29418 + }, + { + "epoch": 0.29419, + "grad_norm": 0.8001984095093255, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 29419 + }, + { + "epoch": 0.2942, + "grad_norm": 0.7115081361423571, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 29420 + }, + { + "epoch": 0.29421, + "grad_norm": 0.6962690573491721, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 29421 + }, + { + "epoch": 0.29422, + "grad_norm": 0.7807381571504582, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 29422 + }, + { + "epoch": 0.29423, + "grad_norm": 0.8697135862092089, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 29423 + }, + { + "epoch": 0.29424, + "grad_norm": 0.8837499048255332, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 29424 + }, + { + "epoch": 0.29425, + "grad_norm": 0.7560250901868305, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 29425 + }, + { + "epoch": 0.29426, + "grad_norm": 0.7125866387982546, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 29426 + }, + { + "epoch": 0.29427, + "grad_norm": 0.7450502256727013, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 29427 + }, + { + "epoch": 0.29428, + "grad_norm": 0.7336590189457762, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 29428 + }, + { + "epoch": 0.29429, + "grad_norm": 0.8395250029308565, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 29429 + }, + { + "epoch": 0.2943, + "grad_norm": 0.8598668184548939, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 29430 + }, + { + "epoch": 0.29431, + "grad_norm": 0.6856413398859864, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 29431 + }, + { + "epoch": 0.29432, + "grad_norm": 0.7418075841971769, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 29432 + }, + { + "epoch": 0.29433, + "grad_norm": 0.8441684806536539, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 29433 + }, + { + "epoch": 0.29434, + "grad_norm": 0.8456196399842097, + "learning_rate": 0.003, + "loss": 4.004, + "step": 29434 + }, + { + "epoch": 0.29435, + "grad_norm": 0.8248257134153639, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 29435 + }, + { + "epoch": 0.29436, + "grad_norm": 0.8252613895763481, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 29436 + }, + { + "epoch": 0.29437, + "grad_norm": 0.8178308339614474, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 29437 + }, + { + "epoch": 0.29438, + "grad_norm": 0.7877369793722209, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 29438 + }, + { + "epoch": 0.29439, + "grad_norm": 0.8534367820299993, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 29439 + }, + { + "epoch": 0.2944, + "grad_norm": 0.8689955391923704, + "learning_rate": 0.003, + "loss": 4.034, + "step": 29440 + }, + { + "epoch": 0.29441, + "grad_norm": 0.9467218036700301, + "learning_rate": 0.003, + "loss": 3.995, + "step": 29441 + }, + { + "epoch": 0.29442, + "grad_norm": 0.831793660150527, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 29442 + }, + { + "epoch": 0.29443, + "grad_norm": 0.7473504423301638, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 29443 + }, + { + "epoch": 0.29444, + "grad_norm": 0.7391161984159491, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 29444 + }, + { + "epoch": 0.29445, + "grad_norm": 0.8136469698437692, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 29445 + }, + { + "epoch": 0.29446, + "grad_norm": 0.7677758227333669, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29446 + }, + { + "epoch": 0.29447, + "grad_norm": 0.6559823079952446, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 29447 + }, + { + "epoch": 0.29448, + "grad_norm": 0.6370504309996226, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 29448 + }, + { + "epoch": 0.29449, + "grad_norm": 0.8016419842085114, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 29449 + }, + { + "epoch": 0.2945, + "grad_norm": 0.868634485864223, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 29450 + }, + { + "epoch": 0.29451, + "grad_norm": 0.8884946242744792, + "learning_rate": 0.003, + "loss": 4.0394, + "step": 29451 + }, + { + "epoch": 0.29452, + "grad_norm": 0.8227242702726834, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 29452 + }, + { + "epoch": 0.29453, + "grad_norm": 0.8258356164641012, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 29453 + }, + { + "epoch": 0.29454, + "grad_norm": 0.7500544417689461, + "learning_rate": 0.003, + "loss": 3.977, + "step": 29454 + }, + { + "epoch": 0.29455, + "grad_norm": 0.8030802809450363, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 29455 + }, + { + "epoch": 0.29456, + "grad_norm": 0.646629847634173, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 29456 + }, + { + "epoch": 0.29457, + "grad_norm": 0.6313331181320871, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 29457 + }, + { + "epoch": 0.29458, + "grad_norm": 0.5386081243789821, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 29458 + }, + { + "epoch": 0.29459, + "grad_norm": 0.5641301986839843, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 29459 + }, + { + "epoch": 0.2946, + "grad_norm": 0.5575313409619456, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 29460 + }, + { + "epoch": 0.29461, + "grad_norm": 0.6507590781454899, + "learning_rate": 0.003, + "loss": 3.989, + "step": 29461 + }, + { + "epoch": 0.29462, + "grad_norm": 0.7476385905756581, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 29462 + }, + { + "epoch": 0.29463, + "grad_norm": 0.7183342359242553, + "learning_rate": 0.003, + "loss": 3.995, + "step": 29463 + }, + { + "epoch": 0.29464, + "grad_norm": 0.655873266361002, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 29464 + }, + { + "epoch": 0.29465, + "grad_norm": 0.6056521432662407, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 29465 + }, + { + "epoch": 0.29466, + "grad_norm": 0.6484174190689189, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 29466 + }, + { + "epoch": 0.29467, + "grad_norm": 0.6960663462145484, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 29467 + }, + { + "epoch": 0.29468, + "grad_norm": 0.8323370525390595, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 29468 + }, + { + "epoch": 0.29469, + "grad_norm": 0.8183860993887317, + "learning_rate": 0.003, + "loss": 3.988, + "step": 29469 + }, + { + "epoch": 0.2947, + "grad_norm": 0.8138324438239292, + "learning_rate": 0.003, + "loss": 3.989, + "step": 29470 + }, + { + "epoch": 0.29471, + "grad_norm": 0.9329241906246988, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 29471 + }, + { + "epoch": 0.29472, + "grad_norm": 0.8886599665223485, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 29472 + }, + { + "epoch": 0.29473, + "grad_norm": 0.9134360051035469, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 29473 + }, + { + "epoch": 0.29474, + "grad_norm": 0.9268959777328842, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 29474 + }, + { + "epoch": 0.29475, + "grad_norm": 0.7613205628627435, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 29475 + }, + { + "epoch": 0.29476, + "grad_norm": 0.7362780043007031, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 29476 + }, + { + "epoch": 0.29477, + "grad_norm": 0.5853332957964141, + "learning_rate": 0.003, + "loss": 3.982, + "step": 29477 + }, + { + "epoch": 0.29478, + "grad_norm": 0.6543904925245833, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 29478 + }, + { + "epoch": 0.29479, + "grad_norm": 0.751304312115912, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 29479 + }, + { + "epoch": 0.2948, + "grad_norm": 0.8863310290854154, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 29480 + }, + { + "epoch": 0.29481, + "grad_norm": 1.0462078596801536, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 29481 + }, + { + "epoch": 0.29482, + "grad_norm": 0.8587525204100414, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 29482 + }, + { + "epoch": 0.29483, + "grad_norm": 0.8165185219074571, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 29483 + }, + { + "epoch": 0.29484, + "grad_norm": 0.9038550659853178, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 29484 + }, + { + "epoch": 0.29485, + "grad_norm": 0.9196076504545966, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 29485 + }, + { + "epoch": 0.29486, + "grad_norm": 0.8048977236079428, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 29486 + }, + { + "epoch": 0.29487, + "grad_norm": 0.8300195648232473, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 29487 + }, + { + "epoch": 0.29488, + "grad_norm": 0.7289571253888071, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 29488 + }, + { + "epoch": 0.29489, + "grad_norm": 0.6484668473176788, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 29489 + }, + { + "epoch": 0.2949, + "grad_norm": 0.5990774763479585, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 29490 + }, + { + "epoch": 0.29491, + "grad_norm": 0.5801533755938516, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 29491 + }, + { + "epoch": 0.29492, + "grad_norm": 0.634405863878058, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 29492 + }, + { + "epoch": 0.29493, + "grad_norm": 0.5333724530682008, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 29493 + }, + { + "epoch": 0.29494, + "grad_norm": 0.628031276709099, + "learning_rate": 0.003, + "loss": 4.028, + "step": 29494 + }, + { + "epoch": 0.29495, + "grad_norm": 0.7393522333128765, + "learning_rate": 0.003, + "loss": 3.97, + "step": 29495 + }, + { + "epoch": 0.29496, + "grad_norm": 0.8328818988194251, + "learning_rate": 0.003, + "loss": 3.99, + "step": 29496 + }, + { + "epoch": 0.29497, + "grad_norm": 0.7684462170440163, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 29497 + }, + { + "epoch": 0.29498, + "grad_norm": 0.7366285012897219, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 29498 + }, + { + "epoch": 0.29499, + "grad_norm": 0.8256358408701273, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 29499 + }, + { + "epoch": 0.295, + "grad_norm": 0.8618351304133307, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 29500 + }, + { + "epoch": 0.29501, + "grad_norm": 1.0449354929099237, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 29501 + }, + { + "epoch": 0.29502, + "grad_norm": 0.9831932758023524, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 29502 + }, + { + "epoch": 0.29503, + "grad_norm": 0.9826430737702081, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 29503 + }, + { + "epoch": 0.29504, + "grad_norm": 0.7625867221847135, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 29504 + }, + { + "epoch": 0.29505, + "grad_norm": 0.6451434890662036, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 29505 + }, + { + "epoch": 0.29506, + "grad_norm": 0.6245474432783473, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 29506 + }, + { + "epoch": 0.29507, + "grad_norm": 0.6250919229474827, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 29507 + }, + { + "epoch": 0.29508, + "grad_norm": 0.7511442541319775, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 29508 + }, + { + "epoch": 0.29509, + "grad_norm": 0.7261267465355057, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 29509 + }, + { + "epoch": 0.2951, + "grad_norm": 0.6781129911861081, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 29510 + }, + { + "epoch": 0.29511, + "grad_norm": 0.7498921376959476, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 29511 + }, + { + "epoch": 0.29512, + "grad_norm": 0.8063490865437305, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 29512 + }, + { + "epoch": 0.29513, + "grad_norm": 0.8083724152554892, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 29513 + }, + { + "epoch": 0.29514, + "grad_norm": 0.9476095465278849, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 29514 + }, + { + "epoch": 0.29515, + "grad_norm": 1.0475291153540138, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 29515 + }, + { + "epoch": 0.29516, + "grad_norm": 0.8258178020533948, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 29516 + }, + { + "epoch": 0.29517, + "grad_norm": 0.8273379248832958, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 29517 + }, + { + "epoch": 0.29518, + "grad_norm": 0.8053630717239681, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 29518 + }, + { + "epoch": 0.29519, + "grad_norm": 0.7876168813094698, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 29519 + }, + { + "epoch": 0.2952, + "grad_norm": 0.7930570031043228, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 29520 + }, + { + "epoch": 0.29521, + "grad_norm": 0.7942641265491818, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 29521 + }, + { + "epoch": 0.29522, + "grad_norm": 0.8374019740810305, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 29522 + }, + { + "epoch": 0.29523, + "grad_norm": 0.8591504813487886, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 29523 + }, + { + "epoch": 0.29524, + "grad_norm": 0.8546246031480071, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 29524 + }, + { + "epoch": 0.29525, + "grad_norm": 0.8179702219050314, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 29525 + }, + { + "epoch": 0.29526, + "grad_norm": 0.7866071483683965, + "learning_rate": 0.003, + "loss": 3.992, + "step": 29526 + }, + { + "epoch": 0.29527, + "grad_norm": 0.7724647641954753, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 29527 + }, + { + "epoch": 0.29528, + "grad_norm": 0.7120141243000305, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 29528 + }, + { + "epoch": 0.29529, + "grad_norm": 0.7691773094505362, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 29529 + }, + { + "epoch": 0.2953, + "grad_norm": 0.88021215615408, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 29530 + }, + { + "epoch": 0.29531, + "grad_norm": 0.8041811926440507, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29531 + }, + { + "epoch": 0.29532, + "grad_norm": 0.6530287507883612, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 29532 + }, + { + "epoch": 0.29533, + "grad_norm": 0.6614003997449974, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 29533 + }, + { + "epoch": 0.29534, + "grad_norm": 0.6840342247267199, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 29534 + }, + { + "epoch": 0.29535, + "grad_norm": 0.6542515665242611, + "learning_rate": 0.003, + "loss": 4.027, + "step": 29535 + }, + { + "epoch": 0.29536, + "grad_norm": 0.5816606765476221, + "learning_rate": 0.003, + "loss": 3.9453, + "step": 29536 + }, + { + "epoch": 0.29537, + "grad_norm": 0.5451584633839403, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 29537 + }, + { + "epoch": 0.29538, + "grad_norm": 0.6482117175552952, + "learning_rate": 0.003, + "loss": 3.967, + "step": 29538 + }, + { + "epoch": 0.29539, + "grad_norm": 0.7361722758649331, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 29539 + }, + { + "epoch": 0.2954, + "grad_norm": 0.8795824569551812, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 29540 + }, + { + "epoch": 0.29541, + "grad_norm": 1.0032498711493818, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 29541 + }, + { + "epoch": 0.29542, + "grad_norm": 1.1181231757989576, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 29542 + }, + { + "epoch": 0.29543, + "grad_norm": 0.738741244647667, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 29543 + }, + { + "epoch": 0.29544, + "grad_norm": 0.6809067043207209, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 29544 + }, + { + "epoch": 0.29545, + "grad_norm": 0.6144358339240239, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 29545 + }, + { + "epoch": 0.29546, + "grad_norm": 0.7209968735228225, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 29546 + }, + { + "epoch": 0.29547, + "grad_norm": 0.9104099215450205, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 29547 + }, + { + "epoch": 0.29548, + "grad_norm": 0.9649937162192005, + "learning_rate": 0.003, + "loss": 3.97, + "step": 29548 + }, + { + "epoch": 0.29549, + "grad_norm": 0.772459788447225, + "learning_rate": 0.003, + "loss": 3.983, + "step": 29549 + }, + { + "epoch": 0.2955, + "grad_norm": 0.6227108522003986, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 29550 + }, + { + "epoch": 0.29551, + "grad_norm": 0.6592052214587667, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 29551 + }, + { + "epoch": 0.29552, + "grad_norm": 0.7308858643393363, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 29552 + }, + { + "epoch": 0.29553, + "grad_norm": 0.6389425309513579, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 29553 + }, + { + "epoch": 0.29554, + "grad_norm": 0.6795397697918207, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 29554 + }, + { + "epoch": 0.29555, + "grad_norm": 0.5398545555696026, + "learning_rate": 0.003, + "loss": 3.9438, + "step": 29555 + }, + { + "epoch": 0.29556, + "grad_norm": 0.5816956197223969, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 29556 + }, + { + "epoch": 0.29557, + "grad_norm": 0.5641118016221, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 29557 + }, + { + "epoch": 0.29558, + "grad_norm": 0.527704060667275, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 29558 + }, + { + "epoch": 0.29559, + "grad_norm": 0.5102719880402166, + "learning_rate": 0.003, + "loss": 3.9418, + "step": 29559 + }, + { + "epoch": 0.2956, + "grad_norm": 0.4478657000211583, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 29560 + }, + { + "epoch": 0.29561, + "grad_norm": 0.5207753672505658, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 29561 + }, + { + "epoch": 0.29562, + "grad_norm": 0.5922455949920297, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 29562 + }, + { + "epoch": 0.29563, + "grad_norm": 0.7156624257165822, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 29563 + }, + { + "epoch": 0.29564, + "grad_norm": 0.753175992305969, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 29564 + }, + { + "epoch": 0.29565, + "grad_norm": 0.7342165038217876, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 29565 + }, + { + "epoch": 0.29566, + "grad_norm": 0.8229672257662402, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 29566 + }, + { + "epoch": 0.29567, + "grad_norm": 0.8281754975513217, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 29567 + }, + { + "epoch": 0.29568, + "grad_norm": 0.7186786986779531, + "learning_rate": 0.003, + "loss": 3.9457, + "step": 29568 + }, + { + "epoch": 0.29569, + "grad_norm": 0.7167119377461322, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 29569 + }, + { + "epoch": 0.2957, + "grad_norm": 0.664860384882884, + "learning_rate": 0.003, + "loss": 3.9317, + "step": 29570 + }, + { + "epoch": 0.29571, + "grad_norm": 0.7920163669590603, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 29571 + }, + { + "epoch": 0.29572, + "grad_norm": 0.8564746347625782, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 29572 + }, + { + "epoch": 0.29573, + "grad_norm": 0.9801983105194155, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 29573 + }, + { + "epoch": 0.29574, + "grad_norm": 1.1819031302076932, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 29574 + }, + { + "epoch": 0.29575, + "grad_norm": 0.7482227462146718, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 29575 + }, + { + "epoch": 0.29576, + "grad_norm": 0.7586247688879294, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 29576 + }, + { + "epoch": 0.29577, + "grad_norm": 0.89219403402389, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 29577 + }, + { + "epoch": 0.29578, + "grad_norm": 1.2435148600789485, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 29578 + }, + { + "epoch": 0.29579, + "grad_norm": 0.798844198177648, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 29579 + }, + { + "epoch": 0.2958, + "grad_norm": 0.8936280644691177, + "learning_rate": 0.003, + "loss": 4.004, + "step": 29580 + }, + { + "epoch": 0.29581, + "grad_norm": 0.85486489424356, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 29581 + }, + { + "epoch": 0.29582, + "grad_norm": 1.0330445163823685, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 29582 + }, + { + "epoch": 0.29583, + "grad_norm": 1.0173794299002994, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 29583 + }, + { + "epoch": 0.29584, + "grad_norm": 1.1731925182518548, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 29584 + }, + { + "epoch": 0.29585, + "grad_norm": 0.8732294986374076, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 29585 + }, + { + "epoch": 0.29586, + "grad_norm": 0.8992731358460216, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 29586 + }, + { + "epoch": 0.29587, + "grad_norm": 0.873570422212814, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 29587 + }, + { + "epoch": 0.29588, + "grad_norm": 0.8813148126524596, + "learning_rate": 0.003, + "loss": 4.0345, + "step": 29588 + }, + { + "epoch": 0.29589, + "grad_norm": 1.0418276815591347, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 29589 + }, + { + "epoch": 0.2959, + "grad_norm": 1.0176800083822517, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 29590 + }, + { + "epoch": 0.29591, + "grad_norm": 0.9775208675933053, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 29591 + }, + { + "epoch": 0.29592, + "grad_norm": 0.9040256437938218, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 29592 + }, + { + "epoch": 0.29593, + "grad_norm": 0.9368493205086643, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 29593 + }, + { + "epoch": 0.29594, + "grad_norm": 0.8767679274509116, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 29594 + }, + { + "epoch": 0.29595, + "grad_norm": 0.8129615460110802, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 29595 + }, + { + "epoch": 0.29596, + "grad_norm": 0.8068526496105106, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 29596 + }, + { + "epoch": 0.29597, + "grad_norm": 0.8148390443805089, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 29597 + }, + { + "epoch": 0.29598, + "grad_norm": 0.9465388837981262, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 29598 + }, + { + "epoch": 0.29599, + "grad_norm": 1.0200253184905541, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 29599 + }, + { + "epoch": 0.296, + "grad_norm": 0.9035445898514877, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 29600 + }, + { + "epoch": 0.29601, + "grad_norm": 0.7667617222205572, + "learning_rate": 0.003, + "loss": 4.0376, + "step": 29601 + }, + { + "epoch": 0.29602, + "grad_norm": 0.6757048237334524, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 29602 + }, + { + "epoch": 0.29603, + "grad_norm": 0.6858791313403002, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 29603 + }, + { + "epoch": 0.29604, + "grad_norm": 0.5994515563386855, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29604 + }, + { + "epoch": 0.29605, + "grad_norm": 0.6424227752951778, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 29605 + }, + { + "epoch": 0.29606, + "grad_norm": 0.5940516610168953, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 29606 + }, + { + "epoch": 0.29607, + "grad_norm": 0.5316766301724957, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 29607 + }, + { + "epoch": 0.29608, + "grad_norm": 0.527785348339321, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 29608 + }, + { + "epoch": 0.29609, + "grad_norm": 0.5346711568680041, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 29609 + }, + { + "epoch": 0.2961, + "grad_norm": 0.5994048362564348, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 29610 + }, + { + "epoch": 0.29611, + "grad_norm": 0.6159683021105655, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 29611 + }, + { + "epoch": 0.29612, + "grad_norm": 0.5312191795816005, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 29612 + }, + { + "epoch": 0.29613, + "grad_norm": 0.47123263544941624, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 29613 + }, + { + "epoch": 0.29614, + "grad_norm": 0.38838721564153905, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 29614 + }, + { + "epoch": 0.29615, + "grad_norm": 0.421347863716654, + "learning_rate": 0.003, + "loss": 3.955, + "step": 29615 + }, + { + "epoch": 0.29616, + "grad_norm": 0.43628608705420857, + "learning_rate": 0.003, + "loss": 3.978, + "step": 29616 + }, + { + "epoch": 0.29617, + "grad_norm": 0.4326740729865703, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 29617 + }, + { + "epoch": 0.29618, + "grad_norm": 0.4997106473264682, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 29618 + }, + { + "epoch": 0.29619, + "grad_norm": 0.6647795989102462, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 29619 + }, + { + "epoch": 0.2962, + "grad_norm": 0.874286016130524, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 29620 + }, + { + "epoch": 0.29621, + "grad_norm": 1.0040665630477266, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 29621 + }, + { + "epoch": 0.29622, + "grad_norm": 0.8518869266257407, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 29622 + }, + { + "epoch": 0.29623, + "grad_norm": 0.8076780504807711, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 29623 + }, + { + "epoch": 0.29624, + "grad_norm": 1.0299517211129772, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 29624 + }, + { + "epoch": 0.29625, + "grad_norm": 0.9809715602114228, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 29625 + }, + { + "epoch": 0.29626, + "grad_norm": 0.9932537735764814, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 29626 + }, + { + "epoch": 0.29627, + "grad_norm": 0.9306406958648759, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 29627 + }, + { + "epoch": 0.29628, + "grad_norm": 0.9074858360164636, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 29628 + }, + { + "epoch": 0.29629, + "grad_norm": 0.7352883370975671, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 29629 + }, + { + "epoch": 0.2963, + "grad_norm": 0.6888627688506558, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 29630 + }, + { + "epoch": 0.29631, + "grad_norm": 0.6609755119737164, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 29631 + }, + { + "epoch": 0.29632, + "grad_norm": 0.7278639899626898, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 29632 + }, + { + "epoch": 0.29633, + "grad_norm": 0.8149356724033544, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 29633 + }, + { + "epoch": 0.29634, + "grad_norm": 0.7749147525084842, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 29634 + }, + { + "epoch": 0.29635, + "grad_norm": 0.7755027025417026, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 29635 + }, + { + "epoch": 0.29636, + "grad_norm": 0.7882977531104569, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 29636 + }, + { + "epoch": 0.29637, + "grad_norm": 0.7821194500788078, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 29637 + }, + { + "epoch": 0.29638, + "grad_norm": 0.7937622326738808, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29638 + }, + { + "epoch": 0.29639, + "grad_norm": 0.7265375907747639, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 29639 + }, + { + "epoch": 0.2964, + "grad_norm": 0.6533188663317461, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 29640 + }, + { + "epoch": 0.29641, + "grad_norm": 0.6270583108142597, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 29641 + }, + { + "epoch": 0.29642, + "grad_norm": 0.6291899631841726, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 29642 + }, + { + "epoch": 0.29643, + "grad_norm": 0.7826309047672233, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 29643 + }, + { + "epoch": 0.29644, + "grad_norm": 0.9215288797688121, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 29644 + }, + { + "epoch": 0.29645, + "grad_norm": 0.9811702967821834, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 29645 + }, + { + "epoch": 0.29646, + "grad_norm": 0.9573393878914656, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 29646 + }, + { + "epoch": 0.29647, + "grad_norm": 1.004165824002102, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 29647 + }, + { + "epoch": 0.29648, + "grad_norm": 0.9127354808837105, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 29648 + }, + { + "epoch": 0.29649, + "grad_norm": 0.7404695434416363, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 29649 + }, + { + "epoch": 0.2965, + "grad_norm": 0.7917425405674352, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 29650 + }, + { + "epoch": 0.29651, + "grad_norm": 0.8287035838965096, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 29651 + }, + { + "epoch": 0.29652, + "grad_norm": 0.8660087371236814, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29652 + }, + { + "epoch": 0.29653, + "grad_norm": 0.7560959042843618, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 29653 + }, + { + "epoch": 0.29654, + "grad_norm": 0.7446505504182909, + "learning_rate": 0.003, + "loss": 3.9454, + "step": 29654 + }, + { + "epoch": 0.29655, + "grad_norm": 0.7018874521362057, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 29655 + }, + { + "epoch": 0.29656, + "grad_norm": 0.6690893375656622, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 29656 + }, + { + "epoch": 0.29657, + "grad_norm": 0.7439109376371267, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 29657 + }, + { + "epoch": 0.29658, + "grad_norm": 0.7726764149904598, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 29658 + }, + { + "epoch": 0.29659, + "grad_norm": 0.7199223813184872, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 29659 + }, + { + "epoch": 0.2966, + "grad_norm": 0.6614979891151156, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 29660 + }, + { + "epoch": 0.29661, + "grad_norm": 0.6579601648098715, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 29661 + }, + { + "epoch": 0.29662, + "grad_norm": 0.6586222413700944, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 29662 + }, + { + "epoch": 0.29663, + "grad_norm": 0.6284459750202964, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 29663 + }, + { + "epoch": 0.29664, + "grad_norm": 0.6483087188463251, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 29664 + }, + { + "epoch": 0.29665, + "grad_norm": 0.5906145162820279, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 29665 + }, + { + "epoch": 0.29666, + "grad_norm": 0.6452831316797465, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 29666 + }, + { + "epoch": 0.29667, + "grad_norm": 0.7448100249477243, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 29667 + }, + { + "epoch": 0.29668, + "grad_norm": 0.81208419294903, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 29668 + }, + { + "epoch": 0.29669, + "grad_norm": 0.8468004450513361, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 29669 + }, + { + "epoch": 0.2967, + "grad_norm": 0.9510136574050936, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 29670 + }, + { + "epoch": 0.29671, + "grad_norm": 1.261516194620651, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 29671 + }, + { + "epoch": 0.29672, + "grad_norm": 0.8454621722442284, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 29672 + }, + { + "epoch": 0.29673, + "grad_norm": 0.7779070710087258, + "learning_rate": 0.003, + "loss": 4.015, + "step": 29673 + }, + { + "epoch": 0.29674, + "grad_norm": 0.6504273846332483, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 29674 + }, + { + "epoch": 0.29675, + "grad_norm": 0.6654774183328371, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 29675 + }, + { + "epoch": 0.29676, + "grad_norm": 0.705197363819079, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 29676 + }, + { + "epoch": 0.29677, + "grad_norm": 0.8339648040209175, + "learning_rate": 0.003, + "loss": 3.984, + "step": 29677 + }, + { + "epoch": 0.29678, + "grad_norm": 1.0035691108979268, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 29678 + }, + { + "epoch": 0.29679, + "grad_norm": 0.8763362960258968, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 29679 + }, + { + "epoch": 0.2968, + "grad_norm": 0.8800597524730474, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 29680 + }, + { + "epoch": 0.29681, + "grad_norm": 0.9173894807971946, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 29681 + }, + { + "epoch": 0.29682, + "grad_norm": 0.934617604298728, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 29682 + }, + { + "epoch": 0.29683, + "grad_norm": 0.9067108405448148, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 29683 + }, + { + "epoch": 0.29684, + "grad_norm": 0.7664985915972556, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 29684 + }, + { + "epoch": 0.29685, + "grad_norm": 0.628343852035736, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 29685 + }, + { + "epoch": 0.29686, + "grad_norm": 0.5828864605623776, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 29686 + }, + { + "epoch": 0.29687, + "grad_norm": 0.516974662969522, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 29687 + }, + { + "epoch": 0.29688, + "grad_norm": 0.4645773760606956, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 29688 + }, + { + "epoch": 0.29689, + "grad_norm": 0.4594274139527887, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 29689 + }, + { + "epoch": 0.2969, + "grad_norm": 0.4592655320164949, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 29690 + }, + { + "epoch": 0.29691, + "grad_norm": 0.5447443290955009, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 29691 + }, + { + "epoch": 0.29692, + "grad_norm": 0.6905403551505621, + "learning_rate": 0.003, + "loss": 3.988, + "step": 29692 + }, + { + "epoch": 0.29693, + "grad_norm": 0.8306146504159395, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 29693 + }, + { + "epoch": 0.29694, + "grad_norm": 0.9294449028000646, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 29694 + }, + { + "epoch": 0.29695, + "grad_norm": 0.9034826411881722, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 29695 + }, + { + "epoch": 0.29696, + "grad_norm": 0.7868894085776437, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 29696 + }, + { + "epoch": 0.29697, + "grad_norm": 0.7667332708764847, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 29697 + }, + { + "epoch": 0.29698, + "grad_norm": 0.8629465175388421, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 29698 + }, + { + "epoch": 0.29699, + "grad_norm": 0.93953947103884, + "learning_rate": 0.003, + "loss": 3.964, + "step": 29699 + }, + { + "epoch": 0.297, + "grad_norm": 0.9217357691154957, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 29700 + }, + { + "epoch": 0.29701, + "grad_norm": 0.718764292068909, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 29701 + }, + { + "epoch": 0.29702, + "grad_norm": 0.7747499767512751, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 29702 + }, + { + "epoch": 0.29703, + "grad_norm": 0.9968674689990544, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 29703 + }, + { + "epoch": 0.29704, + "grad_norm": 1.0588830657921084, + "learning_rate": 0.003, + "loss": 4.0398, + "step": 29704 + }, + { + "epoch": 0.29705, + "grad_norm": 0.7703830487588483, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 29705 + }, + { + "epoch": 0.29706, + "grad_norm": 0.6778500401613684, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 29706 + }, + { + "epoch": 0.29707, + "grad_norm": 0.6611762082899271, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 29707 + }, + { + "epoch": 0.29708, + "grad_norm": 0.710560016556752, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 29708 + }, + { + "epoch": 0.29709, + "grad_norm": 0.738410822479719, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 29709 + }, + { + "epoch": 0.2971, + "grad_norm": 0.7297107518701232, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 29710 + }, + { + "epoch": 0.29711, + "grad_norm": 0.7465820975328996, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 29711 + }, + { + "epoch": 0.29712, + "grad_norm": 0.5759461900210708, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 29712 + }, + { + "epoch": 0.29713, + "grad_norm": 0.5590035767387784, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 29713 + }, + { + "epoch": 0.29714, + "grad_norm": 0.6970919360289314, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 29714 + }, + { + "epoch": 0.29715, + "grad_norm": 0.739887029280458, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 29715 + }, + { + "epoch": 0.29716, + "grad_norm": 0.6895264474487269, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 29716 + }, + { + "epoch": 0.29717, + "grad_norm": 0.679865461693243, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 29717 + }, + { + "epoch": 0.29718, + "grad_norm": 0.7643966934174093, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 29718 + }, + { + "epoch": 0.29719, + "grad_norm": 0.7526553085255858, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 29719 + }, + { + "epoch": 0.2972, + "grad_norm": 0.7623027836084848, + "learning_rate": 0.003, + "loss": 3.984, + "step": 29720 + }, + { + "epoch": 0.29721, + "grad_norm": 0.8169529891905184, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 29721 + }, + { + "epoch": 0.29722, + "grad_norm": 0.93010357937726, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 29722 + }, + { + "epoch": 0.29723, + "grad_norm": 0.9931503267800413, + "learning_rate": 0.003, + "loss": 3.971, + "step": 29723 + }, + { + "epoch": 0.29724, + "grad_norm": 0.8299282990934281, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 29724 + }, + { + "epoch": 0.29725, + "grad_norm": 0.957141454694891, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 29725 + }, + { + "epoch": 0.29726, + "grad_norm": 0.9487228772582277, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 29726 + }, + { + "epoch": 0.29727, + "grad_norm": 0.8993539110861714, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 29727 + }, + { + "epoch": 0.29728, + "grad_norm": 0.8872459787741928, + "learning_rate": 0.003, + "loss": 4.027, + "step": 29728 + }, + { + "epoch": 0.29729, + "grad_norm": 0.8102720455776616, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 29729 + }, + { + "epoch": 0.2973, + "grad_norm": 0.82931269740283, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 29730 + }, + { + "epoch": 0.29731, + "grad_norm": 0.8484027158416328, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 29731 + }, + { + "epoch": 0.29732, + "grad_norm": 0.7507999923542177, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 29732 + }, + { + "epoch": 0.29733, + "grad_norm": 0.7412503509007977, + "learning_rate": 0.003, + "loss": 3.989, + "step": 29733 + }, + { + "epoch": 0.29734, + "grad_norm": 0.7812241470133353, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 29734 + }, + { + "epoch": 0.29735, + "grad_norm": 0.8164430629229563, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 29735 + }, + { + "epoch": 0.29736, + "grad_norm": 0.7590847351970819, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 29736 + }, + { + "epoch": 0.29737, + "grad_norm": 0.685911283811069, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 29737 + }, + { + "epoch": 0.29738, + "grad_norm": 0.697152367788975, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 29738 + }, + { + "epoch": 0.29739, + "grad_norm": 0.7550988612653562, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 29739 + }, + { + "epoch": 0.2974, + "grad_norm": 0.7981586280350925, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 29740 + }, + { + "epoch": 0.29741, + "grad_norm": 0.9343722234619154, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 29741 + }, + { + "epoch": 0.29742, + "grad_norm": 0.974310823744553, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 29742 + }, + { + "epoch": 0.29743, + "grad_norm": 1.0747110670586926, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 29743 + }, + { + "epoch": 0.29744, + "grad_norm": 0.8982224951498265, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 29744 + }, + { + "epoch": 0.29745, + "grad_norm": 0.6579506161751608, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 29745 + }, + { + "epoch": 0.29746, + "grad_norm": 0.6812610787652615, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 29746 + }, + { + "epoch": 0.29747, + "grad_norm": 0.8651463706038001, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 29747 + }, + { + "epoch": 0.29748, + "grad_norm": 0.9253598656089176, + "learning_rate": 0.003, + "loss": 4.027, + "step": 29748 + }, + { + "epoch": 0.29749, + "grad_norm": 0.8138993977645291, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 29749 + }, + { + "epoch": 0.2975, + "grad_norm": 0.7363080761278973, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 29750 + }, + { + "epoch": 0.29751, + "grad_norm": 0.7165403476644904, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 29751 + }, + { + "epoch": 0.29752, + "grad_norm": 0.6546254787441949, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 29752 + }, + { + "epoch": 0.29753, + "grad_norm": 0.743742582295265, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 29753 + }, + { + "epoch": 0.29754, + "grad_norm": 0.7692897115266751, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 29754 + }, + { + "epoch": 0.29755, + "grad_norm": 0.7591159214755335, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 29755 + }, + { + "epoch": 0.29756, + "grad_norm": 0.723883616857687, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 29756 + }, + { + "epoch": 0.29757, + "grad_norm": 0.796044022829638, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 29757 + }, + { + "epoch": 0.29758, + "grad_norm": 0.8726691158096135, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 29758 + }, + { + "epoch": 0.29759, + "grad_norm": 0.7960958376948193, + "learning_rate": 0.003, + "loss": 3.989, + "step": 29759 + }, + { + "epoch": 0.2976, + "grad_norm": 0.7119457585575291, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 29760 + }, + { + "epoch": 0.29761, + "grad_norm": 0.5398501777963406, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 29761 + }, + { + "epoch": 0.29762, + "grad_norm": 0.5110101629751014, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 29762 + }, + { + "epoch": 0.29763, + "grad_norm": 0.5199410530705035, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 29763 + }, + { + "epoch": 0.29764, + "grad_norm": 0.5545142992273853, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 29764 + }, + { + "epoch": 0.29765, + "grad_norm": 0.6023641555049718, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 29765 + }, + { + "epoch": 0.29766, + "grad_norm": 0.7377381329332523, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 29766 + }, + { + "epoch": 0.29767, + "grad_norm": 0.9444678387359768, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 29767 + }, + { + "epoch": 0.29768, + "grad_norm": 1.0996395306960653, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 29768 + }, + { + "epoch": 0.29769, + "grad_norm": 0.7199449965551312, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 29769 + }, + { + "epoch": 0.2977, + "grad_norm": 0.6909216234370619, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 29770 + }, + { + "epoch": 0.29771, + "grad_norm": 0.7742813471375707, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 29771 + }, + { + "epoch": 0.29772, + "grad_norm": 0.696602087157422, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 29772 + }, + { + "epoch": 0.29773, + "grad_norm": 0.6684624776010183, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 29773 + }, + { + "epoch": 0.29774, + "grad_norm": 0.7388295380501227, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 29774 + }, + { + "epoch": 0.29775, + "grad_norm": 0.7443167203313764, + "learning_rate": 0.003, + "loss": 4.007, + "step": 29775 + }, + { + "epoch": 0.29776, + "grad_norm": 0.6859672188523025, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 29776 + }, + { + "epoch": 0.29777, + "grad_norm": 0.742213831616602, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 29777 + }, + { + "epoch": 0.29778, + "grad_norm": 0.8155189966828785, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 29778 + }, + { + "epoch": 0.29779, + "grad_norm": 0.8255706285861468, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 29779 + }, + { + "epoch": 0.2978, + "grad_norm": 0.6516250920087489, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 29780 + }, + { + "epoch": 0.29781, + "grad_norm": 0.7648502930947454, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 29781 + }, + { + "epoch": 0.29782, + "grad_norm": 0.7835275518884292, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 29782 + }, + { + "epoch": 0.29783, + "grad_norm": 0.8755118478529397, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 29783 + }, + { + "epoch": 0.29784, + "grad_norm": 0.843461145729687, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 29784 + }, + { + "epoch": 0.29785, + "grad_norm": 0.869611595452407, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 29785 + }, + { + "epoch": 0.29786, + "grad_norm": 0.9293911407064324, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 29786 + }, + { + "epoch": 0.29787, + "grad_norm": 1.1101582497799025, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 29787 + }, + { + "epoch": 0.29788, + "grad_norm": 0.8799060878811077, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 29788 + }, + { + "epoch": 0.29789, + "grad_norm": 0.7766557815341788, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 29789 + }, + { + "epoch": 0.2979, + "grad_norm": 0.8113171290826647, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 29790 + }, + { + "epoch": 0.29791, + "grad_norm": 0.8118338852339844, + "learning_rate": 0.003, + "loss": 3.982, + "step": 29791 + }, + { + "epoch": 0.29792, + "grad_norm": 0.8951365799250071, + "learning_rate": 0.003, + "loss": 4.0473, + "step": 29792 + }, + { + "epoch": 0.29793, + "grad_norm": 1.0800647633257747, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 29793 + }, + { + "epoch": 0.29794, + "grad_norm": 0.8821377217638536, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 29794 + }, + { + "epoch": 0.29795, + "grad_norm": 0.8100173827094217, + "learning_rate": 0.003, + "loss": 4.038, + "step": 29795 + }, + { + "epoch": 0.29796, + "grad_norm": 0.7485260632558662, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 29796 + }, + { + "epoch": 0.29797, + "grad_norm": 0.6502638881027336, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 29797 + }, + { + "epoch": 0.29798, + "grad_norm": 0.7286660454143361, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 29798 + }, + { + "epoch": 0.29799, + "grad_norm": 0.7253378398293994, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 29799 + }, + { + "epoch": 0.298, + "grad_norm": 0.6664954023963797, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 29800 + }, + { + "epoch": 0.29801, + "grad_norm": 0.551336917506765, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 29801 + }, + { + "epoch": 0.29802, + "grad_norm": 0.6120397327974645, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 29802 + }, + { + "epoch": 0.29803, + "grad_norm": 0.6377280557086408, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 29803 + }, + { + "epoch": 0.29804, + "grad_norm": 0.7356826289958472, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 29804 + }, + { + "epoch": 0.29805, + "grad_norm": 0.8001641852645136, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 29805 + }, + { + "epoch": 0.29806, + "grad_norm": 0.7713137211425939, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 29806 + }, + { + "epoch": 0.29807, + "grad_norm": 0.7540311952074719, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 29807 + }, + { + "epoch": 0.29808, + "grad_norm": 0.7890565266989574, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 29808 + }, + { + "epoch": 0.29809, + "grad_norm": 0.8793971024656345, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 29809 + }, + { + "epoch": 0.2981, + "grad_norm": 1.0182228225199674, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 29810 + }, + { + "epoch": 0.29811, + "grad_norm": 1.0460931506951183, + "learning_rate": 0.003, + "loss": 4.018, + "step": 29811 + }, + { + "epoch": 0.29812, + "grad_norm": 0.9629139595253717, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 29812 + }, + { + "epoch": 0.29813, + "grad_norm": 0.9611559273793256, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 29813 + }, + { + "epoch": 0.29814, + "grad_norm": 0.8450720936980821, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 29814 + }, + { + "epoch": 0.29815, + "grad_norm": 0.7121651050137616, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 29815 + }, + { + "epoch": 0.29816, + "grad_norm": 0.6813900464698338, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 29816 + }, + { + "epoch": 0.29817, + "grad_norm": 0.7234756751103034, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 29817 + }, + { + "epoch": 0.29818, + "grad_norm": 0.6001751025730061, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 29818 + }, + { + "epoch": 0.29819, + "grad_norm": 0.5351554065071149, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 29819 + }, + { + "epoch": 0.2982, + "grad_norm": 0.5069134463353596, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 29820 + }, + { + "epoch": 0.29821, + "grad_norm": 0.5568818782170477, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 29821 + }, + { + "epoch": 0.29822, + "grad_norm": 0.6392747738172316, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 29822 + }, + { + "epoch": 0.29823, + "grad_norm": 0.7147164695854845, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 29823 + }, + { + "epoch": 0.29824, + "grad_norm": 0.7272362440859038, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 29824 + }, + { + "epoch": 0.29825, + "grad_norm": 0.696030332248936, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 29825 + }, + { + "epoch": 0.29826, + "grad_norm": 0.8155502741526491, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 29826 + }, + { + "epoch": 0.29827, + "grad_norm": 1.0358443932792887, + "learning_rate": 0.003, + "loss": 3.988, + "step": 29827 + }, + { + "epoch": 0.29828, + "grad_norm": 1.0311337704983996, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 29828 + }, + { + "epoch": 0.29829, + "grad_norm": 0.851019468958363, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 29829 + }, + { + "epoch": 0.2983, + "grad_norm": 0.8246711812756398, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 29830 + }, + { + "epoch": 0.29831, + "grad_norm": 0.8346979207926464, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 29831 + }, + { + "epoch": 0.29832, + "grad_norm": 0.7598594052057411, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 29832 + }, + { + "epoch": 0.29833, + "grad_norm": 0.7217303224750166, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 29833 + }, + { + "epoch": 0.29834, + "grad_norm": 0.7752469483204097, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 29834 + }, + { + "epoch": 0.29835, + "grad_norm": 0.7732583147542446, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 29835 + }, + { + "epoch": 0.29836, + "grad_norm": 0.7514043684133527, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 29836 + }, + { + "epoch": 0.29837, + "grad_norm": 0.7054481133833758, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 29837 + }, + { + "epoch": 0.29838, + "grad_norm": 0.8239677203756355, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 29838 + }, + { + "epoch": 0.29839, + "grad_norm": 0.8861161483655612, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 29839 + }, + { + "epoch": 0.2984, + "grad_norm": 0.8823849480327627, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 29840 + }, + { + "epoch": 0.29841, + "grad_norm": 0.7715167873920117, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 29841 + }, + { + "epoch": 0.29842, + "grad_norm": 0.6395394902051389, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 29842 + }, + { + "epoch": 0.29843, + "grad_norm": 0.6254653629264932, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 29843 + }, + { + "epoch": 0.29844, + "grad_norm": 0.7039573913111156, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 29844 + }, + { + "epoch": 0.29845, + "grad_norm": 0.7713179288407747, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 29845 + }, + { + "epoch": 0.29846, + "grad_norm": 0.6662895861301096, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 29846 + }, + { + "epoch": 0.29847, + "grad_norm": 0.5378946645708659, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 29847 + }, + { + "epoch": 0.29848, + "grad_norm": 0.7212337379019694, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 29848 + }, + { + "epoch": 0.29849, + "grad_norm": 0.823174713138789, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 29849 + }, + { + "epoch": 0.2985, + "grad_norm": 0.8376561344659573, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 29850 + }, + { + "epoch": 0.29851, + "grad_norm": 0.781208331614673, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 29851 + }, + { + "epoch": 0.29852, + "grad_norm": 0.7779444427745209, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 29852 + }, + { + "epoch": 0.29853, + "grad_norm": 0.7811332102716815, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 29853 + }, + { + "epoch": 0.29854, + "grad_norm": 0.8163451375595107, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 29854 + }, + { + "epoch": 0.29855, + "grad_norm": 0.8063527633326215, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 29855 + }, + { + "epoch": 0.29856, + "grad_norm": 0.749782293756497, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 29856 + }, + { + "epoch": 0.29857, + "grad_norm": 0.7847219010404428, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 29857 + }, + { + "epoch": 0.29858, + "grad_norm": 0.8112312612384585, + "learning_rate": 0.003, + "loss": 3.954, + "step": 29858 + }, + { + "epoch": 0.29859, + "grad_norm": 0.7977156908803495, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 29859 + }, + { + "epoch": 0.2986, + "grad_norm": 0.7441249288931017, + "learning_rate": 0.003, + "loss": 3.977, + "step": 29860 + }, + { + "epoch": 0.29861, + "grad_norm": 0.5900532617874586, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 29861 + }, + { + "epoch": 0.29862, + "grad_norm": 0.5453965802708699, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 29862 + }, + { + "epoch": 0.29863, + "grad_norm": 0.5733954142779315, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 29863 + }, + { + "epoch": 0.29864, + "grad_norm": 0.5797062108343227, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 29864 + }, + { + "epoch": 0.29865, + "grad_norm": 0.5343831268766733, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 29865 + }, + { + "epoch": 0.29866, + "grad_norm": 0.5683150840740924, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 29866 + }, + { + "epoch": 0.29867, + "grad_norm": 0.6740929510733804, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 29867 + }, + { + "epoch": 0.29868, + "grad_norm": 0.7534978948702152, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 29868 + }, + { + "epoch": 0.29869, + "grad_norm": 0.8956693224669559, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 29869 + }, + { + "epoch": 0.2987, + "grad_norm": 0.9109563696714692, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 29870 + }, + { + "epoch": 0.29871, + "grad_norm": 0.8175407817448251, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 29871 + }, + { + "epoch": 0.29872, + "grad_norm": 0.7525562467295968, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 29872 + }, + { + "epoch": 0.29873, + "grad_norm": 0.7746843476131947, + "learning_rate": 0.003, + "loss": 3.976, + "step": 29873 + }, + { + "epoch": 0.29874, + "grad_norm": 0.6962069145189886, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 29874 + }, + { + "epoch": 0.29875, + "grad_norm": 0.7899097140525796, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 29875 + }, + { + "epoch": 0.29876, + "grad_norm": 0.8252771958708744, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 29876 + }, + { + "epoch": 0.29877, + "grad_norm": 0.8523416137546495, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 29877 + }, + { + "epoch": 0.29878, + "grad_norm": 1.005132915228002, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 29878 + }, + { + "epoch": 0.29879, + "grad_norm": 0.9988923967765199, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 29879 + }, + { + "epoch": 0.2988, + "grad_norm": 0.9906501355430091, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 29880 + }, + { + "epoch": 0.29881, + "grad_norm": 0.9782186927533378, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 29881 + }, + { + "epoch": 0.29882, + "grad_norm": 1.1734301109755123, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 29882 + }, + { + "epoch": 0.29883, + "grad_norm": 0.9448907029818533, + "learning_rate": 0.003, + "loss": 4.038, + "step": 29883 + }, + { + "epoch": 0.29884, + "grad_norm": 0.9565100652443743, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 29884 + }, + { + "epoch": 0.29885, + "grad_norm": 1.1813015098696584, + "learning_rate": 0.003, + "loss": 4.01, + "step": 29885 + }, + { + "epoch": 0.29886, + "grad_norm": 0.9206819900887435, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 29886 + }, + { + "epoch": 0.29887, + "grad_norm": 0.8669482815662971, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 29887 + }, + { + "epoch": 0.29888, + "grad_norm": 0.8485900620262089, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 29888 + }, + { + "epoch": 0.29889, + "grad_norm": 0.7799559304696378, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 29889 + }, + { + "epoch": 0.2989, + "grad_norm": 0.8466748662837718, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 29890 + }, + { + "epoch": 0.29891, + "grad_norm": 0.8094465550522999, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 29891 + }, + { + "epoch": 0.29892, + "grad_norm": 0.8360536697742259, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 29892 + }, + { + "epoch": 0.29893, + "grad_norm": 0.8267815306501373, + "learning_rate": 0.003, + "loss": 4.006, + "step": 29893 + }, + { + "epoch": 0.29894, + "grad_norm": 0.7265244158612475, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 29894 + }, + { + "epoch": 0.29895, + "grad_norm": 0.682813947668754, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 29895 + }, + { + "epoch": 0.29896, + "grad_norm": 0.6837822502598946, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 29896 + }, + { + "epoch": 0.29897, + "grad_norm": 0.6069522036715657, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 29897 + }, + { + "epoch": 0.29898, + "grad_norm": 0.6155851105959538, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 29898 + }, + { + "epoch": 0.29899, + "grad_norm": 0.5326525812109024, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 29899 + }, + { + "epoch": 0.299, + "grad_norm": 0.5020706306192345, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 29900 + }, + { + "epoch": 0.29901, + "grad_norm": 0.5195723300261978, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 29901 + }, + { + "epoch": 0.29902, + "grad_norm": 0.5017749504037728, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 29902 + }, + { + "epoch": 0.29903, + "grad_norm": 0.4838890551285558, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 29903 + }, + { + "epoch": 0.29904, + "grad_norm": 0.5098617923565378, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 29904 + }, + { + "epoch": 0.29905, + "grad_norm": 0.5459549412581719, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 29905 + }, + { + "epoch": 0.29906, + "grad_norm": 0.5683064591043331, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 29906 + }, + { + "epoch": 0.29907, + "grad_norm": 0.673588071366819, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 29907 + }, + { + "epoch": 0.29908, + "grad_norm": 0.8545487715067771, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 29908 + }, + { + "epoch": 0.29909, + "grad_norm": 0.9584328644870139, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 29909 + }, + { + "epoch": 0.2991, + "grad_norm": 0.8487022685345835, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 29910 + }, + { + "epoch": 0.29911, + "grad_norm": 0.6671214485510703, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 29911 + }, + { + "epoch": 0.29912, + "grad_norm": 0.6965769906876333, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 29912 + }, + { + "epoch": 0.29913, + "grad_norm": 0.8224583185993534, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 29913 + }, + { + "epoch": 0.29914, + "grad_norm": 0.8086832692972399, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 29914 + }, + { + "epoch": 0.29915, + "grad_norm": 0.7625321255044515, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 29915 + }, + { + "epoch": 0.29916, + "grad_norm": 0.7358821159863004, + "learning_rate": 0.003, + "loss": 3.932, + "step": 29916 + }, + { + "epoch": 0.29917, + "grad_norm": 0.6778727161838243, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 29917 + }, + { + "epoch": 0.29918, + "grad_norm": 0.7282667152715312, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 29918 + }, + { + "epoch": 0.29919, + "grad_norm": 0.8532876703853164, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 29919 + }, + { + "epoch": 0.2992, + "grad_norm": 0.8315021840046077, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 29920 + }, + { + "epoch": 0.29921, + "grad_norm": 0.8130483075802699, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 29921 + }, + { + "epoch": 0.29922, + "grad_norm": 0.7288596430281918, + "learning_rate": 0.003, + "loss": 3.984, + "step": 29922 + }, + { + "epoch": 0.29923, + "grad_norm": 0.7597996264620208, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 29923 + }, + { + "epoch": 0.29924, + "grad_norm": 0.7490788411228562, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 29924 + }, + { + "epoch": 0.29925, + "grad_norm": 0.6779343239742595, + "learning_rate": 0.003, + "loss": 4.001, + "step": 29925 + }, + { + "epoch": 0.29926, + "grad_norm": 0.6193226083977592, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 29926 + }, + { + "epoch": 0.29927, + "grad_norm": 0.6896405006924692, + "learning_rate": 0.003, + "loss": 3.974, + "step": 29927 + }, + { + "epoch": 0.29928, + "grad_norm": 0.8953588092209958, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 29928 + }, + { + "epoch": 0.29929, + "grad_norm": 0.9833964066728615, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 29929 + }, + { + "epoch": 0.2993, + "grad_norm": 0.948981612115989, + "learning_rate": 0.003, + "loss": 4.002, + "step": 29930 + }, + { + "epoch": 0.29931, + "grad_norm": 0.9773204691787324, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 29931 + }, + { + "epoch": 0.29932, + "grad_norm": 1.0068389210531719, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 29932 + }, + { + "epoch": 0.29933, + "grad_norm": 1.1049364231476986, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29933 + }, + { + "epoch": 0.29934, + "grad_norm": 1.01303703100907, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 29934 + }, + { + "epoch": 0.29935, + "grad_norm": 1.0259061322359435, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 29935 + }, + { + "epoch": 0.29936, + "grad_norm": 1.0150262078724557, + "learning_rate": 0.003, + "loss": 4.0377, + "step": 29936 + }, + { + "epoch": 0.29937, + "grad_norm": 0.7934712260274286, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 29937 + }, + { + "epoch": 0.29938, + "grad_norm": 0.8132568610281233, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 29938 + }, + { + "epoch": 0.29939, + "grad_norm": 0.8328962200361225, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 29939 + }, + { + "epoch": 0.2994, + "grad_norm": 0.8480631399508928, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 29940 + }, + { + "epoch": 0.29941, + "grad_norm": 0.9127326581169811, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 29941 + }, + { + "epoch": 0.29942, + "grad_norm": 0.9293548087777963, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 29942 + }, + { + "epoch": 0.29943, + "grad_norm": 0.8236012832218548, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 29943 + }, + { + "epoch": 0.29944, + "grad_norm": 0.7734359466307473, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 29944 + }, + { + "epoch": 0.29945, + "grad_norm": 0.8052196682507663, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 29945 + }, + { + "epoch": 0.29946, + "grad_norm": 0.8348393037143126, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 29946 + }, + { + "epoch": 0.29947, + "grad_norm": 0.7420155977145342, + "learning_rate": 0.003, + "loss": 4.024, + "step": 29947 + }, + { + "epoch": 0.29948, + "grad_norm": 0.9122731899247507, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 29948 + }, + { + "epoch": 0.29949, + "grad_norm": 0.9153472761886788, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 29949 + }, + { + "epoch": 0.2995, + "grad_norm": 0.9006739484445714, + "learning_rate": 0.003, + "loss": 4.015, + "step": 29950 + }, + { + "epoch": 0.29951, + "grad_norm": 0.7749346974244455, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 29951 + }, + { + "epoch": 0.29952, + "grad_norm": 0.6641537056200056, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 29952 + }, + { + "epoch": 0.29953, + "grad_norm": 0.6137146757682171, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 29953 + }, + { + "epoch": 0.29954, + "grad_norm": 0.5908363717155409, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 29954 + }, + { + "epoch": 0.29955, + "grad_norm": 0.6421303009549055, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 29955 + }, + { + "epoch": 0.29956, + "grad_norm": 0.7065969273872108, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 29956 + }, + { + "epoch": 0.29957, + "grad_norm": 0.799738512855166, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 29957 + }, + { + "epoch": 0.29958, + "grad_norm": 0.8805343139585355, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 29958 + }, + { + "epoch": 0.29959, + "grad_norm": 0.8500799317694965, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 29959 + }, + { + "epoch": 0.2996, + "grad_norm": 0.8320107342063822, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 29960 + }, + { + "epoch": 0.29961, + "grad_norm": 0.650275990225366, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 29961 + }, + { + "epoch": 0.29962, + "grad_norm": 0.6157198222174834, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 29962 + }, + { + "epoch": 0.29963, + "grad_norm": 0.6275753518076715, + "learning_rate": 0.003, + "loss": 4.004, + "step": 29963 + }, + { + "epoch": 0.29964, + "grad_norm": 0.6118200959736985, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 29964 + }, + { + "epoch": 0.29965, + "grad_norm": 0.6158635964613296, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 29965 + }, + { + "epoch": 0.29966, + "grad_norm": 0.5678701639069615, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 29966 + }, + { + "epoch": 0.29967, + "grad_norm": 0.5695842740543302, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 29967 + }, + { + "epoch": 0.29968, + "grad_norm": 0.5680943633987341, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 29968 + }, + { + "epoch": 0.29969, + "grad_norm": 0.6811690294474869, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 29969 + }, + { + "epoch": 0.2997, + "grad_norm": 0.7351754171668877, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 29970 + }, + { + "epoch": 0.29971, + "grad_norm": 0.7307533973739845, + "learning_rate": 0.003, + "loss": 3.99, + "step": 29971 + }, + { + "epoch": 0.29972, + "grad_norm": 0.5978912472457922, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 29972 + }, + { + "epoch": 0.29973, + "grad_norm": 0.5663576701913347, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 29973 + }, + { + "epoch": 0.29974, + "grad_norm": 0.6858449628049071, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 29974 + }, + { + "epoch": 0.29975, + "grad_norm": 0.8460837877471055, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 29975 + }, + { + "epoch": 0.29976, + "grad_norm": 0.9645363837223622, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 29976 + }, + { + "epoch": 0.29977, + "grad_norm": 0.9175833418299107, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 29977 + }, + { + "epoch": 0.29978, + "grad_norm": 0.8062356643422123, + "learning_rate": 0.003, + "loss": 3.996, + "step": 29978 + }, + { + "epoch": 0.29979, + "grad_norm": 0.7833934163647951, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 29979 + }, + { + "epoch": 0.2998, + "grad_norm": 0.7178758045432687, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 29980 + }, + { + "epoch": 0.29981, + "grad_norm": 0.7238550649354063, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 29981 + }, + { + "epoch": 0.29982, + "grad_norm": 0.7209982382078313, + "learning_rate": 0.003, + "loss": 3.982, + "step": 29982 + }, + { + "epoch": 0.29983, + "grad_norm": 0.8493669139626769, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 29983 + }, + { + "epoch": 0.29984, + "grad_norm": 1.0205416454284553, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 29984 + }, + { + "epoch": 0.29985, + "grad_norm": 1.3209738389440902, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 29985 + }, + { + "epoch": 0.29986, + "grad_norm": 0.671292398200283, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 29986 + }, + { + "epoch": 0.29987, + "grad_norm": 0.7078327902826304, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 29987 + }, + { + "epoch": 0.29988, + "grad_norm": 0.6990390562355835, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 29988 + }, + { + "epoch": 0.29989, + "grad_norm": 0.7507410933347642, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 29989 + }, + { + "epoch": 0.2999, + "grad_norm": 0.7353559156741715, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 29990 + }, + { + "epoch": 0.29991, + "grad_norm": 0.6595195896586125, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 29991 + }, + { + "epoch": 0.29992, + "grad_norm": 0.6696604286038637, + "learning_rate": 0.003, + "loss": 3.985, + "step": 29992 + }, + { + "epoch": 0.29993, + "grad_norm": 0.6488629145775184, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 29993 + }, + { + "epoch": 0.29994, + "grad_norm": 0.5865466894865808, + "learning_rate": 0.003, + "loss": 4.015, + "step": 29994 + }, + { + "epoch": 0.29995, + "grad_norm": 0.668952246791295, + "learning_rate": 0.003, + "loss": 3.962, + "step": 29995 + }, + { + "epoch": 0.29996, + "grad_norm": 0.6699254122936233, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 29996 + }, + { + "epoch": 0.29997, + "grad_norm": 0.6001259533182426, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 29997 + }, + { + "epoch": 0.29998, + "grad_norm": 0.6766369150905024, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 29998 + }, + { + "epoch": 0.29999, + "grad_norm": 0.653497472484544, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 29999 + }, + { + "epoch": 0.3, + "grad_norm": 0.6439845299577579, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 30000 + }, + { + "epoch": 0.30001, + "grad_norm": 0.6961620837780017, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 30001 + }, + { + "epoch": 0.30002, + "grad_norm": 0.7927142914022126, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 30002 + }, + { + "epoch": 0.30003, + "grad_norm": 0.8630995122578984, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 30003 + }, + { + "epoch": 0.30004, + "grad_norm": 0.8436375008293103, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30004 + }, + { + "epoch": 0.30005, + "grad_norm": 0.7943299670327625, + "learning_rate": 0.003, + "loss": 3.949, + "step": 30005 + }, + { + "epoch": 0.30006, + "grad_norm": 0.8346944480688383, + "learning_rate": 0.003, + "loss": 3.983, + "step": 30006 + }, + { + "epoch": 0.30007, + "grad_norm": 0.7620790090402403, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 30007 + }, + { + "epoch": 0.30008, + "grad_norm": 0.7656112649339123, + "learning_rate": 0.003, + "loss": 3.975, + "step": 30008 + }, + { + "epoch": 0.30009, + "grad_norm": 0.9039714040428786, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 30009 + }, + { + "epoch": 0.3001, + "grad_norm": 1.0398134236121974, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 30010 + }, + { + "epoch": 0.30011, + "grad_norm": 0.9850697250214592, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 30011 + }, + { + "epoch": 0.30012, + "grad_norm": 1.0054117896516999, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 30012 + }, + { + "epoch": 0.30013, + "grad_norm": 0.8935533217195935, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 30013 + }, + { + "epoch": 0.30014, + "grad_norm": 0.6575162457074837, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 30014 + }, + { + "epoch": 0.30015, + "grad_norm": 0.6075439446857566, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 30015 + }, + { + "epoch": 0.30016, + "grad_norm": 0.5598576994118596, + "learning_rate": 0.003, + "loss": 4.027, + "step": 30016 + }, + { + "epoch": 0.30017, + "grad_norm": 0.70260923707676, + "learning_rate": 0.003, + "loss": 3.954, + "step": 30017 + }, + { + "epoch": 0.30018, + "grad_norm": 0.8270458784228392, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 30018 + }, + { + "epoch": 0.30019, + "grad_norm": 0.9424396564219576, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 30019 + }, + { + "epoch": 0.3002, + "grad_norm": 0.9468102876764259, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 30020 + }, + { + "epoch": 0.30021, + "grad_norm": 0.9956375488578177, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 30021 + }, + { + "epoch": 0.30022, + "grad_norm": 1.093371697970419, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 30022 + }, + { + "epoch": 0.30023, + "grad_norm": 0.8628666017660762, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 30023 + }, + { + "epoch": 0.30024, + "grad_norm": 1.0473125042265812, + "learning_rate": 0.003, + "loss": 4.0407, + "step": 30024 + }, + { + "epoch": 0.30025, + "grad_norm": 1.10055573360341, + "learning_rate": 0.003, + "loss": 4.032, + "step": 30025 + }, + { + "epoch": 0.30026, + "grad_norm": 0.8007996969655163, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 30026 + }, + { + "epoch": 0.30027, + "grad_norm": 0.7595666334287954, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 30027 + }, + { + "epoch": 0.30028, + "grad_norm": 0.6858356010831742, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 30028 + }, + { + "epoch": 0.30029, + "grad_norm": 0.5910142993724752, + "learning_rate": 0.003, + "loss": 3.988, + "step": 30029 + }, + { + "epoch": 0.3003, + "grad_norm": 0.6214657810026825, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 30030 + }, + { + "epoch": 0.30031, + "grad_norm": 0.5573475351741258, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 30031 + }, + { + "epoch": 0.30032, + "grad_norm": 0.5550170572140689, + "learning_rate": 0.003, + "loss": 3.985, + "step": 30032 + }, + { + "epoch": 0.30033, + "grad_norm": 0.5017560362096154, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 30033 + }, + { + "epoch": 0.30034, + "grad_norm": 0.556883049810585, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 30034 + }, + { + "epoch": 0.30035, + "grad_norm": 0.7367456828414336, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 30035 + }, + { + "epoch": 0.30036, + "grad_norm": 0.784226284854732, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 30036 + }, + { + "epoch": 0.30037, + "grad_norm": 0.7665932267805667, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 30037 + }, + { + "epoch": 0.30038, + "grad_norm": 0.8302525227557221, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 30038 + }, + { + "epoch": 0.30039, + "grad_norm": 0.9612219344373203, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 30039 + }, + { + "epoch": 0.3004, + "grad_norm": 0.9619389938414729, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 30040 + }, + { + "epoch": 0.30041, + "grad_norm": 0.7546992358989103, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 30041 + }, + { + "epoch": 0.30042, + "grad_norm": 0.6622313289362246, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 30042 + }, + { + "epoch": 0.30043, + "grad_norm": 0.6924771716129254, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 30043 + }, + { + "epoch": 0.30044, + "grad_norm": 0.7501660266377205, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 30044 + }, + { + "epoch": 0.30045, + "grad_norm": 0.7156603068125221, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 30045 + }, + { + "epoch": 0.30046, + "grad_norm": 0.7024657872215043, + "learning_rate": 0.003, + "loss": 4.001, + "step": 30046 + }, + { + "epoch": 0.30047, + "grad_norm": 0.6811128944204005, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 30047 + }, + { + "epoch": 0.30048, + "grad_norm": 0.7653874491084857, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 30048 + }, + { + "epoch": 0.30049, + "grad_norm": 0.8700390538657732, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 30049 + }, + { + "epoch": 0.3005, + "grad_norm": 0.9087275859223113, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 30050 + }, + { + "epoch": 0.30051, + "grad_norm": 0.9715808568261085, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 30051 + }, + { + "epoch": 0.30052, + "grad_norm": 0.7987233572849599, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 30052 + }, + { + "epoch": 0.30053, + "grad_norm": 0.6951798035274899, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 30053 + }, + { + "epoch": 0.30054, + "grad_norm": 0.7199929162716221, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 30054 + }, + { + "epoch": 0.30055, + "grad_norm": 0.7469349513144613, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 30055 + }, + { + "epoch": 0.30056, + "grad_norm": 0.7461775162311183, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 30056 + }, + { + "epoch": 0.30057, + "grad_norm": 0.7767639610584712, + "learning_rate": 0.003, + "loss": 3.983, + "step": 30057 + }, + { + "epoch": 0.30058, + "grad_norm": 0.8803067163821293, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 30058 + }, + { + "epoch": 0.30059, + "grad_norm": 0.8638694354531968, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 30059 + }, + { + "epoch": 0.3006, + "grad_norm": 0.8119399577347823, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 30060 + }, + { + "epoch": 0.30061, + "grad_norm": 0.8562517383885385, + "learning_rate": 0.003, + "loss": 3.994, + "step": 30061 + }, + { + "epoch": 0.30062, + "grad_norm": 0.7930379262016196, + "learning_rate": 0.003, + "loss": 3.984, + "step": 30062 + }, + { + "epoch": 0.30063, + "grad_norm": 0.8161180023442862, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 30063 + }, + { + "epoch": 0.30064, + "grad_norm": 0.7561554880401449, + "learning_rate": 0.003, + "loss": 3.997, + "step": 30064 + }, + { + "epoch": 0.30065, + "grad_norm": 0.6891166672889701, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 30065 + }, + { + "epoch": 0.30066, + "grad_norm": 0.6692723894346025, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 30066 + }, + { + "epoch": 0.30067, + "grad_norm": 0.7551600328988065, + "learning_rate": 0.003, + "loss": 3.981, + "step": 30067 + }, + { + "epoch": 0.30068, + "grad_norm": 0.8419954523051967, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 30068 + }, + { + "epoch": 0.30069, + "grad_norm": 0.9447764710293198, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 30069 + }, + { + "epoch": 0.3007, + "grad_norm": 0.9235885036918561, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 30070 + }, + { + "epoch": 0.30071, + "grad_norm": 0.7754393198497436, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 30071 + }, + { + "epoch": 0.30072, + "grad_norm": 0.7975645105482116, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 30072 + }, + { + "epoch": 0.30073, + "grad_norm": 0.718113078737638, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 30073 + }, + { + "epoch": 0.30074, + "grad_norm": 0.7243475443241448, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 30074 + }, + { + "epoch": 0.30075, + "grad_norm": 0.7552917214838335, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 30075 + }, + { + "epoch": 0.30076, + "grad_norm": 0.6766982925395884, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 30076 + }, + { + "epoch": 0.30077, + "grad_norm": 0.6132188417847071, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 30077 + }, + { + "epoch": 0.30078, + "grad_norm": 0.596816326745439, + "learning_rate": 0.003, + "loss": 3.973, + "step": 30078 + }, + { + "epoch": 0.30079, + "grad_norm": 0.5477780622722308, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 30079 + }, + { + "epoch": 0.3008, + "grad_norm": 0.6057529450414799, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 30080 + }, + { + "epoch": 0.30081, + "grad_norm": 0.6855077680054497, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 30081 + }, + { + "epoch": 0.30082, + "grad_norm": 0.7618026968749335, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 30082 + }, + { + "epoch": 0.30083, + "grad_norm": 0.8559015450339608, + "learning_rate": 0.003, + "loss": 3.963, + "step": 30083 + }, + { + "epoch": 0.30084, + "grad_norm": 0.935756362212325, + "learning_rate": 0.003, + "loss": 4.022, + "step": 30084 + }, + { + "epoch": 0.30085, + "grad_norm": 0.8600747748969391, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 30085 + }, + { + "epoch": 0.30086, + "grad_norm": 0.6961485301018311, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 30086 + }, + { + "epoch": 0.30087, + "grad_norm": 0.6768972646255654, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 30087 + }, + { + "epoch": 0.30088, + "grad_norm": 0.6932940233967243, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 30088 + }, + { + "epoch": 0.30089, + "grad_norm": 0.6501458236420654, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 30089 + }, + { + "epoch": 0.3009, + "grad_norm": 0.6083744278969858, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 30090 + }, + { + "epoch": 0.30091, + "grad_norm": 0.643250365881438, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 30091 + }, + { + "epoch": 0.30092, + "grad_norm": 0.6958887441386606, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 30092 + }, + { + "epoch": 0.30093, + "grad_norm": 0.6955255586375872, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 30093 + }, + { + "epoch": 0.30094, + "grad_norm": 0.7137429065143671, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30094 + }, + { + "epoch": 0.30095, + "grad_norm": 0.7957135412209311, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 30095 + }, + { + "epoch": 0.30096, + "grad_norm": 0.8169152779252569, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 30096 + }, + { + "epoch": 0.30097, + "grad_norm": 0.8205624812458764, + "learning_rate": 0.003, + "loss": 4.013, + "step": 30097 + }, + { + "epoch": 0.30098, + "grad_norm": 0.883870559965492, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 30098 + }, + { + "epoch": 0.30099, + "grad_norm": 0.9516097668672426, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 30099 + }, + { + "epoch": 0.301, + "grad_norm": 1.0238648730413704, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 30100 + }, + { + "epoch": 0.30101, + "grad_norm": 0.9088651172281471, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 30101 + }, + { + "epoch": 0.30102, + "grad_norm": 0.7703966697130435, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 30102 + }, + { + "epoch": 0.30103, + "grad_norm": 0.7082746439692948, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 30103 + }, + { + "epoch": 0.30104, + "grad_norm": 0.6970098219456945, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 30104 + }, + { + "epoch": 0.30105, + "grad_norm": 0.6974627377489221, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 30105 + }, + { + "epoch": 0.30106, + "grad_norm": 0.7444957780822393, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 30106 + }, + { + "epoch": 0.30107, + "grad_norm": 0.8738742606377102, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 30107 + }, + { + "epoch": 0.30108, + "grad_norm": 1.0531614025311264, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 30108 + }, + { + "epoch": 0.30109, + "grad_norm": 1.1353790410385634, + "learning_rate": 0.003, + "loss": 3.983, + "step": 30109 + }, + { + "epoch": 0.3011, + "grad_norm": 0.7583438814947177, + "learning_rate": 0.003, + "loss": 3.995, + "step": 30110 + }, + { + "epoch": 0.30111, + "grad_norm": 0.629322998867449, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30111 + }, + { + "epoch": 0.30112, + "grad_norm": 0.5648745967691933, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 30112 + }, + { + "epoch": 0.30113, + "grad_norm": 0.5497813901137515, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 30113 + }, + { + "epoch": 0.30114, + "grad_norm": 0.5806342749753122, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 30114 + }, + { + "epoch": 0.30115, + "grad_norm": 0.6562337018016443, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 30115 + }, + { + "epoch": 0.30116, + "grad_norm": 0.7789470323034725, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 30116 + }, + { + "epoch": 0.30117, + "grad_norm": 0.9506711430776227, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 30117 + }, + { + "epoch": 0.30118, + "grad_norm": 0.8745608085792244, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 30118 + }, + { + "epoch": 0.30119, + "grad_norm": 0.7163970200362251, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30119 + }, + { + "epoch": 0.3012, + "grad_norm": 0.7759328977001063, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 30120 + }, + { + "epoch": 0.30121, + "grad_norm": 0.9273416721026887, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 30121 + }, + { + "epoch": 0.30122, + "grad_norm": 1.0012687680974193, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 30122 + }, + { + "epoch": 0.30123, + "grad_norm": 0.912654123837061, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 30123 + }, + { + "epoch": 0.30124, + "grad_norm": 0.8957113394984386, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 30124 + }, + { + "epoch": 0.30125, + "grad_norm": 0.9034291266233373, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 30125 + }, + { + "epoch": 0.30126, + "grad_norm": 0.8322413439340592, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 30126 + }, + { + "epoch": 0.30127, + "grad_norm": 0.77921797063722, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 30127 + }, + { + "epoch": 0.30128, + "grad_norm": 0.6658265911706874, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 30128 + }, + { + "epoch": 0.30129, + "grad_norm": 0.675903222010711, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 30129 + }, + { + "epoch": 0.3013, + "grad_norm": 0.7682623811697116, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 30130 + }, + { + "epoch": 0.30131, + "grad_norm": 0.7869339003761044, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 30131 + }, + { + "epoch": 0.30132, + "grad_norm": 0.7013744387312081, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 30132 + }, + { + "epoch": 0.30133, + "grad_norm": 0.6701938916211517, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 30133 + }, + { + "epoch": 0.30134, + "grad_norm": 0.7587445719050533, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 30134 + }, + { + "epoch": 0.30135, + "grad_norm": 0.886052602108762, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 30135 + }, + { + "epoch": 0.30136, + "grad_norm": 1.0172229843055278, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 30136 + }, + { + "epoch": 0.30137, + "grad_norm": 0.8946011400924254, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 30137 + }, + { + "epoch": 0.30138, + "grad_norm": 0.8472766082236914, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 30138 + }, + { + "epoch": 0.30139, + "grad_norm": 0.7872966805264777, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 30139 + }, + { + "epoch": 0.3014, + "grad_norm": 0.7841967562664939, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 30140 + }, + { + "epoch": 0.30141, + "grad_norm": 0.6709980154475167, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 30141 + }, + { + "epoch": 0.30142, + "grad_norm": 0.6577574662959877, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 30142 + }, + { + "epoch": 0.30143, + "grad_norm": 0.610256277658657, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 30143 + }, + { + "epoch": 0.30144, + "grad_norm": 0.6309288024746091, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 30144 + }, + { + "epoch": 0.30145, + "grad_norm": 0.592944281372117, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 30145 + }, + { + "epoch": 0.30146, + "grad_norm": 0.5945483608568751, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 30146 + }, + { + "epoch": 0.30147, + "grad_norm": 0.5797720078794766, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 30147 + }, + { + "epoch": 0.30148, + "grad_norm": 0.5800543676625537, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 30148 + }, + { + "epoch": 0.30149, + "grad_norm": 0.583074117818598, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 30149 + }, + { + "epoch": 0.3015, + "grad_norm": 0.6615627709155782, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 30150 + }, + { + "epoch": 0.30151, + "grad_norm": 0.7101674876258556, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 30151 + }, + { + "epoch": 0.30152, + "grad_norm": 0.7757336814678487, + "learning_rate": 0.003, + "loss": 3.975, + "step": 30152 + }, + { + "epoch": 0.30153, + "grad_norm": 0.7355361006196114, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 30153 + }, + { + "epoch": 0.30154, + "grad_norm": 0.7603263387452995, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 30154 + }, + { + "epoch": 0.30155, + "grad_norm": 1.0069063173882915, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 30155 + }, + { + "epoch": 0.30156, + "grad_norm": 1.0271949852095934, + "learning_rate": 0.003, + "loss": 4.0, + "step": 30156 + }, + { + "epoch": 0.30157, + "grad_norm": 1.0258399641300866, + "learning_rate": 0.003, + "loss": 3.996, + "step": 30157 + }, + { + "epoch": 0.30158, + "grad_norm": 1.0499561518995617, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 30158 + }, + { + "epoch": 0.30159, + "grad_norm": 0.9531170448665813, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 30159 + }, + { + "epoch": 0.3016, + "grad_norm": 0.8373319023763507, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 30160 + }, + { + "epoch": 0.30161, + "grad_norm": 0.8609778612777761, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 30161 + }, + { + "epoch": 0.30162, + "grad_norm": 0.852555835861929, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 30162 + }, + { + "epoch": 0.30163, + "grad_norm": 0.8680997978274002, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 30163 + }, + { + "epoch": 0.30164, + "grad_norm": 1.1123603681186351, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 30164 + }, + { + "epoch": 0.30165, + "grad_norm": 1.0008788705774294, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 30165 + }, + { + "epoch": 0.30166, + "grad_norm": 0.9933395763638699, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 30166 + }, + { + "epoch": 0.30167, + "grad_norm": 0.9865730162447819, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 30167 + }, + { + "epoch": 0.30168, + "grad_norm": 0.8012517845142783, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 30168 + }, + { + "epoch": 0.30169, + "grad_norm": 0.7843605840120103, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 30169 + }, + { + "epoch": 0.3017, + "grad_norm": 0.8441895305379521, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 30170 + }, + { + "epoch": 0.30171, + "grad_norm": 0.8508792484711775, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 30171 + }, + { + "epoch": 0.30172, + "grad_norm": 0.9040343864330485, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 30172 + }, + { + "epoch": 0.30173, + "grad_norm": 0.8112085614213105, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 30173 + }, + { + "epoch": 0.30174, + "grad_norm": 0.7617684975033521, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 30174 + }, + { + "epoch": 0.30175, + "grad_norm": 0.703484897090898, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30175 + }, + { + "epoch": 0.30176, + "grad_norm": 0.7325298338639081, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 30176 + }, + { + "epoch": 0.30177, + "grad_norm": 0.7215420454294346, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 30177 + }, + { + "epoch": 0.30178, + "grad_norm": 0.6671169680241664, + "learning_rate": 0.003, + "loss": 3.991, + "step": 30178 + }, + { + "epoch": 0.30179, + "grad_norm": 0.6068590137371891, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 30179 + }, + { + "epoch": 0.3018, + "grad_norm": 0.5328173774324283, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 30180 + }, + { + "epoch": 0.30181, + "grad_norm": 0.5088918196128543, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 30181 + }, + { + "epoch": 0.30182, + "grad_norm": 0.5309115219326468, + "learning_rate": 0.003, + "loss": 4.014, + "step": 30182 + }, + { + "epoch": 0.30183, + "grad_norm": 0.5975275633484763, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 30183 + }, + { + "epoch": 0.30184, + "grad_norm": 0.7134282199462779, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 30184 + }, + { + "epoch": 0.30185, + "grad_norm": 0.8049890129988431, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 30185 + }, + { + "epoch": 0.30186, + "grad_norm": 0.8778645821270376, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 30186 + }, + { + "epoch": 0.30187, + "grad_norm": 0.8607102385188174, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 30187 + }, + { + "epoch": 0.30188, + "grad_norm": 0.8031971657720082, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 30188 + }, + { + "epoch": 0.30189, + "grad_norm": 0.8843687126555072, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 30189 + }, + { + "epoch": 0.3019, + "grad_norm": 0.9094504380237711, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 30190 + }, + { + "epoch": 0.30191, + "grad_norm": 0.7725643779261337, + "learning_rate": 0.003, + "loss": 3.994, + "step": 30191 + }, + { + "epoch": 0.30192, + "grad_norm": 0.672242861353203, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 30192 + }, + { + "epoch": 0.30193, + "grad_norm": 0.6589218451290599, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 30193 + }, + { + "epoch": 0.30194, + "grad_norm": 0.6358621301509384, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 30194 + }, + { + "epoch": 0.30195, + "grad_norm": 0.6754381885152086, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 30195 + }, + { + "epoch": 0.30196, + "grad_norm": 0.75582368640997, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 30196 + }, + { + "epoch": 0.30197, + "grad_norm": 0.78253885295307, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 30197 + }, + { + "epoch": 0.30198, + "grad_norm": 0.7617020851883801, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 30198 + }, + { + "epoch": 0.30199, + "grad_norm": 0.7815906908388953, + "learning_rate": 0.003, + "loss": 3.993, + "step": 30199 + }, + { + "epoch": 0.302, + "grad_norm": 0.6840348076981995, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 30200 + }, + { + "epoch": 0.30201, + "grad_norm": 0.6665389298509593, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 30201 + }, + { + "epoch": 0.30202, + "grad_norm": 0.6323200090900477, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 30202 + }, + { + "epoch": 0.30203, + "grad_norm": 0.6250762600307734, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 30203 + }, + { + "epoch": 0.30204, + "grad_norm": 0.6711241990314311, + "learning_rate": 0.003, + "loss": 3.983, + "step": 30204 + }, + { + "epoch": 0.30205, + "grad_norm": 0.7103582867126483, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 30205 + }, + { + "epoch": 0.30206, + "grad_norm": 0.6888448041925725, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 30206 + }, + { + "epoch": 0.30207, + "grad_norm": 0.6966891240579459, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 30207 + }, + { + "epoch": 0.30208, + "grad_norm": 0.7713621571166682, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 30208 + }, + { + "epoch": 0.30209, + "grad_norm": 0.7202194396400461, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 30209 + }, + { + "epoch": 0.3021, + "grad_norm": 0.7801646221801394, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 30210 + }, + { + "epoch": 0.30211, + "grad_norm": 0.8573604549669821, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 30211 + }, + { + "epoch": 0.30212, + "grad_norm": 0.9436672626016859, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 30212 + }, + { + "epoch": 0.30213, + "grad_norm": 0.8813963447986417, + "learning_rate": 0.003, + "loss": 3.976, + "step": 30213 + }, + { + "epoch": 0.30214, + "grad_norm": 0.7922618719008149, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 30214 + }, + { + "epoch": 0.30215, + "grad_norm": 0.7955631840372563, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 30215 + }, + { + "epoch": 0.30216, + "grad_norm": 0.8695886570388854, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 30216 + }, + { + "epoch": 0.30217, + "grad_norm": 0.8622060907774539, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 30217 + }, + { + "epoch": 0.30218, + "grad_norm": 0.7942732288646652, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 30218 + }, + { + "epoch": 0.30219, + "grad_norm": 0.7629338873666387, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 30219 + }, + { + "epoch": 0.3022, + "grad_norm": 0.7352666623546226, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 30220 + }, + { + "epoch": 0.30221, + "grad_norm": 0.854008105052202, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 30221 + }, + { + "epoch": 0.30222, + "grad_norm": 1.0360232447567759, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 30222 + }, + { + "epoch": 0.30223, + "grad_norm": 1.231533582913215, + "learning_rate": 0.003, + "loss": 4.014, + "step": 30223 + }, + { + "epoch": 0.30224, + "grad_norm": 0.8494582019109707, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 30224 + }, + { + "epoch": 0.30225, + "grad_norm": 0.7296891289997036, + "learning_rate": 0.003, + "loss": 3.979, + "step": 30225 + }, + { + "epoch": 0.30226, + "grad_norm": 0.6861972440233933, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 30226 + }, + { + "epoch": 0.30227, + "grad_norm": 0.7004309934556224, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 30227 + }, + { + "epoch": 0.30228, + "grad_norm": 0.727748364505684, + "learning_rate": 0.003, + "loss": 4.007, + "step": 30228 + }, + { + "epoch": 0.30229, + "grad_norm": 0.808248287307748, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 30229 + }, + { + "epoch": 0.3023, + "grad_norm": 0.9117001883045328, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 30230 + }, + { + "epoch": 0.30231, + "grad_norm": 1.0076688387809654, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 30231 + }, + { + "epoch": 0.30232, + "grad_norm": 0.8710063695910324, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 30232 + }, + { + "epoch": 0.30233, + "grad_norm": 0.7451878967568036, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 30233 + }, + { + "epoch": 0.30234, + "grad_norm": 0.7853977869286831, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 30234 + }, + { + "epoch": 0.30235, + "grad_norm": 0.7067033824646978, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 30235 + }, + { + "epoch": 0.30236, + "grad_norm": 0.6974435159819622, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 30236 + }, + { + "epoch": 0.30237, + "grad_norm": 0.6331566349621223, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 30237 + }, + { + "epoch": 0.30238, + "grad_norm": 0.6188604851995884, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 30238 + }, + { + "epoch": 0.30239, + "grad_norm": 0.5605869396078661, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 30239 + }, + { + "epoch": 0.3024, + "grad_norm": 0.5714466679990843, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 30240 + }, + { + "epoch": 0.30241, + "grad_norm": 0.573233859618262, + "learning_rate": 0.003, + "loss": 3.992, + "step": 30241 + }, + { + "epoch": 0.30242, + "grad_norm": 0.6232775846064799, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 30242 + }, + { + "epoch": 0.30243, + "grad_norm": 0.6592046920603459, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 30243 + }, + { + "epoch": 0.30244, + "grad_norm": 0.576682608137744, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 30244 + }, + { + "epoch": 0.30245, + "grad_norm": 0.5632588984071242, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 30245 + }, + { + "epoch": 0.30246, + "grad_norm": 0.6072451195616769, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 30246 + }, + { + "epoch": 0.30247, + "grad_norm": 0.6827813256941104, + "learning_rate": 0.003, + "loss": 3.997, + "step": 30247 + }, + { + "epoch": 0.30248, + "grad_norm": 0.6830571013362517, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 30248 + }, + { + "epoch": 0.30249, + "grad_norm": 0.7720516794370517, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 30249 + }, + { + "epoch": 0.3025, + "grad_norm": 0.8507153765243477, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 30250 + }, + { + "epoch": 0.30251, + "grad_norm": 1.0564032327871855, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 30251 + }, + { + "epoch": 0.30252, + "grad_norm": 1.2191758893384426, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 30252 + }, + { + "epoch": 0.30253, + "grad_norm": 0.6956946832259574, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 30253 + }, + { + "epoch": 0.30254, + "grad_norm": 0.6352133690079766, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 30254 + }, + { + "epoch": 0.30255, + "grad_norm": 0.7548707348000436, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 30255 + }, + { + "epoch": 0.30256, + "grad_norm": 0.9160925570378481, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 30256 + }, + { + "epoch": 0.30257, + "grad_norm": 1.0685154662533536, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 30257 + }, + { + "epoch": 0.30258, + "grad_norm": 0.797932325758701, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 30258 + }, + { + "epoch": 0.30259, + "grad_norm": 0.8396141947368342, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 30259 + }, + { + "epoch": 0.3026, + "grad_norm": 0.9607493041307061, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 30260 + }, + { + "epoch": 0.30261, + "grad_norm": 1.0651402290414085, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 30261 + }, + { + "epoch": 0.30262, + "grad_norm": 0.8307552845809926, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30262 + }, + { + "epoch": 0.30263, + "grad_norm": 0.630350032680271, + "learning_rate": 0.003, + "loss": 3.976, + "step": 30263 + }, + { + "epoch": 0.30264, + "grad_norm": 0.6462488556754056, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 30264 + }, + { + "epoch": 0.30265, + "grad_norm": 0.7457437122609921, + "learning_rate": 0.003, + "loss": 4.0414, + "step": 30265 + }, + { + "epoch": 0.30266, + "grad_norm": 0.8639443166033877, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 30266 + }, + { + "epoch": 0.30267, + "grad_norm": 0.91810616381469, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 30267 + }, + { + "epoch": 0.30268, + "grad_norm": 0.8245401104003856, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 30268 + }, + { + "epoch": 0.30269, + "grad_norm": 0.7850125819364692, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 30269 + }, + { + "epoch": 0.3027, + "grad_norm": 0.8210028603736007, + "learning_rate": 0.003, + "loss": 3.999, + "step": 30270 + }, + { + "epoch": 0.30271, + "grad_norm": 0.9151327477587726, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 30271 + }, + { + "epoch": 0.30272, + "grad_norm": 0.8063743813636524, + "learning_rate": 0.003, + "loss": 4.0444, + "step": 30272 + }, + { + "epoch": 0.30273, + "grad_norm": 0.6918603280098903, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 30273 + }, + { + "epoch": 0.30274, + "grad_norm": 0.6713281180406461, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 30274 + }, + { + "epoch": 0.30275, + "grad_norm": 0.719721011721959, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 30275 + }, + { + "epoch": 0.30276, + "grad_norm": 0.8089255724326119, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 30276 + }, + { + "epoch": 0.30277, + "grad_norm": 0.8222632857206488, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 30277 + }, + { + "epoch": 0.30278, + "grad_norm": 0.8370229680471781, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 30278 + }, + { + "epoch": 0.30279, + "grad_norm": 0.729685637564891, + "learning_rate": 0.003, + "loss": 4.0285, + "step": 30279 + }, + { + "epoch": 0.3028, + "grad_norm": 0.6971049931545138, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 30280 + }, + { + "epoch": 0.30281, + "grad_norm": 0.7439095584979601, + "learning_rate": 0.003, + "loss": 3.948, + "step": 30281 + }, + { + "epoch": 0.30282, + "grad_norm": 0.7566913471533806, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 30282 + }, + { + "epoch": 0.30283, + "grad_norm": 0.6998986627655563, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 30283 + }, + { + "epoch": 0.30284, + "grad_norm": 0.721540928518684, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 30284 + }, + { + "epoch": 0.30285, + "grad_norm": 0.6725844626934668, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 30285 + }, + { + "epoch": 0.30286, + "grad_norm": 0.640693391789034, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 30286 + }, + { + "epoch": 0.30287, + "grad_norm": 0.5881875698586054, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 30287 + }, + { + "epoch": 0.30288, + "grad_norm": 0.546310962595517, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 30288 + }, + { + "epoch": 0.30289, + "grad_norm": 0.5568163418858326, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 30289 + }, + { + "epoch": 0.3029, + "grad_norm": 0.5866135054953716, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 30290 + }, + { + "epoch": 0.30291, + "grad_norm": 0.6809348019895972, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 30291 + }, + { + "epoch": 0.30292, + "grad_norm": 0.7400906554904669, + "learning_rate": 0.003, + "loss": 3.997, + "step": 30292 + }, + { + "epoch": 0.30293, + "grad_norm": 0.7274196940679686, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 30293 + }, + { + "epoch": 0.30294, + "grad_norm": 0.7500785904030032, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 30294 + }, + { + "epoch": 0.30295, + "grad_norm": 0.7172583443002554, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 30295 + }, + { + "epoch": 0.30296, + "grad_norm": 0.727311969027601, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 30296 + }, + { + "epoch": 0.30297, + "grad_norm": 0.7967816155204687, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 30297 + }, + { + "epoch": 0.30298, + "grad_norm": 0.9451261568585126, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 30298 + }, + { + "epoch": 0.30299, + "grad_norm": 1.0637776826888272, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 30299 + }, + { + "epoch": 0.303, + "grad_norm": 0.8388328340078741, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 30300 + }, + { + "epoch": 0.30301, + "grad_norm": 0.778034922506815, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 30301 + }, + { + "epoch": 0.30302, + "grad_norm": 0.6745840577668889, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 30302 + }, + { + "epoch": 0.30303, + "grad_norm": 0.7845784426653298, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 30303 + }, + { + "epoch": 0.30304, + "grad_norm": 0.8633623754016247, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 30304 + }, + { + "epoch": 0.30305, + "grad_norm": 0.8880286127801011, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 30305 + }, + { + "epoch": 0.30306, + "grad_norm": 0.9269619923835647, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 30306 + }, + { + "epoch": 0.30307, + "grad_norm": 0.9377992413154499, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 30307 + }, + { + "epoch": 0.30308, + "grad_norm": 0.8817889366996027, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 30308 + }, + { + "epoch": 0.30309, + "grad_norm": 0.9226527738704356, + "learning_rate": 0.003, + "loss": 4.018, + "step": 30309 + }, + { + "epoch": 0.3031, + "grad_norm": 0.9500812372739582, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 30310 + }, + { + "epoch": 0.30311, + "grad_norm": 0.8166077507734797, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 30311 + }, + { + "epoch": 0.30312, + "grad_norm": 0.7295558623337065, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 30312 + }, + { + "epoch": 0.30313, + "grad_norm": 0.7375025993468542, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 30313 + }, + { + "epoch": 0.30314, + "grad_norm": 0.6732998354570615, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 30314 + }, + { + "epoch": 0.30315, + "grad_norm": 0.7490441816375534, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 30315 + }, + { + "epoch": 0.30316, + "grad_norm": 0.7487801978400414, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 30316 + }, + { + "epoch": 0.30317, + "grad_norm": 0.8916739394931498, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 30317 + }, + { + "epoch": 0.30318, + "grad_norm": 1.0639841744716507, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 30318 + }, + { + "epoch": 0.30319, + "grad_norm": 1.0716091710378446, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 30319 + }, + { + "epoch": 0.3032, + "grad_norm": 0.8335211956813008, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 30320 + }, + { + "epoch": 0.30321, + "grad_norm": 0.7010097932729493, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 30321 + }, + { + "epoch": 0.30322, + "grad_norm": 0.7407931243395433, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 30322 + }, + { + "epoch": 0.30323, + "grad_norm": 0.8608802772682541, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 30323 + }, + { + "epoch": 0.30324, + "grad_norm": 0.8792687520986479, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 30324 + }, + { + "epoch": 0.30325, + "grad_norm": 0.7051142181285136, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 30325 + }, + { + "epoch": 0.30326, + "grad_norm": 0.7337788660168967, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 30326 + }, + { + "epoch": 0.30327, + "grad_norm": 0.7079642624614239, + "learning_rate": 0.003, + "loss": 3.988, + "step": 30327 + }, + { + "epoch": 0.30328, + "grad_norm": 0.5809232543598487, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 30328 + }, + { + "epoch": 0.30329, + "grad_norm": 0.6230241979496085, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 30329 + }, + { + "epoch": 0.3033, + "grad_norm": 0.7405799180536967, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 30330 + }, + { + "epoch": 0.30331, + "grad_norm": 0.8013948705197442, + "learning_rate": 0.003, + "loss": 4.0378, + "step": 30331 + }, + { + "epoch": 0.30332, + "grad_norm": 0.901208130346914, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 30332 + }, + { + "epoch": 0.30333, + "grad_norm": 0.9144170262523762, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 30333 + }, + { + "epoch": 0.30334, + "grad_norm": 0.8747259131882942, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 30334 + }, + { + "epoch": 0.30335, + "grad_norm": 0.7584683271843689, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 30335 + }, + { + "epoch": 0.30336, + "grad_norm": 0.8197345597947936, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 30336 + }, + { + "epoch": 0.30337, + "grad_norm": 0.9894948076922317, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 30337 + }, + { + "epoch": 0.30338, + "grad_norm": 0.9233719635669718, + "learning_rate": 0.003, + "loss": 4.0315, + "step": 30338 + }, + { + "epoch": 0.30339, + "grad_norm": 0.8714384055163963, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 30339 + }, + { + "epoch": 0.3034, + "grad_norm": 0.8372642289121145, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 30340 + }, + { + "epoch": 0.30341, + "grad_norm": 0.7637640138876546, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 30341 + }, + { + "epoch": 0.30342, + "grad_norm": 0.7754293379596865, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 30342 + }, + { + "epoch": 0.30343, + "grad_norm": 0.72873076009515, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 30343 + }, + { + "epoch": 0.30344, + "grad_norm": 0.735999789664492, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 30344 + }, + { + "epoch": 0.30345, + "grad_norm": 0.7563549747781702, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 30345 + }, + { + "epoch": 0.30346, + "grad_norm": 0.7842150430909243, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 30346 + }, + { + "epoch": 0.30347, + "grad_norm": 0.7366719335775181, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 30347 + }, + { + "epoch": 0.30348, + "grad_norm": 0.6684980113815568, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 30348 + }, + { + "epoch": 0.30349, + "grad_norm": 0.6402896820618401, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 30349 + }, + { + "epoch": 0.3035, + "grad_norm": 0.607805528624534, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 30350 + }, + { + "epoch": 0.30351, + "grad_norm": 0.5694647937935263, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 30351 + }, + { + "epoch": 0.30352, + "grad_norm": 0.5445567774706037, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 30352 + }, + { + "epoch": 0.30353, + "grad_norm": 0.6120223919163131, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 30353 + }, + { + "epoch": 0.30354, + "grad_norm": 0.696614152460824, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 30354 + }, + { + "epoch": 0.30355, + "grad_norm": 0.7912784120427387, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 30355 + }, + { + "epoch": 0.30356, + "grad_norm": 0.8139954917110194, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 30356 + }, + { + "epoch": 0.30357, + "grad_norm": 0.8230942161469375, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 30357 + }, + { + "epoch": 0.30358, + "grad_norm": 0.6664137934713343, + "learning_rate": 0.003, + "loss": 3.982, + "step": 30358 + }, + { + "epoch": 0.30359, + "grad_norm": 0.6089565157206566, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 30359 + }, + { + "epoch": 0.3036, + "grad_norm": 0.8682829147667909, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 30360 + }, + { + "epoch": 0.30361, + "grad_norm": 0.9401420120479703, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 30361 + }, + { + "epoch": 0.30362, + "grad_norm": 0.8787146701572801, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 30362 + }, + { + "epoch": 0.30363, + "grad_norm": 0.8229713666144822, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 30363 + }, + { + "epoch": 0.30364, + "grad_norm": 0.6914814338814482, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 30364 + }, + { + "epoch": 0.30365, + "grad_norm": 0.7614481817038172, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 30365 + }, + { + "epoch": 0.30366, + "grad_norm": 0.874436276766447, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 30366 + }, + { + "epoch": 0.30367, + "grad_norm": 0.8242119023445588, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 30367 + }, + { + "epoch": 0.30368, + "grad_norm": 0.723852915258188, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 30368 + }, + { + "epoch": 0.30369, + "grad_norm": 0.6975892172111728, + "learning_rate": 0.003, + "loss": 3.99, + "step": 30369 + }, + { + "epoch": 0.3037, + "grad_norm": 0.7900764590844794, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 30370 + }, + { + "epoch": 0.30371, + "grad_norm": 0.8215778789114833, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 30371 + }, + { + "epoch": 0.30372, + "grad_norm": 0.8382556646017757, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 30372 + }, + { + "epoch": 0.30373, + "grad_norm": 0.8868345588028, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 30373 + }, + { + "epoch": 0.30374, + "grad_norm": 0.8963195963750618, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 30374 + }, + { + "epoch": 0.30375, + "grad_norm": 0.8966577038593307, + "learning_rate": 0.003, + "loss": 3.995, + "step": 30375 + }, + { + "epoch": 0.30376, + "grad_norm": 0.8927856322795489, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 30376 + }, + { + "epoch": 0.30377, + "grad_norm": 0.8131138604435358, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 30377 + }, + { + "epoch": 0.30378, + "grad_norm": 0.742159465417321, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 30378 + }, + { + "epoch": 0.30379, + "grad_norm": 0.6539574451577013, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 30379 + }, + { + "epoch": 0.3038, + "grad_norm": 0.62054562575057, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 30380 + }, + { + "epoch": 0.30381, + "grad_norm": 0.6467161730035869, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 30381 + }, + { + "epoch": 0.30382, + "grad_norm": 0.637798540044917, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 30382 + }, + { + "epoch": 0.30383, + "grad_norm": 0.6989487354412616, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 30383 + }, + { + "epoch": 0.30384, + "grad_norm": 0.719341414242135, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 30384 + }, + { + "epoch": 0.30385, + "grad_norm": 0.7258537521962368, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 30385 + }, + { + "epoch": 0.30386, + "grad_norm": 0.7719752748201413, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 30386 + }, + { + "epoch": 0.30387, + "grad_norm": 0.7268651517591151, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 30387 + }, + { + "epoch": 0.30388, + "grad_norm": 0.7515262790849709, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 30388 + }, + { + "epoch": 0.30389, + "grad_norm": 0.7020585436614162, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 30389 + }, + { + "epoch": 0.3039, + "grad_norm": 0.7682328286005603, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 30390 + }, + { + "epoch": 0.30391, + "grad_norm": 0.7703674897094804, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 30391 + }, + { + "epoch": 0.30392, + "grad_norm": 0.8014075008608218, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 30392 + }, + { + "epoch": 0.30393, + "grad_norm": 0.7742939942893362, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 30393 + }, + { + "epoch": 0.30394, + "grad_norm": 0.782916710547356, + "learning_rate": 0.003, + "loss": 3.966, + "step": 30394 + }, + { + "epoch": 0.30395, + "grad_norm": 0.8319926003372111, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 30395 + }, + { + "epoch": 0.30396, + "grad_norm": 0.7945832024683313, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 30396 + }, + { + "epoch": 0.30397, + "grad_norm": 0.7459453526807791, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 30397 + }, + { + "epoch": 0.30398, + "grad_norm": 0.7036579605797674, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 30398 + }, + { + "epoch": 0.30399, + "grad_norm": 0.7543590230008844, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 30399 + }, + { + "epoch": 0.304, + "grad_norm": 0.9458863740031132, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 30400 + }, + { + "epoch": 0.30401, + "grad_norm": 1.0157024878293295, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 30401 + }, + { + "epoch": 0.30402, + "grad_norm": 0.9439862876031628, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 30402 + }, + { + "epoch": 0.30403, + "grad_norm": 0.8254188134428447, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 30403 + }, + { + "epoch": 0.30404, + "grad_norm": 0.7411037642235838, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 30404 + }, + { + "epoch": 0.30405, + "grad_norm": 0.7316751391791152, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 30405 + }, + { + "epoch": 0.30406, + "grad_norm": 0.719923904213756, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 30406 + }, + { + "epoch": 0.30407, + "grad_norm": 0.6852915706808647, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 30407 + }, + { + "epoch": 0.30408, + "grad_norm": 0.6837254106646511, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 30408 + }, + { + "epoch": 0.30409, + "grad_norm": 0.6438863555123915, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 30409 + }, + { + "epoch": 0.3041, + "grad_norm": 0.6634595172891123, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 30410 + }, + { + "epoch": 0.30411, + "grad_norm": 0.7192344312329422, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 30411 + }, + { + "epoch": 0.30412, + "grad_norm": 0.6784883035365626, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 30412 + }, + { + "epoch": 0.30413, + "grad_norm": 0.6318700972234115, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 30413 + }, + { + "epoch": 0.30414, + "grad_norm": 0.5647447197764598, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 30414 + }, + { + "epoch": 0.30415, + "grad_norm": 0.6149542264372235, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 30415 + }, + { + "epoch": 0.30416, + "grad_norm": 0.7598019969987748, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 30416 + }, + { + "epoch": 0.30417, + "grad_norm": 0.8309079719261804, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 30417 + }, + { + "epoch": 0.30418, + "grad_norm": 0.7957173045736348, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 30418 + }, + { + "epoch": 0.30419, + "grad_norm": 0.8988311798378131, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 30419 + }, + { + "epoch": 0.3042, + "grad_norm": 0.8891951599718357, + "learning_rate": 0.003, + "loss": 4.016, + "step": 30420 + }, + { + "epoch": 0.30421, + "grad_norm": 0.9624432867557753, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 30421 + }, + { + "epoch": 0.30422, + "grad_norm": 1.075364329772268, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 30422 + }, + { + "epoch": 0.30423, + "grad_norm": 0.9690566822452692, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 30423 + }, + { + "epoch": 0.30424, + "grad_norm": 0.8569504733284528, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 30424 + }, + { + "epoch": 0.30425, + "grad_norm": 0.7117275255413328, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 30425 + }, + { + "epoch": 0.30426, + "grad_norm": 0.6022382114216845, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 30426 + }, + { + "epoch": 0.30427, + "grad_norm": 0.6371144188047071, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 30427 + }, + { + "epoch": 0.30428, + "grad_norm": 0.6247842664899425, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 30428 + }, + { + "epoch": 0.30429, + "grad_norm": 0.5846985664585957, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 30429 + }, + { + "epoch": 0.3043, + "grad_norm": 0.5652131441682319, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 30430 + }, + { + "epoch": 0.30431, + "grad_norm": 0.5694857559143872, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 30431 + }, + { + "epoch": 0.30432, + "grad_norm": 0.6666639808125907, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 30432 + }, + { + "epoch": 0.30433, + "grad_norm": 0.704378024472217, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 30433 + }, + { + "epoch": 0.30434, + "grad_norm": 0.6724909780189356, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 30434 + }, + { + "epoch": 0.30435, + "grad_norm": 0.7184448742604159, + "learning_rate": 0.003, + "loss": 4.004, + "step": 30435 + }, + { + "epoch": 0.30436, + "grad_norm": 0.9564608169179571, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 30436 + }, + { + "epoch": 0.30437, + "grad_norm": 1.1478806435601931, + "learning_rate": 0.003, + "loss": 3.995, + "step": 30437 + }, + { + "epoch": 0.30438, + "grad_norm": 0.817051025076118, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 30438 + }, + { + "epoch": 0.30439, + "grad_norm": 0.7829886010840665, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 30439 + }, + { + "epoch": 0.3044, + "grad_norm": 0.7398824071255317, + "learning_rate": 0.003, + "loss": 3.9416, + "step": 30440 + }, + { + "epoch": 0.30441, + "grad_norm": 0.8293201229893381, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 30441 + }, + { + "epoch": 0.30442, + "grad_norm": 0.6987389158805482, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 30442 + }, + { + "epoch": 0.30443, + "grad_norm": 0.7708363134216236, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 30443 + }, + { + "epoch": 0.30444, + "grad_norm": 0.8384568912624613, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 30444 + }, + { + "epoch": 0.30445, + "grad_norm": 0.8702544261694563, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 30445 + }, + { + "epoch": 0.30446, + "grad_norm": 0.8851010630654101, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 30446 + }, + { + "epoch": 0.30447, + "grad_norm": 0.8959577313315167, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 30447 + }, + { + "epoch": 0.30448, + "grad_norm": 0.7350927496802678, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 30448 + }, + { + "epoch": 0.30449, + "grad_norm": 0.7873675799620394, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 30449 + }, + { + "epoch": 0.3045, + "grad_norm": 0.6942835271565708, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 30450 + }, + { + "epoch": 0.30451, + "grad_norm": 0.7590577833659338, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 30451 + }, + { + "epoch": 0.30452, + "grad_norm": 0.7617503396196624, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 30452 + }, + { + "epoch": 0.30453, + "grad_norm": 0.7919874803791441, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 30453 + }, + { + "epoch": 0.30454, + "grad_norm": 0.7018136140251534, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 30454 + }, + { + "epoch": 0.30455, + "grad_norm": 0.7532064542867395, + "learning_rate": 0.003, + "loss": 4.011, + "step": 30455 + }, + { + "epoch": 0.30456, + "grad_norm": 0.9077557476396381, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 30456 + }, + { + "epoch": 0.30457, + "grad_norm": 1.0647838452309741, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 30457 + }, + { + "epoch": 0.30458, + "grad_norm": 0.9352025529738925, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 30458 + }, + { + "epoch": 0.30459, + "grad_norm": 0.7978658202402334, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 30459 + }, + { + "epoch": 0.3046, + "grad_norm": 0.6248211890338108, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 30460 + }, + { + "epoch": 0.30461, + "grad_norm": 0.6127713874370663, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 30461 + }, + { + "epoch": 0.30462, + "grad_norm": 0.6664070642379021, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 30462 + }, + { + "epoch": 0.30463, + "grad_norm": 0.6547666874390031, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 30463 + }, + { + "epoch": 0.30464, + "grad_norm": 0.603053650081104, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 30464 + }, + { + "epoch": 0.30465, + "grad_norm": 0.6142018680352122, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 30465 + }, + { + "epoch": 0.30466, + "grad_norm": 0.6888952216104639, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 30466 + }, + { + "epoch": 0.30467, + "grad_norm": 0.6112565805315746, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 30467 + }, + { + "epoch": 0.30468, + "grad_norm": 0.6610740191766377, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 30468 + }, + { + "epoch": 0.30469, + "grad_norm": 0.7382722414905526, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 30469 + }, + { + "epoch": 0.3047, + "grad_norm": 0.9221692605905458, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 30470 + }, + { + "epoch": 0.30471, + "grad_norm": 1.0970663684743867, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 30471 + }, + { + "epoch": 0.30472, + "grad_norm": 0.8491615910829545, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 30472 + }, + { + "epoch": 0.30473, + "grad_norm": 0.6225344473489038, + "learning_rate": 0.003, + "loss": 3.978, + "step": 30473 + }, + { + "epoch": 0.30474, + "grad_norm": 0.6508937077225777, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 30474 + }, + { + "epoch": 0.30475, + "grad_norm": 0.6924269193939822, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 30475 + }, + { + "epoch": 0.30476, + "grad_norm": 0.7945376664587855, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 30476 + }, + { + "epoch": 0.30477, + "grad_norm": 0.8070915776005653, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 30477 + }, + { + "epoch": 0.30478, + "grad_norm": 0.7675514140587927, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 30478 + }, + { + "epoch": 0.30479, + "grad_norm": 0.6512937387222626, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 30479 + }, + { + "epoch": 0.3048, + "grad_norm": 0.6114911121271223, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 30480 + }, + { + "epoch": 0.30481, + "grad_norm": 0.6035677764604246, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 30481 + }, + { + "epoch": 0.30482, + "grad_norm": 0.6501098029377615, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 30482 + }, + { + "epoch": 0.30483, + "grad_norm": 0.7960296074975804, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 30483 + }, + { + "epoch": 0.30484, + "grad_norm": 0.9702667620573546, + "learning_rate": 0.003, + "loss": 3.968, + "step": 30484 + }, + { + "epoch": 0.30485, + "grad_norm": 1.016796578991124, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 30485 + }, + { + "epoch": 0.30486, + "grad_norm": 0.8682294970803536, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 30486 + }, + { + "epoch": 0.30487, + "grad_norm": 0.8509523165889972, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 30487 + }, + { + "epoch": 0.30488, + "grad_norm": 1.0243863371389228, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 30488 + }, + { + "epoch": 0.30489, + "grad_norm": 1.0732516805389856, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 30489 + }, + { + "epoch": 0.3049, + "grad_norm": 0.950767903119956, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 30490 + }, + { + "epoch": 0.30491, + "grad_norm": 1.0437037405513, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 30491 + }, + { + "epoch": 0.30492, + "grad_norm": 0.9925594666637526, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 30492 + }, + { + "epoch": 0.30493, + "grad_norm": 0.8248631547862751, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 30493 + }, + { + "epoch": 0.30494, + "grad_norm": 0.9183806106707204, + "learning_rate": 0.003, + "loss": 4.0312, + "step": 30494 + }, + { + "epoch": 0.30495, + "grad_norm": 0.8544655546006908, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 30495 + }, + { + "epoch": 0.30496, + "grad_norm": 0.8929832023954728, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 30496 + }, + { + "epoch": 0.30497, + "grad_norm": 0.8896215103454512, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 30497 + }, + { + "epoch": 0.30498, + "grad_norm": 0.8922026182597443, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 30498 + }, + { + "epoch": 0.30499, + "grad_norm": 0.9816106209213087, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 30499 + }, + { + "epoch": 0.305, + "grad_norm": 1.0014084860764458, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 30500 + }, + { + "epoch": 0.30501, + "grad_norm": 0.8732336252664387, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 30501 + }, + { + "epoch": 0.30502, + "grad_norm": 0.8197068770422145, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 30502 + }, + { + "epoch": 0.30503, + "grad_norm": 0.7546792795061639, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 30503 + }, + { + "epoch": 0.30504, + "grad_norm": 0.7998331548335118, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 30504 + }, + { + "epoch": 0.30505, + "grad_norm": 0.8723953831671868, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 30505 + }, + { + "epoch": 0.30506, + "grad_norm": 0.8201298761842966, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 30506 + }, + { + "epoch": 0.30507, + "grad_norm": 0.8290374910047598, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 30507 + }, + { + "epoch": 0.30508, + "grad_norm": 0.7497116760715206, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 30508 + }, + { + "epoch": 0.30509, + "grad_norm": 0.6855258572650695, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 30509 + }, + { + "epoch": 0.3051, + "grad_norm": 0.6629131310536791, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 30510 + }, + { + "epoch": 0.30511, + "grad_norm": 0.6586226115820276, + "learning_rate": 0.003, + "loss": 3.986, + "step": 30511 + }, + { + "epoch": 0.30512, + "grad_norm": 0.63480835036755, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 30512 + }, + { + "epoch": 0.30513, + "grad_norm": 0.618632528868831, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 30513 + }, + { + "epoch": 0.30514, + "grad_norm": 0.5212830511599078, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 30514 + }, + { + "epoch": 0.30515, + "grad_norm": 0.5146570052385326, + "learning_rate": 0.003, + "loss": 3.982, + "step": 30515 + }, + { + "epoch": 0.30516, + "grad_norm": 0.464686775711039, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 30516 + }, + { + "epoch": 0.30517, + "grad_norm": 0.5482967752058603, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 30517 + }, + { + "epoch": 0.30518, + "grad_norm": 0.6477514183085147, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 30518 + }, + { + "epoch": 0.30519, + "grad_norm": 0.7231891726029941, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 30519 + }, + { + "epoch": 0.3052, + "grad_norm": 0.757437745512234, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 30520 + }, + { + "epoch": 0.30521, + "grad_norm": 0.7849990498962371, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 30521 + }, + { + "epoch": 0.30522, + "grad_norm": 0.8023429591346013, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 30522 + }, + { + "epoch": 0.30523, + "grad_norm": 0.7779414577636519, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 30523 + }, + { + "epoch": 0.30524, + "grad_norm": 0.8303759734663723, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 30524 + }, + { + "epoch": 0.30525, + "grad_norm": 0.9024385889384204, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 30525 + }, + { + "epoch": 0.30526, + "grad_norm": 0.9127655810730083, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30526 + }, + { + "epoch": 0.30527, + "grad_norm": 0.9127109744471084, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 30527 + }, + { + "epoch": 0.30528, + "grad_norm": 0.6960646333316547, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 30528 + }, + { + "epoch": 0.30529, + "grad_norm": 0.5500280355711663, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 30529 + }, + { + "epoch": 0.3053, + "grad_norm": 0.60973269961039, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 30530 + }, + { + "epoch": 0.30531, + "grad_norm": 0.6514824908392165, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 30531 + }, + { + "epoch": 0.30532, + "grad_norm": 0.668050565284454, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 30532 + }, + { + "epoch": 0.30533, + "grad_norm": 0.6262965147447288, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 30533 + }, + { + "epoch": 0.30534, + "grad_norm": 0.5136265256574624, + "learning_rate": 0.003, + "loss": 3.9447, + "step": 30534 + }, + { + "epoch": 0.30535, + "grad_norm": 0.5409034967701386, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 30535 + }, + { + "epoch": 0.30536, + "grad_norm": 0.619252925995063, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 30536 + }, + { + "epoch": 0.30537, + "grad_norm": 0.6905972422063201, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 30537 + }, + { + "epoch": 0.30538, + "grad_norm": 0.7940954821360773, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 30538 + }, + { + "epoch": 0.30539, + "grad_norm": 0.7858279221860777, + "learning_rate": 0.003, + "loss": 3.985, + "step": 30539 + }, + { + "epoch": 0.3054, + "grad_norm": 0.7199603379298749, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 30540 + }, + { + "epoch": 0.30541, + "grad_norm": 0.7062534107947707, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 30541 + }, + { + "epoch": 0.30542, + "grad_norm": 0.7119566660147312, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 30542 + }, + { + "epoch": 0.30543, + "grad_norm": 0.6883032341883801, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 30543 + }, + { + "epoch": 0.30544, + "grad_norm": 0.7088973964835837, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 30544 + }, + { + "epoch": 0.30545, + "grad_norm": 0.6842850327447385, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 30545 + }, + { + "epoch": 0.30546, + "grad_norm": 0.720538336399598, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 30546 + }, + { + "epoch": 0.30547, + "grad_norm": 0.7777939920750976, + "learning_rate": 0.003, + "loss": 3.993, + "step": 30547 + }, + { + "epoch": 0.30548, + "grad_norm": 0.8882715700856114, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 30548 + }, + { + "epoch": 0.30549, + "grad_norm": 0.9054438471059387, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 30549 + }, + { + "epoch": 0.3055, + "grad_norm": 0.960886250157442, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 30550 + }, + { + "epoch": 0.30551, + "grad_norm": 0.9603135673653792, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 30551 + }, + { + "epoch": 0.30552, + "grad_norm": 0.7608362494334971, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 30552 + }, + { + "epoch": 0.30553, + "grad_norm": 0.8009209595915183, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 30553 + }, + { + "epoch": 0.30554, + "grad_norm": 0.8497052262764814, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 30554 + }, + { + "epoch": 0.30555, + "grad_norm": 0.7621969367000314, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 30555 + }, + { + "epoch": 0.30556, + "grad_norm": 0.7452919688333302, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 30556 + }, + { + "epoch": 0.30557, + "grad_norm": 0.6987684925871963, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 30557 + }, + { + "epoch": 0.30558, + "grad_norm": 0.7577647828133331, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 30558 + }, + { + "epoch": 0.30559, + "grad_norm": 0.8039542484531786, + "learning_rate": 0.003, + "loss": 3.982, + "step": 30559 + }, + { + "epoch": 0.3056, + "grad_norm": 0.841966397507911, + "learning_rate": 0.003, + "loss": 4.0459, + "step": 30560 + }, + { + "epoch": 0.30561, + "grad_norm": 0.9375072891893887, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 30561 + }, + { + "epoch": 0.30562, + "grad_norm": 0.9044804152450688, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 30562 + }, + { + "epoch": 0.30563, + "grad_norm": 1.0037821587759157, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 30563 + }, + { + "epoch": 0.30564, + "grad_norm": 1.0611549207190816, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 30564 + }, + { + "epoch": 0.30565, + "grad_norm": 1.1490756136842053, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 30565 + }, + { + "epoch": 0.30566, + "grad_norm": 0.9567084851955284, + "learning_rate": 0.003, + "loss": 3.99, + "step": 30566 + }, + { + "epoch": 0.30567, + "grad_norm": 0.9241482809284041, + "learning_rate": 0.003, + "loss": 4.015, + "step": 30567 + }, + { + "epoch": 0.30568, + "grad_norm": 0.8621312715655938, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 30568 + }, + { + "epoch": 0.30569, + "grad_norm": 0.8782104795937866, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 30569 + }, + { + "epoch": 0.3057, + "grad_norm": 0.9730583670206575, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 30570 + }, + { + "epoch": 0.30571, + "grad_norm": 1.0394885617563123, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 30571 + }, + { + "epoch": 0.30572, + "grad_norm": 0.9979639026169809, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 30572 + }, + { + "epoch": 0.30573, + "grad_norm": 0.682376939407034, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 30573 + }, + { + "epoch": 0.30574, + "grad_norm": 0.677780281818173, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 30574 + }, + { + "epoch": 0.30575, + "grad_norm": 0.6324916800109034, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 30575 + }, + { + "epoch": 0.30576, + "grad_norm": 0.6291560655507238, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 30576 + }, + { + "epoch": 0.30577, + "grad_norm": 0.6207159724558523, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 30577 + }, + { + "epoch": 0.30578, + "grad_norm": 0.5632454601689261, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 30578 + }, + { + "epoch": 0.30579, + "grad_norm": 0.49993600787083753, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 30579 + }, + { + "epoch": 0.3058, + "grad_norm": 0.46261088572631986, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 30580 + }, + { + "epoch": 0.30581, + "grad_norm": 0.4773617615512335, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 30581 + }, + { + "epoch": 0.30582, + "grad_norm": 0.5073507900468668, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 30582 + }, + { + "epoch": 0.30583, + "grad_norm": 0.5455607445250386, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 30583 + }, + { + "epoch": 0.30584, + "grad_norm": 0.5969637917886709, + "learning_rate": 0.003, + "loss": 3.994, + "step": 30584 + }, + { + "epoch": 0.30585, + "grad_norm": 0.6044823600677658, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 30585 + }, + { + "epoch": 0.30586, + "grad_norm": 0.7007221921459803, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 30586 + }, + { + "epoch": 0.30587, + "grad_norm": 0.9189838989879998, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 30587 + }, + { + "epoch": 0.30588, + "grad_norm": 1.0753930672675291, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 30588 + }, + { + "epoch": 0.30589, + "grad_norm": 0.8448568038751147, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 30589 + }, + { + "epoch": 0.3059, + "grad_norm": 0.6852285049721157, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 30590 + }, + { + "epoch": 0.30591, + "grad_norm": 0.5607239693367306, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 30591 + }, + { + "epoch": 0.30592, + "grad_norm": 0.5640461622319046, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 30592 + }, + { + "epoch": 0.30593, + "grad_norm": 0.6637392520831543, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 30593 + }, + { + "epoch": 0.30594, + "grad_norm": 0.6744385173754779, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 30594 + }, + { + "epoch": 0.30595, + "grad_norm": 0.6870027766088453, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 30595 + }, + { + "epoch": 0.30596, + "grad_norm": 0.766688851583262, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 30596 + }, + { + "epoch": 0.30597, + "grad_norm": 0.8951442669769769, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 30597 + }, + { + "epoch": 0.30598, + "grad_norm": 0.9205535699221554, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 30598 + }, + { + "epoch": 0.30599, + "grad_norm": 0.8441430971545685, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 30599 + }, + { + "epoch": 0.306, + "grad_norm": 0.8429248032313209, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 30600 + }, + { + "epoch": 0.30601, + "grad_norm": 0.7731015785129376, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 30601 + }, + { + "epoch": 0.30602, + "grad_norm": 0.8071706179617744, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 30602 + }, + { + "epoch": 0.30603, + "grad_norm": 0.8247962762912825, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 30603 + }, + { + "epoch": 0.30604, + "grad_norm": 0.735219941231736, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 30604 + }, + { + "epoch": 0.30605, + "grad_norm": 0.7173829072264436, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 30605 + }, + { + "epoch": 0.30606, + "grad_norm": 0.7231326159423177, + "learning_rate": 0.003, + "loss": 4.0305, + "step": 30606 + }, + { + "epoch": 0.30607, + "grad_norm": 0.6903896614599815, + "learning_rate": 0.003, + "loss": 3.981, + "step": 30607 + }, + { + "epoch": 0.30608, + "grad_norm": 0.7516998559058492, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 30608 + }, + { + "epoch": 0.30609, + "grad_norm": 0.807855539899781, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 30609 + }, + { + "epoch": 0.3061, + "grad_norm": 0.7988448506842961, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 30610 + }, + { + "epoch": 0.30611, + "grad_norm": 0.8931259521126967, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 30611 + }, + { + "epoch": 0.30612, + "grad_norm": 1.0115491822739755, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 30612 + }, + { + "epoch": 0.30613, + "grad_norm": 1.0502434810783, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 30613 + }, + { + "epoch": 0.30614, + "grad_norm": 1.1121797722162716, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 30614 + }, + { + "epoch": 0.30615, + "grad_norm": 0.9493032478673614, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 30615 + }, + { + "epoch": 0.30616, + "grad_norm": 0.8247369507155404, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 30616 + }, + { + "epoch": 0.30617, + "grad_norm": 0.713717370218384, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 30617 + }, + { + "epoch": 0.30618, + "grad_norm": 0.6353259187885189, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 30618 + }, + { + "epoch": 0.30619, + "grad_norm": 0.5969541690411722, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 30619 + }, + { + "epoch": 0.3062, + "grad_norm": 0.5959963457314573, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 30620 + }, + { + "epoch": 0.30621, + "grad_norm": 0.6441296276560793, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 30621 + }, + { + "epoch": 0.30622, + "grad_norm": 0.7071536275036495, + "learning_rate": 0.003, + "loss": 3.991, + "step": 30622 + }, + { + "epoch": 0.30623, + "grad_norm": 0.7864772924502808, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 30623 + }, + { + "epoch": 0.30624, + "grad_norm": 0.9626747122601466, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 30624 + }, + { + "epoch": 0.30625, + "grad_norm": 1.101235439654116, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 30625 + }, + { + "epoch": 0.30626, + "grad_norm": 0.8940753974710034, + "learning_rate": 0.003, + "loss": 4.004, + "step": 30626 + }, + { + "epoch": 0.30627, + "grad_norm": 0.8935517222751053, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 30627 + }, + { + "epoch": 0.30628, + "grad_norm": 0.7692802181433714, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 30628 + }, + { + "epoch": 0.30629, + "grad_norm": 0.7835982423431614, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 30629 + }, + { + "epoch": 0.3063, + "grad_norm": 0.7018180323991023, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 30630 + }, + { + "epoch": 0.30631, + "grad_norm": 0.8526973575709412, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 30631 + }, + { + "epoch": 0.30632, + "grad_norm": 0.8645635593708805, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 30632 + }, + { + "epoch": 0.30633, + "grad_norm": 0.8755686308308475, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 30633 + }, + { + "epoch": 0.30634, + "grad_norm": 0.761495591097166, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 30634 + }, + { + "epoch": 0.30635, + "grad_norm": 0.5731003877928362, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 30635 + }, + { + "epoch": 0.30636, + "grad_norm": 0.5993092675064751, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 30636 + }, + { + "epoch": 0.30637, + "grad_norm": 0.6445200750462965, + "learning_rate": 0.003, + "loss": 4.003, + "step": 30637 + }, + { + "epoch": 0.30638, + "grad_norm": 0.6379234704942951, + "learning_rate": 0.003, + "loss": 3.9392, + "step": 30638 + }, + { + "epoch": 0.30639, + "grad_norm": 0.6292251387671278, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 30639 + }, + { + "epoch": 0.3064, + "grad_norm": 0.5780931448911601, + "learning_rate": 0.003, + "loss": 4.005, + "step": 30640 + }, + { + "epoch": 0.30641, + "grad_norm": 0.5588888953567035, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 30641 + }, + { + "epoch": 0.30642, + "grad_norm": 0.5286726542568452, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 30642 + }, + { + "epoch": 0.30643, + "grad_norm": 0.4910225429164421, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 30643 + }, + { + "epoch": 0.30644, + "grad_norm": 0.5028305796210831, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 30644 + }, + { + "epoch": 0.30645, + "grad_norm": 0.6037012841225609, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 30645 + }, + { + "epoch": 0.30646, + "grad_norm": 0.8688753947294329, + "learning_rate": 0.003, + "loss": 3.979, + "step": 30646 + }, + { + "epoch": 0.30647, + "grad_norm": 1.0623999563409818, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 30647 + }, + { + "epoch": 0.30648, + "grad_norm": 0.7895859038074331, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 30648 + }, + { + "epoch": 0.30649, + "grad_norm": 0.5590779726262355, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 30649 + }, + { + "epoch": 0.3065, + "grad_norm": 0.582410527978398, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 30650 + }, + { + "epoch": 0.30651, + "grad_norm": 0.7459279074648794, + "learning_rate": 0.003, + "loss": 4.009, + "step": 30651 + }, + { + "epoch": 0.30652, + "grad_norm": 0.9066946829410395, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 30652 + }, + { + "epoch": 0.30653, + "grad_norm": 0.9078862617966512, + "learning_rate": 0.003, + "loss": 3.966, + "step": 30653 + }, + { + "epoch": 0.30654, + "grad_norm": 0.8701536260805128, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 30654 + }, + { + "epoch": 0.30655, + "grad_norm": 0.9354310281368848, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 30655 + }, + { + "epoch": 0.30656, + "grad_norm": 1.1525676722605926, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 30656 + }, + { + "epoch": 0.30657, + "grad_norm": 0.7195315561198123, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 30657 + }, + { + "epoch": 0.30658, + "grad_norm": 0.8039361139098583, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 30658 + }, + { + "epoch": 0.30659, + "grad_norm": 0.8736685413979213, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 30659 + }, + { + "epoch": 0.3066, + "grad_norm": 0.9170773440293993, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 30660 + }, + { + "epoch": 0.30661, + "grad_norm": 1.014333689165197, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 30661 + }, + { + "epoch": 0.30662, + "grad_norm": 0.8805387113504283, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 30662 + }, + { + "epoch": 0.30663, + "grad_norm": 0.8214415602170925, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 30663 + }, + { + "epoch": 0.30664, + "grad_norm": 0.7740871251860846, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 30664 + }, + { + "epoch": 0.30665, + "grad_norm": 0.8644288739461621, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 30665 + }, + { + "epoch": 0.30666, + "grad_norm": 0.853096676319712, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 30666 + }, + { + "epoch": 0.30667, + "grad_norm": 0.8903486095130998, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 30667 + }, + { + "epoch": 0.30668, + "grad_norm": 0.8864617077098622, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 30668 + }, + { + "epoch": 0.30669, + "grad_norm": 0.7213976981594967, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 30669 + }, + { + "epoch": 0.3067, + "grad_norm": 0.6723520418514083, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 30670 + }, + { + "epoch": 0.30671, + "grad_norm": 0.731445863413075, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30671 + }, + { + "epoch": 0.30672, + "grad_norm": 0.7307122473719292, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 30672 + }, + { + "epoch": 0.30673, + "grad_norm": 0.820800239360758, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 30673 + }, + { + "epoch": 0.30674, + "grad_norm": 0.9457568454512717, + "learning_rate": 0.003, + "loss": 4.025, + "step": 30674 + }, + { + "epoch": 0.30675, + "grad_norm": 1.1082855789368358, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 30675 + }, + { + "epoch": 0.30676, + "grad_norm": 0.8133211415330222, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 30676 + }, + { + "epoch": 0.30677, + "grad_norm": 0.7216471324058984, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 30677 + }, + { + "epoch": 0.30678, + "grad_norm": 0.7236964915221028, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 30678 + }, + { + "epoch": 0.30679, + "grad_norm": 0.8402228080355194, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 30679 + }, + { + "epoch": 0.3068, + "grad_norm": 0.9696013748127934, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 30680 + }, + { + "epoch": 0.30681, + "grad_norm": 1.0396919592384963, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 30681 + }, + { + "epoch": 0.30682, + "grad_norm": 0.9029384003150246, + "learning_rate": 0.003, + "loss": 4.021, + "step": 30682 + }, + { + "epoch": 0.30683, + "grad_norm": 0.7630967602572553, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 30683 + }, + { + "epoch": 0.30684, + "grad_norm": 0.6468703796228881, + "learning_rate": 0.003, + "loss": 4.008, + "step": 30684 + }, + { + "epoch": 0.30685, + "grad_norm": 0.5989506174290103, + "learning_rate": 0.003, + "loss": 3.992, + "step": 30685 + }, + { + "epoch": 0.30686, + "grad_norm": 0.5990397682501772, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 30686 + }, + { + "epoch": 0.30687, + "grad_norm": 0.5406483011707794, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 30687 + }, + { + "epoch": 0.30688, + "grad_norm": 0.5051240207018093, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 30688 + }, + { + "epoch": 0.30689, + "grad_norm": 0.4450098165811868, + "learning_rate": 0.003, + "loss": 3.971, + "step": 30689 + }, + { + "epoch": 0.3069, + "grad_norm": 0.49367792678178085, + "learning_rate": 0.003, + "loss": 3.976, + "step": 30690 + }, + { + "epoch": 0.30691, + "grad_norm": 0.5554113475139516, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 30691 + }, + { + "epoch": 0.30692, + "grad_norm": 0.6946447233070503, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 30692 + }, + { + "epoch": 0.30693, + "grad_norm": 0.8552230292919261, + "learning_rate": 0.003, + "loss": 3.973, + "step": 30693 + }, + { + "epoch": 0.30694, + "grad_norm": 1.0506366434748922, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 30694 + }, + { + "epoch": 0.30695, + "grad_norm": 0.9455739189121428, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 30695 + }, + { + "epoch": 0.30696, + "grad_norm": 0.7764988493143486, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 30696 + }, + { + "epoch": 0.30697, + "grad_norm": 0.7020573009728572, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 30697 + }, + { + "epoch": 0.30698, + "grad_norm": 0.7054225129510691, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 30698 + }, + { + "epoch": 0.30699, + "grad_norm": 0.7368409951286747, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 30699 + }, + { + "epoch": 0.307, + "grad_norm": 0.7393581972123561, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 30700 + }, + { + "epoch": 0.30701, + "grad_norm": 0.7636404102806547, + "learning_rate": 0.003, + "loss": 4.026, + "step": 30701 + }, + { + "epoch": 0.30702, + "grad_norm": 0.7878370322396612, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 30702 + }, + { + "epoch": 0.30703, + "grad_norm": 0.7649680224755199, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 30703 + }, + { + "epoch": 0.30704, + "grad_norm": 0.7178068224230073, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 30704 + }, + { + "epoch": 0.30705, + "grad_norm": 0.6495548638661814, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 30705 + }, + { + "epoch": 0.30706, + "grad_norm": 0.6320208901030873, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 30706 + }, + { + "epoch": 0.30707, + "grad_norm": 0.6505108245041519, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 30707 + }, + { + "epoch": 0.30708, + "grad_norm": 0.6986202321529241, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 30708 + }, + { + "epoch": 0.30709, + "grad_norm": 0.7995403049720159, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 30709 + }, + { + "epoch": 0.3071, + "grad_norm": 0.9344941385715615, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 30710 + }, + { + "epoch": 0.30711, + "grad_norm": 0.9783870639475792, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 30711 + }, + { + "epoch": 0.30712, + "grad_norm": 1.0412898297847997, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 30712 + }, + { + "epoch": 0.30713, + "grad_norm": 0.9880450669317549, + "learning_rate": 0.003, + "loss": 4.038, + "step": 30713 + }, + { + "epoch": 0.30714, + "grad_norm": 0.9691705501412148, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 30714 + }, + { + "epoch": 0.30715, + "grad_norm": 0.824790516111437, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 30715 + }, + { + "epoch": 0.30716, + "grad_norm": 0.6749483299185871, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 30716 + }, + { + "epoch": 0.30717, + "grad_norm": 0.7256769956074393, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 30717 + }, + { + "epoch": 0.30718, + "grad_norm": 0.7684068519577141, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 30718 + }, + { + "epoch": 0.30719, + "grad_norm": 0.6895188056197686, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 30719 + }, + { + "epoch": 0.3072, + "grad_norm": 0.6012110133676017, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 30720 + }, + { + "epoch": 0.30721, + "grad_norm": 0.6071719753375663, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 30721 + }, + { + "epoch": 0.30722, + "grad_norm": 0.6893284615138549, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 30722 + }, + { + "epoch": 0.30723, + "grad_norm": 0.706160281409326, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 30723 + }, + { + "epoch": 0.30724, + "grad_norm": 0.7627385489598731, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 30724 + }, + { + "epoch": 0.30725, + "grad_norm": 0.9336786348471461, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 30725 + }, + { + "epoch": 0.30726, + "grad_norm": 0.8901891231861997, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 30726 + }, + { + "epoch": 0.30727, + "grad_norm": 0.8970241526975681, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 30727 + }, + { + "epoch": 0.30728, + "grad_norm": 0.8615042564690399, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 30728 + }, + { + "epoch": 0.30729, + "grad_norm": 0.8089376230541266, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 30729 + }, + { + "epoch": 0.3073, + "grad_norm": 0.8428545580183702, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 30730 + }, + { + "epoch": 0.30731, + "grad_norm": 0.8927383083122206, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 30731 + }, + { + "epoch": 0.30732, + "grad_norm": 0.8926153499255044, + "learning_rate": 0.003, + "loss": 3.9401, + "step": 30732 + }, + { + "epoch": 0.30733, + "grad_norm": 0.9450136871236534, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 30733 + }, + { + "epoch": 0.30734, + "grad_norm": 1.0153351013260354, + "learning_rate": 0.003, + "loss": 4.006, + "step": 30734 + }, + { + "epoch": 0.30735, + "grad_norm": 0.9757999339254169, + "learning_rate": 0.003, + "loss": 4.007, + "step": 30735 + }, + { + "epoch": 0.30736, + "grad_norm": 0.8661504775887081, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 30736 + }, + { + "epoch": 0.30737, + "grad_norm": 0.8064094938237627, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 30737 + }, + { + "epoch": 0.30738, + "grad_norm": 0.8076288329990192, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 30738 + }, + { + "epoch": 0.30739, + "grad_norm": 0.8103021212042958, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 30739 + }, + { + "epoch": 0.3074, + "grad_norm": 0.798038012822872, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 30740 + }, + { + "epoch": 0.30741, + "grad_norm": 0.7107875778199099, + "learning_rate": 0.003, + "loss": 4.026, + "step": 30741 + }, + { + "epoch": 0.30742, + "grad_norm": 0.7448881512713699, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 30742 + }, + { + "epoch": 0.30743, + "grad_norm": 0.6961798751218881, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 30743 + }, + { + "epoch": 0.30744, + "grad_norm": 0.7536631245132617, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 30744 + }, + { + "epoch": 0.30745, + "grad_norm": 0.8776494492570116, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 30745 + }, + { + "epoch": 0.30746, + "grad_norm": 0.988324752996671, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 30746 + }, + { + "epoch": 0.30747, + "grad_norm": 1.036556194736396, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 30747 + }, + { + "epoch": 0.30748, + "grad_norm": 1.0022053587766218, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 30748 + }, + { + "epoch": 0.30749, + "grad_norm": 0.9622909586752783, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 30749 + }, + { + "epoch": 0.3075, + "grad_norm": 0.8083326481715944, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 30750 + }, + { + "epoch": 0.30751, + "grad_norm": 0.7539088201837407, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 30751 + }, + { + "epoch": 0.30752, + "grad_norm": 0.694679917030286, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 30752 + }, + { + "epoch": 0.30753, + "grad_norm": 0.7071602442954869, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 30753 + }, + { + "epoch": 0.30754, + "grad_norm": 0.7597200997006622, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 30754 + }, + { + "epoch": 0.30755, + "grad_norm": 0.7832533157945087, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 30755 + }, + { + "epoch": 0.30756, + "grad_norm": 0.646026362902485, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 30756 + }, + { + "epoch": 0.30757, + "grad_norm": 0.6391302993977227, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 30757 + }, + { + "epoch": 0.30758, + "grad_norm": 0.6925569795014315, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 30758 + }, + { + "epoch": 0.30759, + "grad_norm": 0.6347963833815734, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 30759 + }, + { + "epoch": 0.3076, + "grad_norm": 0.5663238994663099, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 30760 + }, + { + "epoch": 0.30761, + "grad_norm": 0.5480878201713055, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 30761 + }, + { + "epoch": 0.30762, + "grad_norm": 0.471544392892976, + "learning_rate": 0.003, + "loss": 3.977, + "step": 30762 + }, + { + "epoch": 0.30763, + "grad_norm": 0.45220883381943294, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 30763 + }, + { + "epoch": 0.30764, + "grad_norm": 0.46800678321010536, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 30764 + }, + { + "epoch": 0.30765, + "grad_norm": 0.5092158196562409, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 30765 + }, + { + "epoch": 0.30766, + "grad_norm": 0.594445388112815, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 30766 + }, + { + "epoch": 0.30767, + "grad_norm": 0.6658372539975604, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 30767 + }, + { + "epoch": 0.30768, + "grad_norm": 0.7361448054675229, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 30768 + }, + { + "epoch": 0.30769, + "grad_norm": 0.8386428529627002, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 30769 + }, + { + "epoch": 0.3077, + "grad_norm": 0.9509137596860218, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 30770 + }, + { + "epoch": 0.30771, + "grad_norm": 0.8883914889206476, + "learning_rate": 0.003, + "loss": 3.96, + "step": 30771 + }, + { + "epoch": 0.30772, + "grad_norm": 0.8626947344858577, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 30772 + }, + { + "epoch": 0.30773, + "grad_norm": 0.8971582146783255, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 30773 + }, + { + "epoch": 0.30774, + "grad_norm": 0.815566654976192, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 30774 + }, + { + "epoch": 0.30775, + "grad_norm": 1.006307513330294, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 30775 + }, + { + "epoch": 0.30776, + "grad_norm": 1.091953541889265, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 30776 + }, + { + "epoch": 0.30777, + "grad_norm": 0.8554694298163634, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 30777 + }, + { + "epoch": 0.30778, + "grad_norm": 0.8073110145829112, + "learning_rate": 0.003, + "loss": 3.998, + "step": 30778 + }, + { + "epoch": 0.30779, + "grad_norm": 0.788051851619689, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 30779 + }, + { + "epoch": 0.3078, + "grad_norm": 0.7460065753886247, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 30780 + }, + { + "epoch": 0.30781, + "grad_norm": 0.7076510181754162, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 30781 + }, + { + "epoch": 0.30782, + "grad_norm": 0.7927635518795428, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 30782 + }, + { + "epoch": 0.30783, + "grad_norm": 0.8886491182003751, + "learning_rate": 0.003, + "loss": 4.012, + "step": 30783 + }, + { + "epoch": 0.30784, + "grad_norm": 1.0043292503297545, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 30784 + }, + { + "epoch": 0.30785, + "grad_norm": 0.9318381883220621, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 30785 + }, + { + "epoch": 0.30786, + "grad_norm": 0.7703932511972829, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 30786 + }, + { + "epoch": 0.30787, + "grad_norm": 0.7414113391475076, + "learning_rate": 0.003, + "loss": 3.988, + "step": 30787 + }, + { + "epoch": 0.30788, + "grad_norm": 0.7158695987657876, + "learning_rate": 0.003, + "loss": 3.972, + "step": 30788 + }, + { + "epoch": 0.30789, + "grad_norm": 0.6783863483846845, + "learning_rate": 0.003, + "loss": 3.977, + "step": 30789 + }, + { + "epoch": 0.3079, + "grad_norm": 0.5470978357662233, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 30790 + }, + { + "epoch": 0.30791, + "grad_norm": 0.48506843708737957, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 30791 + }, + { + "epoch": 0.30792, + "grad_norm": 0.478432740993811, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 30792 + }, + { + "epoch": 0.30793, + "grad_norm": 0.484439203100026, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 30793 + }, + { + "epoch": 0.30794, + "grad_norm": 0.4764252784180965, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 30794 + }, + { + "epoch": 0.30795, + "grad_norm": 0.5635671065526068, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 30795 + }, + { + "epoch": 0.30796, + "grad_norm": 0.6655032556406659, + "learning_rate": 0.003, + "loss": 4.01, + "step": 30796 + }, + { + "epoch": 0.30797, + "grad_norm": 0.7183492201266389, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 30797 + }, + { + "epoch": 0.30798, + "grad_norm": 0.6899053913254888, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 30798 + }, + { + "epoch": 0.30799, + "grad_norm": 0.7854857758954151, + "learning_rate": 0.003, + "loss": 4.021, + "step": 30799 + }, + { + "epoch": 0.308, + "grad_norm": 0.9345616019967291, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 30800 + }, + { + "epoch": 0.30801, + "grad_norm": 0.8743403058555902, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 30801 + }, + { + "epoch": 0.30802, + "grad_norm": 0.8276096751809391, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 30802 + }, + { + "epoch": 0.30803, + "grad_norm": 0.8924351174791869, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 30803 + }, + { + "epoch": 0.30804, + "grad_norm": 0.9989610320385091, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 30804 + }, + { + "epoch": 0.30805, + "grad_norm": 0.9640124090074264, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 30805 + }, + { + "epoch": 0.30806, + "grad_norm": 0.8727559474032243, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 30806 + }, + { + "epoch": 0.30807, + "grad_norm": 0.8960842888195992, + "learning_rate": 0.003, + "loss": 4.0, + "step": 30807 + }, + { + "epoch": 0.30808, + "grad_norm": 0.7412842004559712, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 30808 + }, + { + "epoch": 0.30809, + "grad_norm": 0.6959689589109903, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 30809 + }, + { + "epoch": 0.3081, + "grad_norm": 0.8043300025861204, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 30810 + }, + { + "epoch": 0.30811, + "grad_norm": 0.9429014599895525, + "learning_rate": 0.003, + "loss": 3.997, + "step": 30811 + }, + { + "epoch": 0.30812, + "grad_norm": 0.9864189610061014, + "learning_rate": 0.003, + "loss": 4.0401, + "step": 30812 + }, + { + "epoch": 0.30813, + "grad_norm": 0.8744307227456716, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 30813 + }, + { + "epoch": 0.30814, + "grad_norm": 0.7530953281451768, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 30814 + }, + { + "epoch": 0.30815, + "grad_norm": 0.6573116024335557, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 30815 + }, + { + "epoch": 0.30816, + "grad_norm": 0.7477891127130518, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 30816 + }, + { + "epoch": 0.30817, + "grad_norm": 0.6622300592854143, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 30817 + }, + { + "epoch": 0.30818, + "grad_norm": 0.6695190304142338, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 30818 + }, + { + "epoch": 0.30819, + "grad_norm": 0.7046291037020351, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 30819 + }, + { + "epoch": 0.3082, + "grad_norm": 0.7586522722401247, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 30820 + }, + { + "epoch": 0.30821, + "grad_norm": 0.7657395092545133, + "learning_rate": 0.003, + "loss": 4.009, + "step": 30821 + }, + { + "epoch": 0.30822, + "grad_norm": 0.8098190189683012, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 30822 + }, + { + "epoch": 0.30823, + "grad_norm": 0.9559160184638817, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30823 + }, + { + "epoch": 0.30824, + "grad_norm": 0.9338107480199543, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 30824 + }, + { + "epoch": 0.30825, + "grad_norm": 0.6941893499631111, + "learning_rate": 0.003, + "loss": 3.997, + "step": 30825 + }, + { + "epoch": 0.30826, + "grad_norm": 0.7415062514742594, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 30826 + }, + { + "epoch": 0.30827, + "grad_norm": 0.7059515144171957, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 30827 + }, + { + "epoch": 0.30828, + "grad_norm": 0.6293185920235896, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 30828 + }, + { + "epoch": 0.30829, + "grad_norm": 0.5269646811275389, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 30829 + }, + { + "epoch": 0.3083, + "grad_norm": 0.5984707298148768, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 30830 + }, + { + "epoch": 0.30831, + "grad_norm": 0.5881102003305178, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 30831 + }, + { + "epoch": 0.30832, + "grad_norm": 0.6673847784291703, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 30832 + }, + { + "epoch": 0.30833, + "grad_norm": 0.7386120790205787, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 30833 + }, + { + "epoch": 0.30834, + "grad_norm": 0.7445061181262189, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 30834 + }, + { + "epoch": 0.30835, + "grad_norm": 0.7823731482984568, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 30835 + }, + { + "epoch": 0.30836, + "grad_norm": 0.8192056346606221, + "learning_rate": 0.003, + "loss": 3.983, + "step": 30836 + }, + { + "epoch": 0.30837, + "grad_norm": 0.8690897068884951, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 30837 + }, + { + "epoch": 0.30838, + "grad_norm": 1.0238078134471014, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 30838 + }, + { + "epoch": 0.30839, + "grad_norm": 1.186196037618417, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 30839 + }, + { + "epoch": 0.3084, + "grad_norm": 0.9684846226474659, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 30840 + }, + { + "epoch": 0.30841, + "grad_norm": 0.977994948391264, + "learning_rate": 0.003, + "loss": 3.993, + "step": 30841 + }, + { + "epoch": 0.30842, + "grad_norm": 0.8668369762644222, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 30842 + }, + { + "epoch": 0.30843, + "grad_norm": 0.9221025916779777, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 30843 + }, + { + "epoch": 0.30844, + "grad_norm": 0.9399494363662143, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 30844 + }, + { + "epoch": 0.30845, + "grad_norm": 0.944362165592193, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 30845 + }, + { + "epoch": 0.30846, + "grad_norm": 0.8903060748723495, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 30846 + }, + { + "epoch": 0.30847, + "grad_norm": 0.8115839606639849, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 30847 + }, + { + "epoch": 0.30848, + "grad_norm": 0.7251091259426559, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 30848 + }, + { + "epoch": 0.30849, + "grad_norm": 0.6202246501123133, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 30849 + }, + { + "epoch": 0.3085, + "grad_norm": 0.5535773676027965, + "learning_rate": 0.003, + "loss": 3.991, + "step": 30850 + }, + { + "epoch": 0.30851, + "grad_norm": 0.5839452931953595, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 30851 + }, + { + "epoch": 0.30852, + "grad_norm": 0.6266451274801941, + "learning_rate": 0.003, + "loss": 3.979, + "step": 30852 + }, + { + "epoch": 0.30853, + "grad_norm": 0.6860487333819505, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 30853 + }, + { + "epoch": 0.30854, + "grad_norm": 0.6810543003027691, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 30854 + }, + { + "epoch": 0.30855, + "grad_norm": 0.7245448550648762, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 30855 + }, + { + "epoch": 0.30856, + "grad_norm": 0.708327683291528, + "learning_rate": 0.003, + "loss": 3.968, + "step": 30856 + }, + { + "epoch": 0.30857, + "grad_norm": 0.768868754371261, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 30857 + }, + { + "epoch": 0.30858, + "grad_norm": 0.7915077075756518, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 30858 + }, + { + "epoch": 0.30859, + "grad_norm": 0.829494143055885, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 30859 + }, + { + "epoch": 0.3086, + "grad_norm": 0.878059929311821, + "learning_rate": 0.003, + "loss": 3.983, + "step": 30860 + }, + { + "epoch": 0.30861, + "grad_norm": 0.8316789266870879, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 30861 + }, + { + "epoch": 0.30862, + "grad_norm": 0.8223487209017436, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 30862 + }, + { + "epoch": 0.30863, + "grad_norm": 0.6889043081967137, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 30863 + }, + { + "epoch": 0.30864, + "grad_norm": 0.6508898549552122, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 30864 + }, + { + "epoch": 0.30865, + "grad_norm": 0.6399003603711942, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 30865 + }, + { + "epoch": 0.30866, + "grad_norm": 0.6117288905064052, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 30866 + }, + { + "epoch": 0.30867, + "grad_norm": 0.6135171524896568, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 30867 + }, + { + "epoch": 0.30868, + "grad_norm": 0.6383054201706818, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 30868 + }, + { + "epoch": 0.30869, + "grad_norm": 0.6887211576740753, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 30869 + }, + { + "epoch": 0.3087, + "grad_norm": 0.7292019559421877, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 30870 + }, + { + "epoch": 0.30871, + "grad_norm": 0.7344240657974, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 30871 + }, + { + "epoch": 0.30872, + "grad_norm": 0.7042818575887413, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 30872 + }, + { + "epoch": 0.30873, + "grad_norm": 0.6582339307186149, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 30873 + }, + { + "epoch": 0.30874, + "grad_norm": 0.6444385965698475, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 30874 + }, + { + "epoch": 0.30875, + "grad_norm": 0.7452227688578491, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 30875 + }, + { + "epoch": 0.30876, + "grad_norm": 0.7892811433877215, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 30876 + }, + { + "epoch": 0.30877, + "grad_norm": 0.8066861409566186, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 30877 + }, + { + "epoch": 0.30878, + "grad_norm": 0.6686422781478969, + "learning_rate": 0.003, + "loss": 3.968, + "step": 30878 + }, + { + "epoch": 0.30879, + "grad_norm": 0.7081127773861078, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 30879 + }, + { + "epoch": 0.3088, + "grad_norm": 0.824667547901821, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 30880 + }, + { + "epoch": 0.30881, + "grad_norm": 0.823936369311444, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 30881 + }, + { + "epoch": 0.30882, + "grad_norm": 0.8508252372100455, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 30882 + }, + { + "epoch": 0.30883, + "grad_norm": 0.8417387347030129, + "learning_rate": 0.003, + "loss": 3.978, + "step": 30883 + }, + { + "epoch": 0.30884, + "grad_norm": 0.7570454219583144, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 30884 + }, + { + "epoch": 0.30885, + "grad_norm": 0.9057156851760496, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 30885 + }, + { + "epoch": 0.30886, + "grad_norm": 1.026724857349007, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 30886 + }, + { + "epoch": 0.30887, + "grad_norm": 1.0353158555213589, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 30887 + }, + { + "epoch": 0.30888, + "grad_norm": 1.0051199275222262, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 30888 + }, + { + "epoch": 0.30889, + "grad_norm": 0.9322372132822229, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 30889 + }, + { + "epoch": 0.3089, + "grad_norm": 0.8927679346566775, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 30890 + }, + { + "epoch": 0.30891, + "grad_norm": 0.9305786703405863, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 30891 + }, + { + "epoch": 0.30892, + "grad_norm": 0.7856998512773387, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 30892 + }, + { + "epoch": 0.30893, + "grad_norm": 0.7207612766032403, + "learning_rate": 0.003, + "loss": 3.987, + "step": 30893 + }, + { + "epoch": 0.30894, + "grad_norm": 0.7971467126746534, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 30894 + }, + { + "epoch": 0.30895, + "grad_norm": 0.8262938657252148, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 30895 + }, + { + "epoch": 0.30896, + "grad_norm": 0.7194576821078449, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 30896 + }, + { + "epoch": 0.30897, + "grad_norm": 0.7119499846612924, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 30897 + }, + { + "epoch": 0.30898, + "grad_norm": 0.8117671317175896, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 30898 + }, + { + "epoch": 0.30899, + "grad_norm": 0.7909032223690344, + "learning_rate": 0.003, + "loss": 3.9444, + "step": 30899 + }, + { + "epoch": 0.309, + "grad_norm": 0.8547954836721654, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 30900 + }, + { + "epoch": 0.30901, + "grad_norm": 0.9993606101132511, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 30901 + }, + { + "epoch": 0.30902, + "grad_norm": 1.1539127234040136, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 30902 + }, + { + "epoch": 0.30903, + "grad_norm": 0.8633151781392071, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 30903 + }, + { + "epoch": 0.30904, + "grad_norm": 0.78294396645113, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 30904 + }, + { + "epoch": 0.30905, + "grad_norm": 0.8010404980763203, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 30905 + }, + { + "epoch": 0.30906, + "grad_norm": 0.7636728108502587, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 30906 + }, + { + "epoch": 0.30907, + "grad_norm": 0.725270559869079, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 30907 + }, + { + "epoch": 0.30908, + "grad_norm": 0.5888532227087266, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 30908 + }, + { + "epoch": 0.30909, + "grad_norm": 0.6176898726607019, + "learning_rate": 0.003, + "loss": 3.968, + "step": 30909 + }, + { + "epoch": 0.3091, + "grad_norm": 0.6204150305415197, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 30910 + }, + { + "epoch": 0.30911, + "grad_norm": 0.6712598327830015, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 30911 + }, + { + "epoch": 0.30912, + "grad_norm": 0.7538370642438768, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 30912 + }, + { + "epoch": 0.30913, + "grad_norm": 0.7620896545144883, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 30913 + }, + { + "epoch": 0.30914, + "grad_norm": 0.6420803561932852, + "learning_rate": 0.003, + "loss": 3.9454, + "step": 30914 + }, + { + "epoch": 0.30915, + "grad_norm": 0.6008552466212268, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 30915 + }, + { + "epoch": 0.30916, + "grad_norm": 0.6496316515472571, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 30916 + }, + { + "epoch": 0.30917, + "grad_norm": 0.6752896368388958, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 30917 + }, + { + "epoch": 0.30918, + "grad_norm": 0.7270811260132688, + "learning_rate": 0.003, + "loss": 3.999, + "step": 30918 + }, + { + "epoch": 0.30919, + "grad_norm": 0.7020729436094609, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 30919 + }, + { + "epoch": 0.3092, + "grad_norm": 0.7584227711226508, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 30920 + }, + { + "epoch": 0.30921, + "grad_norm": 0.8761955309279029, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 30921 + }, + { + "epoch": 0.30922, + "grad_norm": 0.9064670188551969, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 30922 + }, + { + "epoch": 0.30923, + "grad_norm": 0.7812076463106065, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 30923 + }, + { + "epoch": 0.30924, + "grad_norm": 0.6552460810987472, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 30924 + }, + { + "epoch": 0.30925, + "grad_norm": 0.56413478907913, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 30925 + }, + { + "epoch": 0.30926, + "grad_norm": 0.5472465240664309, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 30926 + }, + { + "epoch": 0.30927, + "grad_norm": 0.5783421935646936, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 30927 + }, + { + "epoch": 0.30928, + "grad_norm": 0.5901376116737042, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 30928 + }, + { + "epoch": 0.30929, + "grad_norm": 0.5920980750942595, + "learning_rate": 0.003, + "loss": 3.943, + "step": 30929 + }, + { + "epoch": 0.3093, + "grad_norm": 0.5862109925435804, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 30930 + }, + { + "epoch": 0.30931, + "grad_norm": 0.6386865450351624, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 30931 + }, + { + "epoch": 0.30932, + "grad_norm": 0.8199398379395937, + "learning_rate": 0.003, + "loss": 3.973, + "step": 30932 + }, + { + "epoch": 0.30933, + "grad_norm": 0.9842177696393363, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 30933 + }, + { + "epoch": 0.30934, + "grad_norm": 0.8602728558618893, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 30934 + }, + { + "epoch": 0.30935, + "grad_norm": 0.7643040088881828, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 30935 + }, + { + "epoch": 0.30936, + "grad_norm": 0.8405141785008783, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 30936 + }, + { + "epoch": 0.30937, + "grad_norm": 0.9359351078141632, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 30937 + }, + { + "epoch": 0.30938, + "grad_norm": 0.9163198618558511, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 30938 + }, + { + "epoch": 0.30939, + "grad_norm": 1.012121577973415, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 30939 + }, + { + "epoch": 0.3094, + "grad_norm": 1.0358480054667218, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 30940 + }, + { + "epoch": 0.30941, + "grad_norm": 0.9509378285120733, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 30941 + }, + { + "epoch": 0.30942, + "grad_norm": 0.9325564529664953, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 30942 + }, + { + "epoch": 0.30943, + "grad_norm": 0.815081794312892, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 30943 + }, + { + "epoch": 0.30944, + "grad_norm": 0.758935596954412, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 30944 + }, + { + "epoch": 0.30945, + "grad_norm": 0.7470402451835138, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 30945 + }, + { + "epoch": 0.30946, + "grad_norm": 0.7977076830887783, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 30946 + }, + { + "epoch": 0.30947, + "grad_norm": 0.9354970698778866, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 30947 + }, + { + "epoch": 0.30948, + "grad_norm": 1.087533923562198, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 30948 + }, + { + "epoch": 0.30949, + "grad_norm": 0.8379640492220042, + "learning_rate": 0.003, + "loss": 3.992, + "step": 30949 + }, + { + "epoch": 0.3095, + "grad_norm": 0.8887376104008275, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 30950 + }, + { + "epoch": 0.30951, + "grad_norm": 0.8832454932332675, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 30951 + }, + { + "epoch": 0.30952, + "grad_norm": 0.7625761683966944, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 30952 + }, + { + "epoch": 0.30953, + "grad_norm": 0.7202787489946189, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 30953 + }, + { + "epoch": 0.30954, + "grad_norm": 0.7652101521720668, + "learning_rate": 0.003, + "loss": 4.008, + "step": 30954 + }, + { + "epoch": 0.30955, + "grad_norm": 0.7273726193474388, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 30955 + }, + { + "epoch": 0.30956, + "grad_norm": 0.7852771384803174, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 30956 + }, + { + "epoch": 0.30957, + "grad_norm": 0.8409192463892418, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 30957 + }, + { + "epoch": 0.30958, + "grad_norm": 0.7570399425768247, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 30958 + }, + { + "epoch": 0.30959, + "grad_norm": 0.7051269919110721, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 30959 + }, + { + "epoch": 0.3096, + "grad_norm": 0.726173370390267, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 30960 + }, + { + "epoch": 0.30961, + "grad_norm": 0.7644655877783314, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 30961 + }, + { + "epoch": 0.30962, + "grad_norm": 0.6911040848607353, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 30962 + }, + { + "epoch": 0.30963, + "grad_norm": 0.6912791773070437, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 30963 + }, + { + "epoch": 0.30964, + "grad_norm": 0.6305348695828923, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 30964 + }, + { + "epoch": 0.30965, + "grad_norm": 0.6234430125198522, + "learning_rate": 0.003, + "loss": 3.982, + "step": 30965 + }, + { + "epoch": 0.30966, + "grad_norm": 0.5463422669338437, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 30966 + }, + { + "epoch": 0.30967, + "grad_norm": 0.7005006795898046, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 30967 + }, + { + "epoch": 0.30968, + "grad_norm": 0.6784303420367543, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 30968 + }, + { + "epoch": 0.30969, + "grad_norm": 0.6900849940068466, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 30969 + }, + { + "epoch": 0.3097, + "grad_norm": 0.7139837526150639, + "learning_rate": 0.003, + "loss": 3.986, + "step": 30970 + }, + { + "epoch": 0.30971, + "grad_norm": 0.6921671625999491, + "learning_rate": 0.003, + "loss": 3.992, + "step": 30971 + }, + { + "epoch": 0.30972, + "grad_norm": 0.5777126899833834, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 30972 + }, + { + "epoch": 0.30973, + "grad_norm": 0.5116483363303546, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 30973 + }, + { + "epoch": 0.30974, + "grad_norm": 0.5231063588265776, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 30974 + }, + { + "epoch": 0.30975, + "grad_norm": 0.512588805624258, + "learning_rate": 0.003, + "loss": 3.989, + "step": 30975 + }, + { + "epoch": 0.30976, + "grad_norm": 0.48837964380794646, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 30976 + }, + { + "epoch": 0.30977, + "grad_norm": 0.5006049054281733, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 30977 + }, + { + "epoch": 0.30978, + "grad_norm": 0.5080561394891504, + "learning_rate": 0.003, + "loss": 3.96, + "step": 30978 + }, + { + "epoch": 0.30979, + "grad_norm": 0.5500733613801865, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 30979 + }, + { + "epoch": 0.3098, + "grad_norm": 0.5954533094166145, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 30980 + }, + { + "epoch": 0.30981, + "grad_norm": 0.6266408088148157, + "learning_rate": 0.003, + "loss": 3.964, + "step": 30981 + }, + { + "epoch": 0.30982, + "grad_norm": 0.6393516103566865, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 30982 + }, + { + "epoch": 0.30983, + "grad_norm": 0.6178388547031101, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 30983 + }, + { + "epoch": 0.30984, + "grad_norm": 0.7365172390301868, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 30984 + }, + { + "epoch": 0.30985, + "grad_norm": 0.8914566710286954, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 30985 + }, + { + "epoch": 0.30986, + "grad_norm": 1.0529705317085079, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 30986 + }, + { + "epoch": 0.30987, + "grad_norm": 1.0553081053877205, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 30987 + }, + { + "epoch": 0.30988, + "grad_norm": 1.0549793220908974, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 30988 + }, + { + "epoch": 0.30989, + "grad_norm": 1.0726091839564194, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 30989 + }, + { + "epoch": 0.3099, + "grad_norm": 1.0027626839448294, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 30990 + }, + { + "epoch": 0.30991, + "grad_norm": 1.0133547226973723, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 30991 + }, + { + "epoch": 0.30992, + "grad_norm": 1.091077080267084, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 30992 + }, + { + "epoch": 0.30993, + "grad_norm": 1.0397333232469463, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 30993 + }, + { + "epoch": 0.30994, + "grad_norm": 1.1434665720454784, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 30994 + }, + { + "epoch": 0.30995, + "grad_norm": 1.0047836878082292, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 30995 + }, + { + "epoch": 0.30996, + "grad_norm": 0.8776729379578516, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 30996 + }, + { + "epoch": 0.30997, + "grad_norm": 0.7412495675348416, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 30997 + }, + { + "epoch": 0.30998, + "grad_norm": 0.7603342604615354, + "learning_rate": 0.003, + "loss": 4.014, + "step": 30998 + }, + { + "epoch": 0.30999, + "grad_norm": 0.742987999544527, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 30999 + }, + { + "epoch": 0.31, + "grad_norm": 0.892032946407184, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 31000 + }, + { + "epoch": 0.31001, + "grad_norm": 1.0251362323789461, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 31001 + }, + { + "epoch": 0.31002, + "grad_norm": 1.037382908430389, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 31002 + }, + { + "epoch": 0.31003, + "grad_norm": 0.8428336685536183, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 31003 + }, + { + "epoch": 0.31004, + "grad_norm": 0.7790588340273729, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 31004 + }, + { + "epoch": 0.31005, + "grad_norm": 0.6939174720006682, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 31005 + }, + { + "epoch": 0.31006, + "grad_norm": 0.6815563120240608, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 31006 + }, + { + "epoch": 0.31007, + "grad_norm": 0.7211023447710457, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31007 + }, + { + "epoch": 0.31008, + "grad_norm": 0.8706198283467371, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 31008 + }, + { + "epoch": 0.31009, + "grad_norm": 0.9188496397835638, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 31009 + }, + { + "epoch": 0.3101, + "grad_norm": 0.735008613429964, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 31010 + }, + { + "epoch": 0.31011, + "grad_norm": 0.7163346006493824, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 31011 + }, + { + "epoch": 0.31012, + "grad_norm": 0.707314733363306, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 31012 + }, + { + "epoch": 0.31013, + "grad_norm": 0.7010198000476481, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 31013 + }, + { + "epoch": 0.31014, + "grad_norm": 0.6617487818801648, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 31014 + }, + { + "epoch": 0.31015, + "grad_norm": 0.7284296876329664, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 31015 + }, + { + "epoch": 0.31016, + "grad_norm": 0.8141072279425294, + "learning_rate": 0.003, + "loss": 3.99, + "step": 31016 + }, + { + "epoch": 0.31017, + "grad_norm": 0.8623394254316862, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 31017 + }, + { + "epoch": 0.31018, + "grad_norm": 0.9708469856488574, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 31018 + }, + { + "epoch": 0.31019, + "grad_norm": 0.9660303934580433, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 31019 + }, + { + "epoch": 0.3102, + "grad_norm": 0.7097193248396648, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 31020 + }, + { + "epoch": 0.31021, + "grad_norm": 0.5837426565221706, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 31021 + }, + { + "epoch": 0.31022, + "grad_norm": 0.6025523483697562, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 31022 + }, + { + "epoch": 0.31023, + "grad_norm": 0.6075812061917369, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 31023 + }, + { + "epoch": 0.31024, + "grad_norm": 0.6369909293838648, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 31024 + }, + { + "epoch": 0.31025, + "grad_norm": 0.5512037838422542, + "learning_rate": 0.003, + "loss": 3.97, + "step": 31025 + }, + { + "epoch": 0.31026, + "grad_norm": 0.5949523582251947, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 31026 + }, + { + "epoch": 0.31027, + "grad_norm": 0.600171396509389, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 31027 + }, + { + "epoch": 0.31028, + "grad_norm": 0.5791913680302758, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 31028 + }, + { + "epoch": 0.31029, + "grad_norm": 0.6020730660573462, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 31029 + }, + { + "epoch": 0.3103, + "grad_norm": 0.610500318201987, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 31030 + }, + { + "epoch": 0.31031, + "grad_norm": 0.6069023867888088, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 31031 + }, + { + "epoch": 0.31032, + "grad_norm": 0.6038985518899808, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 31032 + }, + { + "epoch": 0.31033, + "grad_norm": 0.6562975210311561, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 31033 + }, + { + "epoch": 0.31034, + "grad_norm": 0.7665503857487248, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 31034 + }, + { + "epoch": 0.31035, + "grad_norm": 0.7870407490279994, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 31035 + }, + { + "epoch": 0.31036, + "grad_norm": 0.8072532133352139, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 31036 + }, + { + "epoch": 0.31037, + "grad_norm": 0.7412549456958156, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 31037 + }, + { + "epoch": 0.31038, + "grad_norm": 0.6047437724859202, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 31038 + }, + { + "epoch": 0.31039, + "grad_norm": 0.7166702215517011, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 31039 + }, + { + "epoch": 0.3104, + "grad_norm": 0.8018940280215994, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 31040 + }, + { + "epoch": 0.31041, + "grad_norm": 0.8438168343081196, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 31041 + }, + { + "epoch": 0.31042, + "grad_norm": 0.9740284747852674, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 31042 + }, + { + "epoch": 0.31043, + "grad_norm": 1.004253416873787, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 31043 + }, + { + "epoch": 0.31044, + "grad_norm": 0.8483485345281917, + "learning_rate": 0.003, + "loss": 3.986, + "step": 31044 + }, + { + "epoch": 0.31045, + "grad_norm": 0.8846235409188978, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 31045 + }, + { + "epoch": 0.31046, + "grad_norm": 0.9783677225305171, + "learning_rate": 0.003, + "loss": 4.017, + "step": 31046 + }, + { + "epoch": 0.31047, + "grad_norm": 0.9013481225634851, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 31047 + }, + { + "epoch": 0.31048, + "grad_norm": 0.8881178124653204, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 31048 + }, + { + "epoch": 0.31049, + "grad_norm": 1.084083166086201, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 31049 + }, + { + "epoch": 0.3105, + "grad_norm": 1.1566512159442122, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 31050 + }, + { + "epoch": 0.31051, + "grad_norm": 0.9116597725539374, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 31051 + }, + { + "epoch": 0.31052, + "grad_norm": 0.8197323702332161, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 31052 + }, + { + "epoch": 0.31053, + "grad_norm": 0.7770020122994689, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 31053 + }, + { + "epoch": 0.31054, + "grad_norm": 0.8081955975717172, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 31054 + }, + { + "epoch": 0.31055, + "grad_norm": 0.8489959100120036, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 31055 + }, + { + "epoch": 0.31056, + "grad_norm": 0.9168751865064607, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 31056 + }, + { + "epoch": 0.31057, + "grad_norm": 0.9885522070996664, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 31057 + }, + { + "epoch": 0.31058, + "grad_norm": 1.0175946739579327, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 31058 + }, + { + "epoch": 0.31059, + "grad_norm": 0.9668083762055864, + "learning_rate": 0.003, + "loss": 4.015, + "step": 31059 + }, + { + "epoch": 0.3106, + "grad_norm": 0.9138226511409553, + "learning_rate": 0.003, + "loss": 4.009, + "step": 31060 + }, + { + "epoch": 0.31061, + "grad_norm": 0.9033576665848958, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 31061 + }, + { + "epoch": 0.31062, + "grad_norm": 0.8417371361722844, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 31062 + }, + { + "epoch": 0.31063, + "grad_norm": 0.7915261475412989, + "learning_rate": 0.003, + "loss": 3.978, + "step": 31063 + }, + { + "epoch": 0.31064, + "grad_norm": 0.6342381464083057, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 31064 + }, + { + "epoch": 0.31065, + "grad_norm": 0.6380475911912336, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 31065 + }, + { + "epoch": 0.31066, + "grad_norm": 0.6996304011580692, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 31066 + }, + { + "epoch": 0.31067, + "grad_norm": 0.763103887683148, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 31067 + }, + { + "epoch": 0.31068, + "grad_norm": 0.7771072935990033, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 31068 + }, + { + "epoch": 0.31069, + "grad_norm": 0.8476884582753759, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 31069 + }, + { + "epoch": 0.3107, + "grad_norm": 0.9134573038931794, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 31070 + }, + { + "epoch": 0.31071, + "grad_norm": 0.807487803089054, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 31071 + }, + { + "epoch": 0.31072, + "grad_norm": 0.6071067140028252, + "learning_rate": 0.003, + "loss": 4.007, + "step": 31072 + }, + { + "epoch": 0.31073, + "grad_norm": 0.5717598947170975, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 31073 + }, + { + "epoch": 0.31074, + "grad_norm": 0.5620436107135749, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 31074 + }, + { + "epoch": 0.31075, + "grad_norm": 0.5558033664221015, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 31075 + }, + { + "epoch": 0.31076, + "grad_norm": 0.5817960063617481, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 31076 + }, + { + "epoch": 0.31077, + "grad_norm": 0.568855258608951, + "learning_rate": 0.003, + "loss": 3.983, + "step": 31077 + }, + { + "epoch": 0.31078, + "grad_norm": 0.5015667707921183, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 31078 + }, + { + "epoch": 0.31079, + "grad_norm": 0.41176489072752104, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 31079 + }, + { + "epoch": 0.3108, + "grad_norm": 0.41671339415715425, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 31080 + }, + { + "epoch": 0.31081, + "grad_norm": 0.4202252705173639, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 31081 + }, + { + "epoch": 0.31082, + "grad_norm": 0.483814472549708, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 31082 + }, + { + "epoch": 0.31083, + "grad_norm": 0.5209707248679237, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 31083 + }, + { + "epoch": 0.31084, + "grad_norm": 0.5987800240551254, + "learning_rate": 0.003, + "loss": 3.954, + "step": 31084 + }, + { + "epoch": 0.31085, + "grad_norm": 0.7007578801751667, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 31085 + }, + { + "epoch": 0.31086, + "grad_norm": 0.7422208889764312, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 31086 + }, + { + "epoch": 0.31087, + "grad_norm": 0.7302547877657372, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 31087 + }, + { + "epoch": 0.31088, + "grad_norm": 0.7407409442354445, + "learning_rate": 0.003, + "loss": 3.964, + "step": 31088 + }, + { + "epoch": 0.31089, + "grad_norm": 0.8629039661087534, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 31089 + }, + { + "epoch": 0.3109, + "grad_norm": 0.9748009768852208, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 31090 + }, + { + "epoch": 0.31091, + "grad_norm": 0.9112517492307415, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 31091 + }, + { + "epoch": 0.31092, + "grad_norm": 0.7757871717279935, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 31092 + }, + { + "epoch": 0.31093, + "grad_norm": 0.7329941960835301, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 31093 + }, + { + "epoch": 0.31094, + "grad_norm": 0.8191425411576044, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 31094 + }, + { + "epoch": 0.31095, + "grad_norm": 0.8373626499073836, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 31095 + }, + { + "epoch": 0.31096, + "grad_norm": 0.8676170680684256, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 31096 + }, + { + "epoch": 0.31097, + "grad_norm": 0.9210658720509372, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 31097 + }, + { + "epoch": 0.31098, + "grad_norm": 0.9932621198907831, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 31098 + }, + { + "epoch": 0.31099, + "grad_norm": 0.9441507759204244, + "learning_rate": 0.003, + "loss": 4.029, + "step": 31099 + }, + { + "epoch": 0.311, + "grad_norm": 1.0272779212876861, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 31100 + }, + { + "epoch": 0.31101, + "grad_norm": 1.1026270276640335, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 31101 + }, + { + "epoch": 0.31102, + "grad_norm": 1.023270527349598, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 31102 + }, + { + "epoch": 0.31103, + "grad_norm": 0.9873744644680716, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 31103 + }, + { + "epoch": 0.31104, + "grad_norm": 0.8996128786185218, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 31104 + }, + { + "epoch": 0.31105, + "grad_norm": 0.8637705933742399, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 31105 + }, + { + "epoch": 0.31106, + "grad_norm": 0.853660042050497, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 31106 + }, + { + "epoch": 0.31107, + "grad_norm": 0.8827047488630894, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 31107 + }, + { + "epoch": 0.31108, + "grad_norm": 0.8454018606364125, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 31108 + }, + { + "epoch": 0.31109, + "grad_norm": 0.8002752228561896, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 31109 + }, + { + "epoch": 0.3111, + "grad_norm": 0.9473242882007612, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 31110 + }, + { + "epoch": 0.31111, + "grad_norm": 1.0104892830998513, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 31111 + }, + { + "epoch": 0.31112, + "grad_norm": 0.8417413668442639, + "learning_rate": 0.003, + "loss": 4.0311, + "step": 31112 + }, + { + "epoch": 0.31113, + "grad_norm": 0.774505466543662, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 31113 + }, + { + "epoch": 0.31114, + "grad_norm": 0.6347477020157858, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 31114 + }, + { + "epoch": 0.31115, + "grad_norm": 0.6377369830359464, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 31115 + }, + { + "epoch": 0.31116, + "grad_norm": 0.5967930877468212, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 31116 + }, + { + "epoch": 0.31117, + "grad_norm": 0.5304355085301679, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 31117 + }, + { + "epoch": 0.31118, + "grad_norm": 0.47221464466551893, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 31118 + }, + { + "epoch": 0.31119, + "grad_norm": 0.520412205078515, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 31119 + }, + { + "epoch": 0.3112, + "grad_norm": 0.6103084916434756, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 31120 + }, + { + "epoch": 0.31121, + "grad_norm": 0.6252023070723031, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 31121 + }, + { + "epoch": 0.31122, + "grad_norm": 0.6750299417680928, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 31122 + }, + { + "epoch": 0.31123, + "grad_norm": 0.6785474660879094, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 31123 + }, + { + "epoch": 0.31124, + "grad_norm": 0.7139567581594098, + "learning_rate": 0.003, + "loss": 4.0, + "step": 31124 + }, + { + "epoch": 0.31125, + "grad_norm": 0.705954636916463, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 31125 + }, + { + "epoch": 0.31126, + "grad_norm": 0.580445755298196, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 31126 + }, + { + "epoch": 0.31127, + "grad_norm": 0.6504006545832669, + "learning_rate": 0.003, + "loss": 4.023, + "step": 31127 + }, + { + "epoch": 0.31128, + "grad_norm": 0.6921326412471471, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 31128 + }, + { + "epoch": 0.31129, + "grad_norm": 0.7076236722539305, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 31129 + }, + { + "epoch": 0.3113, + "grad_norm": 0.6702242768626082, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 31130 + }, + { + "epoch": 0.31131, + "grad_norm": 0.5794574105563192, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 31131 + }, + { + "epoch": 0.31132, + "grad_norm": 0.6308071255564441, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 31132 + }, + { + "epoch": 0.31133, + "grad_norm": 0.6243307789421675, + "learning_rate": 0.003, + "loss": 3.982, + "step": 31133 + }, + { + "epoch": 0.31134, + "grad_norm": 0.6795732807917276, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 31134 + }, + { + "epoch": 0.31135, + "grad_norm": 0.7797814868545369, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 31135 + }, + { + "epoch": 0.31136, + "grad_norm": 0.9303876148753916, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 31136 + }, + { + "epoch": 0.31137, + "grad_norm": 0.971601488836015, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 31137 + }, + { + "epoch": 0.31138, + "grad_norm": 0.8579823999188827, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 31138 + }, + { + "epoch": 0.31139, + "grad_norm": 0.748465705605302, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 31139 + }, + { + "epoch": 0.3114, + "grad_norm": 0.7514453106652781, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 31140 + }, + { + "epoch": 0.31141, + "grad_norm": 0.6232619612999749, + "learning_rate": 0.003, + "loss": 4.005, + "step": 31141 + }, + { + "epoch": 0.31142, + "grad_norm": 0.6801491503019156, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 31142 + }, + { + "epoch": 0.31143, + "grad_norm": 0.6930519747570384, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 31143 + }, + { + "epoch": 0.31144, + "grad_norm": 0.8389808100182685, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 31144 + }, + { + "epoch": 0.31145, + "grad_norm": 0.8111475200419408, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 31145 + }, + { + "epoch": 0.31146, + "grad_norm": 0.7912574785992091, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 31146 + }, + { + "epoch": 0.31147, + "grad_norm": 0.7951383093360291, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 31147 + }, + { + "epoch": 0.31148, + "grad_norm": 0.7668790456269193, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 31148 + }, + { + "epoch": 0.31149, + "grad_norm": 0.7386113822542406, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 31149 + }, + { + "epoch": 0.3115, + "grad_norm": 0.8918086635223185, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 31150 + }, + { + "epoch": 0.31151, + "grad_norm": 1.1127839417983332, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 31151 + }, + { + "epoch": 0.31152, + "grad_norm": 0.8590234846430357, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 31152 + }, + { + "epoch": 0.31153, + "grad_norm": 0.7234684877634749, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 31153 + }, + { + "epoch": 0.31154, + "grad_norm": 0.6404047158180917, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 31154 + }, + { + "epoch": 0.31155, + "grad_norm": 0.7178211492383965, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 31155 + }, + { + "epoch": 0.31156, + "grad_norm": 0.7052412052886947, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 31156 + }, + { + "epoch": 0.31157, + "grad_norm": 0.7460560582282273, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 31157 + }, + { + "epoch": 0.31158, + "grad_norm": 0.7182673790548774, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 31158 + }, + { + "epoch": 0.31159, + "grad_norm": 0.8189437620395265, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 31159 + }, + { + "epoch": 0.3116, + "grad_norm": 0.8141689058250565, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 31160 + }, + { + "epoch": 0.31161, + "grad_norm": 0.8073976563105081, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 31161 + }, + { + "epoch": 0.31162, + "grad_norm": 0.7725865453650663, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 31162 + }, + { + "epoch": 0.31163, + "grad_norm": 0.7692336959603697, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 31163 + }, + { + "epoch": 0.31164, + "grad_norm": 0.9670066956538248, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 31164 + }, + { + "epoch": 0.31165, + "grad_norm": 1.1425214611599732, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 31165 + }, + { + "epoch": 0.31166, + "grad_norm": 0.8911368973853409, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 31166 + }, + { + "epoch": 0.31167, + "grad_norm": 0.881646036821957, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 31167 + }, + { + "epoch": 0.31168, + "grad_norm": 0.7956890791946036, + "learning_rate": 0.003, + "loss": 3.977, + "step": 31168 + }, + { + "epoch": 0.31169, + "grad_norm": 0.6995224020080626, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 31169 + }, + { + "epoch": 0.3117, + "grad_norm": 0.733578827379169, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 31170 + }, + { + "epoch": 0.31171, + "grad_norm": 0.8056421321041363, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 31171 + }, + { + "epoch": 0.31172, + "grad_norm": 0.8996329637525974, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 31172 + }, + { + "epoch": 0.31173, + "grad_norm": 0.7995083890132276, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 31173 + }, + { + "epoch": 0.31174, + "grad_norm": 0.722590194004554, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 31174 + }, + { + "epoch": 0.31175, + "grad_norm": 0.7290952562206231, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 31175 + }, + { + "epoch": 0.31176, + "grad_norm": 0.7693270196621661, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 31176 + }, + { + "epoch": 0.31177, + "grad_norm": 0.8344402465659211, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 31177 + }, + { + "epoch": 0.31178, + "grad_norm": 0.7583001881919242, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 31178 + }, + { + "epoch": 0.31179, + "grad_norm": 0.7211133388643656, + "learning_rate": 0.003, + "loss": 4.009, + "step": 31179 + }, + { + "epoch": 0.3118, + "grad_norm": 0.6919928179301874, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 31180 + }, + { + "epoch": 0.31181, + "grad_norm": 0.6617039927169659, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 31181 + }, + { + "epoch": 0.31182, + "grad_norm": 0.6809962049041225, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 31182 + }, + { + "epoch": 0.31183, + "grad_norm": 0.7515299635621419, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 31183 + }, + { + "epoch": 0.31184, + "grad_norm": 0.8322221542927184, + "learning_rate": 0.003, + "loss": 3.978, + "step": 31184 + }, + { + "epoch": 0.31185, + "grad_norm": 1.0092301241748114, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 31185 + }, + { + "epoch": 0.31186, + "grad_norm": 0.9222968402401555, + "learning_rate": 0.003, + "loss": 4.02, + "step": 31186 + }, + { + "epoch": 0.31187, + "grad_norm": 0.7630815951172742, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 31187 + }, + { + "epoch": 0.31188, + "grad_norm": 0.692148028787573, + "learning_rate": 0.003, + "loss": 3.935, + "step": 31188 + }, + { + "epoch": 0.31189, + "grad_norm": 0.7484203215652013, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 31189 + }, + { + "epoch": 0.3119, + "grad_norm": 0.8096005051226944, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 31190 + }, + { + "epoch": 0.31191, + "grad_norm": 0.7917749962596342, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 31191 + }, + { + "epoch": 0.31192, + "grad_norm": 0.7752294444286661, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 31192 + }, + { + "epoch": 0.31193, + "grad_norm": 0.6758852801818436, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 31193 + }, + { + "epoch": 0.31194, + "grad_norm": 0.6215431752160729, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 31194 + }, + { + "epoch": 0.31195, + "grad_norm": 0.6883857093978883, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 31195 + }, + { + "epoch": 0.31196, + "grad_norm": 0.8807591249991233, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 31196 + }, + { + "epoch": 0.31197, + "grad_norm": 0.9936392967331825, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 31197 + }, + { + "epoch": 0.31198, + "grad_norm": 0.9480951991738189, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 31198 + }, + { + "epoch": 0.31199, + "grad_norm": 0.9193063020053618, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 31199 + }, + { + "epoch": 0.312, + "grad_norm": 1.0480181949605034, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 31200 + }, + { + "epoch": 0.31201, + "grad_norm": 0.8381968640176936, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 31201 + }, + { + "epoch": 0.31202, + "grad_norm": 0.852978141414907, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 31202 + }, + { + "epoch": 0.31203, + "grad_norm": 0.8605375731252733, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 31203 + }, + { + "epoch": 0.31204, + "grad_norm": 0.9017501875424654, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 31204 + }, + { + "epoch": 0.31205, + "grad_norm": 0.9072627759805277, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 31205 + }, + { + "epoch": 0.31206, + "grad_norm": 0.8682383270444185, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 31206 + }, + { + "epoch": 0.31207, + "grad_norm": 0.8280108394045752, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 31207 + }, + { + "epoch": 0.31208, + "grad_norm": 0.7367098158728816, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 31208 + }, + { + "epoch": 0.31209, + "grad_norm": 0.7369861033709204, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 31209 + }, + { + "epoch": 0.3121, + "grad_norm": 0.7219546169892839, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 31210 + }, + { + "epoch": 0.31211, + "grad_norm": 0.7149250134429247, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 31211 + }, + { + "epoch": 0.31212, + "grad_norm": 0.7628513696329641, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 31212 + }, + { + "epoch": 0.31213, + "grad_norm": 0.7235104196056313, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 31213 + }, + { + "epoch": 0.31214, + "grad_norm": 0.7044345614501731, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 31214 + }, + { + "epoch": 0.31215, + "grad_norm": 0.8176178810543412, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 31215 + }, + { + "epoch": 0.31216, + "grad_norm": 0.9040353562762125, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 31216 + }, + { + "epoch": 0.31217, + "grad_norm": 1.1176679603994912, + "learning_rate": 0.003, + "loss": 3.98, + "step": 31217 + }, + { + "epoch": 0.31218, + "grad_norm": 0.8942900616505125, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 31218 + }, + { + "epoch": 0.31219, + "grad_norm": 0.8201732672950159, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 31219 + }, + { + "epoch": 0.3122, + "grad_norm": 0.8378839202662954, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 31220 + }, + { + "epoch": 0.31221, + "grad_norm": 0.8216610690839802, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 31221 + }, + { + "epoch": 0.31222, + "grad_norm": 0.8547287241253633, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 31222 + }, + { + "epoch": 0.31223, + "grad_norm": 0.8569161322452963, + "learning_rate": 0.003, + "loss": 3.995, + "step": 31223 + }, + { + "epoch": 0.31224, + "grad_norm": 0.7888421077793456, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 31224 + }, + { + "epoch": 0.31225, + "grad_norm": 0.6659497466346352, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 31225 + }, + { + "epoch": 0.31226, + "grad_norm": 0.6420988395738844, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 31226 + }, + { + "epoch": 0.31227, + "grad_norm": 0.6239062812933156, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 31227 + }, + { + "epoch": 0.31228, + "grad_norm": 0.6024392874457212, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 31228 + }, + { + "epoch": 0.31229, + "grad_norm": 0.6204636365746076, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 31229 + }, + { + "epoch": 0.3123, + "grad_norm": 0.6491289821564004, + "learning_rate": 0.003, + "loss": 4.0375, + "step": 31230 + }, + { + "epoch": 0.31231, + "grad_norm": 0.5870638797683769, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 31231 + }, + { + "epoch": 0.31232, + "grad_norm": 0.5446564431115671, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 31232 + }, + { + "epoch": 0.31233, + "grad_norm": 0.6159017026854212, + "learning_rate": 0.003, + "loss": 3.989, + "step": 31233 + }, + { + "epoch": 0.31234, + "grad_norm": 0.5849450330183834, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 31234 + }, + { + "epoch": 0.31235, + "grad_norm": 0.67334333752276, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 31235 + }, + { + "epoch": 0.31236, + "grad_norm": 0.8146023095434207, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 31236 + }, + { + "epoch": 0.31237, + "grad_norm": 1.0321186420193489, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 31237 + }, + { + "epoch": 0.31238, + "grad_norm": 1.0062411226533292, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 31238 + }, + { + "epoch": 0.31239, + "grad_norm": 0.7842469504456087, + "learning_rate": 0.003, + "loss": 3.946, + "step": 31239 + }, + { + "epoch": 0.3124, + "grad_norm": 0.5994928687889175, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 31240 + }, + { + "epoch": 0.31241, + "grad_norm": 0.8179209972537306, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 31241 + }, + { + "epoch": 0.31242, + "grad_norm": 0.842405895886417, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 31242 + }, + { + "epoch": 0.31243, + "grad_norm": 0.8105159701480366, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 31243 + }, + { + "epoch": 0.31244, + "grad_norm": 0.6563693321024223, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 31244 + }, + { + "epoch": 0.31245, + "grad_norm": 0.6801200610841829, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 31245 + }, + { + "epoch": 0.31246, + "grad_norm": 0.7014949780323506, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 31246 + }, + { + "epoch": 0.31247, + "grad_norm": 0.8067283868289886, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 31247 + }, + { + "epoch": 0.31248, + "grad_norm": 0.7566525372041998, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 31248 + }, + { + "epoch": 0.31249, + "grad_norm": 0.6681254232024721, + "learning_rate": 0.003, + "loss": 3.987, + "step": 31249 + }, + { + "epoch": 0.3125, + "grad_norm": 0.6382480227903837, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 31250 + }, + { + "epoch": 0.31251, + "grad_norm": 0.7814717148285947, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 31251 + }, + { + "epoch": 0.31252, + "grad_norm": 0.8188384848067881, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 31252 + }, + { + "epoch": 0.31253, + "grad_norm": 0.917811418923036, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 31253 + }, + { + "epoch": 0.31254, + "grad_norm": 0.8647757053669729, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 31254 + }, + { + "epoch": 0.31255, + "grad_norm": 0.8677798755561084, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 31255 + }, + { + "epoch": 0.31256, + "grad_norm": 0.8430219025286795, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 31256 + }, + { + "epoch": 0.31257, + "grad_norm": 0.7966005349993389, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 31257 + }, + { + "epoch": 0.31258, + "grad_norm": 0.7385290185554112, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 31258 + }, + { + "epoch": 0.31259, + "grad_norm": 0.6834527514202022, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 31259 + }, + { + "epoch": 0.3126, + "grad_norm": 0.6737847774162654, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 31260 + }, + { + "epoch": 0.31261, + "grad_norm": 0.7578028275349046, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 31261 + }, + { + "epoch": 0.31262, + "grad_norm": 0.8551331679365847, + "learning_rate": 0.003, + "loss": 4.014, + "step": 31262 + }, + { + "epoch": 0.31263, + "grad_norm": 0.9317795791279307, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 31263 + }, + { + "epoch": 0.31264, + "grad_norm": 0.9944148161684465, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 31264 + }, + { + "epoch": 0.31265, + "grad_norm": 1.0129046487332742, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 31265 + }, + { + "epoch": 0.31266, + "grad_norm": 0.859186037374196, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 31266 + }, + { + "epoch": 0.31267, + "grad_norm": 0.7916206307472495, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 31267 + }, + { + "epoch": 0.31268, + "grad_norm": 0.8134058448377764, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 31268 + }, + { + "epoch": 0.31269, + "grad_norm": 0.8645563557431563, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 31269 + }, + { + "epoch": 0.3127, + "grad_norm": 0.9076786997696017, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 31270 + }, + { + "epoch": 0.31271, + "grad_norm": 0.8753549154311137, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 31271 + }, + { + "epoch": 0.31272, + "grad_norm": 0.8523073817406822, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 31272 + }, + { + "epoch": 0.31273, + "grad_norm": 0.9136009651111546, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 31273 + }, + { + "epoch": 0.31274, + "grad_norm": 1.0330660474779985, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 31274 + }, + { + "epoch": 0.31275, + "grad_norm": 0.8243375879682232, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 31275 + }, + { + "epoch": 0.31276, + "grad_norm": 0.7463576287870658, + "learning_rate": 0.003, + "loss": 3.999, + "step": 31276 + }, + { + "epoch": 0.31277, + "grad_norm": 0.6615346898339981, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 31277 + }, + { + "epoch": 0.31278, + "grad_norm": 0.6874510732452611, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 31278 + }, + { + "epoch": 0.31279, + "grad_norm": 0.6663098398806636, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 31279 + }, + { + "epoch": 0.3128, + "grad_norm": 0.6167591945239604, + "learning_rate": 0.003, + "loss": 3.981, + "step": 31280 + }, + { + "epoch": 0.31281, + "grad_norm": 0.5926105429595095, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 31281 + }, + { + "epoch": 0.31282, + "grad_norm": 0.5880779065259081, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 31282 + }, + { + "epoch": 0.31283, + "grad_norm": 0.6246671150484144, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 31283 + }, + { + "epoch": 0.31284, + "grad_norm": 0.5967397943809154, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 31284 + }, + { + "epoch": 0.31285, + "grad_norm": 0.6896814224998208, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 31285 + }, + { + "epoch": 0.31286, + "grad_norm": 0.7422451647130234, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 31286 + }, + { + "epoch": 0.31287, + "grad_norm": 0.7423079880050177, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 31287 + }, + { + "epoch": 0.31288, + "grad_norm": 0.7455264497750159, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 31288 + }, + { + "epoch": 0.31289, + "grad_norm": 0.7592124364252195, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 31289 + }, + { + "epoch": 0.3129, + "grad_norm": 0.8105540122118177, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 31290 + }, + { + "epoch": 0.31291, + "grad_norm": 0.7819204872789103, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 31291 + }, + { + "epoch": 0.31292, + "grad_norm": 0.6853183144653008, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 31292 + }, + { + "epoch": 0.31293, + "grad_norm": 0.754579441641701, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31293 + }, + { + "epoch": 0.31294, + "grad_norm": 0.6894962975637483, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 31294 + }, + { + "epoch": 0.31295, + "grad_norm": 0.6581588122094043, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 31295 + }, + { + "epoch": 0.31296, + "grad_norm": 0.7831018228737575, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 31296 + }, + { + "epoch": 0.31297, + "grad_norm": 0.9525056876339869, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 31297 + }, + { + "epoch": 0.31298, + "grad_norm": 0.9896804784823345, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 31298 + }, + { + "epoch": 0.31299, + "grad_norm": 0.8541313751538362, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 31299 + }, + { + "epoch": 0.313, + "grad_norm": 0.8229624360571415, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 31300 + }, + { + "epoch": 0.31301, + "grad_norm": 0.7906233865241191, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 31301 + }, + { + "epoch": 0.31302, + "grad_norm": 0.9263476014119403, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 31302 + }, + { + "epoch": 0.31303, + "grad_norm": 0.9162730190548893, + "learning_rate": 0.003, + "loss": 3.968, + "step": 31303 + }, + { + "epoch": 0.31304, + "grad_norm": 0.8738858648391773, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 31304 + }, + { + "epoch": 0.31305, + "grad_norm": 0.6977745393049133, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 31305 + }, + { + "epoch": 0.31306, + "grad_norm": 0.7758484146320918, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 31306 + }, + { + "epoch": 0.31307, + "grad_norm": 0.8529540148446213, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 31307 + }, + { + "epoch": 0.31308, + "grad_norm": 0.8282510283003609, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 31308 + }, + { + "epoch": 0.31309, + "grad_norm": 0.7614545453508965, + "learning_rate": 0.003, + "loss": 4.006, + "step": 31309 + }, + { + "epoch": 0.3131, + "grad_norm": 0.7869246289143194, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 31310 + }, + { + "epoch": 0.31311, + "grad_norm": 0.7645787959482204, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 31311 + }, + { + "epoch": 0.31312, + "grad_norm": 0.7197707648423978, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 31312 + }, + { + "epoch": 0.31313, + "grad_norm": 0.6924685877932654, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 31313 + }, + { + "epoch": 0.31314, + "grad_norm": 0.7270347718299432, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 31314 + }, + { + "epoch": 0.31315, + "grad_norm": 0.8273626542693924, + "learning_rate": 0.003, + "loss": 4.0436, + "step": 31315 + }, + { + "epoch": 0.31316, + "grad_norm": 0.7704462606534558, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 31316 + }, + { + "epoch": 0.31317, + "grad_norm": 0.6770556910524984, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 31317 + }, + { + "epoch": 0.31318, + "grad_norm": 0.6399123852662121, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 31318 + }, + { + "epoch": 0.31319, + "grad_norm": 0.6317380595652456, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 31319 + }, + { + "epoch": 0.3132, + "grad_norm": 0.6809558409410508, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 31320 + }, + { + "epoch": 0.31321, + "grad_norm": 0.6836112652104607, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 31321 + }, + { + "epoch": 0.31322, + "grad_norm": 0.7404355589088928, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 31322 + }, + { + "epoch": 0.31323, + "grad_norm": 0.7853047156095747, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 31323 + }, + { + "epoch": 0.31324, + "grad_norm": 0.85071920457623, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 31324 + }, + { + "epoch": 0.31325, + "grad_norm": 0.9857725462354703, + "learning_rate": 0.003, + "loss": 3.979, + "step": 31325 + }, + { + "epoch": 0.31326, + "grad_norm": 1.0597244782426325, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 31326 + }, + { + "epoch": 0.31327, + "grad_norm": 1.0078557491971243, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 31327 + }, + { + "epoch": 0.31328, + "grad_norm": 0.9106410905294389, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 31328 + }, + { + "epoch": 0.31329, + "grad_norm": 0.8495271751307613, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 31329 + }, + { + "epoch": 0.3133, + "grad_norm": 0.8507272135903999, + "learning_rate": 0.003, + "loss": 4.0, + "step": 31330 + }, + { + "epoch": 0.31331, + "grad_norm": 0.7446327561370498, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 31331 + }, + { + "epoch": 0.31332, + "grad_norm": 0.834295932974183, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 31332 + }, + { + "epoch": 0.31333, + "grad_norm": 0.833275822777871, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 31333 + }, + { + "epoch": 0.31334, + "grad_norm": 0.9081556798155092, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 31334 + }, + { + "epoch": 0.31335, + "grad_norm": 0.9079651721544032, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 31335 + }, + { + "epoch": 0.31336, + "grad_norm": 0.8347876696407904, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 31336 + }, + { + "epoch": 0.31337, + "grad_norm": 0.6393881678981963, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 31337 + }, + { + "epoch": 0.31338, + "grad_norm": 0.78188935729861, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 31338 + }, + { + "epoch": 0.31339, + "grad_norm": 0.6238878223213176, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 31339 + }, + { + "epoch": 0.3134, + "grad_norm": 0.6962261093674061, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 31340 + }, + { + "epoch": 0.31341, + "grad_norm": 0.7373523399376508, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 31341 + }, + { + "epoch": 0.31342, + "grad_norm": 0.6960319143131153, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 31342 + }, + { + "epoch": 0.31343, + "grad_norm": 0.7373292740848139, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 31343 + }, + { + "epoch": 0.31344, + "grad_norm": 0.695506686174735, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 31344 + }, + { + "epoch": 0.31345, + "grad_norm": 0.7410018311366112, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 31345 + }, + { + "epoch": 0.31346, + "grad_norm": 0.7192146389783168, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 31346 + }, + { + "epoch": 0.31347, + "grad_norm": 0.8316189615502158, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 31347 + }, + { + "epoch": 0.31348, + "grad_norm": 1.0255703044763556, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 31348 + }, + { + "epoch": 0.31349, + "grad_norm": 1.1545883642534018, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 31349 + }, + { + "epoch": 0.3135, + "grad_norm": 0.6343100063057794, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31350 + }, + { + "epoch": 0.31351, + "grad_norm": 0.704550022813636, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 31351 + }, + { + "epoch": 0.31352, + "grad_norm": 0.7626361805951364, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 31352 + }, + { + "epoch": 0.31353, + "grad_norm": 0.8677383865194459, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 31353 + }, + { + "epoch": 0.31354, + "grad_norm": 0.9633279644801048, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 31354 + }, + { + "epoch": 0.31355, + "grad_norm": 0.8680560310864954, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 31355 + }, + { + "epoch": 0.31356, + "grad_norm": 0.9000464345248032, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 31356 + }, + { + "epoch": 0.31357, + "grad_norm": 1.0874121712674591, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 31357 + }, + { + "epoch": 0.31358, + "grad_norm": 0.8963841158042741, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 31358 + }, + { + "epoch": 0.31359, + "grad_norm": 0.8803283969845263, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 31359 + }, + { + "epoch": 0.3136, + "grad_norm": 0.7390497418926945, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 31360 + }, + { + "epoch": 0.31361, + "grad_norm": 0.6367074460073798, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 31361 + }, + { + "epoch": 0.31362, + "grad_norm": 0.6280875571285174, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 31362 + }, + { + "epoch": 0.31363, + "grad_norm": 0.5551384044849028, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 31363 + }, + { + "epoch": 0.31364, + "grad_norm": 0.5469698066762005, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 31364 + }, + { + "epoch": 0.31365, + "grad_norm": 0.5222809129109387, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 31365 + }, + { + "epoch": 0.31366, + "grad_norm": 0.5036662432692862, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 31366 + }, + { + "epoch": 0.31367, + "grad_norm": 0.5251315961528108, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 31367 + }, + { + "epoch": 0.31368, + "grad_norm": 0.6142158444790445, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 31368 + }, + { + "epoch": 0.31369, + "grad_norm": 0.6436648444641934, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 31369 + }, + { + "epoch": 0.3137, + "grad_norm": 0.5954540680296275, + "learning_rate": 0.003, + "loss": 4.013, + "step": 31370 + }, + { + "epoch": 0.31371, + "grad_norm": 0.6077766403411022, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 31371 + }, + { + "epoch": 0.31372, + "grad_norm": 0.6431409538604632, + "learning_rate": 0.003, + "loss": 4.011, + "step": 31372 + }, + { + "epoch": 0.31373, + "grad_norm": 0.6300108477049061, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 31373 + }, + { + "epoch": 0.31374, + "grad_norm": 0.6674042914191126, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 31374 + }, + { + "epoch": 0.31375, + "grad_norm": 0.746777922672568, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 31375 + }, + { + "epoch": 0.31376, + "grad_norm": 0.8658337765492937, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 31376 + }, + { + "epoch": 0.31377, + "grad_norm": 0.9048865228504357, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 31377 + }, + { + "epoch": 0.31378, + "grad_norm": 0.8717398398344398, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 31378 + }, + { + "epoch": 0.31379, + "grad_norm": 1.0211123561413629, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 31379 + }, + { + "epoch": 0.3138, + "grad_norm": 0.8210614853684512, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 31380 + }, + { + "epoch": 0.31381, + "grad_norm": 0.6995844820575052, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 31381 + }, + { + "epoch": 0.31382, + "grad_norm": 0.7797920388054954, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 31382 + }, + { + "epoch": 0.31383, + "grad_norm": 0.8955156643331705, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 31383 + }, + { + "epoch": 0.31384, + "grad_norm": 0.9734473003972014, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 31384 + }, + { + "epoch": 0.31385, + "grad_norm": 0.9112546057243767, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 31385 + }, + { + "epoch": 0.31386, + "grad_norm": 0.9702349587879817, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 31386 + }, + { + "epoch": 0.31387, + "grad_norm": 0.9840014697169992, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 31387 + }, + { + "epoch": 0.31388, + "grad_norm": 0.950119747256484, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 31388 + }, + { + "epoch": 0.31389, + "grad_norm": 1.0582725919483054, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 31389 + }, + { + "epoch": 0.3139, + "grad_norm": 1.052721981163448, + "learning_rate": 0.003, + "loss": 4.0271, + "step": 31390 + }, + { + "epoch": 0.31391, + "grad_norm": 1.0267078346929206, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 31391 + }, + { + "epoch": 0.31392, + "grad_norm": 0.9977702077778096, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 31392 + }, + { + "epoch": 0.31393, + "grad_norm": 1.0140620705343522, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 31393 + }, + { + "epoch": 0.31394, + "grad_norm": 1.0376001412674865, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 31394 + }, + { + "epoch": 0.31395, + "grad_norm": 0.8299751430411503, + "learning_rate": 0.003, + "loss": 3.996, + "step": 31395 + }, + { + "epoch": 0.31396, + "grad_norm": 0.7542044734690915, + "learning_rate": 0.003, + "loss": 4.005, + "step": 31396 + }, + { + "epoch": 0.31397, + "grad_norm": 0.6892539078535996, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 31397 + }, + { + "epoch": 0.31398, + "grad_norm": 0.7829593562903362, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 31398 + }, + { + "epoch": 0.31399, + "grad_norm": 0.818690610804301, + "learning_rate": 0.003, + "loss": 3.999, + "step": 31399 + }, + { + "epoch": 0.314, + "grad_norm": 0.7111850342884106, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 31400 + }, + { + "epoch": 0.31401, + "grad_norm": 0.7405841650275177, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 31401 + }, + { + "epoch": 0.31402, + "grad_norm": 0.6847160142667071, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 31402 + }, + { + "epoch": 0.31403, + "grad_norm": 0.6684190541333563, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 31403 + }, + { + "epoch": 0.31404, + "grad_norm": 0.6822518437806139, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 31404 + }, + { + "epoch": 0.31405, + "grad_norm": 0.6836497805397812, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 31405 + }, + { + "epoch": 0.31406, + "grad_norm": 0.6371991382253589, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 31406 + }, + { + "epoch": 0.31407, + "grad_norm": 0.5214872265034362, + "learning_rate": 0.003, + "loss": 3.976, + "step": 31407 + }, + { + "epoch": 0.31408, + "grad_norm": 0.5613025660553628, + "learning_rate": 0.003, + "loss": 4.0, + "step": 31408 + }, + { + "epoch": 0.31409, + "grad_norm": 0.5684870419281113, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 31409 + }, + { + "epoch": 0.3141, + "grad_norm": 0.5114994914460816, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 31410 + }, + { + "epoch": 0.31411, + "grad_norm": 0.4687235497528338, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 31411 + }, + { + "epoch": 0.31412, + "grad_norm": 0.48272295214514177, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 31412 + }, + { + "epoch": 0.31413, + "grad_norm": 0.5343670550815725, + "learning_rate": 0.003, + "loss": 3.9436, + "step": 31413 + }, + { + "epoch": 0.31414, + "grad_norm": 0.5273019492949782, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 31414 + }, + { + "epoch": 0.31415, + "grad_norm": 0.6698685725571566, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 31415 + }, + { + "epoch": 0.31416, + "grad_norm": 0.8104351564788229, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 31416 + }, + { + "epoch": 0.31417, + "grad_norm": 0.9463965293362269, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 31417 + }, + { + "epoch": 0.31418, + "grad_norm": 1.098656270252246, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 31418 + }, + { + "epoch": 0.31419, + "grad_norm": 0.8896883392421535, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 31419 + }, + { + "epoch": 0.3142, + "grad_norm": 0.6806767049867196, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 31420 + }, + { + "epoch": 0.31421, + "grad_norm": 0.5663326821613401, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 31421 + }, + { + "epoch": 0.31422, + "grad_norm": 0.5128326010744997, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 31422 + }, + { + "epoch": 0.31423, + "grad_norm": 0.5118688654361037, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 31423 + }, + { + "epoch": 0.31424, + "grad_norm": 0.5807882060317309, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 31424 + }, + { + "epoch": 0.31425, + "grad_norm": 0.6000616815680035, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 31425 + }, + { + "epoch": 0.31426, + "grad_norm": 0.7177786075337692, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 31426 + }, + { + "epoch": 0.31427, + "grad_norm": 0.8860952887898316, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 31427 + }, + { + "epoch": 0.31428, + "grad_norm": 0.9590539750200853, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 31428 + }, + { + "epoch": 0.31429, + "grad_norm": 0.8024720438965185, + "learning_rate": 0.003, + "loss": 3.988, + "step": 31429 + }, + { + "epoch": 0.3143, + "grad_norm": 0.5998887528725171, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 31430 + }, + { + "epoch": 0.31431, + "grad_norm": 0.698264090789974, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 31431 + }, + { + "epoch": 0.31432, + "grad_norm": 0.7768809918667693, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 31432 + }, + { + "epoch": 0.31433, + "grad_norm": 0.8125504820255965, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 31433 + }, + { + "epoch": 0.31434, + "grad_norm": 0.8431934178688821, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 31434 + }, + { + "epoch": 0.31435, + "grad_norm": 0.8363288828718941, + "learning_rate": 0.003, + "loss": 3.995, + "step": 31435 + }, + { + "epoch": 0.31436, + "grad_norm": 0.881064212790058, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 31436 + }, + { + "epoch": 0.31437, + "grad_norm": 0.9568249873741644, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 31437 + }, + { + "epoch": 0.31438, + "grad_norm": 0.9207152721295703, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 31438 + }, + { + "epoch": 0.31439, + "grad_norm": 0.8559859857184334, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 31439 + }, + { + "epoch": 0.3144, + "grad_norm": 0.7340888204257001, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 31440 + }, + { + "epoch": 0.31441, + "grad_norm": 0.7330228314098725, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 31441 + }, + { + "epoch": 0.31442, + "grad_norm": 0.6970473424938174, + "learning_rate": 0.003, + "loss": 4.008, + "step": 31442 + }, + { + "epoch": 0.31443, + "grad_norm": 0.7622403757876408, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 31443 + }, + { + "epoch": 0.31444, + "grad_norm": 0.7972753373362061, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 31444 + }, + { + "epoch": 0.31445, + "grad_norm": 0.8672339686620824, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 31445 + }, + { + "epoch": 0.31446, + "grad_norm": 0.9556892996125186, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 31446 + }, + { + "epoch": 0.31447, + "grad_norm": 1.0318618529093746, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 31447 + }, + { + "epoch": 0.31448, + "grad_norm": 0.9319323125910052, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 31448 + }, + { + "epoch": 0.31449, + "grad_norm": 0.7802505920662188, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 31449 + }, + { + "epoch": 0.3145, + "grad_norm": 0.685519241333618, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 31450 + }, + { + "epoch": 0.31451, + "grad_norm": 0.6019177377848499, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 31451 + }, + { + "epoch": 0.31452, + "grad_norm": 0.607547186427337, + "learning_rate": 0.003, + "loss": 4.013, + "step": 31452 + }, + { + "epoch": 0.31453, + "grad_norm": 0.6390458934465084, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 31453 + }, + { + "epoch": 0.31454, + "grad_norm": 0.739010803752754, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 31454 + }, + { + "epoch": 0.31455, + "grad_norm": 0.7220607869329706, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 31455 + }, + { + "epoch": 0.31456, + "grad_norm": 0.6077977251564978, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 31456 + }, + { + "epoch": 0.31457, + "grad_norm": 0.5905266255801905, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 31457 + }, + { + "epoch": 0.31458, + "grad_norm": 0.6568308378039422, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 31458 + }, + { + "epoch": 0.31459, + "grad_norm": 0.7387396108798714, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 31459 + }, + { + "epoch": 0.3146, + "grad_norm": 0.8185754828753187, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 31460 + }, + { + "epoch": 0.31461, + "grad_norm": 0.9064007691948931, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 31461 + }, + { + "epoch": 0.31462, + "grad_norm": 1.0189206817273913, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 31462 + }, + { + "epoch": 0.31463, + "grad_norm": 1.0909548400570848, + "learning_rate": 0.003, + "loss": 3.969, + "step": 31463 + }, + { + "epoch": 0.31464, + "grad_norm": 0.9836482725798298, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 31464 + }, + { + "epoch": 0.31465, + "grad_norm": 0.741599651813708, + "learning_rate": 0.003, + "loss": 3.971, + "step": 31465 + }, + { + "epoch": 0.31466, + "grad_norm": 0.8082455486580193, + "learning_rate": 0.003, + "loss": 3.999, + "step": 31466 + }, + { + "epoch": 0.31467, + "grad_norm": 0.820831111061453, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 31467 + }, + { + "epoch": 0.31468, + "grad_norm": 0.8618567403331017, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 31468 + }, + { + "epoch": 0.31469, + "grad_norm": 1.0166385955137918, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 31469 + }, + { + "epoch": 0.3147, + "grad_norm": 0.9905363251863806, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 31470 + }, + { + "epoch": 0.31471, + "grad_norm": 0.997062723687687, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 31471 + }, + { + "epoch": 0.31472, + "grad_norm": 1.0081140341236159, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 31472 + }, + { + "epoch": 0.31473, + "grad_norm": 0.9307327657660751, + "learning_rate": 0.003, + "loss": 4.005, + "step": 31473 + }, + { + "epoch": 0.31474, + "grad_norm": 0.8649145399395152, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 31474 + }, + { + "epoch": 0.31475, + "grad_norm": 0.8940046602049555, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 31475 + }, + { + "epoch": 0.31476, + "grad_norm": 0.8231651061021514, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 31476 + }, + { + "epoch": 0.31477, + "grad_norm": 0.7193922264543358, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 31477 + }, + { + "epoch": 0.31478, + "grad_norm": 0.7413302897334093, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 31478 + }, + { + "epoch": 0.31479, + "grad_norm": 0.7355544875849359, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 31479 + }, + { + "epoch": 0.3148, + "grad_norm": 0.7462187605074948, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 31480 + }, + { + "epoch": 0.31481, + "grad_norm": 0.8124814037065167, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 31481 + }, + { + "epoch": 0.31482, + "grad_norm": 0.8320632424670663, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 31482 + }, + { + "epoch": 0.31483, + "grad_norm": 0.8031448764438583, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 31483 + }, + { + "epoch": 0.31484, + "grad_norm": 0.9834082520832036, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 31484 + }, + { + "epoch": 0.31485, + "grad_norm": 0.927194572431671, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 31485 + }, + { + "epoch": 0.31486, + "grad_norm": 0.7983412534133011, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 31486 + }, + { + "epoch": 0.31487, + "grad_norm": 0.7751991807452033, + "learning_rate": 0.003, + "loss": 4.002, + "step": 31487 + }, + { + "epoch": 0.31488, + "grad_norm": 0.8630349287442013, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 31488 + }, + { + "epoch": 0.31489, + "grad_norm": 0.8937912517142013, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 31489 + }, + { + "epoch": 0.3149, + "grad_norm": 0.8903544988193766, + "learning_rate": 0.003, + "loss": 4.038, + "step": 31490 + }, + { + "epoch": 0.31491, + "grad_norm": 0.7648402879237263, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 31491 + }, + { + "epoch": 0.31492, + "grad_norm": 0.5689107179238628, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 31492 + }, + { + "epoch": 0.31493, + "grad_norm": 0.6040345297780466, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 31493 + }, + { + "epoch": 0.31494, + "grad_norm": 0.7112361422266272, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 31494 + }, + { + "epoch": 0.31495, + "grad_norm": 0.8448476473346469, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 31495 + }, + { + "epoch": 0.31496, + "grad_norm": 0.7265492709554301, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 31496 + }, + { + "epoch": 0.31497, + "grad_norm": 0.6812321439194725, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 31497 + }, + { + "epoch": 0.31498, + "grad_norm": 0.6226055503262878, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 31498 + }, + { + "epoch": 0.31499, + "grad_norm": 0.6550250417220864, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 31499 + }, + { + "epoch": 0.315, + "grad_norm": 0.6881400169003802, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 31500 + }, + { + "epoch": 0.31501, + "grad_norm": 0.559875933503326, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 31501 + }, + { + "epoch": 0.31502, + "grad_norm": 0.5788781661214076, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 31502 + }, + { + "epoch": 0.31503, + "grad_norm": 0.6034132667817462, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 31503 + }, + { + "epoch": 0.31504, + "grad_norm": 0.5485430819657694, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 31504 + }, + { + "epoch": 0.31505, + "grad_norm": 0.4821972341680985, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 31505 + }, + { + "epoch": 0.31506, + "grad_norm": 0.5221959365479588, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 31506 + }, + { + "epoch": 0.31507, + "grad_norm": 0.7507344363214228, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 31507 + }, + { + "epoch": 0.31508, + "grad_norm": 0.8899348468542817, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 31508 + }, + { + "epoch": 0.31509, + "grad_norm": 0.9868277311273859, + "learning_rate": 0.003, + "loss": 3.993, + "step": 31509 + }, + { + "epoch": 0.3151, + "grad_norm": 0.9276345231302366, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 31510 + }, + { + "epoch": 0.31511, + "grad_norm": 0.8289993893312083, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 31511 + }, + { + "epoch": 0.31512, + "grad_norm": 0.7818769200018721, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 31512 + }, + { + "epoch": 0.31513, + "grad_norm": 0.7924851976894758, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 31513 + }, + { + "epoch": 0.31514, + "grad_norm": 0.9034918317733082, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 31514 + }, + { + "epoch": 0.31515, + "grad_norm": 0.9165010623475257, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 31515 + }, + { + "epoch": 0.31516, + "grad_norm": 0.8751710625991085, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 31516 + }, + { + "epoch": 0.31517, + "grad_norm": 0.8692014014076025, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 31517 + }, + { + "epoch": 0.31518, + "grad_norm": 1.0179412617165808, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 31518 + }, + { + "epoch": 0.31519, + "grad_norm": 1.0433337160948926, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 31519 + }, + { + "epoch": 0.3152, + "grad_norm": 0.91173444737685, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 31520 + }, + { + "epoch": 0.31521, + "grad_norm": 0.9170418621429762, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 31521 + }, + { + "epoch": 0.31522, + "grad_norm": 0.8741935726343047, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 31522 + }, + { + "epoch": 0.31523, + "grad_norm": 0.8061203889846585, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 31523 + }, + { + "epoch": 0.31524, + "grad_norm": 0.7715736140861419, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 31524 + }, + { + "epoch": 0.31525, + "grad_norm": 0.6388755081223233, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 31525 + }, + { + "epoch": 0.31526, + "grad_norm": 0.6394521115592122, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 31526 + }, + { + "epoch": 0.31527, + "grad_norm": 0.5771358248473228, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 31527 + }, + { + "epoch": 0.31528, + "grad_norm": 0.6181694092168625, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 31528 + }, + { + "epoch": 0.31529, + "grad_norm": 0.6624125989262303, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 31529 + }, + { + "epoch": 0.3153, + "grad_norm": 0.7630191754604813, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 31530 + }, + { + "epoch": 0.31531, + "grad_norm": 0.7825965330751214, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 31531 + }, + { + "epoch": 0.31532, + "grad_norm": 0.9169353483547562, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 31532 + }, + { + "epoch": 0.31533, + "grad_norm": 1.1359397932664732, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 31533 + }, + { + "epoch": 0.31534, + "grad_norm": 0.7260055599191184, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 31534 + }, + { + "epoch": 0.31535, + "grad_norm": 0.5809530213670794, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 31535 + }, + { + "epoch": 0.31536, + "grad_norm": 0.6149318494495714, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 31536 + }, + { + "epoch": 0.31537, + "grad_norm": 0.6354217972951728, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 31537 + }, + { + "epoch": 0.31538, + "grad_norm": 0.681685163661406, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 31538 + }, + { + "epoch": 0.31539, + "grad_norm": 0.6542666807531605, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 31539 + }, + { + "epoch": 0.3154, + "grad_norm": 0.5935524130091183, + "learning_rate": 0.003, + "loss": 3.98, + "step": 31540 + }, + { + "epoch": 0.31541, + "grad_norm": 0.6192733889412796, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 31541 + }, + { + "epoch": 0.31542, + "grad_norm": 0.616998945863303, + "learning_rate": 0.003, + "loss": 3.9364, + "step": 31542 + }, + { + "epoch": 0.31543, + "grad_norm": 0.6356156121487547, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 31543 + }, + { + "epoch": 0.31544, + "grad_norm": 0.7226690301456123, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 31544 + }, + { + "epoch": 0.31545, + "grad_norm": 0.8519139929310533, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 31545 + }, + { + "epoch": 0.31546, + "grad_norm": 1.012647088946187, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 31546 + }, + { + "epoch": 0.31547, + "grad_norm": 1.0657092315920003, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 31547 + }, + { + "epoch": 0.31548, + "grad_norm": 0.8639909223977952, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 31548 + }, + { + "epoch": 0.31549, + "grad_norm": 0.7126183580991681, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 31549 + }, + { + "epoch": 0.3155, + "grad_norm": 0.6224735158883888, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 31550 + }, + { + "epoch": 0.31551, + "grad_norm": 0.611971121561259, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 31551 + }, + { + "epoch": 0.31552, + "grad_norm": 0.6594435458409803, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 31552 + }, + { + "epoch": 0.31553, + "grad_norm": 0.7151887436703107, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 31553 + }, + { + "epoch": 0.31554, + "grad_norm": 0.8228555853463446, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 31554 + }, + { + "epoch": 0.31555, + "grad_norm": 0.9700982267873098, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 31555 + }, + { + "epoch": 0.31556, + "grad_norm": 1.0083374104422629, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 31556 + }, + { + "epoch": 0.31557, + "grad_norm": 0.9510369933002458, + "learning_rate": 0.003, + "loss": 3.999, + "step": 31557 + }, + { + "epoch": 0.31558, + "grad_norm": 0.8926714914274031, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 31558 + }, + { + "epoch": 0.31559, + "grad_norm": 0.7467355889446711, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 31559 + }, + { + "epoch": 0.3156, + "grad_norm": 0.7943578594626417, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 31560 + }, + { + "epoch": 0.31561, + "grad_norm": 0.8163658662323992, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 31561 + }, + { + "epoch": 0.31562, + "grad_norm": 0.8197045286280007, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 31562 + }, + { + "epoch": 0.31563, + "grad_norm": 0.9156554830087428, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 31563 + }, + { + "epoch": 0.31564, + "grad_norm": 0.7863511888057314, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 31564 + }, + { + "epoch": 0.31565, + "grad_norm": 0.7627726291994781, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 31565 + }, + { + "epoch": 0.31566, + "grad_norm": 0.8861703526952687, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 31566 + }, + { + "epoch": 0.31567, + "grad_norm": 0.9488708746665759, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 31567 + }, + { + "epoch": 0.31568, + "grad_norm": 0.7814226082607978, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 31568 + }, + { + "epoch": 0.31569, + "grad_norm": 0.8309062424170264, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 31569 + }, + { + "epoch": 0.3157, + "grad_norm": 0.904223244007755, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 31570 + }, + { + "epoch": 0.31571, + "grad_norm": 0.8507970335500323, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 31571 + }, + { + "epoch": 0.31572, + "grad_norm": 0.7567489447648231, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 31572 + }, + { + "epoch": 0.31573, + "grad_norm": 0.6092994777413153, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 31573 + }, + { + "epoch": 0.31574, + "grad_norm": 0.616994040941264, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 31574 + }, + { + "epoch": 0.31575, + "grad_norm": 0.6391751565234467, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 31575 + }, + { + "epoch": 0.31576, + "grad_norm": 0.7347668555402371, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 31576 + }, + { + "epoch": 0.31577, + "grad_norm": 0.7384074322237403, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 31577 + }, + { + "epoch": 0.31578, + "grad_norm": 0.7382396846330965, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 31578 + }, + { + "epoch": 0.31579, + "grad_norm": 0.6099623650407281, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 31579 + }, + { + "epoch": 0.3158, + "grad_norm": 0.5208843555357507, + "learning_rate": 0.003, + "loss": 4.014, + "step": 31580 + }, + { + "epoch": 0.31581, + "grad_norm": 0.5522060389189684, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 31581 + }, + { + "epoch": 0.31582, + "grad_norm": 0.5110300366878285, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 31582 + }, + { + "epoch": 0.31583, + "grad_norm": 0.5376313015059565, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 31583 + }, + { + "epoch": 0.31584, + "grad_norm": 0.6053891016121475, + "learning_rate": 0.003, + "loss": 3.973, + "step": 31584 + }, + { + "epoch": 0.31585, + "grad_norm": 0.6528793372495584, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 31585 + }, + { + "epoch": 0.31586, + "grad_norm": 0.7046005783715049, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 31586 + }, + { + "epoch": 0.31587, + "grad_norm": 0.7346487825861463, + "learning_rate": 0.003, + "loss": 4.004, + "step": 31587 + }, + { + "epoch": 0.31588, + "grad_norm": 0.7380227077955139, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 31588 + }, + { + "epoch": 0.31589, + "grad_norm": 0.7979502414765581, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 31589 + }, + { + "epoch": 0.3159, + "grad_norm": 0.8630974344337722, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 31590 + }, + { + "epoch": 0.31591, + "grad_norm": 0.8335326364655101, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 31591 + }, + { + "epoch": 0.31592, + "grad_norm": 0.895731483677092, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 31592 + }, + { + "epoch": 0.31593, + "grad_norm": 0.915391286162728, + "learning_rate": 0.003, + "loss": 3.983, + "step": 31593 + }, + { + "epoch": 0.31594, + "grad_norm": 0.9114330718330896, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 31594 + }, + { + "epoch": 0.31595, + "grad_norm": 0.9030584236752632, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 31595 + }, + { + "epoch": 0.31596, + "grad_norm": 0.9082556669693908, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 31596 + }, + { + "epoch": 0.31597, + "grad_norm": 0.8952931172532499, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 31597 + }, + { + "epoch": 0.31598, + "grad_norm": 0.9035599671363576, + "learning_rate": 0.003, + "loss": 4.001, + "step": 31598 + }, + { + "epoch": 0.31599, + "grad_norm": 0.9522963064889485, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 31599 + }, + { + "epoch": 0.316, + "grad_norm": 0.9151650588806586, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 31600 + }, + { + "epoch": 0.31601, + "grad_norm": 0.9180591781610943, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 31601 + }, + { + "epoch": 0.31602, + "grad_norm": 1.032843862685059, + "learning_rate": 0.003, + "loss": 4.022, + "step": 31602 + }, + { + "epoch": 0.31603, + "grad_norm": 0.9985996273409476, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 31603 + }, + { + "epoch": 0.31604, + "grad_norm": 0.9834314010252817, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 31604 + }, + { + "epoch": 0.31605, + "grad_norm": 0.9383579548717883, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 31605 + }, + { + "epoch": 0.31606, + "grad_norm": 0.8785412202281785, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 31606 + }, + { + "epoch": 0.31607, + "grad_norm": 0.8818118561051456, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 31607 + }, + { + "epoch": 0.31608, + "grad_norm": 0.8521837055944489, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 31608 + }, + { + "epoch": 0.31609, + "grad_norm": 0.9261523088352447, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 31609 + }, + { + "epoch": 0.3161, + "grad_norm": 0.8460047222979186, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 31610 + }, + { + "epoch": 0.31611, + "grad_norm": 0.7256080250430393, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 31611 + }, + { + "epoch": 0.31612, + "grad_norm": 0.6979398038282285, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 31612 + }, + { + "epoch": 0.31613, + "grad_norm": 0.587825603102284, + "learning_rate": 0.003, + "loss": 3.993, + "step": 31613 + }, + { + "epoch": 0.31614, + "grad_norm": 0.6541064306646046, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 31614 + }, + { + "epoch": 0.31615, + "grad_norm": 0.5839361613962366, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 31615 + }, + { + "epoch": 0.31616, + "grad_norm": 0.602234561251948, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 31616 + }, + { + "epoch": 0.31617, + "grad_norm": 0.6368297279718695, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 31617 + }, + { + "epoch": 0.31618, + "grad_norm": 0.6416059758023578, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 31618 + }, + { + "epoch": 0.31619, + "grad_norm": 0.5933359639086812, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 31619 + }, + { + "epoch": 0.3162, + "grad_norm": 0.5194745046617828, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 31620 + }, + { + "epoch": 0.31621, + "grad_norm": 0.6436436117104847, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 31621 + }, + { + "epoch": 0.31622, + "grad_norm": 0.8140077487296052, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 31622 + }, + { + "epoch": 0.31623, + "grad_norm": 0.7850128011337874, + "learning_rate": 0.003, + "loss": 3.968, + "step": 31623 + }, + { + "epoch": 0.31624, + "grad_norm": 0.7807296206422635, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 31624 + }, + { + "epoch": 0.31625, + "grad_norm": 0.7332132688650277, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 31625 + }, + { + "epoch": 0.31626, + "grad_norm": 0.617828700394776, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 31626 + }, + { + "epoch": 0.31627, + "grad_norm": 0.553429834786404, + "learning_rate": 0.003, + "loss": 3.984, + "step": 31627 + }, + { + "epoch": 0.31628, + "grad_norm": 0.6808719091537541, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 31628 + }, + { + "epoch": 0.31629, + "grad_norm": 0.8356916506054359, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 31629 + }, + { + "epoch": 0.3163, + "grad_norm": 0.8715769679736283, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 31630 + }, + { + "epoch": 0.31631, + "grad_norm": 0.8803693461204645, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 31631 + }, + { + "epoch": 0.31632, + "grad_norm": 0.8030775299088697, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 31632 + }, + { + "epoch": 0.31633, + "grad_norm": 0.6265069866951528, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 31633 + }, + { + "epoch": 0.31634, + "grad_norm": 0.5801693706819351, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 31634 + }, + { + "epoch": 0.31635, + "grad_norm": 0.6654431114898016, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 31635 + }, + { + "epoch": 0.31636, + "grad_norm": 0.7249951891961991, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 31636 + }, + { + "epoch": 0.31637, + "grad_norm": 0.7084824928688149, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 31637 + }, + { + "epoch": 0.31638, + "grad_norm": 0.6266845398104454, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 31638 + }, + { + "epoch": 0.31639, + "grad_norm": 0.7137882058403564, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 31639 + }, + { + "epoch": 0.3164, + "grad_norm": 0.8185513709805086, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 31640 + }, + { + "epoch": 0.31641, + "grad_norm": 0.8204254744785495, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 31641 + }, + { + "epoch": 0.31642, + "grad_norm": 0.680327918636531, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 31642 + }, + { + "epoch": 0.31643, + "grad_norm": 0.6625796391243503, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 31643 + }, + { + "epoch": 0.31644, + "grad_norm": 0.7526608314163746, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 31644 + }, + { + "epoch": 0.31645, + "grad_norm": 0.8466259866715139, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 31645 + }, + { + "epoch": 0.31646, + "grad_norm": 0.8581277233398487, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 31646 + }, + { + "epoch": 0.31647, + "grad_norm": 0.8992015407212358, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 31647 + }, + { + "epoch": 0.31648, + "grad_norm": 0.9446731367669894, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 31648 + }, + { + "epoch": 0.31649, + "grad_norm": 1.0722795826637974, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 31649 + }, + { + "epoch": 0.3165, + "grad_norm": 1.2187969067043816, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 31650 + }, + { + "epoch": 0.31651, + "grad_norm": 0.8829786796232042, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 31651 + }, + { + "epoch": 0.31652, + "grad_norm": 0.7721362719969883, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 31652 + }, + { + "epoch": 0.31653, + "grad_norm": 0.7290513923434807, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 31653 + }, + { + "epoch": 0.31654, + "grad_norm": 0.7394451380266387, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 31654 + }, + { + "epoch": 0.31655, + "grad_norm": 0.8368481431175087, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 31655 + }, + { + "epoch": 0.31656, + "grad_norm": 1.0720632968747617, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 31656 + }, + { + "epoch": 0.31657, + "grad_norm": 0.9984621167835941, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 31657 + }, + { + "epoch": 0.31658, + "grad_norm": 0.82048642429402, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 31658 + }, + { + "epoch": 0.31659, + "grad_norm": 0.733473610565061, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 31659 + }, + { + "epoch": 0.3166, + "grad_norm": 0.7307774042844122, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 31660 + }, + { + "epoch": 0.31661, + "grad_norm": 0.625402202691473, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 31661 + }, + { + "epoch": 0.31662, + "grad_norm": 0.6757545260920541, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 31662 + }, + { + "epoch": 0.31663, + "grad_norm": 0.7505205476741217, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 31663 + }, + { + "epoch": 0.31664, + "grad_norm": 0.8620634668317341, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 31664 + }, + { + "epoch": 0.31665, + "grad_norm": 0.9476131590409607, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 31665 + }, + { + "epoch": 0.31666, + "grad_norm": 0.803602805665305, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 31666 + }, + { + "epoch": 0.31667, + "grad_norm": 0.7245946778223203, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 31667 + }, + { + "epoch": 0.31668, + "grad_norm": 0.7638272975632923, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 31668 + }, + { + "epoch": 0.31669, + "grad_norm": 0.8680100872130795, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 31669 + }, + { + "epoch": 0.3167, + "grad_norm": 0.8051901530683964, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 31670 + }, + { + "epoch": 0.31671, + "grad_norm": 0.7615353110695724, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 31671 + }, + { + "epoch": 0.31672, + "grad_norm": 0.7055375633874734, + "learning_rate": 0.003, + "loss": 3.968, + "step": 31672 + }, + { + "epoch": 0.31673, + "grad_norm": 0.7085202399153819, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 31673 + }, + { + "epoch": 0.31674, + "grad_norm": 0.7269867738671417, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 31674 + }, + { + "epoch": 0.31675, + "grad_norm": 0.7852127393952418, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 31675 + }, + { + "epoch": 0.31676, + "grad_norm": 0.8803980529612471, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 31676 + }, + { + "epoch": 0.31677, + "grad_norm": 0.9910604768765962, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 31677 + }, + { + "epoch": 0.31678, + "grad_norm": 1.0640465451051797, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 31678 + }, + { + "epoch": 0.31679, + "grad_norm": 0.9617480163039466, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 31679 + }, + { + "epoch": 0.3168, + "grad_norm": 0.8283244288519587, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 31680 + }, + { + "epoch": 0.31681, + "grad_norm": 0.7209196146474455, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 31681 + }, + { + "epoch": 0.31682, + "grad_norm": 0.7709556085340009, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 31682 + }, + { + "epoch": 0.31683, + "grad_norm": 0.9185632226587691, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 31683 + }, + { + "epoch": 0.31684, + "grad_norm": 0.9641614364805692, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 31684 + }, + { + "epoch": 0.31685, + "grad_norm": 0.8399319184869729, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 31685 + }, + { + "epoch": 0.31686, + "grad_norm": 0.6028451365317037, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31686 + }, + { + "epoch": 0.31687, + "grad_norm": 0.556102347856903, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 31687 + }, + { + "epoch": 0.31688, + "grad_norm": 0.7046483473582005, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 31688 + }, + { + "epoch": 0.31689, + "grad_norm": 0.7970744827646861, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 31689 + }, + { + "epoch": 0.3169, + "grad_norm": 0.802758202798388, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 31690 + }, + { + "epoch": 0.31691, + "grad_norm": 0.7166853570575534, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 31691 + }, + { + "epoch": 0.31692, + "grad_norm": 0.5381838225627598, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 31692 + }, + { + "epoch": 0.31693, + "grad_norm": 0.5440537717261943, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 31693 + }, + { + "epoch": 0.31694, + "grad_norm": 0.6566204614397217, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 31694 + }, + { + "epoch": 0.31695, + "grad_norm": 0.7718834277082167, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 31695 + }, + { + "epoch": 0.31696, + "grad_norm": 0.8365482155595287, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 31696 + }, + { + "epoch": 0.31697, + "grad_norm": 0.7304725524682135, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 31697 + }, + { + "epoch": 0.31698, + "grad_norm": 0.5575525592084005, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 31698 + }, + { + "epoch": 0.31699, + "grad_norm": 0.707107903122326, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 31699 + }, + { + "epoch": 0.317, + "grad_norm": 0.6803438893192706, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 31700 + }, + { + "epoch": 0.31701, + "grad_norm": 0.5618063165885792, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 31701 + }, + { + "epoch": 0.31702, + "grad_norm": 0.5381287784804072, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31702 + }, + { + "epoch": 0.31703, + "grad_norm": 0.6364165926233148, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 31703 + }, + { + "epoch": 0.31704, + "grad_norm": 0.767060132456975, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 31704 + }, + { + "epoch": 0.31705, + "grad_norm": 0.83969315784373, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 31705 + }, + { + "epoch": 0.31706, + "grad_norm": 0.8152581287709355, + "learning_rate": 0.003, + "loss": 4.021, + "step": 31706 + }, + { + "epoch": 0.31707, + "grad_norm": 0.7284207485841185, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 31707 + }, + { + "epoch": 0.31708, + "grad_norm": 0.6576246522820981, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 31708 + }, + { + "epoch": 0.31709, + "grad_norm": 0.7196063600900479, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 31709 + }, + { + "epoch": 0.3171, + "grad_norm": 0.6951519148938367, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 31710 + }, + { + "epoch": 0.31711, + "grad_norm": 0.7496936598850871, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 31711 + }, + { + "epoch": 0.31712, + "grad_norm": 0.9339326025480439, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 31712 + }, + { + "epoch": 0.31713, + "grad_norm": 1.058084119806949, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 31713 + }, + { + "epoch": 0.31714, + "grad_norm": 0.9431614384976073, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 31714 + }, + { + "epoch": 0.31715, + "grad_norm": 0.9132855353932288, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 31715 + }, + { + "epoch": 0.31716, + "grad_norm": 0.9039592159931148, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 31716 + }, + { + "epoch": 0.31717, + "grad_norm": 0.9072610648109333, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 31717 + }, + { + "epoch": 0.31718, + "grad_norm": 0.9463703092368142, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 31718 + }, + { + "epoch": 0.31719, + "grad_norm": 1.0151713822845205, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 31719 + }, + { + "epoch": 0.3172, + "grad_norm": 1.1240811045804207, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 31720 + }, + { + "epoch": 0.31721, + "grad_norm": 0.9460854421652679, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 31721 + }, + { + "epoch": 0.31722, + "grad_norm": 0.7842157349186878, + "learning_rate": 0.003, + "loss": 3.99, + "step": 31722 + }, + { + "epoch": 0.31723, + "grad_norm": 0.7459101046506807, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 31723 + }, + { + "epoch": 0.31724, + "grad_norm": 0.8647891869050701, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 31724 + }, + { + "epoch": 0.31725, + "grad_norm": 0.8893497658794384, + "learning_rate": 0.003, + "loss": 4.013, + "step": 31725 + }, + { + "epoch": 0.31726, + "grad_norm": 0.8834111020780769, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 31726 + }, + { + "epoch": 0.31727, + "grad_norm": 0.8701043078441293, + "learning_rate": 0.003, + "loss": 4.0, + "step": 31727 + }, + { + "epoch": 0.31728, + "grad_norm": 0.7919147894745585, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 31728 + }, + { + "epoch": 0.31729, + "grad_norm": 0.6733231451249339, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 31729 + }, + { + "epoch": 0.3173, + "grad_norm": 0.6638877314356568, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 31730 + }, + { + "epoch": 0.31731, + "grad_norm": 0.7426874909158137, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 31731 + }, + { + "epoch": 0.31732, + "grad_norm": 0.7840651490621319, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 31732 + }, + { + "epoch": 0.31733, + "grad_norm": 0.758644657888431, + "learning_rate": 0.003, + "loss": 4.008, + "step": 31733 + }, + { + "epoch": 0.31734, + "grad_norm": 0.809808812411536, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 31734 + }, + { + "epoch": 0.31735, + "grad_norm": 0.8055785455021943, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 31735 + }, + { + "epoch": 0.31736, + "grad_norm": 0.7333244560907675, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 31736 + }, + { + "epoch": 0.31737, + "grad_norm": 0.6467748012612559, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 31737 + }, + { + "epoch": 0.31738, + "grad_norm": 0.6733874498651308, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 31738 + }, + { + "epoch": 0.31739, + "grad_norm": 0.8978206581555167, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 31739 + }, + { + "epoch": 0.3174, + "grad_norm": 1.1596882906562225, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 31740 + }, + { + "epoch": 0.31741, + "grad_norm": 0.6890302751600853, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 31741 + }, + { + "epoch": 0.31742, + "grad_norm": 0.5749663397755224, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 31742 + }, + { + "epoch": 0.31743, + "grad_norm": 0.6416937709159793, + "learning_rate": 0.003, + "loss": 4.006, + "step": 31743 + }, + { + "epoch": 0.31744, + "grad_norm": 0.6928774240587394, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 31744 + }, + { + "epoch": 0.31745, + "grad_norm": 0.6461224617231287, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 31745 + }, + { + "epoch": 0.31746, + "grad_norm": 0.5294897979478193, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 31746 + }, + { + "epoch": 0.31747, + "grad_norm": 0.6109955473331491, + "learning_rate": 0.003, + "loss": 3.9368, + "step": 31747 + }, + { + "epoch": 0.31748, + "grad_norm": 0.667202889014503, + "learning_rate": 0.003, + "loss": 3.981, + "step": 31748 + }, + { + "epoch": 0.31749, + "grad_norm": 0.726310167443708, + "learning_rate": 0.003, + "loss": 3.979, + "step": 31749 + }, + { + "epoch": 0.3175, + "grad_norm": 0.674210400615161, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 31750 + }, + { + "epoch": 0.31751, + "grad_norm": 0.5483529366476108, + "learning_rate": 0.003, + "loss": 3.995, + "step": 31751 + }, + { + "epoch": 0.31752, + "grad_norm": 0.5818319577149661, + "learning_rate": 0.003, + "loss": 3.979, + "step": 31752 + }, + { + "epoch": 0.31753, + "grad_norm": 0.6871662583028016, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 31753 + }, + { + "epoch": 0.31754, + "grad_norm": 0.7582251506166802, + "learning_rate": 0.003, + "loss": 3.9343, + "step": 31754 + }, + { + "epoch": 0.31755, + "grad_norm": 0.7446718270823062, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 31755 + }, + { + "epoch": 0.31756, + "grad_norm": 0.6799208391182578, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 31756 + }, + { + "epoch": 0.31757, + "grad_norm": 0.7683049264898139, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 31757 + }, + { + "epoch": 0.31758, + "grad_norm": 0.8473704024914789, + "learning_rate": 0.003, + "loss": 3.979, + "step": 31758 + }, + { + "epoch": 0.31759, + "grad_norm": 0.8855736114211855, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 31759 + }, + { + "epoch": 0.3176, + "grad_norm": 0.8763259596099651, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 31760 + }, + { + "epoch": 0.31761, + "grad_norm": 0.922468696831155, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 31761 + }, + { + "epoch": 0.31762, + "grad_norm": 0.8496654696092035, + "learning_rate": 0.003, + "loss": 4.004, + "step": 31762 + }, + { + "epoch": 0.31763, + "grad_norm": 0.8126182933869873, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 31763 + }, + { + "epoch": 0.31764, + "grad_norm": 0.8355755870673461, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 31764 + }, + { + "epoch": 0.31765, + "grad_norm": 0.9601035752268813, + "learning_rate": 0.003, + "loss": 4.008, + "step": 31765 + }, + { + "epoch": 0.31766, + "grad_norm": 0.9125482994242565, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 31766 + }, + { + "epoch": 0.31767, + "grad_norm": 0.7449452933236884, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 31767 + }, + { + "epoch": 0.31768, + "grad_norm": 0.642398080768342, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 31768 + }, + { + "epoch": 0.31769, + "grad_norm": 0.6003950884964313, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 31769 + }, + { + "epoch": 0.3177, + "grad_norm": 0.563668306906634, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 31770 + }, + { + "epoch": 0.31771, + "grad_norm": 0.5883985786067228, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 31771 + }, + { + "epoch": 0.31772, + "grad_norm": 0.6882412965705705, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 31772 + }, + { + "epoch": 0.31773, + "grad_norm": 0.7994707578767958, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31773 + }, + { + "epoch": 0.31774, + "grad_norm": 0.8854270615523627, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 31774 + }, + { + "epoch": 0.31775, + "grad_norm": 0.9521192214198714, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 31775 + }, + { + "epoch": 0.31776, + "grad_norm": 0.9497934371224154, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 31776 + }, + { + "epoch": 0.31777, + "grad_norm": 0.864886787954947, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 31777 + }, + { + "epoch": 0.31778, + "grad_norm": 0.8110492954692471, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 31778 + }, + { + "epoch": 0.31779, + "grad_norm": 0.6791576367006297, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 31779 + }, + { + "epoch": 0.3178, + "grad_norm": 0.636540271412901, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 31780 + }, + { + "epoch": 0.31781, + "grad_norm": 0.5961065219703631, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 31781 + }, + { + "epoch": 0.31782, + "grad_norm": 0.6445038235822039, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 31782 + }, + { + "epoch": 0.31783, + "grad_norm": 0.7098589013687561, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 31783 + }, + { + "epoch": 0.31784, + "grad_norm": 0.7532937234601129, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 31784 + }, + { + "epoch": 0.31785, + "grad_norm": 0.8361864525252173, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 31785 + }, + { + "epoch": 0.31786, + "grad_norm": 0.9561270656844395, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 31786 + }, + { + "epoch": 0.31787, + "grad_norm": 1.126342794621896, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 31787 + }, + { + "epoch": 0.31788, + "grad_norm": 0.9616508171721904, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 31788 + }, + { + "epoch": 0.31789, + "grad_norm": 0.9459693916912466, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 31789 + }, + { + "epoch": 0.3179, + "grad_norm": 0.9782029259689415, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 31790 + }, + { + "epoch": 0.31791, + "grad_norm": 0.6885233727994079, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 31791 + }, + { + "epoch": 0.31792, + "grad_norm": 0.6595303510574755, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 31792 + }, + { + "epoch": 0.31793, + "grad_norm": 0.7494809670809002, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 31793 + }, + { + "epoch": 0.31794, + "grad_norm": 0.806226988741871, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 31794 + }, + { + "epoch": 0.31795, + "grad_norm": 0.846767273377786, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 31795 + }, + { + "epoch": 0.31796, + "grad_norm": 0.7687151654272835, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 31796 + }, + { + "epoch": 0.31797, + "grad_norm": 0.7475912505692808, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 31797 + }, + { + "epoch": 0.31798, + "grad_norm": 0.7581317943050715, + "learning_rate": 0.003, + "loss": 4.026, + "step": 31798 + }, + { + "epoch": 0.31799, + "grad_norm": 0.8342468125730619, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 31799 + }, + { + "epoch": 0.318, + "grad_norm": 0.7688606289313725, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 31800 + }, + { + "epoch": 0.31801, + "grad_norm": 0.8215720625172096, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 31801 + }, + { + "epoch": 0.31802, + "grad_norm": 0.9921453163348399, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 31802 + }, + { + "epoch": 0.31803, + "grad_norm": 1.2237429998597678, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 31803 + }, + { + "epoch": 0.31804, + "grad_norm": 0.6725727933014177, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 31804 + }, + { + "epoch": 0.31805, + "grad_norm": 0.677938925966328, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 31805 + }, + { + "epoch": 0.31806, + "grad_norm": 0.7213663719242345, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 31806 + }, + { + "epoch": 0.31807, + "grad_norm": 0.8224762082223116, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 31807 + }, + { + "epoch": 0.31808, + "grad_norm": 0.7141131577818105, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 31808 + }, + { + "epoch": 0.31809, + "grad_norm": 0.571779731972319, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 31809 + }, + { + "epoch": 0.3181, + "grad_norm": 0.6994236171621633, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 31810 + }, + { + "epoch": 0.31811, + "grad_norm": 0.8580255970561842, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 31811 + }, + { + "epoch": 0.31812, + "grad_norm": 0.9051441294580559, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 31812 + }, + { + "epoch": 0.31813, + "grad_norm": 0.8432164353820426, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 31813 + }, + { + "epoch": 0.31814, + "grad_norm": 0.7268091524623942, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 31814 + }, + { + "epoch": 0.31815, + "grad_norm": 0.7716264963910653, + "learning_rate": 0.003, + "loss": 3.994, + "step": 31815 + }, + { + "epoch": 0.31816, + "grad_norm": 0.735685538700645, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 31816 + }, + { + "epoch": 0.31817, + "grad_norm": 0.7530517874322691, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 31817 + }, + { + "epoch": 0.31818, + "grad_norm": 0.8169384335610046, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 31818 + }, + { + "epoch": 0.31819, + "grad_norm": 0.7457511126013204, + "learning_rate": 0.003, + "loss": 3.987, + "step": 31819 + }, + { + "epoch": 0.3182, + "grad_norm": 0.6588909273954243, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 31820 + }, + { + "epoch": 0.31821, + "grad_norm": 0.6777444548637176, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 31821 + }, + { + "epoch": 0.31822, + "grad_norm": 0.7129753255936896, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 31822 + }, + { + "epoch": 0.31823, + "grad_norm": 0.6781534433900759, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 31823 + }, + { + "epoch": 0.31824, + "grad_norm": 0.6912237127555697, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 31824 + }, + { + "epoch": 0.31825, + "grad_norm": 0.7104797147619997, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 31825 + }, + { + "epoch": 0.31826, + "grad_norm": 0.7161501784203648, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 31826 + }, + { + "epoch": 0.31827, + "grad_norm": 0.9207062601881275, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 31827 + }, + { + "epoch": 0.31828, + "grad_norm": 1.1115966128183645, + "learning_rate": 0.003, + "loss": 3.965, + "step": 31828 + }, + { + "epoch": 0.31829, + "grad_norm": 0.7563520002616145, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 31829 + }, + { + "epoch": 0.3183, + "grad_norm": 0.6308265461193289, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 31830 + }, + { + "epoch": 0.31831, + "grad_norm": 0.6599632485346836, + "learning_rate": 0.003, + "loss": 4.011, + "step": 31831 + }, + { + "epoch": 0.31832, + "grad_norm": 0.754387540299049, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 31832 + }, + { + "epoch": 0.31833, + "grad_norm": 0.8221193894587014, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 31833 + }, + { + "epoch": 0.31834, + "grad_norm": 0.9005553543962926, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 31834 + }, + { + "epoch": 0.31835, + "grad_norm": 0.8149946141566686, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 31835 + }, + { + "epoch": 0.31836, + "grad_norm": 0.8483867889315377, + "learning_rate": 0.003, + "loss": 3.993, + "step": 31836 + }, + { + "epoch": 0.31837, + "grad_norm": 0.7644064640178072, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 31837 + }, + { + "epoch": 0.31838, + "grad_norm": 0.7274373116928567, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 31838 + }, + { + "epoch": 0.31839, + "grad_norm": 0.7559585333751206, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 31839 + }, + { + "epoch": 0.3184, + "grad_norm": 0.6832731412252002, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 31840 + }, + { + "epoch": 0.31841, + "grad_norm": 0.739262601953663, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 31841 + }, + { + "epoch": 0.31842, + "grad_norm": 0.7273446328088662, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 31842 + }, + { + "epoch": 0.31843, + "grad_norm": 0.7753395089167503, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 31843 + }, + { + "epoch": 0.31844, + "grad_norm": 0.884222004734516, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 31844 + }, + { + "epoch": 0.31845, + "grad_norm": 0.9061418760946485, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 31845 + }, + { + "epoch": 0.31846, + "grad_norm": 0.8381317957201302, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 31846 + }, + { + "epoch": 0.31847, + "grad_norm": 0.7116319539900883, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 31847 + }, + { + "epoch": 0.31848, + "grad_norm": 0.64817424088107, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 31848 + }, + { + "epoch": 0.31849, + "grad_norm": 0.7088497345668447, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 31849 + }, + { + "epoch": 0.3185, + "grad_norm": 0.7573872487984202, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 31850 + }, + { + "epoch": 0.31851, + "grad_norm": 0.8648385704870547, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 31851 + }, + { + "epoch": 0.31852, + "grad_norm": 1.0809548038196464, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 31852 + }, + { + "epoch": 0.31853, + "grad_norm": 0.9696068528196983, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 31853 + }, + { + "epoch": 0.31854, + "grad_norm": 1.0268696664557517, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 31854 + }, + { + "epoch": 0.31855, + "grad_norm": 0.8468537762813815, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 31855 + }, + { + "epoch": 0.31856, + "grad_norm": 0.7288849010465094, + "learning_rate": 0.003, + "loss": 3.979, + "step": 31856 + }, + { + "epoch": 0.31857, + "grad_norm": 0.6590721380351754, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 31857 + }, + { + "epoch": 0.31858, + "grad_norm": 0.5648333098600069, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 31858 + }, + { + "epoch": 0.31859, + "grad_norm": 0.5473908880678279, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 31859 + }, + { + "epoch": 0.3186, + "grad_norm": 0.48770639799081267, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 31860 + }, + { + "epoch": 0.31861, + "grad_norm": 0.41354843920086226, + "learning_rate": 0.003, + "loss": 3.9393, + "step": 31861 + }, + { + "epoch": 0.31862, + "grad_norm": 0.410822970920087, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 31862 + }, + { + "epoch": 0.31863, + "grad_norm": 0.39256999562039113, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 31863 + }, + { + "epoch": 0.31864, + "grad_norm": 0.4503355022501163, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 31864 + }, + { + "epoch": 0.31865, + "grad_norm": 0.5335505986985645, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 31865 + }, + { + "epoch": 0.31866, + "grad_norm": 0.651358215781505, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 31866 + }, + { + "epoch": 0.31867, + "grad_norm": 0.7061136021450987, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 31867 + }, + { + "epoch": 0.31868, + "grad_norm": 0.781192591014719, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 31868 + }, + { + "epoch": 0.31869, + "grad_norm": 0.9076079392321944, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 31869 + }, + { + "epoch": 0.3187, + "grad_norm": 1.0741606253004352, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 31870 + }, + { + "epoch": 0.31871, + "grad_norm": 0.8768190260973738, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 31871 + }, + { + "epoch": 0.31872, + "grad_norm": 0.8726850202137618, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 31872 + }, + { + "epoch": 0.31873, + "grad_norm": 0.8324251549690226, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 31873 + }, + { + "epoch": 0.31874, + "grad_norm": 0.9326029332344585, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 31874 + }, + { + "epoch": 0.31875, + "grad_norm": 0.9849764956385183, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31875 + }, + { + "epoch": 0.31876, + "grad_norm": 0.8720825705973673, + "learning_rate": 0.003, + "loss": 3.983, + "step": 31876 + }, + { + "epoch": 0.31877, + "grad_norm": 0.9596625070168259, + "learning_rate": 0.003, + "loss": 3.994, + "step": 31877 + }, + { + "epoch": 0.31878, + "grad_norm": 1.164773153735069, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 31878 + }, + { + "epoch": 0.31879, + "grad_norm": 0.918691488333349, + "learning_rate": 0.003, + "loss": 3.996, + "step": 31879 + }, + { + "epoch": 0.3188, + "grad_norm": 0.8547783569788111, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 31880 + }, + { + "epoch": 0.31881, + "grad_norm": 0.8825302240221657, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 31881 + }, + { + "epoch": 0.31882, + "grad_norm": 0.7771271265679142, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 31882 + }, + { + "epoch": 0.31883, + "grad_norm": 0.7869732135304731, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 31883 + }, + { + "epoch": 0.31884, + "grad_norm": 0.7481474489931419, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 31884 + }, + { + "epoch": 0.31885, + "grad_norm": 0.8382067145858256, + "learning_rate": 0.003, + "loss": 3.993, + "step": 31885 + }, + { + "epoch": 0.31886, + "grad_norm": 0.8964604966404649, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 31886 + }, + { + "epoch": 0.31887, + "grad_norm": 0.9238634402691546, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 31887 + }, + { + "epoch": 0.31888, + "grad_norm": 0.8105440535383424, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 31888 + }, + { + "epoch": 0.31889, + "grad_norm": 0.7203985385523838, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 31889 + }, + { + "epoch": 0.3189, + "grad_norm": 0.7304096344170811, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 31890 + }, + { + "epoch": 0.31891, + "grad_norm": 0.6648854885055627, + "learning_rate": 0.003, + "loss": 3.975, + "step": 31891 + }, + { + "epoch": 0.31892, + "grad_norm": 0.7313221702190056, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 31892 + }, + { + "epoch": 0.31893, + "grad_norm": 0.839521105414917, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 31893 + }, + { + "epoch": 0.31894, + "grad_norm": 0.8482078475628754, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 31894 + }, + { + "epoch": 0.31895, + "grad_norm": 0.741941051622253, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 31895 + }, + { + "epoch": 0.31896, + "grad_norm": 0.6654480731937702, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 31896 + }, + { + "epoch": 0.31897, + "grad_norm": 0.6417944826913009, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 31897 + }, + { + "epoch": 0.31898, + "grad_norm": 0.5200785947137824, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 31898 + }, + { + "epoch": 0.31899, + "grad_norm": 0.5383163549098526, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 31899 + }, + { + "epoch": 0.319, + "grad_norm": 0.5478482038221933, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 31900 + }, + { + "epoch": 0.31901, + "grad_norm": 0.6516255500016235, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 31901 + }, + { + "epoch": 0.31902, + "grad_norm": 0.7809076843304884, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 31902 + }, + { + "epoch": 0.31903, + "grad_norm": 0.9470623803802675, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 31903 + }, + { + "epoch": 0.31904, + "grad_norm": 0.9318409658492729, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 31904 + }, + { + "epoch": 0.31905, + "grad_norm": 0.7530259677576897, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 31905 + }, + { + "epoch": 0.31906, + "grad_norm": 0.6818069189038456, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 31906 + }, + { + "epoch": 0.31907, + "grad_norm": 0.6681572861357502, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 31907 + }, + { + "epoch": 0.31908, + "grad_norm": 0.8313026101239096, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 31908 + }, + { + "epoch": 0.31909, + "grad_norm": 0.8596624447857342, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 31909 + }, + { + "epoch": 0.3191, + "grad_norm": 0.8455172277177909, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 31910 + }, + { + "epoch": 0.31911, + "grad_norm": 0.7255770900612719, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 31911 + }, + { + "epoch": 0.31912, + "grad_norm": 0.6798466285073222, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 31912 + }, + { + "epoch": 0.31913, + "grad_norm": 0.7999757929773575, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 31913 + }, + { + "epoch": 0.31914, + "grad_norm": 0.9675730033401978, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 31914 + }, + { + "epoch": 0.31915, + "grad_norm": 1.051956872098646, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 31915 + }, + { + "epoch": 0.31916, + "grad_norm": 0.901572020575101, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 31916 + }, + { + "epoch": 0.31917, + "grad_norm": 0.8095287178693227, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 31917 + }, + { + "epoch": 0.31918, + "grad_norm": 0.8844451675113253, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 31918 + }, + { + "epoch": 0.31919, + "grad_norm": 0.9166728024135784, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 31919 + }, + { + "epoch": 0.3192, + "grad_norm": 0.8525849440292542, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 31920 + }, + { + "epoch": 0.31921, + "grad_norm": 0.8034631947766953, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 31921 + }, + { + "epoch": 0.31922, + "grad_norm": 0.6737821849415834, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 31922 + }, + { + "epoch": 0.31923, + "grad_norm": 0.7086489965309919, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 31923 + }, + { + "epoch": 0.31924, + "grad_norm": 0.7255115193035477, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 31924 + }, + { + "epoch": 0.31925, + "grad_norm": 0.733287596935119, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 31925 + }, + { + "epoch": 0.31926, + "grad_norm": 0.8490318391466672, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 31926 + }, + { + "epoch": 0.31927, + "grad_norm": 0.8847699705273674, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 31927 + }, + { + "epoch": 0.31928, + "grad_norm": 0.9828173361648305, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 31928 + }, + { + "epoch": 0.31929, + "grad_norm": 0.9349530333376154, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 31929 + }, + { + "epoch": 0.3193, + "grad_norm": 0.6948495132674675, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 31930 + }, + { + "epoch": 0.31931, + "grad_norm": 0.6122958716750048, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 31931 + }, + { + "epoch": 0.31932, + "grad_norm": 0.5858671532671529, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 31932 + }, + { + "epoch": 0.31933, + "grad_norm": 0.5390589720118347, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 31933 + }, + { + "epoch": 0.31934, + "grad_norm": 0.5288800520084221, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 31934 + }, + { + "epoch": 0.31935, + "grad_norm": 0.5168050938123749, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 31935 + }, + { + "epoch": 0.31936, + "grad_norm": 0.5050877429759719, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 31936 + }, + { + "epoch": 0.31937, + "grad_norm": 0.5534075649008333, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 31937 + }, + { + "epoch": 0.31938, + "grad_norm": 0.5887924779244667, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 31938 + }, + { + "epoch": 0.31939, + "grad_norm": 0.7480822349019131, + "learning_rate": 0.003, + "loss": 3.967, + "step": 31939 + }, + { + "epoch": 0.3194, + "grad_norm": 0.8840806473608013, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 31940 + }, + { + "epoch": 0.31941, + "grad_norm": 0.8545137802664997, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 31941 + }, + { + "epoch": 0.31942, + "grad_norm": 0.7821126270729682, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 31942 + }, + { + "epoch": 0.31943, + "grad_norm": 0.8042745085526383, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 31943 + }, + { + "epoch": 0.31944, + "grad_norm": 0.8474573688663235, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 31944 + }, + { + "epoch": 0.31945, + "grad_norm": 0.8494883352795717, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 31945 + }, + { + "epoch": 0.31946, + "grad_norm": 0.9115045909421171, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 31946 + }, + { + "epoch": 0.31947, + "grad_norm": 1.0502058004972759, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 31947 + }, + { + "epoch": 0.31948, + "grad_norm": 1.040172827625348, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 31948 + }, + { + "epoch": 0.31949, + "grad_norm": 0.8239026426972503, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 31949 + }, + { + "epoch": 0.3195, + "grad_norm": 0.7288380251245961, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 31950 + }, + { + "epoch": 0.31951, + "grad_norm": 0.7616414151973018, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 31951 + }, + { + "epoch": 0.31952, + "grad_norm": 0.8353074595513478, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 31952 + }, + { + "epoch": 0.31953, + "grad_norm": 0.9085485179035011, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 31953 + }, + { + "epoch": 0.31954, + "grad_norm": 0.9646104243291471, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 31954 + }, + { + "epoch": 0.31955, + "grad_norm": 0.9595084571455332, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 31955 + }, + { + "epoch": 0.31956, + "grad_norm": 0.7517555960095799, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 31956 + }, + { + "epoch": 0.31957, + "grad_norm": 0.7591790477807613, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 31957 + }, + { + "epoch": 0.31958, + "grad_norm": 0.7420872390714406, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 31958 + }, + { + "epoch": 0.31959, + "grad_norm": 0.6192089472262413, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 31959 + }, + { + "epoch": 0.3196, + "grad_norm": 0.6092107216774968, + "learning_rate": 0.003, + "loss": 3.994, + "step": 31960 + }, + { + "epoch": 0.31961, + "grad_norm": 0.663276022245525, + "learning_rate": 0.003, + "loss": 4.007, + "step": 31961 + }, + { + "epoch": 0.31962, + "grad_norm": 0.6350292224597864, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 31962 + }, + { + "epoch": 0.31963, + "grad_norm": 0.6521739150690882, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 31963 + }, + { + "epoch": 0.31964, + "grad_norm": 0.7375276927151186, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 31964 + }, + { + "epoch": 0.31965, + "grad_norm": 0.8342455201422586, + "learning_rate": 0.003, + "loss": 4.0411, + "step": 31965 + }, + { + "epoch": 0.31966, + "grad_norm": 0.8318613506780602, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 31966 + }, + { + "epoch": 0.31967, + "grad_norm": 0.7710382521431954, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 31967 + }, + { + "epoch": 0.31968, + "grad_norm": 0.8108612317895993, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 31968 + }, + { + "epoch": 0.31969, + "grad_norm": 0.9026174948768635, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 31969 + }, + { + "epoch": 0.3197, + "grad_norm": 0.8977451145196578, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 31970 + }, + { + "epoch": 0.31971, + "grad_norm": 0.9433939665981164, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 31971 + }, + { + "epoch": 0.31972, + "grad_norm": 1.0131516754874026, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 31972 + }, + { + "epoch": 0.31973, + "grad_norm": 1.010177658340365, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 31973 + }, + { + "epoch": 0.31974, + "grad_norm": 0.8608741612639113, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 31974 + }, + { + "epoch": 0.31975, + "grad_norm": 0.8026376083078866, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 31975 + }, + { + "epoch": 0.31976, + "grad_norm": 0.8388253988183354, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 31976 + }, + { + "epoch": 0.31977, + "grad_norm": 0.9567429839156887, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 31977 + }, + { + "epoch": 0.31978, + "grad_norm": 0.9882419828434348, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 31978 + }, + { + "epoch": 0.31979, + "grad_norm": 0.8708929425973461, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 31979 + }, + { + "epoch": 0.3198, + "grad_norm": 0.786263777077322, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 31980 + }, + { + "epoch": 0.31981, + "grad_norm": 0.7906251500467005, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 31981 + }, + { + "epoch": 0.31982, + "grad_norm": 0.7197179273914398, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 31982 + }, + { + "epoch": 0.31983, + "grad_norm": 0.747804059466519, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 31983 + }, + { + "epoch": 0.31984, + "grad_norm": 0.6741150032825922, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 31984 + }, + { + "epoch": 0.31985, + "grad_norm": 0.7555939591416013, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 31985 + }, + { + "epoch": 0.31986, + "grad_norm": 0.7205794570032646, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 31986 + }, + { + "epoch": 0.31987, + "grad_norm": 0.7102697386915516, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 31987 + }, + { + "epoch": 0.31988, + "grad_norm": 0.5900077104016238, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 31988 + }, + { + "epoch": 0.31989, + "grad_norm": 0.5619324796483888, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 31989 + }, + { + "epoch": 0.3199, + "grad_norm": 0.6053725517836074, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 31990 + }, + { + "epoch": 0.31991, + "grad_norm": 0.6660525601648876, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 31991 + }, + { + "epoch": 0.31992, + "grad_norm": 0.7551360128235625, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 31992 + }, + { + "epoch": 0.31993, + "grad_norm": 0.7866713416697798, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 31993 + }, + { + "epoch": 0.31994, + "grad_norm": 0.9430081411967484, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 31994 + }, + { + "epoch": 0.31995, + "grad_norm": 0.9746481420365113, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 31995 + }, + { + "epoch": 0.31996, + "grad_norm": 0.8737352844603216, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 31996 + }, + { + "epoch": 0.31997, + "grad_norm": 0.8434946453872948, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 31997 + }, + { + "epoch": 0.31998, + "grad_norm": 0.7197802306636427, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 31998 + }, + { + "epoch": 0.31999, + "grad_norm": 0.7307342984901362, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 31999 + }, + { + "epoch": 0.32, + "grad_norm": 0.6919947895243121, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 32000 + }, + { + "epoch": 0.32001, + "grad_norm": 0.7057231122358727, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 32001 + }, + { + "epoch": 0.32002, + "grad_norm": 0.7128843649070411, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 32002 + }, + { + "epoch": 0.32003, + "grad_norm": 0.7331520675167915, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 32003 + }, + { + "epoch": 0.32004, + "grad_norm": 0.6898010056340559, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 32004 + }, + { + "epoch": 0.32005, + "grad_norm": 0.723251444160145, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 32005 + }, + { + "epoch": 0.32006, + "grad_norm": 0.6737566450023083, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 32006 + }, + { + "epoch": 0.32007, + "grad_norm": 0.6047920646954645, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 32007 + }, + { + "epoch": 0.32008, + "grad_norm": 0.6365291832809371, + "learning_rate": 0.003, + "loss": 3.965, + "step": 32008 + }, + { + "epoch": 0.32009, + "grad_norm": 0.6292304678593418, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 32009 + }, + { + "epoch": 0.3201, + "grad_norm": 0.6232998273136429, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 32010 + }, + { + "epoch": 0.32011, + "grad_norm": 0.6779099305570605, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 32011 + }, + { + "epoch": 0.32012, + "grad_norm": 0.7578252111493488, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 32012 + }, + { + "epoch": 0.32013, + "grad_norm": 0.8154220630321735, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 32013 + }, + { + "epoch": 0.32014, + "grad_norm": 0.8059317296736411, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 32014 + }, + { + "epoch": 0.32015, + "grad_norm": 0.8169640424695499, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 32015 + }, + { + "epoch": 0.32016, + "grad_norm": 0.9465669218270005, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 32016 + }, + { + "epoch": 0.32017, + "grad_norm": 1.0590594791596775, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 32017 + }, + { + "epoch": 0.32018, + "grad_norm": 1.0617033061269046, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 32018 + }, + { + "epoch": 0.32019, + "grad_norm": 0.7816409915229384, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 32019 + }, + { + "epoch": 0.3202, + "grad_norm": 0.6651851074981271, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 32020 + }, + { + "epoch": 0.32021, + "grad_norm": 0.8342959088737365, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 32021 + }, + { + "epoch": 0.32022, + "grad_norm": 0.9361597588004823, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 32022 + }, + { + "epoch": 0.32023, + "grad_norm": 0.9791618667674861, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 32023 + }, + { + "epoch": 0.32024, + "grad_norm": 0.8768546456708405, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 32024 + }, + { + "epoch": 0.32025, + "grad_norm": 0.8323796043077157, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 32025 + }, + { + "epoch": 0.32026, + "grad_norm": 0.863553221700312, + "learning_rate": 0.003, + "loss": 4.005, + "step": 32026 + }, + { + "epoch": 0.32027, + "grad_norm": 0.909280520512369, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 32027 + }, + { + "epoch": 0.32028, + "grad_norm": 0.8087189886605777, + "learning_rate": 0.003, + "loss": 3.99, + "step": 32028 + }, + { + "epoch": 0.32029, + "grad_norm": 0.7677622627836047, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 32029 + }, + { + "epoch": 0.3203, + "grad_norm": 0.7684954992467827, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 32030 + }, + { + "epoch": 0.32031, + "grad_norm": 0.7179150481015238, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 32031 + }, + { + "epoch": 0.32032, + "grad_norm": 0.6628247654027073, + "learning_rate": 0.003, + "loss": 4.014, + "step": 32032 + }, + { + "epoch": 0.32033, + "grad_norm": 0.6643816113048233, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 32033 + }, + { + "epoch": 0.32034, + "grad_norm": 0.555580896851849, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 32034 + }, + { + "epoch": 0.32035, + "grad_norm": 0.6056329038194807, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 32035 + }, + { + "epoch": 0.32036, + "grad_norm": 0.5849728085263639, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 32036 + }, + { + "epoch": 0.32037, + "grad_norm": 0.6797418626849081, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 32037 + }, + { + "epoch": 0.32038, + "grad_norm": 0.7367897834849633, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 32038 + }, + { + "epoch": 0.32039, + "grad_norm": 0.7113406595712753, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 32039 + }, + { + "epoch": 0.3204, + "grad_norm": 0.6919482336380433, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 32040 + }, + { + "epoch": 0.32041, + "grad_norm": 0.6202246912455543, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 32041 + }, + { + "epoch": 0.32042, + "grad_norm": 0.6527695173432879, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 32042 + }, + { + "epoch": 0.32043, + "grad_norm": 0.6871254782101617, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 32043 + }, + { + "epoch": 0.32044, + "grad_norm": 0.7671660381538162, + "learning_rate": 0.003, + "loss": 3.983, + "step": 32044 + }, + { + "epoch": 0.32045, + "grad_norm": 0.7852966152736531, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 32045 + }, + { + "epoch": 0.32046, + "grad_norm": 0.7366954264235952, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 32046 + }, + { + "epoch": 0.32047, + "grad_norm": 0.8420163572449417, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 32047 + }, + { + "epoch": 0.32048, + "grad_norm": 1.0171977286608, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 32048 + }, + { + "epoch": 0.32049, + "grad_norm": 1.0790947825026562, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 32049 + }, + { + "epoch": 0.3205, + "grad_norm": 0.9432026845357054, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32050 + }, + { + "epoch": 0.32051, + "grad_norm": 0.8239237860179928, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 32051 + }, + { + "epoch": 0.32052, + "grad_norm": 0.7240729914279501, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 32052 + }, + { + "epoch": 0.32053, + "grad_norm": 0.6977480943175844, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 32053 + }, + { + "epoch": 0.32054, + "grad_norm": 0.6630680538864455, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 32054 + }, + { + "epoch": 0.32055, + "grad_norm": 0.7021402392911524, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 32055 + }, + { + "epoch": 0.32056, + "grad_norm": 0.7602752683049586, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 32056 + }, + { + "epoch": 0.32057, + "grad_norm": 0.7886077666800678, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 32057 + }, + { + "epoch": 0.32058, + "grad_norm": 0.7573627189033812, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 32058 + }, + { + "epoch": 0.32059, + "grad_norm": 0.7299158492656895, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 32059 + }, + { + "epoch": 0.3206, + "grad_norm": 0.8021752528690828, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 32060 + }, + { + "epoch": 0.32061, + "grad_norm": 0.8920861964532211, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 32061 + }, + { + "epoch": 0.32062, + "grad_norm": 0.9354677508452879, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 32062 + }, + { + "epoch": 0.32063, + "grad_norm": 0.9707348230831223, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 32063 + }, + { + "epoch": 0.32064, + "grad_norm": 0.8450814668923479, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 32064 + }, + { + "epoch": 0.32065, + "grad_norm": 0.7503122230674278, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 32065 + }, + { + "epoch": 0.32066, + "grad_norm": 0.6285757830162781, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 32066 + }, + { + "epoch": 0.32067, + "grad_norm": 0.6159514260711, + "learning_rate": 0.003, + "loss": 3.992, + "step": 32067 + }, + { + "epoch": 0.32068, + "grad_norm": 0.7060756132428246, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 32068 + }, + { + "epoch": 0.32069, + "grad_norm": 0.8048976903449161, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 32069 + }, + { + "epoch": 0.3207, + "grad_norm": 0.783095502233158, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 32070 + }, + { + "epoch": 0.32071, + "grad_norm": 0.8555320079994186, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 32071 + }, + { + "epoch": 0.32072, + "grad_norm": 0.8642259749399621, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 32072 + }, + { + "epoch": 0.32073, + "grad_norm": 0.7538980639617767, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 32073 + }, + { + "epoch": 0.32074, + "grad_norm": 0.7207306286138376, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 32074 + }, + { + "epoch": 0.32075, + "grad_norm": 0.643371824395421, + "learning_rate": 0.003, + "loss": 3.979, + "step": 32075 + }, + { + "epoch": 0.32076, + "grad_norm": 0.7093957655435345, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 32076 + }, + { + "epoch": 0.32077, + "grad_norm": 0.7150201571244981, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 32077 + }, + { + "epoch": 0.32078, + "grad_norm": 0.7079259925112834, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 32078 + }, + { + "epoch": 0.32079, + "grad_norm": 0.7605356719726919, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 32079 + }, + { + "epoch": 0.3208, + "grad_norm": 0.8578510561038674, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 32080 + }, + { + "epoch": 0.32081, + "grad_norm": 0.9007391814211166, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 32081 + }, + { + "epoch": 0.32082, + "grad_norm": 0.9782677692831245, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 32082 + }, + { + "epoch": 0.32083, + "grad_norm": 0.8218006776327383, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 32083 + }, + { + "epoch": 0.32084, + "grad_norm": 0.7198649033780099, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 32084 + }, + { + "epoch": 0.32085, + "grad_norm": 0.7470595593762345, + "learning_rate": 0.003, + "loss": 4.022, + "step": 32085 + }, + { + "epoch": 0.32086, + "grad_norm": 0.7796840996988588, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 32086 + }, + { + "epoch": 0.32087, + "grad_norm": 0.7582187392643048, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 32087 + }, + { + "epoch": 0.32088, + "grad_norm": 0.6851282374313443, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 32088 + }, + { + "epoch": 0.32089, + "grad_norm": 0.6235303128462065, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 32089 + }, + { + "epoch": 0.3209, + "grad_norm": 0.6121910705897056, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 32090 + }, + { + "epoch": 0.32091, + "grad_norm": 0.5867308860178237, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 32091 + }, + { + "epoch": 0.32092, + "grad_norm": 0.5826049500340156, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 32092 + }, + { + "epoch": 0.32093, + "grad_norm": 0.653769189036181, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 32093 + }, + { + "epoch": 0.32094, + "grad_norm": 0.6902932643003685, + "learning_rate": 0.003, + "loss": 3.971, + "step": 32094 + }, + { + "epoch": 0.32095, + "grad_norm": 0.7067654927599522, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 32095 + }, + { + "epoch": 0.32096, + "grad_norm": 0.778042594438611, + "learning_rate": 0.003, + "loss": 3.993, + "step": 32096 + }, + { + "epoch": 0.32097, + "grad_norm": 0.8233398438785431, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 32097 + }, + { + "epoch": 0.32098, + "grad_norm": 0.9187915843414044, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 32098 + }, + { + "epoch": 0.32099, + "grad_norm": 0.9973540043318528, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 32099 + }, + { + "epoch": 0.321, + "grad_norm": 1.1010515236231937, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 32100 + }, + { + "epoch": 0.32101, + "grad_norm": 0.9977328184661128, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 32101 + }, + { + "epoch": 0.32102, + "grad_norm": 1.1048305324264664, + "learning_rate": 0.003, + "loss": 3.96, + "step": 32102 + }, + { + "epoch": 0.32103, + "grad_norm": 0.975817992011615, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 32103 + }, + { + "epoch": 0.32104, + "grad_norm": 0.8559935013685342, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 32104 + }, + { + "epoch": 0.32105, + "grad_norm": 0.9800816648779553, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 32105 + }, + { + "epoch": 0.32106, + "grad_norm": 1.0687439061779822, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 32106 + }, + { + "epoch": 0.32107, + "grad_norm": 0.7354760583151407, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 32107 + }, + { + "epoch": 0.32108, + "grad_norm": 0.7010657371415377, + "learning_rate": 0.003, + "loss": 4.002, + "step": 32108 + }, + { + "epoch": 0.32109, + "grad_norm": 0.6886602984950979, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 32109 + }, + { + "epoch": 0.3211, + "grad_norm": 0.695061869249373, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 32110 + }, + { + "epoch": 0.32111, + "grad_norm": 0.702871882881348, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 32111 + }, + { + "epoch": 0.32112, + "grad_norm": 0.7303891327674774, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 32112 + }, + { + "epoch": 0.32113, + "grad_norm": 0.7845981721025582, + "learning_rate": 0.003, + "loss": 4.011, + "step": 32113 + }, + { + "epoch": 0.32114, + "grad_norm": 0.8872214747300189, + "learning_rate": 0.003, + "loss": 4.035, + "step": 32114 + }, + { + "epoch": 0.32115, + "grad_norm": 0.9407972409480547, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 32115 + }, + { + "epoch": 0.32116, + "grad_norm": 0.889803243305263, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 32116 + }, + { + "epoch": 0.32117, + "grad_norm": 0.8379885307119157, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 32117 + }, + { + "epoch": 0.32118, + "grad_norm": 0.8991202271981841, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 32118 + }, + { + "epoch": 0.32119, + "grad_norm": 0.9177560967530771, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 32119 + }, + { + "epoch": 0.3212, + "grad_norm": 0.8264775406072646, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 32120 + }, + { + "epoch": 0.32121, + "grad_norm": 0.8198321786579157, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 32121 + }, + { + "epoch": 0.32122, + "grad_norm": 0.7934531336147251, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 32122 + }, + { + "epoch": 0.32123, + "grad_norm": 0.6939375737667486, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 32123 + }, + { + "epoch": 0.32124, + "grad_norm": 0.7080254086256264, + "learning_rate": 0.003, + "loss": 4.002, + "step": 32124 + }, + { + "epoch": 0.32125, + "grad_norm": 0.6833901104662135, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 32125 + }, + { + "epoch": 0.32126, + "grad_norm": 0.7042872622138089, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 32126 + }, + { + "epoch": 0.32127, + "grad_norm": 0.7251209800378783, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 32127 + }, + { + "epoch": 0.32128, + "grad_norm": 0.7608615781858414, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 32128 + }, + { + "epoch": 0.32129, + "grad_norm": 0.7846541896367375, + "learning_rate": 0.003, + "loss": 4.016, + "step": 32129 + }, + { + "epoch": 0.3213, + "grad_norm": 0.8308972557294461, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 32130 + }, + { + "epoch": 0.32131, + "grad_norm": 0.7168737664621124, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 32131 + }, + { + "epoch": 0.32132, + "grad_norm": 0.7292270728583966, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 32132 + }, + { + "epoch": 0.32133, + "grad_norm": 0.7466393305823277, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 32133 + }, + { + "epoch": 0.32134, + "grad_norm": 0.6849138716005883, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 32134 + }, + { + "epoch": 0.32135, + "grad_norm": 0.668335695534405, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 32135 + }, + { + "epoch": 0.32136, + "grad_norm": 0.7535378601448575, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 32136 + }, + { + "epoch": 0.32137, + "grad_norm": 0.8057652835384281, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 32137 + }, + { + "epoch": 0.32138, + "grad_norm": 0.8586567491452892, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 32138 + }, + { + "epoch": 0.32139, + "grad_norm": 0.7360838013397533, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 32139 + }, + { + "epoch": 0.3214, + "grad_norm": 0.72252284679597, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 32140 + }, + { + "epoch": 0.32141, + "grad_norm": 0.7629607282669221, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 32141 + }, + { + "epoch": 0.32142, + "grad_norm": 0.7590380386245028, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 32142 + }, + { + "epoch": 0.32143, + "grad_norm": 0.7393899812502166, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 32143 + }, + { + "epoch": 0.32144, + "grad_norm": 0.7903767981612639, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 32144 + }, + { + "epoch": 0.32145, + "grad_norm": 0.8292950590321136, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 32145 + }, + { + "epoch": 0.32146, + "grad_norm": 0.7831110612642261, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 32146 + }, + { + "epoch": 0.32147, + "grad_norm": 0.8324385522980581, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 32147 + }, + { + "epoch": 0.32148, + "grad_norm": 0.8709127470875128, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 32148 + }, + { + "epoch": 0.32149, + "grad_norm": 0.8101534602407837, + "learning_rate": 0.003, + "loss": 3.9358, + "step": 32149 + }, + { + "epoch": 0.3215, + "grad_norm": 0.8583682216654982, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 32150 + }, + { + "epoch": 0.32151, + "grad_norm": 0.976623657256968, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 32151 + }, + { + "epoch": 0.32152, + "grad_norm": 1.0362733320366915, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 32152 + }, + { + "epoch": 0.32153, + "grad_norm": 0.811883174874041, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 32153 + }, + { + "epoch": 0.32154, + "grad_norm": 0.8252019233691393, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 32154 + }, + { + "epoch": 0.32155, + "grad_norm": 0.7487519823551032, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 32155 + }, + { + "epoch": 0.32156, + "grad_norm": 0.7052748994720115, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 32156 + }, + { + "epoch": 0.32157, + "grad_norm": 0.7468697377014418, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 32157 + }, + { + "epoch": 0.32158, + "grad_norm": 0.7403253571849318, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 32158 + }, + { + "epoch": 0.32159, + "grad_norm": 0.7124047446126223, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 32159 + }, + { + "epoch": 0.3216, + "grad_norm": 0.6606541316007202, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 32160 + }, + { + "epoch": 0.32161, + "grad_norm": 0.6570305733833857, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 32161 + }, + { + "epoch": 0.32162, + "grad_norm": 0.6214383292206244, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 32162 + }, + { + "epoch": 0.32163, + "grad_norm": 0.542886299709257, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 32163 + }, + { + "epoch": 0.32164, + "grad_norm": 0.57346491876282, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 32164 + }, + { + "epoch": 0.32165, + "grad_norm": 0.6137070366118975, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 32165 + }, + { + "epoch": 0.32166, + "grad_norm": 0.7506189408564139, + "learning_rate": 0.003, + "loss": 3.987, + "step": 32166 + }, + { + "epoch": 0.32167, + "grad_norm": 0.9094234321283536, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 32167 + }, + { + "epoch": 0.32168, + "grad_norm": 0.963045188717562, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 32168 + }, + { + "epoch": 0.32169, + "grad_norm": 0.8342288369970569, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 32169 + }, + { + "epoch": 0.3217, + "grad_norm": 0.855525117545117, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 32170 + }, + { + "epoch": 0.32171, + "grad_norm": 0.9214549961722142, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 32171 + }, + { + "epoch": 0.32172, + "grad_norm": 0.8426575101381236, + "learning_rate": 0.003, + "loss": 4.02, + "step": 32172 + }, + { + "epoch": 0.32173, + "grad_norm": 0.8273874158496256, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 32173 + }, + { + "epoch": 0.32174, + "grad_norm": 0.89555567107049, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 32174 + }, + { + "epoch": 0.32175, + "grad_norm": 1.0029853556428425, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 32175 + }, + { + "epoch": 0.32176, + "grad_norm": 1.1850559159164176, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 32176 + }, + { + "epoch": 0.32177, + "grad_norm": 0.8808549148531364, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 32177 + }, + { + "epoch": 0.32178, + "grad_norm": 0.8379493903240176, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 32178 + }, + { + "epoch": 0.32179, + "grad_norm": 0.8606349640126761, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 32179 + }, + { + "epoch": 0.3218, + "grad_norm": 0.7637456042416844, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 32180 + }, + { + "epoch": 0.32181, + "grad_norm": 0.7799827149734623, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 32181 + }, + { + "epoch": 0.32182, + "grad_norm": 0.7951259606170238, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 32182 + }, + { + "epoch": 0.32183, + "grad_norm": 0.8594510738127398, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 32183 + }, + { + "epoch": 0.32184, + "grad_norm": 0.8312348466112356, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 32184 + }, + { + "epoch": 0.32185, + "grad_norm": 0.863765874160016, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32185 + }, + { + "epoch": 0.32186, + "grad_norm": 0.8493236103174471, + "learning_rate": 0.003, + "loss": 3.981, + "step": 32186 + }, + { + "epoch": 0.32187, + "grad_norm": 0.9100765112934529, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 32187 + }, + { + "epoch": 0.32188, + "grad_norm": 0.8295268190209965, + "learning_rate": 0.003, + "loss": 4.016, + "step": 32188 + }, + { + "epoch": 0.32189, + "grad_norm": 0.8243792050417827, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 32189 + }, + { + "epoch": 0.3219, + "grad_norm": 0.7791853761109334, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 32190 + }, + { + "epoch": 0.32191, + "grad_norm": 0.6781932542996704, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 32191 + }, + { + "epoch": 0.32192, + "grad_norm": 0.6509509645854071, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 32192 + }, + { + "epoch": 0.32193, + "grad_norm": 0.575626711442176, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 32193 + }, + { + "epoch": 0.32194, + "grad_norm": 0.581216786583278, + "learning_rate": 0.003, + "loss": 3.975, + "step": 32194 + }, + { + "epoch": 0.32195, + "grad_norm": 0.5497202938393349, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 32195 + }, + { + "epoch": 0.32196, + "grad_norm": 0.48520758035346284, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 32196 + }, + { + "epoch": 0.32197, + "grad_norm": 0.5275511914128894, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 32197 + }, + { + "epoch": 0.32198, + "grad_norm": 0.49597963073426593, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 32198 + }, + { + "epoch": 0.32199, + "grad_norm": 0.6241004794020556, + "learning_rate": 0.003, + "loss": 3.992, + "step": 32199 + }, + { + "epoch": 0.322, + "grad_norm": 0.7491787342743061, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 32200 + }, + { + "epoch": 0.32201, + "grad_norm": 0.7689128447089781, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 32201 + }, + { + "epoch": 0.32202, + "grad_norm": 0.8550747229943633, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32202 + }, + { + "epoch": 0.32203, + "grad_norm": 0.96760722939227, + "learning_rate": 0.003, + "loss": 3.96, + "step": 32203 + }, + { + "epoch": 0.32204, + "grad_norm": 1.2770805816849433, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 32204 + }, + { + "epoch": 0.32205, + "grad_norm": 0.6930796924417849, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 32205 + }, + { + "epoch": 0.32206, + "grad_norm": 0.7144387678644208, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 32206 + }, + { + "epoch": 0.32207, + "grad_norm": 0.7074703534649351, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 32207 + }, + { + "epoch": 0.32208, + "grad_norm": 0.7038933420468924, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 32208 + }, + { + "epoch": 0.32209, + "grad_norm": 0.8023913324883862, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 32209 + }, + { + "epoch": 0.3221, + "grad_norm": 0.9153420133328457, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 32210 + }, + { + "epoch": 0.32211, + "grad_norm": 1.0133385919426525, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 32211 + }, + { + "epoch": 0.32212, + "grad_norm": 0.9921427461168435, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 32212 + }, + { + "epoch": 0.32213, + "grad_norm": 1.1363744936756854, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32213 + }, + { + "epoch": 0.32214, + "grad_norm": 0.8308024977353505, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 32214 + }, + { + "epoch": 0.32215, + "grad_norm": 0.7440304955216566, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 32215 + }, + { + "epoch": 0.32216, + "grad_norm": 0.7399906450468634, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 32216 + }, + { + "epoch": 0.32217, + "grad_norm": 0.7535709577191061, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 32217 + }, + { + "epoch": 0.32218, + "grad_norm": 0.7256004874900769, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 32218 + }, + { + "epoch": 0.32219, + "grad_norm": 0.6483195309291235, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 32219 + }, + { + "epoch": 0.3222, + "grad_norm": 0.6334468153104299, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 32220 + }, + { + "epoch": 0.32221, + "grad_norm": 0.6944086824157021, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 32221 + }, + { + "epoch": 0.32222, + "grad_norm": 0.6764486068264335, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 32222 + }, + { + "epoch": 0.32223, + "grad_norm": 0.6012498546237487, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 32223 + }, + { + "epoch": 0.32224, + "grad_norm": 0.7258786178607293, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 32224 + }, + { + "epoch": 0.32225, + "grad_norm": 0.8056690293126079, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 32225 + }, + { + "epoch": 0.32226, + "grad_norm": 0.7291285292331786, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 32226 + }, + { + "epoch": 0.32227, + "grad_norm": 0.7200754290524488, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 32227 + }, + { + "epoch": 0.32228, + "grad_norm": 0.793178367776835, + "learning_rate": 0.003, + "loss": 3.988, + "step": 32228 + }, + { + "epoch": 0.32229, + "grad_norm": 0.7319580567102799, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 32229 + }, + { + "epoch": 0.3223, + "grad_norm": 0.7560453683381249, + "learning_rate": 0.003, + "loss": 3.977, + "step": 32230 + }, + { + "epoch": 0.32231, + "grad_norm": 0.7099472682060974, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 32231 + }, + { + "epoch": 0.32232, + "grad_norm": 0.7384968292641299, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 32232 + }, + { + "epoch": 0.32233, + "grad_norm": 0.7190063395179155, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 32233 + }, + { + "epoch": 0.32234, + "grad_norm": 0.7885346053929547, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 32234 + }, + { + "epoch": 0.32235, + "grad_norm": 0.7116238772511645, + "learning_rate": 0.003, + "loss": 3.9425, + "step": 32235 + }, + { + "epoch": 0.32236, + "grad_norm": 0.6015147502996774, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 32236 + }, + { + "epoch": 0.32237, + "grad_norm": 0.5670178271237986, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 32237 + }, + { + "epoch": 0.32238, + "grad_norm": 0.5843968765166297, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 32238 + }, + { + "epoch": 0.32239, + "grad_norm": 0.5543354260691333, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 32239 + }, + { + "epoch": 0.3224, + "grad_norm": 0.5386946204741965, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 32240 + }, + { + "epoch": 0.32241, + "grad_norm": 0.5836258236901998, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 32241 + }, + { + "epoch": 0.32242, + "grad_norm": 0.6406763329308478, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 32242 + }, + { + "epoch": 0.32243, + "grad_norm": 0.7663045571883297, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 32243 + }, + { + "epoch": 0.32244, + "grad_norm": 1.117772720935684, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 32244 + }, + { + "epoch": 0.32245, + "grad_norm": 1.1130940453565807, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 32245 + }, + { + "epoch": 0.32246, + "grad_norm": 0.7896564623280325, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 32246 + }, + { + "epoch": 0.32247, + "grad_norm": 0.7556958663299378, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 32247 + }, + { + "epoch": 0.32248, + "grad_norm": 0.7377657900334164, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 32248 + }, + { + "epoch": 0.32249, + "grad_norm": 0.7681178060711691, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 32249 + }, + { + "epoch": 0.3225, + "grad_norm": 0.8339728306607217, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 32250 + }, + { + "epoch": 0.32251, + "grad_norm": 0.949529702845581, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 32251 + }, + { + "epoch": 0.32252, + "grad_norm": 0.9692328989084109, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 32252 + }, + { + "epoch": 0.32253, + "grad_norm": 0.9758487443339704, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 32253 + }, + { + "epoch": 0.32254, + "grad_norm": 1.0145264755662695, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 32254 + }, + { + "epoch": 0.32255, + "grad_norm": 0.7211703063373702, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32255 + }, + { + "epoch": 0.32256, + "grad_norm": 0.7540057257710413, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 32256 + }, + { + "epoch": 0.32257, + "grad_norm": 0.8924245839536613, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 32257 + }, + { + "epoch": 0.32258, + "grad_norm": 1.0627454081759913, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 32258 + }, + { + "epoch": 0.32259, + "grad_norm": 0.7517022980217745, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 32259 + }, + { + "epoch": 0.3226, + "grad_norm": 0.7040706063496647, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 32260 + }, + { + "epoch": 0.32261, + "grad_norm": 0.6777416947327571, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 32261 + }, + { + "epoch": 0.32262, + "grad_norm": 0.7547243953394298, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 32262 + }, + { + "epoch": 0.32263, + "grad_norm": 0.7656669400235931, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 32263 + }, + { + "epoch": 0.32264, + "grad_norm": 0.680134149274127, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 32264 + }, + { + "epoch": 0.32265, + "grad_norm": 0.7891393760339106, + "learning_rate": 0.003, + "loss": 3.986, + "step": 32265 + }, + { + "epoch": 0.32266, + "grad_norm": 0.8023271278395029, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 32266 + }, + { + "epoch": 0.32267, + "grad_norm": 0.6689901192092358, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 32267 + }, + { + "epoch": 0.32268, + "grad_norm": 0.6994988629940893, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 32268 + }, + { + "epoch": 0.32269, + "grad_norm": 0.781045646281159, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32269 + }, + { + "epoch": 0.3227, + "grad_norm": 0.6863749427229356, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 32270 + }, + { + "epoch": 0.32271, + "grad_norm": 0.6994414514533898, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 32271 + }, + { + "epoch": 0.32272, + "grad_norm": 0.7839926635202319, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 32272 + }, + { + "epoch": 0.32273, + "grad_norm": 0.7986548115153075, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 32273 + }, + { + "epoch": 0.32274, + "grad_norm": 0.7813982364706346, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 32274 + }, + { + "epoch": 0.32275, + "grad_norm": 0.8481283139118575, + "learning_rate": 0.003, + "loss": 3.996, + "step": 32275 + }, + { + "epoch": 0.32276, + "grad_norm": 1.02057241283112, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 32276 + }, + { + "epoch": 0.32277, + "grad_norm": 1.0734706390733324, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 32277 + }, + { + "epoch": 0.32278, + "grad_norm": 0.6892337864596693, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 32278 + }, + { + "epoch": 0.32279, + "grad_norm": 0.6084456402193037, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 32279 + }, + { + "epoch": 0.3228, + "grad_norm": 0.6789856330240672, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 32280 + }, + { + "epoch": 0.32281, + "grad_norm": 0.846176776652521, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 32281 + }, + { + "epoch": 0.32282, + "grad_norm": 0.9088549832778551, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 32282 + }, + { + "epoch": 0.32283, + "grad_norm": 0.8838154916099894, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 32283 + }, + { + "epoch": 0.32284, + "grad_norm": 0.8596401145779461, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 32284 + }, + { + "epoch": 0.32285, + "grad_norm": 0.7993215953931886, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 32285 + }, + { + "epoch": 0.32286, + "grad_norm": 0.8954729699300711, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 32286 + }, + { + "epoch": 0.32287, + "grad_norm": 0.9189220655130544, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 32287 + }, + { + "epoch": 0.32288, + "grad_norm": 0.9556461691275862, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 32288 + }, + { + "epoch": 0.32289, + "grad_norm": 0.8080757960761596, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 32289 + }, + { + "epoch": 0.3229, + "grad_norm": 0.8148294709007803, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 32290 + }, + { + "epoch": 0.32291, + "grad_norm": 0.8316042756443994, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 32291 + }, + { + "epoch": 0.32292, + "grad_norm": 0.8086201296853974, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 32292 + }, + { + "epoch": 0.32293, + "grad_norm": 0.7806510104126277, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 32293 + }, + { + "epoch": 0.32294, + "grad_norm": 0.6941826640098954, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 32294 + }, + { + "epoch": 0.32295, + "grad_norm": 0.6848839669689792, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 32295 + }, + { + "epoch": 0.32296, + "grad_norm": 0.6845008717191665, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 32296 + }, + { + "epoch": 0.32297, + "grad_norm": 0.7565957344298504, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 32297 + }, + { + "epoch": 0.32298, + "grad_norm": 0.6965968343153228, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 32298 + }, + { + "epoch": 0.32299, + "grad_norm": 0.73576081081793, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 32299 + }, + { + "epoch": 0.323, + "grad_norm": 0.7764349703903396, + "learning_rate": 0.003, + "loss": 3.988, + "step": 32300 + }, + { + "epoch": 0.32301, + "grad_norm": 0.8519998835615973, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 32301 + }, + { + "epoch": 0.32302, + "grad_norm": 0.8901171650561085, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 32302 + }, + { + "epoch": 0.32303, + "grad_norm": 0.9391071150759094, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 32303 + }, + { + "epoch": 0.32304, + "grad_norm": 0.827468257989338, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 32304 + }, + { + "epoch": 0.32305, + "grad_norm": 0.7953077347596115, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 32305 + }, + { + "epoch": 0.32306, + "grad_norm": 0.6806149616188067, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 32306 + }, + { + "epoch": 0.32307, + "grad_norm": 0.6168767460691202, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 32307 + }, + { + "epoch": 0.32308, + "grad_norm": 0.5893175047110519, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 32308 + }, + { + "epoch": 0.32309, + "grad_norm": 0.5890873385536202, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 32309 + }, + { + "epoch": 0.3231, + "grad_norm": 0.613015305418399, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 32310 + }, + { + "epoch": 0.32311, + "grad_norm": 0.7294369118858977, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 32311 + }, + { + "epoch": 0.32312, + "grad_norm": 0.8133159719788028, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 32312 + }, + { + "epoch": 0.32313, + "grad_norm": 0.9136389271538476, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 32313 + }, + { + "epoch": 0.32314, + "grad_norm": 0.8822934437060629, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 32314 + }, + { + "epoch": 0.32315, + "grad_norm": 0.6949996235436755, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 32315 + }, + { + "epoch": 0.32316, + "grad_norm": 0.6733034549932763, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 32316 + }, + { + "epoch": 0.32317, + "grad_norm": 0.7595130858284708, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 32317 + }, + { + "epoch": 0.32318, + "grad_norm": 0.7249547217623226, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 32318 + }, + { + "epoch": 0.32319, + "grad_norm": 0.7671373554125168, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 32319 + }, + { + "epoch": 0.3232, + "grad_norm": 0.7852772022423561, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 32320 + }, + { + "epoch": 0.32321, + "grad_norm": 0.8280199725762108, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 32321 + }, + { + "epoch": 0.32322, + "grad_norm": 0.7987783302022028, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 32322 + }, + { + "epoch": 0.32323, + "grad_norm": 0.7775368383562887, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32323 + }, + { + "epoch": 0.32324, + "grad_norm": 0.703977109837914, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 32324 + }, + { + "epoch": 0.32325, + "grad_norm": 0.6458656227713594, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 32325 + }, + { + "epoch": 0.32326, + "grad_norm": 0.5905915260310333, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 32326 + }, + { + "epoch": 0.32327, + "grad_norm": 0.6505836527598603, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 32327 + }, + { + "epoch": 0.32328, + "grad_norm": 0.7327771788608834, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 32328 + }, + { + "epoch": 0.32329, + "grad_norm": 0.7665017278810501, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 32329 + }, + { + "epoch": 0.3233, + "grad_norm": 0.8981168307134929, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 32330 + }, + { + "epoch": 0.32331, + "grad_norm": 0.9557815307487552, + "learning_rate": 0.003, + "loss": 3.952, + "step": 32331 + }, + { + "epoch": 0.32332, + "grad_norm": 0.8441964299135168, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 32332 + }, + { + "epoch": 0.32333, + "grad_norm": 0.7312718246273124, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 32333 + }, + { + "epoch": 0.32334, + "grad_norm": 0.9017187046916715, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 32334 + }, + { + "epoch": 0.32335, + "grad_norm": 1.0828626769007104, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 32335 + }, + { + "epoch": 0.32336, + "grad_norm": 0.8566334671086289, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 32336 + }, + { + "epoch": 0.32337, + "grad_norm": 0.7222248138204108, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 32337 + }, + { + "epoch": 0.32338, + "grad_norm": 0.753595893982446, + "learning_rate": 0.003, + "loss": 4.002, + "step": 32338 + }, + { + "epoch": 0.32339, + "grad_norm": 0.7930491931156585, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 32339 + }, + { + "epoch": 0.3234, + "grad_norm": 0.9457262127932005, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 32340 + }, + { + "epoch": 0.32341, + "grad_norm": 0.9303602320202451, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 32341 + }, + { + "epoch": 0.32342, + "grad_norm": 0.8880211945992039, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 32342 + }, + { + "epoch": 0.32343, + "grad_norm": 0.8211579933567111, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 32343 + }, + { + "epoch": 0.32344, + "grad_norm": 0.816673866095842, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 32344 + }, + { + "epoch": 0.32345, + "grad_norm": 0.9870818834196295, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 32345 + }, + { + "epoch": 0.32346, + "grad_norm": 1.1547657593293932, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 32346 + }, + { + "epoch": 0.32347, + "grad_norm": 0.8422312574273185, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 32347 + }, + { + "epoch": 0.32348, + "grad_norm": 0.7477836748946708, + "learning_rate": 0.003, + "loss": 3.997, + "step": 32348 + }, + { + "epoch": 0.32349, + "grad_norm": 0.7031042574337255, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 32349 + }, + { + "epoch": 0.3235, + "grad_norm": 0.7003117929636642, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 32350 + }, + { + "epoch": 0.32351, + "grad_norm": 0.7516852800918602, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 32351 + }, + { + "epoch": 0.32352, + "grad_norm": 0.7537391798830888, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 32352 + }, + { + "epoch": 0.32353, + "grad_norm": 0.7161969280185638, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 32353 + }, + { + "epoch": 0.32354, + "grad_norm": 0.7784297471729035, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 32354 + }, + { + "epoch": 0.32355, + "grad_norm": 0.8069836314251247, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 32355 + }, + { + "epoch": 0.32356, + "grad_norm": 0.8505102455352052, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 32356 + }, + { + "epoch": 0.32357, + "grad_norm": 0.829511253830312, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 32357 + }, + { + "epoch": 0.32358, + "grad_norm": 0.9029712618756314, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 32358 + }, + { + "epoch": 0.32359, + "grad_norm": 1.0393015348767738, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 32359 + }, + { + "epoch": 0.3236, + "grad_norm": 0.8422518748209692, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 32360 + }, + { + "epoch": 0.32361, + "grad_norm": 0.7387391383972032, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 32361 + }, + { + "epoch": 0.32362, + "grad_norm": 0.642351542289895, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 32362 + }, + { + "epoch": 0.32363, + "grad_norm": 0.6936128076744731, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 32363 + }, + { + "epoch": 0.32364, + "grad_norm": 0.726271991815273, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 32364 + }, + { + "epoch": 0.32365, + "grad_norm": 0.751328813090223, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 32365 + }, + { + "epoch": 0.32366, + "grad_norm": 0.8377268727917412, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 32366 + }, + { + "epoch": 0.32367, + "grad_norm": 0.90159122024197, + "learning_rate": 0.003, + "loss": 4.004, + "step": 32367 + }, + { + "epoch": 0.32368, + "grad_norm": 1.0060646848272294, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 32368 + }, + { + "epoch": 0.32369, + "grad_norm": 1.0379737015652666, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 32369 + }, + { + "epoch": 0.3237, + "grad_norm": 0.907160828811077, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 32370 + }, + { + "epoch": 0.32371, + "grad_norm": 0.9287031650122158, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 32371 + }, + { + "epoch": 0.32372, + "grad_norm": 0.7782341014881347, + "learning_rate": 0.003, + "loss": 4.023, + "step": 32372 + }, + { + "epoch": 0.32373, + "grad_norm": 0.6953968908523781, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 32373 + }, + { + "epoch": 0.32374, + "grad_norm": 0.7328823624952594, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 32374 + }, + { + "epoch": 0.32375, + "grad_norm": 0.8832003827958945, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 32375 + }, + { + "epoch": 0.32376, + "grad_norm": 1.0305260183038405, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 32376 + }, + { + "epoch": 0.32377, + "grad_norm": 0.8985159836045642, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 32377 + }, + { + "epoch": 0.32378, + "grad_norm": 0.8008824186078726, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 32378 + }, + { + "epoch": 0.32379, + "grad_norm": 0.7910737987280383, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 32379 + }, + { + "epoch": 0.3238, + "grad_norm": 0.7542223122038602, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 32380 + }, + { + "epoch": 0.32381, + "grad_norm": 0.658554852705772, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 32381 + }, + { + "epoch": 0.32382, + "grad_norm": 0.5706820733481007, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 32382 + }, + { + "epoch": 0.32383, + "grad_norm": 0.5267348778627401, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 32383 + }, + { + "epoch": 0.32384, + "grad_norm": 0.5345279155947545, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 32384 + }, + { + "epoch": 0.32385, + "grad_norm": 0.5166500355875295, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32385 + }, + { + "epoch": 0.32386, + "grad_norm": 0.5307290351967339, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 32386 + }, + { + "epoch": 0.32387, + "grad_norm": 0.6211401756208206, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 32387 + }, + { + "epoch": 0.32388, + "grad_norm": 0.663253472446645, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 32388 + }, + { + "epoch": 0.32389, + "grad_norm": 0.7057823342971332, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 32389 + }, + { + "epoch": 0.3239, + "grad_norm": 0.7847524774772955, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 32390 + }, + { + "epoch": 0.32391, + "grad_norm": 0.8317640843281036, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 32391 + }, + { + "epoch": 0.32392, + "grad_norm": 0.798454539230976, + "learning_rate": 0.003, + "loss": 3.995, + "step": 32392 + }, + { + "epoch": 0.32393, + "grad_norm": 0.6421388513748152, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 32393 + }, + { + "epoch": 0.32394, + "grad_norm": 0.652072510518012, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 32394 + }, + { + "epoch": 0.32395, + "grad_norm": 0.6743499192287827, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 32395 + }, + { + "epoch": 0.32396, + "grad_norm": 0.630625496819289, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 32396 + }, + { + "epoch": 0.32397, + "grad_norm": 0.686201266474919, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 32397 + }, + { + "epoch": 0.32398, + "grad_norm": 0.7565933804030127, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 32398 + }, + { + "epoch": 0.32399, + "grad_norm": 0.7767566208338935, + "learning_rate": 0.003, + "loss": 3.979, + "step": 32399 + }, + { + "epoch": 0.324, + "grad_norm": 0.7777683030670142, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 32400 + }, + { + "epoch": 0.32401, + "grad_norm": 0.8383657245034265, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 32401 + }, + { + "epoch": 0.32402, + "grad_norm": 0.9178987308430016, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 32402 + }, + { + "epoch": 0.32403, + "grad_norm": 0.9043119164900428, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 32403 + }, + { + "epoch": 0.32404, + "grad_norm": 0.8123669265890694, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 32404 + }, + { + "epoch": 0.32405, + "grad_norm": 0.8024145558123593, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 32405 + }, + { + "epoch": 0.32406, + "grad_norm": 0.8644568620308771, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 32406 + }, + { + "epoch": 0.32407, + "grad_norm": 0.8288211517891441, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 32407 + }, + { + "epoch": 0.32408, + "grad_norm": 0.8510114491695564, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 32408 + }, + { + "epoch": 0.32409, + "grad_norm": 0.9477440703501009, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 32409 + }, + { + "epoch": 0.3241, + "grad_norm": 1.0101365821122983, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 32410 + }, + { + "epoch": 0.32411, + "grad_norm": 1.0588791524026242, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 32411 + }, + { + "epoch": 0.32412, + "grad_norm": 0.925161299181364, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 32412 + }, + { + "epoch": 0.32413, + "grad_norm": 0.8345601899203129, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 32413 + }, + { + "epoch": 0.32414, + "grad_norm": 0.7485807631176958, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 32414 + }, + { + "epoch": 0.32415, + "grad_norm": 0.7786855943853043, + "learning_rate": 0.003, + "loss": 3.984, + "step": 32415 + }, + { + "epoch": 0.32416, + "grad_norm": 0.8241591804312011, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 32416 + }, + { + "epoch": 0.32417, + "grad_norm": 0.8607164491584618, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 32417 + }, + { + "epoch": 0.32418, + "grad_norm": 0.8554088909006019, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 32418 + }, + { + "epoch": 0.32419, + "grad_norm": 0.8071557318949425, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 32419 + }, + { + "epoch": 0.3242, + "grad_norm": 0.9105400735233043, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 32420 + }, + { + "epoch": 0.32421, + "grad_norm": 0.9313077515885786, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 32421 + }, + { + "epoch": 0.32422, + "grad_norm": 0.8251779252621109, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 32422 + }, + { + "epoch": 0.32423, + "grad_norm": 0.6741120013536069, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 32423 + }, + { + "epoch": 0.32424, + "grad_norm": 0.7121489402041004, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 32424 + }, + { + "epoch": 0.32425, + "grad_norm": 0.6852797645537898, + "learning_rate": 0.003, + "loss": 4.011, + "step": 32425 + }, + { + "epoch": 0.32426, + "grad_norm": 0.7612382814509199, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 32426 + }, + { + "epoch": 0.32427, + "grad_norm": 0.7196602269914593, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 32427 + }, + { + "epoch": 0.32428, + "grad_norm": 0.7311738372867046, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 32428 + }, + { + "epoch": 0.32429, + "grad_norm": 0.7383375690057334, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 32429 + }, + { + "epoch": 0.3243, + "grad_norm": 0.7902308205529306, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 32430 + }, + { + "epoch": 0.32431, + "grad_norm": 0.7434158224494138, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 32431 + }, + { + "epoch": 0.32432, + "grad_norm": 0.7045931208036031, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 32432 + }, + { + "epoch": 0.32433, + "grad_norm": 0.8271769062079406, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 32433 + }, + { + "epoch": 0.32434, + "grad_norm": 0.8378612227839202, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 32434 + }, + { + "epoch": 0.32435, + "grad_norm": 0.8039293973276502, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 32435 + }, + { + "epoch": 0.32436, + "grad_norm": 0.7585211993308283, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 32436 + }, + { + "epoch": 0.32437, + "grad_norm": 0.7687724385181548, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 32437 + }, + { + "epoch": 0.32438, + "grad_norm": 0.8015090850386876, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 32438 + }, + { + "epoch": 0.32439, + "grad_norm": 0.7881416264277149, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 32439 + }, + { + "epoch": 0.3244, + "grad_norm": 0.7077160663064533, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 32440 + }, + { + "epoch": 0.32441, + "grad_norm": 0.661597879961696, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 32441 + }, + { + "epoch": 0.32442, + "grad_norm": 0.7200443219985317, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 32442 + }, + { + "epoch": 0.32443, + "grad_norm": 0.753753278536807, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 32443 + }, + { + "epoch": 0.32444, + "grad_norm": 0.8171799832010028, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 32444 + }, + { + "epoch": 0.32445, + "grad_norm": 0.8620965578767518, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 32445 + }, + { + "epoch": 0.32446, + "grad_norm": 0.6989261247193281, + "learning_rate": 0.003, + "loss": 3.989, + "step": 32446 + }, + { + "epoch": 0.32447, + "grad_norm": 0.6724783193759587, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 32447 + }, + { + "epoch": 0.32448, + "grad_norm": 0.7124468640808808, + "learning_rate": 0.003, + "loss": 3.997, + "step": 32448 + }, + { + "epoch": 0.32449, + "grad_norm": 0.6604006276149781, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32449 + }, + { + "epoch": 0.3245, + "grad_norm": 0.6582025163519288, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 32450 + }, + { + "epoch": 0.32451, + "grad_norm": 0.7597901951303865, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 32451 + }, + { + "epoch": 0.32452, + "grad_norm": 0.8494818826462961, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 32452 + }, + { + "epoch": 0.32453, + "grad_norm": 0.9850425096492422, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 32453 + }, + { + "epoch": 0.32454, + "grad_norm": 1.081777946141542, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 32454 + }, + { + "epoch": 0.32455, + "grad_norm": 0.7789383530955002, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 32455 + }, + { + "epoch": 0.32456, + "grad_norm": 0.7138706133033073, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 32456 + }, + { + "epoch": 0.32457, + "grad_norm": 0.8628634636551729, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 32457 + }, + { + "epoch": 0.32458, + "grad_norm": 0.9079796519522905, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 32458 + }, + { + "epoch": 0.32459, + "grad_norm": 0.8341698945891635, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 32459 + }, + { + "epoch": 0.3246, + "grad_norm": 0.7819734640578174, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32460 + }, + { + "epoch": 0.32461, + "grad_norm": 0.7817515748233049, + "learning_rate": 0.003, + "loss": 4.01, + "step": 32461 + }, + { + "epoch": 0.32462, + "grad_norm": 0.7932171173821214, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 32462 + }, + { + "epoch": 0.32463, + "grad_norm": 0.7930390219598644, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 32463 + }, + { + "epoch": 0.32464, + "grad_norm": 0.8992266915852529, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 32464 + }, + { + "epoch": 0.32465, + "grad_norm": 0.8127581952753924, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 32465 + }, + { + "epoch": 0.32466, + "grad_norm": 0.869310950684725, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 32466 + }, + { + "epoch": 0.32467, + "grad_norm": 0.8856819963536633, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 32467 + }, + { + "epoch": 0.32468, + "grad_norm": 0.7870726527280452, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 32468 + }, + { + "epoch": 0.32469, + "grad_norm": 0.6881273780727152, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 32469 + }, + { + "epoch": 0.3247, + "grad_norm": 0.5595154131948215, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 32470 + }, + { + "epoch": 0.32471, + "grad_norm": 0.5381383970891414, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 32471 + }, + { + "epoch": 0.32472, + "grad_norm": 0.5797318990986297, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 32472 + }, + { + "epoch": 0.32473, + "grad_norm": 0.604077074477301, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 32473 + }, + { + "epoch": 0.32474, + "grad_norm": 0.5603307977406687, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 32474 + }, + { + "epoch": 0.32475, + "grad_norm": 0.6473386614071096, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 32475 + }, + { + "epoch": 0.32476, + "grad_norm": 0.7264599061029899, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 32476 + }, + { + "epoch": 0.32477, + "grad_norm": 0.766026441074514, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 32477 + }, + { + "epoch": 0.32478, + "grad_norm": 0.7900047840504326, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 32478 + }, + { + "epoch": 0.32479, + "grad_norm": 0.686148988683982, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 32479 + }, + { + "epoch": 0.3248, + "grad_norm": 0.5745973271441049, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 32480 + }, + { + "epoch": 0.32481, + "grad_norm": 0.5947539356337728, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 32481 + }, + { + "epoch": 0.32482, + "grad_norm": 0.6312814418010703, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 32482 + }, + { + "epoch": 0.32483, + "grad_norm": 0.6303468046064825, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 32483 + }, + { + "epoch": 0.32484, + "grad_norm": 0.6880082442858364, + "learning_rate": 0.003, + "loss": 4.006, + "step": 32484 + }, + { + "epoch": 0.32485, + "grad_norm": 0.7473273926294901, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 32485 + }, + { + "epoch": 0.32486, + "grad_norm": 0.7702866883860513, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 32486 + }, + { + "epoch": 0.32487, + "grad_norm": 0.8590030472473376, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 32487 + }, + { + "epoch": 0.32488, + "grad_norm": 0.9544239599661145, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 32488 + }, + { + "epoch": 0.32489, + "grad_norm": 0.8258195668600231, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 32489 + }, + { + "epoch": 0.3249, + "grad_norm": 0.9802675576299881, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 32490 + }, + { + "epoch": 0.32491, + "grad_norm": 1.0382916614166076, + "learning_rate": 0.003, + "loss": 4.029, + "step": 32491 + }, + { + "epoch": 0.32492, + "grad_norm": 0.9477840714192368, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 32492 + }, + { + "epoch": 0.32493, + "grad_norm": 0.7349434985891798, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 32493 + }, + { + "epoch": 0.32494, + "grad_norm": 0.6767965466295527, + "learning_rate": 0.003, + "loss": 3.998, + "step": 32494 + }, + { + "epoch": 0.32495, + "grad_norm": 0.5927987693329156, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 32495 + }, + { + "epoch": 0.32496, + "grad_norm": 0.6886342019975473, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 32496 + }, + { + "epoch": 0.32497, + "grad_norm": 0.7250249633301921, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 32497 + }, + { + "epoch": 0.32498, + "grad_norm": 0.7842032503994344, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 32498 + }, + { + "epoch": 0.32499, + "grad_norm": 0.9232811263190857, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32499 + }, + { + "epoch": 0.325, + "grad_norm": 0.9704345894472105, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 32500 + }, + { + "epoch": 0.32501, + "grad_norm": 0.8151523602922417, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 32501 + }, + { + "epoch": 0.32502, + "grad_norm": 0.7092895479728561, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 32502 + }, + { + "epoch": 0.32503, + "grad_norm": 0.7815819317826079, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 32503 + }, + { + "epoch": 0.32504, + "grad_norm": 0.7700396127039457, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 32504 + }, + { + "epoch": 0.32505, + "grad_norm": 0.8276704568150907, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 32505 + }, + { + "epoch": 0.32506, + "grad_norm": 0.7693300327806609, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 32506 + }, + { + "epoch": 0.32507, + "grad_norm": 0.8619119710707909, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 32507 + }, + { + "epoch": 0.32508, + "grad_norm": 0.8465772139554147, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 32508 + }, + { + "epoch": 0.32509, + "grad_norm": 0.8702317731332714, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 32509 + }, + { + "epoch": 0.3251, + "grad_norm": 0.9964427885498648, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 32510 + }, + { + "epoch": 0.32511, + "grad_norm": 1.1368789418323015, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 32511 + }, + { + "epoch": 0.32512, + "grad_norm": 0.910909418123374, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32512 + }, + { + "epoch": 0.32513, + "grad_norm": 1.0063967216345624, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 32513 + }, + { + "epoch": 0.32514, + "grad_norm": 0.9202585709392497, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 32514 + }, + { + "epoch": 0.32515, + "grad_norm": 0.8421967487340553, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 32515 + }, + { + "epoch": 0.32516, + "grad_norm": 0.7387353402542179, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 32516 + }, + { + "epoch": 0.32517, + "grad_norm": 0.7272856489695355, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 32517 + }, + { + "epoch": 0.32518, + "grad_norm": 0.7545216613778571, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 32518 + }, + { + "epoch": 0.32519, + "grad_norm": 0.7863432527147065, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 32519 + }, + { + "epoch": 0.3252, + "grad_norm": 0.7732397027615565, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 32520 + }, + { + "epoch": 0.32521, + "grad_norm": 0.7703672650650554, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 32521 + }, + { + "epoch": 0.32522, + "grad_norm": 0.8266919858519749, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 32522 + }, + { + "epoch": 0.32523, + "grad_norm": 0.9561964186588527, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 32523 + }, + { + "epoch": 0.32524, + "grad_norm": 0.8705374878065728, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 32524 + }, + { + "epoch": 0.32525, + "grad_norm": 0.775114032483768, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 32525 + }, + { + "epoch": 0.32526, + "grad_norm": 0.7674856189769003, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 32526 + }, + { + "epoch": 0.32527, + "grad_norm": 0.8221190990345003, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 32527 + }, + { + "epoch": 0.32528, + "grad_norm": 0.897965189752583, + "learning_rate": 0.003, + "loss": 3.998, + "step": 32528 + }, + { + "epoch": 0.32529, + "grad_norm": 0.9057855664460374, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 32529 + }, + { + "epoch": 0.3253, + "grad_norm": 0.9169917534988835, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 32530 + }, + { + "epoch": 0.32531, + "grad_norm": 0.7530929522590044, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 32531 + }, + { + "epoch": 0.32532, + "grad_norm": 0.7420035514280899, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 32532 + }, + { + "epoch": 0.32533, + "grad_norm": 0.7179002485612221, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32533 + }, + { + "epoch": 0.32534, + "grad_norm": 0.7262335280983769, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 32534 + }, + { + "epoch": 0.32535, + "grad_norm": 0.6769032546773378, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 32535 + }, + { + "epoch": 0.32536, + "grad_norm": 0.7201775345087419, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 32536 + }, + { + "epoch": 0.32537, + "grad_norm": 0.7941833813661797, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 32537 + }, + { + "epoch": 0.32538, + "grad_norm": 0.8374002222809181, + "learning_rate": 0.003, + "loss": 3.968, + "step": 32538 + }, + { + "epoch": 0.32539, + "grad_norm": 0.8377343653326518, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 32539 + }, + { + "epoch": 0.3254, + "grad_norm": 0.7189859583295255, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 32540 + }, + { + "epoch": 0.32541, + "grad_norm": 0.574077542892145, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 32541 + }, + { + "epoch": 0.32542, + "grad_norm": 0.6138195968046597, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 32542 + }, + { + "epoch": 0.32543, + "grad_norm": 0.6361388956498558, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 32543 + }, + { + "epoch": 0.32544, + "grad_norm": 0.6654996500847127, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 32544 + }, + { + "epoch": 0.32545, + "grad_norm": 0.6883135735202398, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 32545 + }, + { + "epoch": 0.32546, + "grad_norm": 0.7208526920241536, + "learning_rate": 0.003, + "loss": 4.005, + "step": 32546 + }, + { + "epoch": 0.32547, + "grad_norm": 0.7846895437767211, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 32547 + }, + { + "epoch": 0.32548, + "grad_norm": 0.8472992853500051, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 32548 + }, + { + "epoch": 0.32549, + "grad_norm": 0.7452711685303752, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 32549 + }, + { + "epoch": 0.3255, + "grad_norm": 0.7610939667062946, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 32550 + }, + { + "epoch": 0.32551, + "grad_norm": 0.7373693002633842, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 32551 + }, + { + "epoch": 0.32552, + "grad_norm": 0.7716358355177172, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 32552 + }, + { + "epoch": 0.32553, + "grad_norm": 0.7720664450965974, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 32553 + }, + { + "epoch": 0.32554, + "grad_norm": 0.7835966018534156, + "learning_rate": 0.003, + "loss": 4.001, + "step": 32554 + }, + { + "epoch": 0.32555, + "grad_norm": 0.7498509431998892, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 32555 + }, + { + "epoch": 0.32556, + "grad_norm": 0.7588939749022566, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 32556 + }, + { + "epoch": 0.32557, + "grad_norm": 0.8230138832616144, + "learning_rate": 0.003, + "loss": 4.001, + "step": 32557 + }, + { + "epoch": 0.32558, + "grad_norm": 0.8675038635170293, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 32558 + }, + { + "epoch": 0.32559, + "grad_norm": 0.8126974539556449, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 32559 + }, + { + "epoch": 0.3256, + "grad_norm": 0.7900111715408002, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 32560 + }, + { + "epoch": 0.32561, + "grad_norm": 0.7676151572257998, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 32561 + }, + { + "epoch": 0.32562, + "grad_norm": 0.9420319102214667, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 32562 + }, + { + "epoch": 0.32563, + "grad_norm": 1.0618045115065768, + "learning_rate": 0.003, + "loss": 3.996, + "step": 32563 + }, + { + "epoch": 0.32564, + "grad_norm": 1.1081474822266493, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 32564 + }, + { + "epoch": 0.32565, + "grad_norm": 1.017474398280674, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 32565 + }, + { + "epoch": 0.32566, + "grad_norm": 0.8345664787721947, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 32566 + }, + { + "epoch": 0.32567, + "grad_norm": 0.635725498528937, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 32567 + }, + { + "epoch": 0.32568, + "grad_norm": 0.5978964146807338, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 32568 + }, + { + "epoch": 0.32569, + "grad_norm": 0.5646130969445117, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 32569 + }, + { + "epoch": 0.3257, + "grad_norm": 0.5488501200488453, + "learning_rate": 0.003, + "loss": 3.965, + "step": 32570 + }, + { + "epoch": 0.32571, + "grad_norm": 0.5312911292566908, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 32571 + }, + { + "epoch": 0.32572, + "grad_norm": 0.5879349378792414, + "learning_rate": 0.003, + "loss": 3.971, + "step": 32572 + }, + { + "epoch": 0.32573, + "grad_norm": 0.6248413109373242, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 32573 + }, + { + "epoch": 0.32574, + "grad_norm": 0.6009361932585751, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 32574 + }, + { + "epoch": 0.32575, + "grad_norm": 0.556228551178026, + "learning_rate": 0.003, + "loss": 3.975, + "step": 32575 + }, + { + "epoch": 0.32576, + "grad_norm": 0.568831753568843, + "learning_rate": 0.003, + "loss": 3.958, + "step": 32576 + }, + { + "epoch": 0.32577, + "grad_norm": 0.6016614125278934, + "learning_rate": 0.003, + "loss": 3.95, + "step": 32577 + }, + { + "epoch": 0.32578, + "grad_norm": 0.6544799188148172, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 32578 + }, + { + "epoch": 0.32579, + "grad_norm": 0.7516183833776086, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 32579 + }, + { + "epoch": 0.3258, + "grad_norm": 0.823296920210147, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 32580 + }, + { + "epoch": 0.32581, + "grad_norm": 0.7741433239876018, + "learning_rate": 0.003, + "loss": 4.006, + "step": 32581 + }, + { + "epoch": 0.32582, + "grad_norm": 0.6739221901556316, + "learning_rate": 0.003, + "loss": 3.98, + "step": 32582 + }, + { + "epoch": 0.32583, + "grad_norm": 0.7263331575911651, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 32583 + }, + { + "epoch": 0.32584, + "grad_norm": 0.6531545046898188, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 32584 + }, + { + "epoch": 0.32585, + "grad_norm": 0.6520467122732648, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 32585 + }, + { + "epoch": 0.32586, + "grad_norm": 0.7702168490147847, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 32586 + }, + { + "epoch": 0.32587, + "grad_norm": 0.939182169636328, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 32587 + }, + { + "epoch": 0.32588, + "grad_norm": 1.091200826627072, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 32588 + }, + { + "epoch": 0.32589, + "grad_norm": 0.9848214626305346, + "learning_rate": 0.003, + "loss": 3.999, + "step": 32589 + }, + { + "epoch": 0.3259, + "grad_norm": 0.8541590483431238, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 32590 + }, + { + "epoch": 0.32591, + "grad_norm": 0.837996471740042, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 32591 + }, + { + "epoch": 0.32592, + "grad_norm": 0.9637816551165902, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 32592 + }, + { + "epoch": 0.32593, + "grad_norm": 0.964370121944266, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 32593 + }, + { + "epoch": 0.32594, + "grad_norm": 0.9046992143085222, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 32594 + }, + { + "epoch": 0.32595, + "grad_norm": 0.9765974685310298, + "learning_rate": 0.003, + "loss": 4.011, + "step": 32595 + }, + { + "epoch": 0.32596, + "grad_norm": 1.034278397600311, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 32596 + }, + { + "epoch": 0.32597, + "grad_norm": 0.9662307057871596, + "learning_rate": 0.003, + "loss": 4.006, + "step": 32597 + }, + { + "epoch": 0.32598, + "grad_norm": 0.8200320212654384, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 32598 + }, + { + "epoch": 0.32599, + "grad_norm": 0.6311448451552012, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 32599 + }, + { + "epoch": 0.326, + "grad_norm": 0.7610059522607808, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 32600 + }, + { + "epoch": 0.32601, + "grad_norm": 0.7744114518586531, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 32601 + }, + { + "epoch": 0.32602, + "grad_norm": 0.691187551215425, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 32602 + }, + { + "epoch": 0.32603, + "grad_norm": 0.7392032283051292, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 32603 + }, + { + "epoch": 0.32604, + "grad_norm": 0.6954875546014998, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 32604 + }, + { + "epoch": 0.32605, + "grad_norm": 0.6063203733949752, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 32605 + }, + { + "epoch": 0.32606, + "grad_norm": 0.6548198902999814, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 32606 + }, + { + "epoch": 0.32607, + "grad_norm": 0.7782967618399697, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32607 + }, + { + "epoch": 0.32608, + "grad_norm": 0.8073143192933486, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 32608 + }, + { + "epoch": 0.32609, + "grad_norm": 0.8861783855160253, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 32609 + }, + { + "epoch": 0.3261, + "grad_norm": 0.9802171550794772, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 32610 + }, + { + "epoch": 0.32611, + "grad_norm": 0.7923754436558478, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 32611 + }, + { + "epoch": 0.32612, + "grad_norm": 0.8383983561368845, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 32612 + }, + { + "epoch": 0.32613, + "grad_norm": 0.8169685752589327, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 32613 + }, + { + "epoch": 0.32614, + "grad_norm": 0.8220289949443589, + "learning_rate": 0.003, + "loss": 3.995, + "step": 32614 + }, + { + "epoch": 0.32615, + "grad_norm": 0.6520990055932918, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 32615 + }, + { + "epoch": 0.32616, + "grad_norm": 0.5651081985139592, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 32616 + }, + { + "epoch": 0.32617, + "grad_norm": 0.5242948276158677, + "learning_rate": 0.003, + "loss": 4.001, + "step": 32617 + }, + { + "epoch": 0.32618, + "grad_norm": 0.5967560037867855, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 32618 + }, + { + "epoch": 0.32619, + "grad_norm": 0.6043299009374558, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 32619 + }, + { + "epoch": 0.3262, + "grad_norm": 0.6535160741466813, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 32620 + }, + { + "epoch": 0.32621, + "grad_norm": 0.6873004830711132, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 32621 + }, + { + "epoch": 0.32622, + "grad_norm": 0.7842929990961237, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 32622 + }, + { + "epoch": 0.32623, + "grad_norm": 0.9310522337328205, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 32623 + }, + { + "epoch": 0.32624, + "grad_norm": 0.9797735310016932, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 32624 + }, + { + "epoch": 0.32625, + "grad_norm": 1.0564022044442913, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 32625 + }, + { + "epoch": 0.32626, + "grad_norm": 0.9502512077216781, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 32626 + }, + { + "epoch": 0.32627, + "grad_norm": 0.9469407701617725, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 32627 + }, + { + "epoch": 0.32628, + "grad_norm": 0.8630015615562538, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 32628 + }, + { + "epoch": 0.32629, + "grad_norm": 0.7906838373841172, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 32629 + }, + { + "epoch": 0.3263, + "grad_norm": 0.7915542621250243, + "learning_rate": 0.003, + "loss": 3.993, + "step": 32630 + }, + { + "epoch": 0.32631, + "grad_norm": 0.8521982731434345, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 32631 + }, + { + "epoch": 0.32632, + "grad_norm": 0.9010610024792869, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 32632 + }, + { + "epoch": 0.32633, + "grad_norm": 0.895443707997085, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 32633 + }, + { + "epoch": 0.32634, + "grad_norm": 0.7491465218239356, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 32634 + }, + { + "epoch": 0.32635, + "grad_norm": 0.7538237978453719, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 32635 + }, + { + "epoch": 0.32636, + "grad_norm": 0.7592642191884995, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 32636 + }, + { + "epoch": 0.32637, + "grad_norm": 0.7599641095969052, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 32637 + }, + { + "epoch": 0.32638, + "grad_norm": 0.6015931286025786, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 32638 + }, + { + "epoch": 0.32639, + "grad_norm": 0.602194671694923, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 32639 + }, + { + "epoch": 0.3264, + "grad_norm": 0.6814889621346365, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 32640 + }, + { + "epoch": 0.32641, + "grad_norm": 0.7659281319115473, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 32641 + }, + { + "epoch": 0.32642, + "grad_norm": 0.7848535885780742, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 32642 + }, + { + "epoch": 0.32643, + "grad_norm": 0.6968196504826973, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 32643 + }, + { + "epoch": 0.32644, + "grad_norm": 0.9143198627418534, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 32644 + }, + { + "epoch": 0.32645, + "grad_norm": 0.8588449903708697, + "learning_rate": 0.003, + "loss": 4.001, + "step": 32645 + }, + { + "epoch": 0.32646, + "grad_norm": 0.739292764169173, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32646 + }, + { + "epoch": 0.32647, + "grad_norm": 0.7762452413522549, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 32647 + }, + { + "epoch": 0.32648, + "grad_norm": 0.7200075146015519, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 32648 + }, + { + "epoch": 0.32649, + "grad_norm": 0.7389409010583626, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 32649 + }, + { + "epoch": 0.3265, + "grad_norm": 0.812597634821576, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 32650 + }, + { + "epoch": 0.32651, + "grad_norm": 0.8229699775087339, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 32651 + }, + { + "epoch": 0.32652, + "grad_norm": 0.9415109763044698, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 32652 + }, + { + "epoch": 0.32653, + "grad_norm": 0.9532228382683025, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 32653 + }, + { + "epoch": 0.32654, + "grad_norm": 0.8009500035806729, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 32654 + }, + { + "epoch": 0.32655, + "grad_norm": 0.6742494930145473, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 32655 + }, + { + "epoch": 0.32656, + "grad_norm": 0.6265482678664163, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 32656 + }, + { + "epoch": 0.32657, + "grad_norm": 0.6510877591833497, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 32657 + }, + { + "epoch": 0.32658, + "grad_norm": 0.6382876481185773, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 32658 + }, + { + "epoch": 0.32659, + "grad_norm": 0.5837012784083025, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 32659 + }, + { + "epoch": 0.3266, + "grad_norm": 0.6257613876720286, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 32660 + }, + { + "epoch": 0.32661, + "grad_norm": 0.8085948378973382, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 32661 + }, + { + "epoch": 0.32662, + "grad_norm": 0.9760785486181971, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 32662 + }, + { + "epoch": 0.32663, + "grad_norm": 1.1875588011749945, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 32663 + }, + { + "epoch": 0.32664, + "grad_norm": 0.8160627659079428, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 32664 + }, + { + "epoch": 0.32665, + "grad_norm": 0.7884890950615491, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 32665 + }, + { + "epoch": 0.32666, + "grad_norm": 0.7794759709851903, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 32666 + }, + { + "epoch": 0.32667, + "grad_norm": 0.7472157781059815, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 32667 + }, + { + "epoch": 0.32668, + "grad_norm": 0.70935764549178, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32668 + }, + { + "epoch": 0.32669, + "grad_norm": 0.7192571617208463, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 32669 + }, + { + "epoch": 0.3267, + "grad_norm": 0.660937698714747, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 32670 + }, + { + "epoch": 0.32671, + "grad_norm": 0.7238142110186718, + "learning_rate": 0.003, + "loss": 4.008, + "step": 32671 + }, + { + "epoch": 0.32672, + "grad_norm": 0.8000814871241636, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 32672 + }, + { + "epoch": 0.32673, + "grad_norm": 0.8105233755746634, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 32673 + }, + { + "epoch": 0.32674, + "grad_norm": 0.8027544335043886, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 32674 + }, + { + "epoch": 0.32675, + "grad_norm": 0.7321465106414675, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 32675 + }, + { + "epoch": 0.32676, + "grad_norm": 0.7607682940696665, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 32676 + }, + { + "epoch": 0.32677, + "grad_norm": 0.7172644830111563, + "learning_rate": 0.003, + "loss": 3.983, + "step": 32677 + }, + { + "epoch": 0.32678, + "grad_norm": 0.7623233843854913, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 32678 + }, + { + "epoch": 0.32679, + "grad_norm": 0.697543801501382, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 32679 + }, + { + "epoch": 0.3268, + "grad_norm": 0.6968136933428036, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 32680 + }, + { + "epoch": 0.32681, + "grad_norm": 0.6142078008110796, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 32681 + }, + { + "epoch": 0.32682, + "grad_norm": 0.6584048059659453, + "learning_rate": 0.003, + "loss": 3.982, + "step": 32682 + }, + { + "epoch": 0.32683, + "grad_norm": 0.7085796999970607, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 32683 + }, + { + "epoch": 0.32684, + "grad_norm": 0.7240364275741855, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 32684 + }, + { + "epoch": 0.32685, + "grad_norm": 0.8351310028888943, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 32685 + }, + { + "epoch": 0.32686, + "grad_norm": 1.003158394772656, + "learning_rate": 0.003, + "loss": 4.0, + "step": 32686 + }, + { + "epoch": 0.32687, + "grad_norm": 1.1133512147630213, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 32687 + }, + { + "epoch": 0.32688, + "grad_norm": 0.7226821611698088, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 32688 + }, + { + "epoch": 0.32689, + "grad_norm": 0.6816158897284755, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 32689 + }, + { + "epoch": 0.3269, + "grad_norm": 0.8118399560726886, + "learning_rate": 0.003, + "loss": 3.98, + "step": 32690 + }, + { + "epoch": 0.32691, + "grad_norm": 1.0558339638120637, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 32691 + }, + { + "epoch": 0.32692, + "grad_norm": 0.9639290383273068, + "learning_rate": 0.003, + "loss": 3.997, + "step": 32692 + }, + { + "epoch": 0.32693, + "grad_norm": 0.6492115069639569, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 32693 + }, + { + "epoch": 0.32694, + "grad_norm": 0.6198514598916798, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 32694 + }, + { + "epoch": 0.32695, + "grad_norm": 0.7357136587704053, + "learning_rate": 0.003, + "loss": 3.977, + "step": 32695 + }, + { + "epoch": 0.32696, + "grad_norm": 0.7713556005371356, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 32696 + }, + { + "epoch": 0.32697, + "grad_norm": 0.738644065589724, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 32697 + }, + { + "epoch": 0.32698, + "grad_norm": 0.7180035196039992, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 32698 + }, + { + "epoch": 0.32699, + "grad_norm": 0.7286471245107005, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 32699 + }, + { + "epoch": 0.327, + "grad_norm": 0.6791547640764742, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 32700 + }, + { + "epoch": 0.32701, + "grad_norm": 0.6779958221071051, + "learning_rate": 0.003, + "loss": 3.9342, + "step": 32701 + }, + { + "epoch": 0.32702, + "grad_norm": 0.813067870818534, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 32702 + }, + { + "epoch": 0.32703, + "grad_norm": 0.8609445235561121, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 32703 + }, + { + "epoch": 0.32704, + "grad_norm": 0.7840046076471667, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 32704 + }, + { + "epoch": 0.32705, + "grad_norm": 0.779802037387009, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 32705 + }, + { + "epoch": 0.32706, + "grad_norm": 0.8130788498135558, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 32706 + }, + { + "epoch": 0.32707, + "grad_norm": 0.7538769503982768, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 32707 + }, + { + "epoch": 0.32708, + "grad_norm": 0.8426088109928268, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 32708 + }, + { + "epoch": 0.32709, + "grad_norm": 0.9156144554380573, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 32709 + }, + { + "epoch": 0.3271, + "grad_norm": 0.973712286537224, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 32710 + }, + { + "epoch": 0.32711, + "grad_norm": 0.9970891331635603, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 32711 + }, + { + "epoch": 0.32712, + "grad_norm": 1.004459790914804, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 32712 + }, + { + "epoch": 0.32713, + "grad_norm": 0.9749310326943156, + "learning_rate": 0.003, + "loss": 4.001, + "step": 32713 + }, + { + "epoch": 0.32714, + "grad_norm": 1.0356883616378356, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 32714 + }, + { + "epoch": 0.32715, + "grad_norm": 0.9469085433066625, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 32715 + }, + { + "epoch": 0.32716, + "grad_norm": 1.0429898024817852, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 32716 + }, + { + "epoch": 0.32717, + "grad_norm": 1.085257747708577, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 32717 + }, + { + "epoch": 0.32718, + "grad_norm": 0.8220528927180419, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 32718 + }, + { + "epoch": 0.32719, + "grad_norm": 0.7527705781979974, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 32719 + }, + { + "epoch": 0.3272, + "grad_norm": 0.8440673618271902, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 32720 + }, + { + "epoch": 0.32721, + "grad_norm": 0.8659758946543888, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 32721 + }, + { + "epoch": 0.32722, + "grad_norm": 0.8904973869160664, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 32722 + }, + { + "epoch": 0.32723, + "grad_norm": 0.8779549742482545, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 32723 + }, + { + "epoch": 0.32724, + "grad_norm": 0.9007999094894868, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 32724 + }, + { + "epoch": 0.32725, + "grad_norm": 0.9226574213530057, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 32725 + }, + { + "epoch": 0.32726, + "grad_norm": 1.0088991489411456, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 32726 + }, + { + "epoch": 0.32727, + "grad_norm": 0.9039910153155408, + "learning_rate": 0.003, + "loss": 4.008, + "step": 32727 + }, + { + "epoch": 0.32728, + "grad_norm": 0.7465635953939028, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 32728 + }, + { + "epoch": 0.32729, + "grad_norm": 0.6272000658253492, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 32729 + }, + { + "epoch": 0.3273, + "grad_norm": 0.6570622580887004, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 32730 + }, + { + "epoch": 0.32731, + "grad_norm": 0.6670841894675537, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 32731 + }, + { + "epoch": 0.32732, + "grad_norm": 0.6636908556317521, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 32732 + }, + { + "epoch": 0.32733, + "grad_norm": 0.6126559800769142, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 32733 + }, + { + "epoch": 0.32734, + "grad_norm": 0.5241915492810458, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 32734 + }, + { + "epoch": 0.32735, + "grad_norm": 0.44596505940811604, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 32735 + }, + { + "epoch": 0.32736, + "grad_norm": 0.46454637385215247, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 32736 + }, + { + "epoch": 0.32737, + "grad_norm": 0.47698905791931134, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 32737 + }, + { + "epoch": 0.32738, + "grad_norm": 0.48244850622633945, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 32738 + }, + { + "epoch": 0.32739, + "grad_norm": 0.4975821714760253, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 32739 + }, + { + "epoch": 0.3274, + "grad_norm": 0.5553567627061886, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 32740 + }, + { + "epoch": 0.32741, + "grad_norm": 0.6760294970278311, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 32741 + }, + { + "epoch": 0.32742, + "grad_norm": 0.6427076006307416, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 32742 + }, + { + "epoch": 0.32743, + "grad_norm": 0.6398240231469872, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 32743 + }, + { + "epoch": 0.32744, + "grad_norm": 0.6752877757735942, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 32744 + }, + { + "epoch": 0.32745, + "grad_norm": 0.6622967475096224, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 32745 + }, + { + "epoch": 0.32746, + "grad_norm": 0.7053011944258265, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 32746 + }, + { + "epoch": 0.32747, + "grad_norm": 0.7696034503673893, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 32747 + }, + { + "epoch": 0.32748, + "grad_norm": 0.8083224519903692, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 32748 + }, + { + "epoch": 0.32749, + "grad_norm": 0.8315192693798058, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 32749 + }, + { + "epoch": 0.3275, + "grad_norm": 1.0744692217231673, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 32750 + }, + { + "epoch": 0.32751, + "grad_norm": 1.2057842349723689, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 32751 + }, + { + "epoch": 0.32752, + "grad_norm": 0.658661986770472, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 32752 + }, + { + "epoch": 0.32753, + "grad_norm": 0.7303475518764831, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 32753 + }, + { + "epoch": 0.32754, + "grad_norm": 0.8467161249212787, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 32754 + }, + { + "epoch": 0.32755, + "grad_norm": 0.77846019748709, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 32755 + }, + { + "epoch": 0.32756, + "grad_norm": 0.7059768021163294, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 32756 + }, + { + "epoch": 0.32757, + "grad_norm": 0.5812635847444457, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 32757 + }, + { + "epoch": 0.32758, + "grad_norm": 0.569540823135799, + "learning_rate": 0.003, + "loss": 3.962, + "step": 32758 + }, + { + "epoch": 0.32759, + "grad_norm": 0.5822844430356552, + "learning_rate": 0.003, + "loss": 4.008, + "step": 32759 + }, + { + "epoch": 0.3276, + "grad_norm": 0.6127856504854841, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 32760 + }, + { + "epoch": 0.32761, + "grad_norm": 0.7529525677743305, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 32761 + }, + { + "epoch": 0.32762, + "grad_norm": 0.8733353892203578, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 32762 + }, + { + "epoch": 0.32763, + "grad_norm": 0.8095454727680832, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 32763 + }, + { + "epoch": 0.32764, + "grad_norm": 0.7625881927604184, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 32764 + }, + { + "epoch": 0.32765, + "grad_norm": 0.7907769687867172, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 32765 + }, + { + "epoch": 0.32766, + "grad_norm": 0.8695948376279162, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 32766 + }, + { + "epoch": 0.32767, + "grad_norm": 0.8990871206354837, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 32767 + }, + { + "epoch": 0.32768, + "grad_norm": 0.8897510665320264, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 32768 + }, + { + "epoch": 0.32769, + "grad_norm": 0.7792265917150226, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 32769 + }, + { + "epoch": 0.3277, + "grad_norm": 0.6952924106586355, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 32770 + }, + { + "epoch": 0.32771, + "grad_norm": 0.7743041658209251, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 32771 + }, + { + "epoch": 0.32772, + "grad_norm": 0.8771934535595429, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 32772 + }, + { + "epoch": 0.32773, + "grad_norm": 0.9379294431609098, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 32773 + }, + { + "epoch": 0.32774, + "grad_norm": 1.02705578168102, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 32774 + }, + { + "epoch": 0.32775, + "grad_norm": 1.0027161631848378, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 32775 + }, + { + "epoch": 0.32776, + "grad_norm": 0.9158517192323817, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 32776 + }, + { + "epoch": 0.32777, + "grad_norm": 0.8268500634137633, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 32777 + }, + { + "epoch": 0.32778, + "grad_norm": 0.7391341868197885, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 32778 + }, + { + "epoch": 0.32779, + "grad_norm": 0.7506356380589342, + "learning_rate": 0.003, + "loss": 4.019, + "step": 32779 + }, + { + "epoch": 0.3278, + "grad_norm": 0.7725289341406629, + "learning_rate": 0.003, + "loss": 3.981, + "step": 32780 + }, + { + "epoch": 0.32781, + "grad_norm": 0.7589185649981585, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 32781 + }, + { + "epoch": 0.32782, + "grad_norm": 0.7674028601493749, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 32782 + }, + { + "epoch": 0.32783, + "grad_norm": 0.7800471439384058, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 32783 + }, + { + "epoch": 0.32784, + "grad_norm": 0.8205505157004541, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 32784 + }, + { + "epoch": 0.32785, + "grad_norm": 0.8715195784514882, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 32785 + }, + { + "epoch": 0.32786, + "grad_norm": 0.7748269102998888, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 32786 + }, + { + "epoch": 0.32787, + "grad_norm": 0.6396250303567037, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 32787 + }, + { + "epoch": 0.32788, + "grad_norm": 0.6155418017927045, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 32788 + }, + { + "epoch": 0.32789, + "grad_norm": 0.6273951590304467, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 32789 + }, + { + "epoch": 0.3279, + "grad_norm": 0.6646558153458801, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 32790 + }, + { + "epoch": 0.32791, + "grad_norm": 0.8322556212745522, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 32791 + }, + { + "epoch": 0.32792, + "grad_norm": 0.9495019758883747, + "learning_rate": 0.003, + "loss": 3.991, + "step": 32792 + }, + { + "epoch": 0.32793, + "grad_norm": 1.0473282597207223, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 32793 + }, + { + "epoch": 0.32794, + "grad_norm": 0.7496242848459589, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 32794 + }, + { + "epoch": 0.32795, + "grad_norm": 0.5174757962946598, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 32795 + }, + { + "epoch": 0.32796, + "grad_norm": 0.630357721083219, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 32796 + }, + { + "epoch": 0.32797, + "grad_norm": 0.674470593286904, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 32797 + }, + { + "epoch": 0.32798, + "grad_norm": 0.6912145301558088, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 32798 + }, + { + "epoch": 0.32799, + "grad_norm": 0.593254653990585, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 32799 + }, + { + "epoch": 0.328, + "grad_norm": 0.5110596178707499, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 32800 + }, + { + "epoch": 0.32801, + "grad_norm": 0.4872975589494703, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 32801 + }, + { + "epoch": 0.32802, + "grad_norm": 0.4898764446232931, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 32802 + }, + { + "epoch": 0.32803, + "grad_norm": 0.4785069782932609, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 32803 + }, + { + "epoch": 0.32804, + "grad_norm": 0.5538668731671331, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 32804 + }, + { + "epoch": 0.32805, + "grad_norm": 0.5893932323819068, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 32805 + }, + { + "epoch": 0.32806, + "grad_norm": 0.5961510526513236, + "learning_rate": 0.003, + "loss": 3.983, + "step": 32806 + }, + { + "epoch": 0.32807, + "grad_norm": 0.6076000044657588, + "learning_rate": 0.003, + "loss": 4.005, + "step": 32807 + }, + { + "epoch": 0.32808, + "grad_norm": 0.6611915385682422, + "learning_rate": 0.003, + "loss": 4.0507, + "step": 32808 + }, + { + "epoch": 0.32809, + "grad_norm": 0.8153074554816346, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 32809 + }, + { + "epoch": 0.3281, + "grad_norm": 1.0198227779423614, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 32810 + }, + { + "epoch": 0.32811, + "grad_norm": 1.1382244902508003, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 32811 + }, + { + "epoch": 0.32812, + "grad_norm": 0.8060269120413558, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 32812 + }, + { + "epoch": 0.32813, + "grad_norm": 0.9536288958848552, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 32813 + }, + { + "epoch": 0.32814, + "grad_norm": 1.0656292962599896, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 32814 + }, + { + "epoch": 0.32815, + "grad_norm": 1.1472260110505084, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 32815 + }, + { + "epoch": 0.32816, + "grad_norm": 0.7938120183409754, + "learning_rate": 0.003, + "loss": 4.0391, + "step": 32816 + }, + { + "epoch": 0.32817, + "grad_norm": 0.8292145520265127, + "learning_rate": 0.003, + "loss": 3.992, + "step": 32817 + }, + { + "epoch": 0.32818, + "grad_norm": 1.0635000791071225, + "learning_rate": 0.003, + "loss": 3.965, + "step": 32818 + }, + { + "epoch": 0.32819, + "grad_norm": 1.1241194311943843, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 32819 + }, + { + "epoch": 0.3282, + "grad_norm": 1.0090216123533142, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 32820 + }, + { + "epoch": 0.32821, + "grad_norm": 1.0484217463842238, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 32821 + }, + { + "epoch": 0.32822, + "grad_norm": 0.9474729141323978, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 32822 + }, + { + "epoch": 0.32823, + "grad_norm": 0.8782981339240928, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 32823 + }, + { + "epoch": 0.32824, + "grad_norm": 0.7885483678816999, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 32824 + }, + { + "epoch": 0.32825, + "grad_norm": 0.8100173525866303, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 32825 + }, + { + "epoch": 0.32826, + "grad_norm": 0.8158000233797114, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 32826 + }, + { + "epoch": 0.32827, + "grad_norm": 0.7886097246365316, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 32827 + }, + { + "epoch": 0.32828, + "grad_norm": 0.7933787427747924, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 32828 + }, + { + "epoch": 0.32829, + "grad_norm": 0.7745289516126034, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 32829 + }, + { + "epoch": 0.3283, + "grad_norm": 0.7748929786032526, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 32830 + }, + { + "epoch": 0.32831, + "grad_norm": 0.7976453877171631, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 32831 + }, + { + "epoch": 0.32832, + "grad_norm": 0.7439030892660524, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 32832 + }, + { + "epoch": 0.32833, + "grad_norm": 0.6858627633485721, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 32833 + }, + { + "epoch": 0.32834, + "grad_norm": 0.6064320922380688, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 32834 + }, + { + "epoch": 0.32835, + "grad_norm": 0.604532069972935, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 32835 + }, + { + "epoch": 0.32836, + "grad_norm": 0.6448820202993633, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 32836 + }, + { + "epoch": 0.32837, + "grad_norm": 0.6959271912642053, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 32837 + }, + { + "epoch": 0.32838, + "grad_norm": 0.8539452811750249, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 32838 + }, + { + "epoch": 0.32839, + "grad_norm": 0.9192887979314035, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 32839 + }, + { + "epoch": 0.3284, + "grad_norm": 0.9619721800678417, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 32840 + }, + { + "epoch": 0.32841, + "grad_norm": 0.8718316030479287, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 32841 + }, + { + "epoch": 0.32842, + "grad_norm": 0.8983173315787177, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 32842 + }, + { + "epoch": 0.32843, + "grad_norm": 0.7728754754627248, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 32843 + }, + { + "epoch": 0.32844, + "grad_norm": 0.7096640657537352, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 32844 + }, + { + "epoch": 0.32845, + "grad_norm": 0.7282123775632675, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 32845 + }, + { + "epoch": 0.32846, + "grad_norm": 0.7824975149769986, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 32846 + }, + { + "epoch": 0.32847, + "grad_norm": 0.7638934394916168, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 32847 + }, + { + "epoch": 0.32848, + "grad_norm": 0.8526200608708896, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 32848 + }, + { + "epoch": 0.32849, + "grad_norm": 0.8478261637979566, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 32849 + }, + { + "epoch": 0.3285, + "grad_norm": 0.847078084606403, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 32850 + }, + { + "epoch": 0.32851, + "grad_norm": 0.889725117815169, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 32851 + }, + { + "epoch": 0.32852, + "grad_norm": 0.94230388222864, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 32852 + }, + { + "epoch": 0.32853, + "grad_norm": 0.9609973010331806, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 32853 + }, + { + "epoch": 0.32854, + "grad_norm": 0.9036723860780308, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 32854 + }, + { + "epoch": 0.32855, + "grad_norm": 0.8427533576315388, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 32855 + }, + { + "epoch": 0.32856, + "grad_norm": 0.8557968056364358, + "learning_rate": 0.003, + "loss": 3.993, + "step": 32856 + }, + { + "epoch": 0.32857, + "grad_norm": 0.919569487322373, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 32857 + }, + { + "epoch": 0.32858, + "grad_norm": 0.8844206486770467, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 32858 + }, + { + "epoch": 0.32859, + "grad_norm": 0.969482636905511, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 32859 + }, + { + "epoch": 0.3286, + "grad_norm": 0.8834103921981592, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 32860 + }, + { + "epoch": 0.32861, + "grad_norm": 0.8068648944167289, + "learning_rate": 0.003, + "loss": 3.9469, + "step": 32861 + }, + { + "epoch": 0.32862, + "grad_norm": 0.7811208652761631, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 32862 + }, + { + "epoch": 0.32863, + "grad_norm": 0.8054761974792677, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 32863 + }, + { + "epoch": 0.32864, + "grad_norm": 0.6601301359075268, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 32864 + }, + { + "epoch": 0.32865, + "grad_norm": 0.692393456837886, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 32865 + }, + { + "epoch": 0.32866, + "grad_norm": 0.6592350168759628, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 32866 + }, + { + "epoch": 0.32867, + "grad_norm": 0.6026787595381881, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 32867 + }, + { + "epoch": 0.32868, + "grad_norm": 0.5573964608292767, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 32868 + }, + { + "epoch": 0.32869, + "grad_norm": 0.5642363230870255, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 32869 + }, + { + "epoch": 0.3287, + "grad_norm": 0.6510607669499013, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 32870 + }, + { + "epoch": 0.32871, + "grad_norm": 0.759539245227309, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 32871 + }, + { + "epoch": 0.32872, + "grad_norm": 0.8179554060969013, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 32872 + }, + { + "epoch": 0.32873, + "grad_norm": 0.818449677125031, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 32873 + }, + { + "epoch": 0.32874, + "grad_norm": 0.7968757379336414, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 32874 + }, + { + "epoch": 0.32875, + "grad_norm": 0.7137515131216497, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 32875 + }, + { + "epoch": 0.32876, + "grad_norm": 0.6787398825314515, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 32876 + }, + { + "epoch": 0.32877, + "grad_norm": 0.7040318487048864, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 32877 + }, + { + "epoch": 0.32878, + "grad_norm": 0.6845670101589484, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 32878 + }, + { + "epoch": 0.32879, + "grad_norm": 0.6938891002178842, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 32879 + }, + { + "epoch": 0.3288, + "grad_norm": 0.5901797287941712, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 32880 + }, + { + "epoch": 0.32881, + "grad_norm": 0.563991039131691, + "learning_rate": 0.003, + "loss": 3.974, + "step": 32881 + }, + { + "epoch": 0.32882, + "grad_norm": 0.5717107935741441, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 32882 + }, + { + "epoch": 0.32883, + "grad_norm": 0.6334755158485497, + "learning_rate": 0.003, + "loss": 3.979, + "step": 32883 + }, + { + "epoch": 0.32884, + "grad_norm": 0.6062553244274331, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 32884 + }, + { + "epoch": 0.32885, + "grad_norm": 0.5838322109834937, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 32885 + }, + { + "epoch": 0.32886, + "grad_norm": 0.5968506965074145, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 32886 + }, + { + "epoch": 0.32887, + "grad_norm": 0.7760828629626655, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 32887 + }, + { + "epoch": 0.32888, + "grad_norm": 0.9358180281470112, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 32888 + }, + { + "epoch": 0.32889, + "grad_norm": 0.9875447999786845, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 32889 + }, + { + "epoch": 0.3289, + "grad_norm": 1.1880990336504342, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 32890 + }, + { + "epoch": 0.32891, + "grad_norm": 0.9137471473872005, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 32891 + }, + { + "epoch": 0.32892, + "grad_norm": 0.759742964185045, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 32892 + }, + { + "epoch": 0.32893, + "grad_norm": 0.6563468302361619, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 32893 + }, + { + "epoch": 0.32894, + "grad_norm": 0.8378578148272359, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 32894 + }, + { + "epoch": 0.32895, + "grad_norm": 1.0072949072630102, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 32895 + }, + { + "epoch": 0.32896, + "grad_norm": 0.9828249028806013, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 32896 + }, + { + "epoch": 0.32897, + "grad_norm": 0.9034935099804733, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 32897 + }, + { + "epoch": 0.32898, + "grad_norm": 0.8504531320536644, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 32898 + }, + { + "epoch": 0.32899, + "grad_norm": 0.8072342553700766, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 32899 + }, + { + "epoch": 0.329, + "grad_norm": 0.8191013376363482, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 32900 + }, + { + "epoch": 0.32901, + "grad_norm": 0.8278862557095948, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 32901 + }, + { + "epoch": 0.32902, + "grad_norm": 0.7249202885202084, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 32902 + }, + { + "epoch": 0.32903, + "grad_norm": 0.8474894538765883, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 32903 + }, + { + "epoch": 0.32904, + "grad_norm": 1.012332629506725, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 32904 + }, + { + "epoch": 0.32905, + "grad_norm": 1.0232426500324767, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 32905 + }, + { + "epoch": 0.32906, + "grad_norm": 0.9372964998467593, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 32906 + }, + { + "epoch": 0.32907, + "grad_norm": 0.8809699003345762, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 32907 + }, + { + "epoch": 0.32908, + "grad_norm": 0.8205448456651785, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 32908 + }, + { + "epoch": 0.32909, + "grad_norm": 0.7817477055847457, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 32909 + }, + { + "epoch": 0.3291, + "grad_norm": 0.8035281814822548, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 32910 + }, + { + "epoch": 0.32911, + "grad_norm": 0.8065424931105812, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 32911 + }, + { + "epoch": 0.32912, + "grad_norm": 0.8388425017135317, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 32912 + }, + { + "epoch": 0.32913, + "grad_norm": 0.7678604870201496, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 32913 + }, + { + "epoch": 0.32914, + "grad_norm": 0.7199725380981121, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 32914 + }, + { + "epoch": 0.32915, + "grad_norm": 0.7013498711423759, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 32915 + }, + { + "epoch": 0.32916, + "grad_norm": 0.7364452382829556, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 32916 + }, + { + "epoch": 0.32917, + "grad_norm": 0.820966670736273, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 32917 + }, + { + "epoch": 0.32918, + "grad_norm": 0.8495642211987263, + "learning_rate": 0.003, + "loss": 4.002, + "step": 32918 + }, + { + "epoch": 0.32919, + "grad_norm": 0.8931015224296106, + "learning_rate": 0.003, + "loss": 3.976, + "step": 32919 + }, + { + "epoch": 0.3292, + "grad_norm": 0.8771374902324328, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 32920 + }, + { + "epoch": 0.32921, + "grad_norm": 0.9097044323888797, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 32921 + }, + { + "epoch": 0.32922, + "grad_norm": 0.8712813553148209, + "learning_rate": 0.003, + "loss": 3.979, + "step": 32922 + }, + { + "epoch": 0.32923, + "grad_norm": 0.8919656858417108, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 32923 + }, + { + "epoch": 0.32924, + "grad_norm": 0.8895376241723181, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 32924 + }, + { + "epoch": 0.32925, + "grad_norm": 0.8306537295931141, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 32925 + }, + { + "epoch": 0.32926, + "grad_norm": 0.6979311009330385, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 32926 + }, + { + "epoch": 0.32927, + "grad_norm": 0.6625470926453488, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 32927 + }, + { + "epoch": 0.32928, + "grad_norm": 0.5939099215822721, + "learning_rate": 0.003, + "loss": 4.015, + "step": 32928 + }, + { + "epoch": 0.32929, + "grad_norm": 0.5902636662349117, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 32929 + }, + { + "epoch": 0.3293, + "grad_norm": 0.6311925260490501, + "learning_rate": 0.003, + "loss": 3.984, + "step": 32930 + }, + { + "epoch": 0.32931, + "grad_norm": 0.6450706570062175, + "learning_rate": 0.003, + "loss": 3.977, + "step": 32931 + }, + { + "epoch": 0.32932, + "grad_norm": 0.6643533393049675, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 32932 + }, + { + "epoch": 0.32933, + "grad_norm": 0.7254742125328495, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 32933 + }, + { + "epoch": 0.32934, + "grad_norm": 0.7013945611124324, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 32934 + }, + { + "epoch": 0.32935, + "grad_norm": 0.669574892398363, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 32935 + }, + { + "epoch": 0.32936, + "grad_norm": 0.6730013780293104, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 32936 + }, + { + "epoch": 0.32937, + "grad_norm": 0.7210749530333693, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 32937 + }, + { + "epoch": 0.32938, + "grad_norm": 0.8465170030769573, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 32938 + }, + { + "epoch": 0.32939, + "grad_norm": 0.8281701511490523, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 32939 + }, + { + "epoch": 0.3294, + "grad_norm": 0.8074299182191687, + "learning_rate": 0.003, + "loss": 3.993, + "step": 32940 + }, + { + "epoch": 0.32941, + "grad_norm": 0.8372869474647039, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 32941 + }, + { + "epoch": 0.32942, + "grad_norm": 0.8749764436914885, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 32942 + }, + { + "epoch": 0.32943, + "grad_norm": 0.8480142634452706, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 32943 + }, + { + "epoch": 0.32944, + "grad_norm": 0.8017894475635505, + "learning_rate": 0.003, + "loss": 4.007, + "step": 32944 + }, + { + "epoch": 0.32945, + "grad_norm": 0.7767616043774269, + "learning_rate": 0.003, + "loss": 3.978, + "step": 32945 + }, + { + "epoch": 0.32946, + "grad_norm": 0.7506773562620641, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 32946 + }, + { + "epoch": 0.32947, + "grad_norm": 0.6204149279772704, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 32947 + }, + { + "epoch": 0.32948, + "grad_norm": 0.5970844536909191, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 32948 + }, + { + "epoch": 0.32949, + "grad_norm": 0.6374970930252032, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 32949 + }, + { + "epoch": 0.3295, + "grad_norm": 0.6776840703782315, + "learning_rate": 0.003, + "loss": 4.016, + "step": 32950 + }, + { + "epoch": 0.32951, + "grad_norm": 0.6758722809905853, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 32951 + }, + { + "epoch": 0.32952, + "grad_norm": 0.7108000170095626, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 32952 + }, + { + "epoch": 0.32953, + "grad_norm": 0.8364719899353257, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 32953 + }, + { + "epoch": 0.32954, + "grad_norm": 1.075558204340313, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 32954 + }, + { + "epoch": 0.32955, + "grad_norm": 1.1738918934959135, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 32955 + }, + { + "epoch": 0.32956, + "grad_norm": 0.8374815859994187, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 32956 + }, + { + "epoch": 0.32957, + "grad_norm": 0.6459306089248549, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 32957 + }, + { + "epoch": 0.32958, + "grad_norm": 0.6350313489633417, + "learning_rate": 0.003, + "loss": 3.992, + "step": 32958 + }, + { + "epoch": 0.32959, + "grad_norm": 0.6730506100159537, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 32959 + }, + { + "epoch": 0.3296, + "grad_norm": 0.7438327965492839, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 32960 + }, + { + "epoch": 0.32961, + "grad_norm": 0.7776997667953615, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 32961 + }, + { + "epoch": 0.32962, + "grad_norm": 0.746026508861824, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 32962 + }, + { + "epoch": 0.32963, + "grad_norm": 0.7062678385718443, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 32963 + }, + { + "epoch": 0.32964, + "grad_norm": 0.7712644927839718, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 32964 + }, + { + "epoch": 0.32965, + "grad_norm": 0.768029122722783, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 32965 + }, + { + "epoch": 0.32966, + "grad_norm": 0.878309255160749, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 32966 + }, + { + "epoch": 0.32967, + "grad_norm": 0.8626910627767816, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 32967 + }, + { + "epoch": 0.32968, + "grad_norm": 0.8400530647015862, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 32968 + }, + { + "epoch": 0.32969, + "grad_norm": 0.717929598363136, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 32969 + }, + { + "epoch": 0.3297, + "grad_norm": 0.746065545178276, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 32970 + }, + { + "epoch": 0.32971, + "grad_norm": 0.7686865114375607, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 32971 + }, + { + "epoch": 0.32972, + "grad_norm": 0.8342032012534345, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 32972 + }, + { + "epoch": 0.32973, + "grad_norm": 0.8833726948160179, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 32973 + }, + { + "epoch": 0.32974, + "grad_norm": 0.8549094748753961, + "learning_rate": 0.003, + "loss": 3.986, + "step": 32974 + }, + { + "epoch": 0.32975, + "grad_norm": 0.8844315077313208, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 32975 + }, + { + "epoch": 0.32976, + "grad_norm": 0.8764757956256802, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 32976 + }, + { + "epoch": 0.32977, + "grad_norm": 0.8176971227721048, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 32977 + }, + { + "epoch": 0.32978, + "grad_norm": 0.8317657243368861, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 32978 + }, + { + "epoch": 0.32979, + "grad_norm": 0.8487915443895797, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 32979 + }, + { + "epoch": 0.3298, + "grad_norm": 0.683907786230574, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 32980 + }, + { + "epoch": 0.32981, + "grad_norm": 0.7198723629950107, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 32981 + }, + { + "epoch": 0.32982, + "grad_norm": 0.7882712785513853, + "learning_rate": 0.003, + "loss": 3.973, + "step": 32982 + }, + { + "epoch": 0.32983, + "grad_norm": 0.696155544606702, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 32983 + }, + { + "epoch": 0.32984, + "grad_norm": 0.7043988680362341, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 32984 + }, + { + "epoch": 0.32985, + "grad_norm": 0.7447303597173025, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 32985 + }, + { + "epoch": 0.32986, + "grad_norm": 0.7496590999166475, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 32986 + }, + { + "epoch": 0.32987, + "grad_norm": 0.6871320528739147, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 32987 + }, + { + "epoch": 0.32988, + "grad_norm": 0.6695597036595261, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 32988 + }, + { + "epoch": 0.32989, + "grad_norm": 0.7518018635673829, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 32989 + }, + { + "epoch": 0.3299, + "grad_norm": 0.7619326512191543, + "learning_rate": 0.003, + "loss": 3.9338, + "step": 32990 + }, + { + "epoch": 0.32991, + "grad_norm": 0.7376926982398793, + "learning_rate": 0.003, + "loss": 4.001, + "step": 32991 + }, + { + "epoch": 0.32992, + "grad_norm": 0.7447345135107574, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 32992 + }, + { + "epoch": 0.32993, + "grad_norm": 0.7016336670486686, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 32993 + }, + { + "epoch": 0.32994, + "grad_norm": 0.7249893664965987, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 32994 + }, + { + "epoch": 0.32995, + "grad_norm": 0.8652405210781213, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 32995 + }, + { + "epoch": 0.32996, + "grad_norm": 0.9495246901398426, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 32996 + }, + { + "epoch": 0.32997, + "grad_norm": 0.9810155005054407, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 32997 + }, + { + "epoch": 0.32998, + "grad_norm": 0.9804947930017796, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 32998 + }, + { + "epoch": 0.32999, + "grad_norm": 1.0085648054784142, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 32999 + }, + { + "epoch": 0.33, + "grad_norm": 1.0037184464399003, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 33000 + }, + { + "epoch": 0.33001, + "grad_norm": 0.8286316207162157, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 33001 + }, + { + "epoch": 0.33002, + "grad_norm": 0.9527639306679931, + "learning_rate": 0.003, + "loss": 3.978, + "step": 33002 + }, + { + "epoch": 0.33003, + "grad_norm": 0.9392679021561026, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 33003 + }, + { + "epoch": 0.33004, + "grad_norm": 0.8833847300087863, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 33004 + }, + { + "epoch": 0.33005, + "grad_norm": 0.7333404136058012, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 33005 + }, + { + "epoch": 0.33006, + "grad_norm": 0.702576149190601, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 33006 + }, + { + "epoch": 0.33007, + "grad_norm": 0.599771463222031, + "learning_rate": 0.003, + "loss": 4.002, + "step": 33007 + }, + { + "epoch": 0.33008, + "grad_norm": 0.5459532839681039, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 33008 + }, + { + "epoch": 0.33009, + "grad_norm": 0.5375018378452713, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 33009 + }, + { + "epoch": 0.3301, + "grad_norm": 0.5867009335367855, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 33010 + }, + { + "epoch": 0.33011, + "grad_norm": 0.6621027943369704, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 33011 + }, + { + "epoch": 0.33012, + "grad_norm": 0.6860467280436642, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 33012 + }, + { + "epoch": 0.33013, + "grad_norm": 0.7560429279265813, + "learning_rate": 0.003, + "loss": 4.015, + "step": 33013 + }, + { + "epoch": 0.33014, + "grad_norm": 0.8761022375092434, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 33014 + }, + { + "epoch": 0.33015, + "grad_norm": 0.9529925101596914, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 33015 + }, + { + "epoch": 0.33016, + "grad_norm": 0.8877762188261281, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 33016 + }, + { + "epoch": 0.33017, + "grad_norm": 0.8215975936697322, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 33017 + }, + { + "epoch": 0.33018, + "grad_norm": 0.7963499714014775, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 33018 + }, + { + "epoch": 0.33019, + "grad_norm": 0.6678328329624472, + "learning_rate": 0.003, + "loss": 3.978, + "step": 33019 + }, + { + "epoch": 0.3302, + "grad_norm": 0.603111195785538, + "learning_rate": 0.003, + "loss": 3.993, + "step": 33020 + }, + { + "epoch": 0.33021, + "grad_norm": 0.7535613020824236, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 33021 + }, + { + "epoch": 0.33022, + "grad_norm": 0.8839672855799587, + "learning_rate": 0.003, + "loss": 3.979, + "step": 33022 + }, + { + "epoch": 0.33023, + "grad_norm": 0.9145027574003476, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 33023 + }, + { + "epoch": 0.33024, + "grad_norm": 0.8536006091429307, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 33024 + }, + { + "epoch": 0.33025, + "grad_norm": 0.7295089987756002, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 33025 + }, + { + "epoch": 0.33026, + "grad_norm": 0.6832654674068681, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 33026 + }, + { + "epoch": 0.33027, + "grad_norm": 0.7634406939851296, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 33027 + }, + { + "epoch": 0.33028, + "grad_norm": 0.8468403181684964, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 33028 + }, + { + "epoch": 0.33029, + "grad_norm": 0.9459163561033057, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 33029 + }, + { + "epoch": 0.3303, + "grad_norm": 1.1518195154501822, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 33030 + }, + { + "epoch": 0.33031, + "grad_norm": 0.8167427945998331, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 33031 + }, + { + "epoch": 0.33032, + "grad_norm": 0.6560503527872589, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 33032 + }, + { + "epoch": 0.33033, + "grad_norm": 0.7764488870636914, + "learning_rate": 0.003, + "loss": 3.974, + "step": 33033 + }, + { + "epoch": 0.33034, + "grad_norm": 0.7086024991778798, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 33034 + }, + { + "epoch": 0.33035, + "grad_norm": 0.6592372793295366, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 33035 + }, + { + "epoch": 0.33036, + "grad_norm": 0.7174909471469147, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 33036 + }, + { + "epoch": 0.33037, + "grad_norm": 0.7329390466385357, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 33037 + }, + { + "epoch": 0.33038, + "grad_norm": 0.7330259146760113, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 33038 + }, + { + "epoch": 0.33039, + "grad_norm": 0.7052264446275791, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 33039 + }, + { + "epoch": 0.3304, + "grad_norm": 0.7438501622148752, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 33040 + }, + { + "epoch": 0.33041, + "grad_norm": 0.73427022590375, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 33041 + }, + { + "epoch": 0.33042, + "grad_norm": 0.7862128020592497, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 33042 + }, + { + "epoch": 0.33043, + "grad_norm": 0.891503018503518, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 33043 + }, + { + "epoch": 0.33044, + "grad_norm": 0.8317949191768997, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 33044 + }, + { + "epoch": 0.33045, + "grad_norm": 0.774290957207972, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 33045 + }, + { + "epoch": 0.33046, + "grad_norm": 0.8249600993222146, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 33046 + }, + { + "epoch": 0.33047, + "grad_norm": 0.788098151205191, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 33047 + }, + { + "epoch": 0.33048, + "grad_norm": 0.6758700297528755, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 33048 + }, + { + "epoch": 0.33049, + "grad_norm": 0.6754689515397349, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 33049 + }, + { + "epoch": 0.3305, + "grad_norm": 0.6835592961323111, + "learning_rate": 0.003, + "loss": 3.961, + "step": 33050 + }, + { + "epoch": 0.33051, + "grad_norm": 0.6290553248688597, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 33051 + }, + { + "epoch": 0.33052, + "grad_norm": 0.6548195247739973, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 33052 + }, + { + "epoch": 0.33053, + "grad_norm": 0.6903667487910845, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 33053 + }, + { + "epoch": 0.33054, + "grad_norm": 0.6517581199347023, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 33054 + }, + { + "epoch": 0.33055, + "grad_norm": 0.629906636729511, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 33055 + }, + { + "epoch": 0.33056, + "grad_norm": 0.5949159024869884, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 33056 + }, + { + "epoch": 0.33057, + "grad_norm": 0.6862190994045326, + "learning_rate": 0.003, + "loss": 3.98, + "step": 33057 + }, + { + "epoch": 0.33058, + "grad_norm": 0.7630344829044688, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 33058 + }, + { + "epoch": 0.33059, + "grad_norm": 0.7036098719396116, + "learning_rate": 0.003, + "loss": 3.982, + "step": 33059 + }, + { + "epoch": 0.3306, + "grad_norm": 0.7229022883340227, + "learning_rate": 0.003, + "loss": 3.971, + "step": 33060 + }, + { + "epoch": 0.33061, + "grad_norm": 0.8556570908272729, + "learning_rate": 0.003, + "loss": 4.007, + "step": 33061 + }, + { + "epoch": 0.33062, + "grad_norm": 0.9368370861239483, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 33062 + }, + { + "epoch": 0.33063, + "grad_norm": 1.0853435951503383, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 33063 + }, + { + "epoch": 0.33064, + "grad_norm": 1.0148243717825225, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 33064 + }, + { + "epoch": 0.33065, + "grad_norm": 0.9675070025192998, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 33065 + }, + { + "epoch": 0.33066, + "grad_norm": 1.0112742680149398, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 33066 + }, + { + "epoch": 0.33067, + "grad_norm": 0.9980198443859721, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 33067 + }, + { + "epoch": 0.33068, + "grad_norm": 1.0351634510380585, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 33068 + }, + { + "epoch": 0.33069, + "grad_norm": 1.1396994723812135, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 33069 + }, + { + "epoch": 0.3307, + "grad_norm": 0.8004774507362381, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 33070 + }, + { + "epoch": 0.33071, + "grad_norm": 0.7139404933361434, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 33071 + }, + { + "epoch": 0.33072, + "grad_norm": 0.7944123985664007, + "learning_rate": 0.003, + "loss": 4.036, + "step": 33072 + }, + { + "epoch": 0.33073, + "grad_norm": 0.8482671606640935, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 33073 + }, + { + "epoch": 0.33074, + "grad_norm": 0.8575903995796804, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 33074 + }, + { + "epoch": 0.33075, + "grad_norm": 0.9892361501978365, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 33075 + }, + { + "epoch": 0.33076, + "grad_norm": 1.073774359200846, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 33076 + }, + { + "epoch": 0.33077, + "grad_norm": 0.8968371078729452, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 33077 + }, + { + "epoch": 0.33078, + "grad_norm": 0.8313556252810229, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 33078 + }, + { + "epoch": 0.33079, + "grad_norm": 0.7875732422978886, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 33079 + }, + { + "epoch": 0.3308, + "grad_norm": 0.8741856345076098, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 33080 + }, + { + "epoch": 0.33081, + "grad_norm": 0.8827102089250491, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 33081 + }, + { + "epoch": 0.33082, + "grad_norm": 0.8279725156072693, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 33082 + }, + { + "epoch": 0.33083, + "grad_norm": 0.8514964883911131, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 33083 + }, + { + "epoch": 0.33084, + "grad_norm": 0.6688684504748108, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 33084 + }, + { + "epoch": 0.33085, + "grad_norm": 0.670985372755684, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 33085 + }, + { + "epoch": 0.33086, + "grad_norm": 0.7658518658087498, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 33086 + }, + { + "epoch": 0.33087, + "grad_norm": 0.9324758428909802, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 33087 + }, + { + "epoch": 0.33088, + "grad_norm": 1.0371879565680817, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 33088 + }, + { + "epoch": 0.33089, + "grad_norm": 0.8095471541129461, + "learning_rate": 0.003, + "loss": 4.0387, + "step": 33089 + }, + { + "epoch": 0.3309, + "grad_norm": 0.7200572784107176, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 33090 + }, + { + "epoch": 0.33091, + "grad_norm": 0.6864021430529368, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 33091 + }, + { + "epoch": 0.33092, + "grad_norm": 0.6812378099993635, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 33092 + }, + { + "epoch": 0.33093, + "grad_norm": 0.7063693410098989, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 33093 + }, + { + "epoch": 0.33094, + "grad_norm": 0.7258911047723344, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 33094 + }, + { + "epoch": 0.33095, + "grad_norm": 0.6731732154846899, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 33095 + }, + { + "epoch": 0.33096, + "grad_norm": 0.7020712742994422, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 33096 + }, + { + "epoch": 0.33097, + "grad_norm": 0.6654154807512578, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 33097 + }, + { + "epoch": 0.33098, + "grad_norm": 0.6323547634162442, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 33098 + }, + { + "epoch": 0.33099, + "grad_norm": 0.6602496581607338, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 33099 + }, + { + "epoch": 0.331, + "grad_norm": 0.7839554361328291, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 33100 + }, + { + "epoch": 0.33101, + "grad_norm": 0.9238581370940611, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 33101 + }, + { + "epoch": 0.33102, + "grad_norm": 1.018242664838756, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 33102 + }, + { + "epoch": 0.33103, + "grad_norm": 0.9010758592042146, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 33103 + }, + { + "epoch": 0.33104, + "grad_norm": 0.8914671781239023, + "learning_rate": 0.003, + "loss": 3.965, + "step": 33104 + }, + { + "epoch": 0.33105, + "grad_norm": 0.9361857725337834, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 33105 + }, + { + "epoch": 0.33106, + "grad_norm": 0.9021586080208678, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 33106 + }, + { + "epoch": 0.33107, + "grad_norm": 0.7496614200586129, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 33107 + }, + { + "epoch": 0.33108, + "grad_norm": 0.8083140428552238, + "learning_rate": 0.003, + "loss": 3.993, + "step": 33108 + }, + { + "epoch": 0.33109, + "grad_norm": 0.6996581165177119, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 33109 + }, + { + "epoch": 0.3311, + "grad_norm": 0.6687226484632507, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 33110 + }, + { + "epoch": 0.33111, + "grad_norm": 0.7068689187367119, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 33111 + }, + { + "epoch": 0.33112, + "grad_norm": 0.835518109824594, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 33112 + }, + { + "epoch": 0.33113, + "grad_norm": 0.8046669662593011, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 33113 + }, + { + "epoch": 0.33114, + "grad_norm": 0.7276462297919163, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 33114 + }, + { + "epoch": 0.33115, + "grad_norm": 0.8745000706031788, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 33115 + }, + { + "epoch": 0.33116, + "grad_norm": 1.116085915044433, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 33116 + }, + { + "epoch": 0.33117, + "grad_norm": 1.068011209027845, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 33117 + }, + { + "epoch": 0.33118, + "grad_norm": 0.8044588628862963, + "learning_rate": 0.003, + "loss": 4.027, + "step": 33118 + }, + { + "epoch": 0.33119, + "grad_norm": 0.5961754299753291, + "learning_rate": 0.003, + "loss": 3.982, + "step": 33119 + }, + { + "epoch": 0.3312, + "grad_norm": 0.6345196261936661, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 33120 + }, + { + "epoch": 0.33121, + "grad_norm": 0.6645084848399925, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 33121 + }, + { + "epoch": 0.33122, + "grad_norm": 0.64383092914217, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 33122 + }, + { + "epoch": 0.33123, + "grad_norm": 0.616625524766998, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 33123 + }, + { + "epoch": 0.33124, + "grad_norm": 0.6001468221574751, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 33124 + }, + { + "epoch": 0.33125, + "grad_norm": 0.5957809496409816, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 33125 + }, + { + "epoch": 0.33126, + "grad_norm": 0.702724749644638, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 33126 + }, + { + "epoch": 0.33127, + "grad_norm": 0.6694560237739695, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 33127 + }, + { + "epoch": 0.33128, + "grad_norm": 0.7122912942474721, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 33128 + }, + { + "epoch": 0.33129, + "grad_norm": 0.7000663662414407, + "learning_rate": 0.003, + "loss": 4.002, + "step": 33129 + }, + { + "epoch": 0.3313, + "grad_norm": 0.6734568943932131, + "learning_rate": 0.003, + "loss": 4.004, + "step": 33130 + }, + { + "epoch": 0.33131, + "grad_norm": 0.7715872377093371, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 33131 + }, + { + "epoch": 0.33132, + "grad_norm": 0.777827953334238, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 33132 + }, + { + "epoch": 0.33133, + "grad_norm": 0.8596757457212185, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 33133 + }, + { + "epoch": 0.33134, + "grad_norm": 0.8564380010106324, + "learning_rate": 0.003, + "loss": 3.997, + "step": 33134 + }, + { + "epoch": 0.33135, + "grad_norm": 0.8700794180369777, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 33135 + }, + { + "epoch": 0.33136, + "grad_norm": 0.7708627180766798, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 33136 + }, + { + "epoch": 0.33137, + "grad_norm": 0.7639974368266439, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 33137 + }, + { + "epoch": 0.33138, + "grad_norm": 0.6792509565450243, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 33138 + }, + { + "epoch": 0.33139, + "grad_norm": 0.7315773939578657, + "learning_rate": 0.003, + "loss": 3.966, + "step": 33139 + }, + { + "epoch": 0.3314, + "grad_norm": 0.7817437196826839, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 33140 + }, + { + "epoch": 0.33141, + "grad_norm": 0.8161982097996433, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 33141 + }, + { + "epoch": 0.33142, + "grad_norm": 0.8041858070832147, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 33142 + }, + { + "epoch": 0.33143, + "grad_norm": 0.767020400452125, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 33143 + }, + { + "epoch": 0.33144, + "grad_norm": 0.6909221801823985, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 33144 + }, + { + "epoch": 0.33145, + "grad_norm": 0.6708119462731829, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 33145 + }, + { + "epoch": 0.33146, + "grad_norm": 0.5838017624683731, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 33146 + }, + { + "epoch": 0.33147, + "grad_norm": 0.5842263616811327, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 33147 + }, + { + "epoch": 0.33148, + "grad_norm": 0.5834242120177353, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 33148 + }, + { + "epoch": 0.33149, + "grad_norm": 0.555749027813404, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 33149 + }, + { + "epoch": 0.3315, + "grad_norm": 0.509912732260329, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 33150 + }, + { + "epoch": 0.33151, + "grad_norm": 0.5307601849170567, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 33151 + }, + { + "epoch": 0.33152, + "grad_norm": 0.5569313086089185, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 33152 + }, + { + "epoch": 0.33153, + "grad_norm": 0.7099806000831822, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 33153 + }, + { + "epoch": 0.33154, + "grad_norm": 0.8764725853063179, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 33154 + }, + { + "epoch": 0.33155, + "grad_norm": 1.0974792350315805, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 33155 + }, + { + "epoch": 0.33156, + "grad_norm": 0.9555431190709404, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 33156 + }, + { + "epoch": 0.33157, + "grad_norm": 0.8382339737260408, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 33157 + }, + { + "epoch": 0.33158, + "grad_norm": 0.736172168342769, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 33158 + }, + { + "epoch": 0.33159, + "grad_norm": 0.7982148832658039, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 33159 + }, + { + "epoch": 0.3316, + "grad_norm": 0.8611076469869523, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 33160 + }, + { + "epoch": 0.33161, + "grad_norm": 0.803109719999346, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 33161 + }, + { + "epoch": 0.33162, + "grad_norm": 0.8551833858009712, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 33162 + }, + { + "epoch": 0.33163, + "grad_norm": 0.7640017308172187, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 33163 + }, + { + "epoch": 0.33164, + "grad_norm": 0.6864094705113877, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 33164 + }, + { + "epoch": 0.33165, + "grad_norm": 0.6245023894589988, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 33165 + }, + { + "epoch": 0.33166, + "grad_norm": 0.6767737580617446, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 33166 + }, + { + "epoch": 0.33167, + "grad_norm": 0.6649707649316876, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 33167 + }, + { + "epoch": 0.33168, + "grad_norm": 0.7830539680917328, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 33168 + }, + { + "epoch": 0.33169, + "grad_norm": 0.9689113297108701, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 33169 + }, + { + "epoch": 0.3317, + "grad_norm": 1.1508218623905664, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 33170 + }, + { + "epoch": 0.33171, + "grad_norm": 0.7501189264591976, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 33171 + }, + { + "epoch": 0.33172, + "grad_norm": 0.8486880272508689, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 33172 + }, + { + "epoch": 0.33173, + "grad_norm": 0.8809250450562328, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 33173 + }, + { + "epoch": 0.33174, + "grad_norm": 0.829431178545158, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 33174 + }, + { + "epoch": 0.33175, + "grad_norm": 0.8764810417188746, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 33175 + }, + { + "epoch": 0.33176, + "grad_norm": 0.8760721213994513, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 33176 + }, + { + "epoch": 0.33177, + "grad_norm": 0.9292015876359156, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 33177 + }, + { + "epoch": 0.33178, + "grad_norm": 0.8903092218450999, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 33178 + }, + { + "epoch": 0.33179, + "grad_norm": 0.8918335273669638, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 33179 + }, + { + "epoch": 0.3318, + "grad_norm": 0.9675541835778044, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 33180 + }, + { + "epoch": 0.33181, + "grad_norm": 0.8242735077661641, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 33181 + }, + { + "epoch": 0.33182, + "grad_norm": 0.7386242658039714, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 33182 + }, + { + "epoch": 0.33183, + "grad_norm": 0.8365484774280166, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 33183 + }, + { + "epoch": 0.33184, + "grad_norm": 0.7958244761227712, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 33184 + }, + { + "epoch": 0.33185, + "grad_norm": 0.7856195247525567, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 33185 + }, + { + "epoch": 0.33186, + "grad_norm": 0.7860724709830696, + "learning_rate": 0.003, + "loss": 3.995, + "step": 33186 + }, + { + "epoch": 0.33187, + "grad_norm": 0.828249366076115, + "learning_rate": 0.003, + "loss": 3.968, + "step": 33187 + }, + { + "epoch": 0.33188, + "grad_norm": 0.8409910813148642, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 33188 + }, + { + "epoch": 0.33189, + "grad_norm": 0.899244395181846, + "learning_rate": 0.003, + "loss": 3.991, + "step": 33189 + }, + { + "epoch": 0.3319, + "grad_norm": 0.9425924272595715, + "learning_rate": 0.003, + "loss": 4.021, + "step": 33190 + }, + { + "epoch": 0.33191, + "grad_norm": 0.9207111059960593, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 33191 + }, + { + "epoch": 0.33192, + "grad_norm": 0.9156346246445559, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 33192 + }, + { + "epoch": 0.33193, + "grad_norm": 0.9548245306216198, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 33193 + }, + { + "epoch": 0.33194, + "grad_norm": 0.9324985532185877, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 33194 + }, + { + "epoch": 0.33195, + "grad_norm": 0.9405224426405336, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 33195 + }, + { + "epoch": 0.33196, + "grad_norm": 1.0083585327045728, + "learning_rate": 0.003, + "loss": 4.0357, + "step": 33196 + }, + { + "epoch": 0.33197, + "grad_norm": 1.005850498480576, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 33197 + }, + { + "epoch": 0.33198, + "grad_norm": 0.9205948409548798, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 33198 + }, + { + "epoch": 0.33199, + "grad_norm": 0.9607173702189641, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 33199 + }, + { + "epoch": 0.332, + "grad_norm": 0.9605020722862123, + "learning_rate": 0.003, + "loss": 3.986, + "step": 33200 + }, + { + "epoch": 0.33201, + "grad_norm": 0.9666667339878994, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 33201 + }, + { + "epoch": 0.33202, + "grad_norm": 0.8121843786570418, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 33202 + }, + { + "epoch": 0.33203, + "grad_norm": 0.6591019996419614, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 33203 + }, + { + "epoch": 0.33204, + "grad_norm": 0.6256892891043461, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 33204 + }, + { + "epoch": 0.33205, + "grad_norm": 0.5238099467364323, + "learning_rate": 0.003, + "loss": 3.997, + "step": 33205 + }, + { + "epoch": 0.33206, + "grad_norm": 0.5323119673026283, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 33206 + }, + { + "epoch": 0.33207, + "grad_norm": 0.4612795709769452, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 33207 + }, + { + "epoch": 0.33208, + "grad_norm": 0.4275608254262378, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 33208 + }, + { + "epoch": 0.33209, + "grad_norm": 0.3864292091105178, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 33209 + }, + { + "epoch": 0.3321, + "grad_norm": 0.45549872090199073, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 33210 + }, + { + "epoch": 0.33211, + "grad_norm": 0.41006452246737224, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 33211 + }, + { + "epoch": 0.33212, + "grad_norm": 0.4000335743772074, + "learning_rate": 0.003, + "loss": 3.9426, + "step": 33212 + }, + { + "epoch": 0.33213, + "grad_norm": 0.38791003393414947, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 33213 + }, + { + "epoch": 0.33214, + "grad_norm": 0.44663959767937533, + "learning_rate": 0.003, + "loss": 3.971, + "step": 33214 + }, + { + "epoch": 0.33215, + "grad_norm": 0.47621104188203867, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 33215 + }, + { + "epoch": 0.33216, + "grad_norm": 0.5301901940071835, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 33216 + }, + { + "epoch": 0.33217, + "grad_norm": 0.7034701101115683, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 33217 + }, + { + "epoch": 0.33218, + "grad_norm": 0.9349342781452703, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 33218 + }, + { + "epoch": 0.33219, + "grad_norm": 1.073158131600844, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 33219 + }, + { + "epoch": 0.3322, + "grad_norm": 0.9291284511277068, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 33220 + }, + { + "epoch": 0.33221, + "grad_norm": 0.8391527771245669, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 33221 + }, + { + "epoch": 0.33222, + "grad_norm": 0.7228839682946157, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 33222 + }, + { + "epoch": 0.33223, + "grad_norm": 0.6997231457728863, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 33223 + }, + { + "epoch": 0.33224, + "grad_norm": 0.7078852194095031, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 33224 + }, + { + "epoch": 0.33225, + "grad_norm": 0.8886228280740733, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 33225 + }, + { + "epoch": 0.33226, + "grad_norm": 0.9548610113465634, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 33226 + }, + { + "epoch": 0.33227, + "grad_norm": 1.0152229431742046, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 33227 + }, + { + "epoch": 0.33228, + "grad_norm": 0.869415013297261, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 33228 + }, + { + "epoch": 0.33229, + "grad_norm": 0.9096303011762328, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 33229 + }, + { + "epoch": 0.3323, + "grad_norm": 1.0565608077723796, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 33230 + }, + { + "epoch": 0.33231, + "grad_norm": 0.8630444357412161, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 33231 + }, + { + "epoch": 0.33232, + "grad_norm": 0.7531070675815347, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 33232 + }, + { + "epoch": 0.33233, + "grad_norm": 0.7597095599690037, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 33233 + }, + { + "epoch": 0.33234, + "grad_norm": 0.7214062678428342, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 33234 + }, + { + "epoch": 0.33235, + "grad_norm": 0.7462110998692484, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 33235 + }, + { + "epoch": 0.33236, + "grad_norm": 0.7205977408568348, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 33236 + }, + { + "epoch": 0.33237, + "grad_norm": 0.6233686410868369, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 33237 + }, + { + "epoch": 0.33238, + "grad_norm": 0.5578661853156955, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 33238 + }, + { + "epoch": 0.33239, + "grad_norm": 0.6103091558520957, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 33239 + }, + { + "epoch": 0.3324, + "grad_norm": 0.6556396411367605, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 33240 + }, + { + "epoch": 0.33241, + "grad_norm": 0.6501526072717125, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 33241 + }, + { + "epoch": 0.33242, + "grad_norm": 0.6445194706559834, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 33242 + }, + { + "epoch": 0.33243, + "grad_norm": 0.6986656855325606, + "learning_rate": 0.003, + "loss": 3.994, + "step": 33243 + }, + { + "epoch": 0.33244, + "grad_norm": 0.6782061423694925, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 33244 + }, + { + "epoch": 0.33245, + "grad_norm": 0.6011038703484989, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 33245 + }, + { + "epoch": 0.33246, + "grad_norm": 0.6530892936694759, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 33246 + }, + { + "epoch": 0.33247, + "grad_norm": 0.8012650797792122, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 33247 + }, + { + "epoch": 0.33248, + "grad_norm": 1.0994135577306332, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 33248 + }, + { + "epoch": 0.33249, + "grad_norm": 0.9522725215087873, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 33249 + }, + { + "epoch": 0.3325, + "grad_norm": 0.864241067555164, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 33250 + }, + { + "epoch": 0.33251, + "grad_norm": 0.7448870107424003, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 33251 + }, + { + "epoch": 0.33252, + "grad_norm": 0.6972042336042429, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 33252 + }, + { + "epoch": 0.33253, + "grad_norm": 0.8517864788409271, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 33253 + }, + { + "epoch": 0.33254, + "grad_norm": 0.9450185438324774, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 33254 + }, + { + "epoch": 0.33255, + "grad_norm": 1.0075592073065915, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 33255 + }, + { + "epoch": 0.33256, + "grad_norm": 1.120264646975982, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 33256 + }, + { + "epoch": 0.33257, + "grad_norm": 1.1097878549489562, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 33257 + }, + { + "epoch": 0.33258, + "grad_norm": 0.9690067856587664, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 33258 + }, + { + "epoch": 0.33259, + "grad_norm": 0.9024464098866689, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 33259 + }, + { + "epoch": 0.3326, + "grad_norm": 0.9260737620185874, + "learning_rate": 0.003, + "loss": 4.034, + "step": 33260 + }, + { + "epoch": 0.33261, + "grad_norm": 0.9332165203658583, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 33261 + }, + { + "epoch": 0.33262, + "grad_norm": 0.9439087136604378, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 33262 + }, + { + "epoch": 0.33263, + "grad_norm": 0.9273726332509223, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 33263 + }, + { + "epoch": 0.33264, + "grad_norm": 0.7863893171654993, + "learning_rate": 0.003, + "loss": 3.991, + "step": 33264 + }, + { + "epoch": 0.33265, + "grad_norm": 0.7180887670381533, + "learning_rate": 0.003, + "loss": 4.015, + "step": 33265 + }, + { + "epoch": 0.33266, + "grad_norm": 0.7211039877591388, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 33266 + }, + { + "epoch": 0.33267, + "grad_norm": 0.7839625530195892, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 33267 + }, + { + "epoch": 0.33268, + "grad_norm": 0.7361375768651393, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 33268 + }, + { + "epoch": 0.33269, + "grad_norm": 0.7332542244281529, + "learning_rate": 0.003, + "loss": 3.978, + "step": 33269 + }, + { + "epoch": 0.3327, + "grad_norm": 0.8118368025716788, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 33270 + }, + { + "epoch": 0.33271, + "grad_norm": 0.8899139695363684, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 33271 + }, + { + "epoch": 0.33272, + "grad_norm": 0.8476864803783233, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 33272 + }, + { + "epoch": 0.33273, + "grad_norm": 0.6900490652505926, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 33273 + }, + { + "epoch": 0.33274, + "grad_norm": 0.5787638226678878, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 33274 + }, + { + "epoch": 0.33275, + "grad_norm": 0.5439660016087247, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 33275 + }, + { + "epoch": 0.33276, + "grad_norm": 0.5111876662098982, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 33276 + }, + { + "epoch": 0.33277, + "grad_norm": 0.5228537287801889, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 33277 + }, + { + "epoch": 0.33278, + "grad_norm": 0.49294688473802045, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 33278 + }, + { + "epoch": 0.33279, + "grad_norm": 0.5732736506285745, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 33279 + }, + { + "epoch": 0.3328, + "grad_norm": 0.7718527354443603, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 33280 + }, + { + "epoch": 0.33281, + "grad_norm": 0.939750874139315, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 33281 + }, + { + "epoch": 0.33282, + "grad_norm": 0.9310044720690329, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 33282 + }, + { + "epoch": 0.33283, + "grad_norm": 0.764084810298089, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 33283 + }, + { + "epoch": 0.33284, + "grad_norm": 0.6561899740603696, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 33284 + }, + { + "epoch": 0.33285, + "grad_norm": 0.7220465719306081, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 33285 + }, + { + "epoch": 0.33286, + "grad_norm": 0.7288863140087607, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 33286 + }, + { + "epoch": 0.33287, + "grad_norm": 0.6821704272426864, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 33287 + }, + { + "epoch": 0.33288, + "grad_norm": 0.6892028424351793, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 33288 + }, + { + "epoch": 0.33289, + "grad_norm": 0.6991368657799107, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 33289 + }, + { + "epoch": 0.3329, + "grad_norm": 0.753923244606163, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 33290 + }, + { + "epoch": 0.33291, + "grad_norm": 0.8105293902188, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 33291 + }, + { + "epoch": 0.33292, + "grad_norm": 0.7620700576680113, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 33292 + }, + { + "epoch": 0.33293, + "grad_norm": 0.6627253972740726, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 33293 + }, + { + "epoch": 0.33294, + "grad_norm": 0.5927397228666006, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 33294 + }, + { + "epoch": 0.33295, + "grad_norm": 0.5634456494205975, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 33295 + }, + { + "epoch": 0.33296, + "grad_norm": 0.6176616656308944, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 33296 + }, + { + "epoch": 0.33297, + "grad_norm": 0.6614742620194219, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 33297 + }, + { + "epoch": 0.33298, + "grad_norm": 0.7988552669144953, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 33298 + }, + { + "epoch": 0.33299, + "grad_norm": 0.9987200759412306, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 33299 + }, + { + "epoch": 0.333, + "grad_norm": 1.0183342615339381, + "learning_rate": 0.003, + "loss": 3.975, + "step": 33300 + }, + { + "epoch": 0.33301, + "grad_norm": 0.7060160640653994, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 33301 + }, + { + "epoch": 0.33302, + "grad_norm": 0.6717623962629373, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 33302 + }, + { + "epoch": 0.33303, + "grad_norm": 0.6974258863119814, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 33303 + }, + { + "epoch": 0.33304, + "grad_norm": 0.7908366844700394, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 33304 + }, + { + "epoch": 0.33305, + "grad_norm": 0.7720016717691444, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 33305 + }, + { + "epoch": 0.33306, + "grad_norm": 0.7875330473149358, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 33306 + }, + { + "epoch": 0.33307, + "grad_norm": 0.8271515420580527, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 33307 + }, + { + "epoch": 0.33308, + "grad_norm": 0.7312063634099442, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 33308 + }, + { + "epoch": 0.33309, + "grad_norm": 0.8332291584403306, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 33309 + }, + { + "epoch": 0.3331, + "grad_norm": 0.9430116093243258, + "learning_rate": 0.003, + "loss": 4.0433, + "step": 33310 + }, + { + "epoch": 0.33311, + "grad_norm": 1.0767864830433764, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 33311 + }, + { + "epoch": 0.33312, + "grad_norm": 1.089704456613664, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 33312 + }, + { + "epoch": 0.33313, + "grad_norm": 0.9661514602865802, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 33313 + }, + { + "epoch": 0.33314, + "grad_norm": 0.820210461299899, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 33314 + }, + { + "epoch": 0.33315, + "grad_norm": 0.8231170160188902, + "learning_rate": 0.003, + "loss": 3.971, + "step": 33315 + }, + { + "epoch": 0.33316, + "grad_norm": 0.8544494696822411, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 33316 + }, + { + "epoch": 0.33317, + "grad_norm": 0.9527051202528556, + "learning_rate": 0.003, + "loss": 3.969, + "step": 33317 + }, + { + "epoch": 0.33318, + "grad_norm": 1.0074299393321782, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 33318 + }, + { + "epoch": 0.33319, + "grad_norm": 0.9444146230899901, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 33319 + }, + { + "epoch": 0.3332, + "grad_norm": 0.916944056039486, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 33320 + }, + { + "epoch": 0.33321, + "grad_norm": 0.8422577911009209, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 33321 + }, + { + "epoch": 0.33322, + "grad_norm": 0.879529550866377, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 33322 + }, + { + "epoch": 0.33323, + "grad_norm": 0.9096739487042941, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 33323 + }, + { + "epoch": 0.33324, + "grad_norm": 0.8090209154844916, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 33324 + }, + { + "epoch": 0.33325, + "grad_norm": 0.709239774417721, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 33325 + }, + { + "epoch": 0.33326, + "grad_norm": 0.6772873310885107, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 33326 + }, + { + "epoch": 0.33327, + "grad_norm": 0.694400846428523, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 33327 + }, + { + "epoch": 0.33328, + "grad_norm": 0.6487395035152479, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 33328 + }, + { + "epoch": 0.33329, + "grad_norm": 0.6403509081456649, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 33329 + }, + { + "epoch": 0.3333, + "grad_norm": 0.6301479691811448, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 33330 + }, + { + "epoch": 0.33331, + "grad_norm": 0.6369463096145872, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 33331 + }, + { + "epoch": 0.33332, + "grad_norm": 0.7342366144599565, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 33332 + }, + { + "epoch": 0.33333, + "grad_norm": 0.7448651781366991, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 33333 + }, + { + "epoch": 0.33334, + "grad_norm": 0.7680682054647752, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 33334 + }, + { + "epoch": 0.33335, + "grad_norm": 0.7525535228141139, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 33335 + }, + { + "epoch": 0.33336, + "grad_norm": 0.7641301232563465, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 33336 + }, + { + "epoch": 0.33337, + "grad_norm": 0.7939687733203786, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 33337 + }, + { + "epoch": 0.33338, + "grad_norm": 0.7048057029537325, + "learning_rate": 0.003, + "loss": 3.975, + "step": 33338 + }, + { + "epoch": 0.33339, + "grad_norm": 0.8005269505577021, + "learning_rate": 0.003, + "loss": 3.987, + "step": 33339 + }, + { + "epoch": 0.3334, + "grad_norm": 0.8685440017143388, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 33340 + }, + { + "epoch": 0.33341, + "grad_norm": 0.8206766601771464, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 33341 + }, + { + "epoch": 0.33342, + "grad_norm": 0.65062333665852, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 33342 + }, + { + "epoch": 0.33343, + "grad_norm": 0.5555598677001569, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 33343 + }, + { + "epoch": 0.33344, + "grad_norm": 0.5237611296667347, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 33344 + }, + { + "epoch": 0.33345, + "grad_norm": 0.4709908124113265, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 33345 + }, + { + "epoch": 0.33346, + "grad_norm": 0.496514399283154, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 33346 + }, + { + "epoch": 0.33347, + "grad_norm": 0.5985923069115427, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 33347 + }, + { + "epoch": 0.33348, + "grad_norm": 0.7261746395524676, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 33348 + }, + { + "epoch": 0.33349, + "grad_norm": 0.8455338185543173, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 33349 + }, + { + "epoch": 0.3335, + "grad_norm": 0.9580205257866269, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 33350 + }, + { + "epoch": 0.33351, + "grad_norm": 1.0192371597180923, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 33351 + }, + { + "epoch": 0.33352, + "grad_norm": 0.741807673803933, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 33352 + }, + { + "epoch": 0.33353, + "grad_norm": 0.8300153674394402, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 33353 + }, + { + "epoch": 0.33354, + "grad_norm": 0.9633164174570505, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 33354 + }, + { + "epoch": 0.33355, + "grad_norm": 0.8719395015846299, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 33355 + }, + { + "epoch": 0.33356, + "grad_norm": 0.827571808325051, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 33356 + }, + { + "epoch": 0.33357, + "grad_norm": 0.8734926634955492, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 33357 + }, + { + "epoch": 0.33358, + "grad_norm": 0.7931871723970322, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 33358 + }, + { + "epoch": 0.33359, + "grad_norm": 0.8693686410223074, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 33359 + }, + { + "epoch": 0.3336, + "grad_norm": 0.8724737729053846, + "learning_rate": 0.003, + "loss": 4.0352, + "step": 33360 + }, + { + "epoch": 0.33361, + "grad_norm": 0.9108435712066858, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 33361 + }, + { + "epoch": 0.33362, + "grad_norm": 1.0016862265930746, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 33362 + }, + { + "epoch": 0.33363, + "grad_norm": 1.1086216742951625, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 33363 + }, + { + "epoch": 0.33364, + "grad_norm": 0.7805858809257478, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 33364 + }, + { + "epoch": 0.33365, + "grad_norm": 0.876074145909678, + "learning_rate": 0.003, + "loss": 3.979, + "step": 33365 + }, + { + "epoch": 0.33366, + "grad_norm": 0.8654177297517543, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 33366 + }, + { + "epoch": 0.33367, + "grad_norm": 0.8415815460418498, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 33367 + }, + { + "epoch": 0.33368, + "grad_norm": 1.0226111282265258, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 33368 + }, + { + "epoch": 0.33369, + "grad_norm": 0.9666420213226399, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 33369 + }, + { + "epoch": 0.3337, + "grad_norm": 0.8081532502803501, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 33370 + }, + { + "epoch": 0.33371, + "grad_norm": 0.6080101195230717, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 33371 + }, + { + "epoch": 0.33372, + "grad_norm": 0.5447309256865952, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 33372 + }, + { + "epoch": 0.33373, + "grad_norm": 0.548125346804314, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 33373 + }, + { + "epoch": 0.33374, + "grad_norm": 0.6110122848698442, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 33374 + }, + { + "epoch": 0.33375, + "grad_norm": 0.70341029025306, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 33375 + }, + { + "epoch": 0.33376, + "grad_norm": 0.7589187238322244, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 33376 + }, + { + "epoch": 0.33377, + "grad_norm": 0.7696978569213867, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 33377 + }, + { + "epoch": 0.33378, + "grad_norm": 0.756327744702586, + "learning_rate": 0.003, + "loss": 3.992, + "step": 33378 + }, + { + "epoch": 0.33379, + "grad_norm": 0.6477750357517559, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 33379 + }, + { + "epoch": 0.3338, + "grad_norm": 0.6335688699479947, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 33380 + }, + { + "epoch": 0.33381, + "grad_norm": 0.6450097227791179, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 33381 + }, + { + "epoch": 0.33382, + "grad_norm": 0.6946092771120719, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 33382 + }, + { + "epoch": 0.33383, + "grad_norm": 0.7520344187846716, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 33383 + }, + { + "epoch": 0.33384, + "grad_norm": 0.6951293776146402, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 33384 + }, + { + "epoch": 0.33385, + "grad_norm": 0.6566479489181772, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 33385 + }, + { + "epoch": 0.33386, + "grad_norm": 0.639984483247555, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 33386 + }, + { + "epoch": 0.33387, + "grad_norm": 0.6624703075333852, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 33387 + }, + { + "epoch": 0.33388, + "grad_norm": 0.6658992845582978, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 33388 + }, + { + "epoch": 0.33389, + "grad_norm": 0.6608046223693048, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 33389 + }, + { + "epoch": 0.3339, + "grad_norm": 0.686643403805201, + "learning_rate": 0.003, + "loss": 3.994, + "step": 33390 + }, + { + "epoch": 0.33391, + "grad_norm": 0.6277036854472606, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 33391 + }, + { + "epoch": 0.33392, + "grad_norm": 0.5909477377057012, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 33392 + }, + { + "epoch": 0.33393, + "grad_norm": 0.6762310054150892, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 33393 + }, + { + "epoch": 0.33394, + "grad_norm": 0.7942089091010567, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 33394 + }, + { + "epoch": 0.33395, + "grad_norm": 0.8778917287635393, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 33395 + }, + { + "epoch": 0.33396, + "grad_norm": 0.943293602944069, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 33396 + }, + { + "epoch": 0.33397, + "grad_norm": 0.9597365508693335, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 33397 + }, + { + "epoch": 0.33398, + "grad_norm": 0.8526594230744542, + "learning_rate": 0.003, + "loss": 3.969, + "step": 33398 + }, + { + "epoch": 0.33399, + "grad_norm": 0.8207087321735091, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 33399 + }, + { + "epoch": 0.334, + "grad_norm": 0.8961331903841344, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 33400 + }, + { + "epoch": 0.33401, + "grad_norm": 0.9936471313215837, + "learning_rate": 0.003, + "loss": 4.0332, + "step": 33401 + }, + { + "epoch": 0.33402, + "grad_norm": 0.9151602479675194, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 33402 + }, + { + "epoch": 0.33403, + "grad_norm": 0.7833474781692247, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 33403 + }, + { + "epoch": 0.33404, + "grad_norm": 0.7859385225530087, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 33404 + }, + { + "epoch": 0.33405, + "grad_norm": 0.8501417380415625, + "learning_rate": 0.003, + "loss": 3.987, + "step": 33405 + }, + { + "epoch": 0.33406, + "grad_norm": 0.875309707703512, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 33406 + }, + { + "epoch": 0.33407, + "grad_norm": 1.0392062004709068, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 33407 + }, + { + "epoch": 0.33408, + "grad_norm": 0.8972817605540337, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 33408 + }, + { + "epoch": 0.33409, + "grad_norm": 0.8008965884579123, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 33409 + }, + { + "epoch": 0.3341, + "grad_norm": 0.8329919515206095, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 33410 + }, + { + "epoch": 0.33411, + "grad_norm": 0.8364484168432906, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 33411 + }, + { + "epoch": 0.33412, + "grad_norm": 0.7695956791122299, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 33412 + }, + { + "epoch": 0.33413, + "grad_norm": 0.768545332265812, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 33413 + }, + { + "epoch": 0.33414, + "grad_norm": 0.8509063300483094, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 33414 + }, + { + "epoch": 0.33415, + "grad_norm": 0.9210764526099763, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 33415 + }, + { + "epoch": 0.33416, + "grad_norm": 0.872754460525427, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 33416 + }, + { + "epoch": 0.33417, + "grad_norm": 0.8562412897587873, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 33417 + }, + { + "epoch": 0.33418, + "grad_norm": 0.8736150297530427, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 33418 + }, + { + "epoch": 0.33419, + "grad_norm": 0.8066395472469151, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 33419 + }, + { + "epoch": 0.3342, + "grad_norm": 0.788594309941608, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 33420 + }, + { + "epoch": 0.33421, + "grad_norm": 0.7576674378812908, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 33421 + }, + { + "epoch": 0.33422, + "grad_norm": 0.6577461826442177, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 33422 + }, + { + "epoch": 0.33423, + "grad_norm": 0.5222217842729269, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 33423 + }, + { + "epoch": 0.33424, + "grad_norm": 0.5669992279260443, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 33424 + }, + { + "epoch": 0.33425, + "grad_norm": 0.612723222240785, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 33425 + }, + { + "epoch": 0.33426, + "grad_norm": 0.7516595553878264, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 33426 + }, + { + "epoch": 0.33427, + "grad_norm": 0.7887831350777331, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 33427 + }, + { + "epoch": 0.33428, + "grad_norm": 0.7711579989438189, + "learning_rate": 0.003, + "loss": 3.978, + "step": 33428 + }, + { + "epoch": 0.33429, + "grad_norm": 0.7520477856612278, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 33429 + }, + { + "epoch": 0.3343, + "grad_norm": 0.7344427684981555, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 33430 + }, + { + "epoch": 0.33431, + "grad_norm": 0.7687925208901221, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 33431 + }, + { + "epoch": 0.33432, + "grad_norm": 1.0333675285496633, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 33432 + }, + { + "epoch": 0.33433, + "grad_norm": 1.154082025923331, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 33433 + }, + { + "epoch": 0.33434, + "grad_norm": 0.6185126189122564, + "learning_rate": 0.003, + "loss": 3.989, + "step": 33434 + }, + { + "epoch": 0.33435, + "grad_norm": 0.6349147063083439, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 33435 + }, + { + "epoch": 0.33436, + "grad_norm": 0.7593343149731988, + "learning_rate": 0.003, + "loss": 3.998, + "step": 33436 + }, + { + "epoch": 0.33437, + "grad_norm": 0.8051079376348318, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 33437 + }, + { + "epoch": 0.33438, + "grad_norm": 0.872622161397441, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 33438 + }, + { + "epoch": 0.33439, + "grad_norm": 0.851726859968353, + "learning_rate": 0.003, + "loss": 3.978, + "step": 33439 + }, + { + "epoch": 0.3344, + "grad_norm": 0.746266901700009, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 33440 + }, + { + "epoch": 0.33441, + "grad_norm": 0.7364925950814779, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 33441 + }, + { + "epoch": 0.33442, + "grad_norm": 0.8629447002828757, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 33442 + }, + { + "epoch": 0.33443, + "grad_norm": 0.7955542904055274, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 33443 + }, + { + "epoch": 0.33444, + "grad_norm": 0.8634148749957642, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 33444 + }, + { + "epoch": 0.33445, + "grad_norm": 0.8888000537236965, + "learning_rate": 0.003, + "loss": 3.993, + "step": 33445 + }, + { + "epoch": 0.33446, + "grad_norm": 0.8595697339418874, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 33446 + }, + { + "epoch": 0.33447, + "grad_norm": 0.9178640219492382, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 33447 + }, + { + "epoch": 0.33448, + "grad_norm": 0.9420775896455555, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 33448 + }, + { + "epoch": 0.33449, + "grad_norm": 0.7795425766611449, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 33449 + }, + { + "epoch": 0.3345, + "grad_norm": 0.7727874947004291, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 33450 + }, + { + "epoch": 0.33451, + "grad_norm": 0.7478845717413845, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 33451 + }, + { + "epoch": 0.33452, + "grad_norm": 0.6441826806187052, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 33452 + }, + { + "epoch": 0.33453, + "grad_norm": 0.6593406637519617, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 33453 + }, + { + "epoch": 0.33454, + "grad_norm": 0.7078968632056589, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 33454 + }, + { + "epoch": 0.33455, + "grad_norm": 0.7966359058913215, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 33455 + }, + { + "epoch": 0.33456, + "grad_norm": 0.8900316463403269, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 33456 + }, + { + "epoch": 0.33457, + "grad_norm": 0.8761939399890278, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 33457 + }, + { + "epoch": 0.33458, + "grad_norm": 0.8637189036479411, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 33458 + }, + { + "epoch": 0.33459, + "grad_norm": 0.7713606412931694, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 33459 + }, + { + "epoch": 0.3346, + "grad_norm": 0.7521872509675822, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 33460 + }, + { + "epoch": 0.33461, + "grad_norm": 0.6947250506650834, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 33461 + }, + { + "epoch": 0.33462, + "grad_norm": 0.7843411512955639, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 33462 + }, + { + "epoch": 0.33463, + "grad_norm": 0.7963691166796438, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 33463 + }, + { + "epoch": 0.33464, + "grad_norm": 0.7988258334330762, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 33464 + }, + { + "epoch": 0.33465, + "grad_norm": 0.8280608070890821, + "learning_rate": 0.003, + "loss": 3.996, + "step": 33465 + }, + { + "epoch": 0.33466, + "grad_norm": 0.8271955430760831, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 33466 + }, + { + "epoch": 0.33467, + "grad_norm": 0.9728311059588309, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 33467 + }, + { + "epoch": 0.33468, + "grad_norm": 1.0142743131183518, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 33468 + }, + { + "epoch": 0.33469, + "grad_norm": 0.913340944069229, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 33469 + }, + { + "epoch": 0.3347, + "grad_norm": 0.8102616039327969, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 33470 + }, + { + "epoch": 0.33471, + "grad_norm": 0.7660612011474013, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 33471 + }, + { + "epoch": 0.33472, + "grad_norm": 0.7157882117539357, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 33472 + }, + { + "epoch": 0.33473, + "grad_norm": 0.6442225476759049, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 33473 + }, + { + "epoch": 0.33474, + "grad_norm": 0.6013590596644791, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 33474 + }, + { + "epoch": 0.33475, + "grad_norm": 0.6603819669760107, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 33475 + }, + { + "epoch": 0.33476, + "grad_norm": 0.7538478805602378, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 33476 + }, + { + "epoch": 0.33477, + "grad_norm": 0.8070075285727636, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 33477 + }, + { + "epoch": 0.33478, + "grad_norm": 0.9128685973527001, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 33478 + }, + { + "epoch": 0.33479, + "grad_norm": 0.9695415401450634, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 33479 + }, + { + "epoch": 0.3348, + "grad_norm": 0.8555537013836366, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 33480 + }, + { + "epoch": 0.33481, + "grad_norm": 0.8452097483971536, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 33481 + }, + { + "epoch": 0.33482, + "grad_norm": 0.8584182354735395, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 33482 + }, + { + "epoch": 0.33483, + "grad_norm": 0.6649335794543317, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 33483 + }, + { + "epoch": 0.33484, + "grad_norm": 0.7003702235708922, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 33484 + }, + { + "epoch": 0.33485, + "grad_norm": 0.6466961918872908, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 33485 + }, + { + "epoch": 0.33486, + "grad_norm": 0.5774694357771835, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 33486 + }, + { + "epoch": 0.33487, + "grad_norm": 0.4699672266026504, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 33487 + }, + { + "epoch": 0.33488, + "grad_norm": 0.46180387374022847, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 33488 + }, + { + "epoch": 0.33489, + "grad_norm": 0.45096682972082425, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 33489 + }, + { + "epoch": 0.3349, + "grad_norm": 0.48356117468450643, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 33490 + }, + { + "epoch": 0.33491, + "grad_norm": 0.5455944378791592, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 33491 + }, + { + "epoch": 0.33492, + "grad_norm": 0.6419513799880926, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 33492 + }, + { + "epoch": 0.33493, + "grad_norm": 0.7696628956904631, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 33493 + }, + { + "epoch": 0.33494, + "grad_norm": 1.0112313213961643, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 33494 + }, + { + "epoch": 0.33495, + "grad_norm": 1.073073647915534, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 33495 + }, + { + "epoch": 0.33496, + "grad_norm": 0.8270615586998828, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 33496 + }, + { + "epoch": 0.33497, + "grad_norm": 0.7594739041715237, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 33497 + }, + { + "epoch": 0.33498, + "grad_norm": 0.7740272924273907, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 33498 + }, + { + "epoch": 0.33499, + "grad_norm": 0.8308325456592837, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 33499 + }, + { + "epoch": 0.335, + "grad_norm": 0.8501729234731347, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 33500 + }, + { + "epoch": 0.33501, + "grad_norm": 0.8498030011295786, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 33501 + }, + { + "epoch": 0.33502, + "grad_norm": 1.029766549000508, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 33502 + }, + { + "epoch": 0.33503, + "grad_norm": 1.0497660780001805, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 33503 + }, + { + "epoch": 0.33504, + "grad_norm": 0.9611967725845255, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 33504 + }, + { + "epoch": 0.33505, + "grad_norm": 0.9551252998765479, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 33505 + }, + { + "epoch": 0.33506, + "grad_norm": 0.9092778442301125, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 33506 + }, + { + "epoch": 0.33507, + "grad_norm": 0.815782955131056, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 33507 + }, + { + "epoch": 0.33508, + "grad_norm": 0.8801320628478314, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 33508 + }, + { + "epoch": 0.33509, + "grad_norm": 0.7881502608436183, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 33509 + }, + { + "epoch": 0.3351, + "grad_norm": 0.7109952917867544, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 33510 + }, + { + "epoch": 0.33511, + "grad_norm": 0.6824641835911958, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 33511 + }, + { + "epoch": 0.33512, + "grad_norm": 0.6229603434706246, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 33512 + }, + { + "epoch": 0.33513, + "grad_norm": 0.6742282230576788, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 33513 + }, + { + "epoch": 0.33514, + "grad_norm": 0.8402341113408649, + "learning_rate": 0.003, + "loss": 4.005, + "step": 33514 + }, + { + "epoch": 0.33515, + "grad_norm": 1.002052606551875, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 33515 + }, + { + "epoch": 0.33516, + "grad_norm": 1.1270543494624434, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 33516 + }, + { + "epoch": 0.33517, + "grad_norm": 0.9130317431032554, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 33517 + }, + { + "epoch": 0.33518, + "grad_norm": 0.938508050202798, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 33518 + }, + { + "epoch": 0.33519, + "grad_norm": 0.9562499957683066, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 33519 + }, + { + "epoch": 0.3352, + "grad_norm": 0.9835734830222058, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 33520 + }, + { + "epoch": 0.33521, + "grad_norm": 0.9535684402736997, + "learning_rate": 0.003, + "loss": 4.014, + "step": 33521 + }, + { + "epoch": 0.33522, + "grad_norm": 0.7370417563877537, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 33522 + }, + { + "epoch": 0.33523, + "grad_norm": 0.6559058056855313, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 33523 + }, + { + "epoch": 0.33524, + "grad_norm": 0.6059160219414161, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 33524 + }, + { + "epoch": 0.33525, + "grad_norm": 0.6154685546152311, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 33525 + }, + { + "epoch": 0.33526, + "grad_norm": 0.6271760584397947, + "learning_rate": 0.003, + "loss": 3.995, + "step": 33526 + }, + { + "epoch": 0.33527, + "grad_norm": 0.6033221808121506, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 33527 + }, + { + "epoch": 0.33528, + "grad_norm": 0.6025190887653719, + "learning_rate": 0.003, + "loss": 3.992, + "step": 33528 + }, + { + "epoch": 0.33529, + "grad_norm": 0.6505341630835115, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 33529 + }, + { + "epoch": 0.3353, + "grad_norm": 0.6709447462085464, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 33530 + }, + { + "epoch": 0.33531, + "grad_norm": 0.8457388875567553, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 33531 + }, + { + "epoch": 0.33532, + "grad_norm": 1.1002603421924626, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 33532 + }, + { + "epoch": 0.33533, + "grad_norm": 0.9625056461282303, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 33533 + }, + { + "epoch": 0.33534, + "grad_norm": 0.9875558788881231, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 33534 + }, + { + "epoch": 0.33535, + "grad_norm": 0.7798084712064736, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 33535 + }, + { + "epoch": 0.33536, + "grad_norm": 0.6597985147452244, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 33536 + }, + { + "epoch": 0.33537, + "grad_norm": 0.7166061646713058, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 33537 + }, + { + "epoch": 0.33538, + "grad_norm": 0.7214439464094405, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 33538 + }, + { + "epoch": 0.33539, + "grad_norm": 0.6562148882593254, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 33539 + }, + { + "epoch": 0.3354, + "grad_norm": 0.6653645140461291, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 33540 + }, + { + "epoch": 0.33541, + "grad_norm": 0.6226353364397047, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 33541 + }, + { + "epoch": 0.33542, + "grad_norm": 0.558588712705203, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 33542 + }, + { + "epoch": 0.33543, + "grad_norm": 0.5128692077204258, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 33543 + }, + { + "epoch": 0.33544, + "grad_norm": 0.5890241088541315, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 33544 + }, + { + "epoch": 0.33545, + "grad_norm": 0.5870937069857444, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 33545 + }, + { + "epoch": 0.33546, + "grad_norm": 0.5533030049635697, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 33546 + }, + { + "epoch": 0.33547, + "grad_norm": 0.5633951444067933, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 33547 + }, + { + "epoch": 0.33548, + "grad_norm": 0.6552526711532735, + "learning_rate": 0.003, + "loss": 3.97, + "step": 33548 + }, + { + "epoch": 0.33549, + "grad_norm": 0.7034810549896099, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 33549 + }, + { + "epoch": 0.3355, + "grad_norm": 0.6298518134598897, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 33550 + }, + { + "epoch": 0.33551, + "grad_norm": 0.6392307268998529, + "learning_rate": 0.003, + "loss": 3.968, + "step": 33551 + }, + { + "epoch": 0.33552, + "grad_norm": 0.7919433812507498, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 33552 + }, + { + "epoch": 0.33553, + "grad_norm": 0.8593819239834136, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 33553 + }, + { + "epoch": 0.33554, + "grad_norm": 0.8994681453023957, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 33554 + }, + { + "epoch": 0.33555, + "grad_norm": 0.825050995747293, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 33555 + }, + { + "epoch": 0.33556, + "grad_norm": 0.7028933280875055, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 33556 + }, + { + "epoch": 0.33557, + "grad_norm": 0.6395750076451303, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 33557 + }, + { + "epoch": 0.33558, + "grad_norm": 0.6652266720145046, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 33558 + }, + { + "epoch": 0.33559, + "grad_norm": 0.8377404649058883, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 33559 + }, + { + "epoch": 0.3356, + "grad_norm": 0.8690386958560038, + "learning_rate": 0.003, + "loss": 4.006, + "step": 33560 + }, + { + "epoch": 0.33561, + "grad_norm": 1.000352474857566, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 33561 + }, + { + "epoch": 0.33562, + "grad_norm": 0.9601239893944286, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 33562 + }, + { + "epoch": 0.33563, + "grad_norm": 0.7972771933410638, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 33563 + }, + { + "epoch": 0.33564, + "grad_norm": 0.6741156768711467, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 33564 + }, + { + "epoch": 0.33565, + "grad_norm": 0.5927941643881357, + "learning_rate": 0.003, + "loss": 3.976, + "step": 33565 + }, + { + "epoch": 0.33566, + "grad_norm": 0.5398926313176862, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 33566 + }, + { + "epoch": 0.33567, + "grad_norm": 0.5127262961782355, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 33567 + }, + { + "epoch": 0.33568, + "grad_norm": 0.5499141936961487, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 33568 + }, + { + "epoch": 0.33569, + "grad_norm": 0.5720258304253508, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 33569 + }, + { + "epoch": 0.3357, + "grad_norm": 0.623547487312547, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 33570 + }, + { + "epoch": 0.33571, + "grad_norm": 0.7115805392948182, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 33571 + }, + { + "epoch": 0.33572, + "grad_norm": 0.7871958732020214, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 33572 + }, + { + "epoch": 0.33573, + "grad_norm": 0.7636535010447562, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 33573 + }, + { + "epoch": 0.33574, + "grad_norm": 0.7752227039350121, + "learning_rate": 0.003, + "loss": 3.987, + "step": 33574 + }, + { + "epoch": 0.33575, + "grad_norm": 0.7890890875739599, + "learning_rate": 0.003, + "loss": 3.995, + "step": 33575 + }, + { + "epoch": 0.33576, + "grad_norm": 0.7240201333056064, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 33576 + }, + { + "epoch": 0.33577, + "grad_norm": 1.01581550447934, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 33577 + }, + { + "epoch": 0.33578, + "grad_norm": 1.1327339447254117, + "learning_rate": 0.003, + "loss": 4.0489, + "step": 33578 + }, + { + "epoch": 0.33579, + "grad_norm": 0.8085202082705915, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 33579 + }, + { + "epoch": 0.3358, + "grad_norm": 0.7335945976355239, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 33580 + }, + { + "epoch": 0.33581, + "grad_norm": 0.7750465758082072, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 33581 + }, + { + "epoch": 0.33582, + "grad_norm": 0.8500303595628221, + "learning_rate": 0.003, + "loss": 4.013, + "step": 33582 + }, + { + "epoch": 0.33583, + "grad_norm": 0.8534257181328379, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 33583 + }, + { + "epoch": 0.33584, + "grad_norm": 1.153011260189563, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 33584 + }, + { + "epoch": 0.33585, + "grad_norm": 1.0677400666426535, + "learning_rate": 0.003, + "loss": 3.96, + "step": 33585 + }, + { + "epoch": 0.33586, + "grad_norm": 0.9215942913770053, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 33586 + }, + { + "epoch": 0.33587, + "grad_norm": 0.8582917787901244, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 33587 + }, + { + "epoch": 0.33588, + "grad_norm": 0.6942730180871068, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 33588 + }, + { + "epoch": 0.33589, + "grad_norm": 0.7048554011240629, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 33589 + }, + { + "epoch": 0.3359, + "grad_norm": 0.7854323156036319, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 33590 + }, + { + "epoch": 0.33591, + "grad_norm": 0.8970120729644949, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 33591 + }, + { + "epoch": 0.33592, + "grad_norm": 0.8637283446785131, + "learning_rate": 0.003, + "loss": 3.99, + "step": 33592 + }, + { + "epoch": 0.33593, + "grad_norm": 0.9197991892353133, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 33593 + }, + { + "epoch": 0.33594, + "grad_norm": 1.039555354952665, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 33594 + }, + { + "epoch": 0.33595, + "grad_norm": 0.9642206593176849, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 33595 + }, + { + "epoch": 0.33596, + "grad_norm": 0.901565008470563, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 33596 + }, + { + "epoch": 0.33597, + "grad_norm": 0.9537479110248879, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 33597 + }, + { + "epoch": 0.33598, + "grad_norm": 0.9175572163285055, + "learning_rate": 0.003, + "loss": 3.999, + "step": 33598 + }, + { + "epoch": 0.33599, + "grad_norm": 0.8452731857243038, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 33599 + }, + { + "epoch": 0.336, + "grad_norm": 0.7420105823669453, + "learning_rate": 0.003, + "loss": 4.006, + "step": 33600 + }, + { + "epoch": 0.33601, + "grad_norm": 0.6931556049952425, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 33601 + }, + { + "epoch": 0.33602, + "grad_norm": 0.6922549537654166, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 33602 + }, + { + "epoch": 0.33603, + "grad_norm": 0.7465051681495524, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 33603 + }, + { + "epoch": 0.33604, + "grad_norm": 0.7936206573906655, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 33604 + }, + { + "epoch": 0.33605, + "grad_norm": 0.8327344064138966, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 33605 + }, + { + "epoch": 0.33606, + "grad_norm": 0.825012180301588, + "learning_rate": 0.003, + "loss": 4.009, + "step": 33606 + }, + { + "epoch": 0.33607, + "grad_norm": 0.8490283312394542, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 33607 + }, + { + "epoch": 0.33608, + "grad_norm": 0.7355910026275033, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 33608 + }, + { + "epoch": 0.33609, + "grad_norm": 0.6525637234372414, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 33609 + }, + { + "epoch": 0.3361, + "grad_norm": 0.5846660271716403, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 33610 + }, + { + "epoch": 0.33611, + "grad_norm": 0.6016644221138858, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 33611 + }, + { + "epoch": 0.33612, + "grad_norm": 0.5920118263631005, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 33612 + }, + { + "epoch": 0.33613, + "grad_norm": 0.5961932545047682, + "learning_rate": 0.003, + "loss": 3.991, + "step": 33613 + }, + { + "epoch": 0.33614, + "grad_norm": 0.6517363267782237, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 33614 + }, + { + "epoch": 0.33615, + "grad_norm": 0.6832987558132714, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 33615 + }, + { + "epoch": 0.33616, + "grad_norm": 0.7181808404795543, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 33616 + }, + { + "epoch": 0.33617, + "grad_norm": 0.8081437353069952, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 33617 + }, + { + "epoch": 0.33618, + "grad_norm": 0.6995283913426225, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 33618 + }, + { + "epoch": 0.33619, + "grad_norm": 0.7507604444082118, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 33619 + }, + { + "epoch": 0.3362, + "grad_norm": 0.8561912678528447, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 33620 + }, + { + "epoch": 0.33621, + "grad_norm": 0.9962586691329925, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 33621 + }, + { + "epoch": 0.33622, + "grad_norm": 1.2228454898958465, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 33622 + }, + { + "epoch": 0.33623, + "grad_norm": 0.777578724248978, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 33623 + }, + { + "epoch": 0.33624, + "grad_norm": 0.7796837719921059, + "learning_rate": 0.003, + "loss": 3.978, + "step": 33624 + }, + { + "epoch": 0.33625, + "grad_norm": 0.7496136631432694, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 33625 + }, + { + "epoch": 0.33626, + "grad_norm": 0.7224852565910445, + "learning_rate": 0.003, + "loss": 3.978, + "step": 33626 + }, + { + "epoch": 0.33627, + "grad_norm": 0.7869867572774111, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 33627 + }, + { + "epoch": 0.33628, + "grad_norm": 0.8074823235096419, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 33628 + }, + { + "epoch": 0.33629, + "grad_norm": 0.7809204631077574, + "learning_rate": 0.003, + "loss": 3.989, + "step": 33629 + }, + { + "epoch": 0.3363, + "grad_norm": 0.7843917493380893, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 33630 + }, + { + "epoch": 0.33631, + "grad_norm": 0.8309817372524265, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 33631 + }, + { + "epoch": 0.33632, + "grad_norm": 0.9617817543100508, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 33632 + }, + { + "epoch": 0.33633, + "grad_norm": 0.9928702378110471, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 33633 + }, + { + "epoch": 0.33634, + "grad_norm": 0.9144093082735797, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 33634 + }, + { + "epoch": 0.33635, + "grad_norm": 0.9657802548683388, + "learning_rate": 0.003, + "loss": 3.981, + "step": 33635 + }, + { + "epoch": 0.33636, + "grad_norm": 0.9372756735450768, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 33636 + }, + { + "epoch": 0.33637, + "grad_norm": 0.8649407689782636, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 33637 + }, + { + "epoch": 0.33638, + "grad_norm": 0.8980582552291673, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 33638 + }, + { + "epoch": 0.33639, + "grad_norm": 0.881177818815781, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 33639 + }, + { + "epoch": 0.3364, + "grad_norm": 0.7605738910514735, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 33640 + }, + { + "epoch": 0.33641, + "grad_norm": 0.681444628465705, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 33641 + }, + { + "epoch": 0.33642, + "grad_norm": 0.7182672811168369, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 33642 + }, + { + "epoch": 0.33643, + "grad_norm": 0.8406626202247018, + "learning_rate": 0.003, + "loss": 3.97, + "step": 33643 + }, + { + "epoch": 0.33644, + "grad_norm": 0.9047231908659292, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 33644 + }, + { + "epoch": 0.33645, + "grad_norm": 0.9176897632384198, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 33645 + }, + { + "epoch": 0.33646, + "grad_norm": 0.8795298865976076, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 33646 + }, + { + "epoch": 0.33647, + "grad_norm": 0.9207827924377354, + "learning_rate": 0.003, + "loss": 3.992, + "step": 33647 + }, + { + "epoch": 0.33648, + "grad_norm": 0.9519540558815363, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 33648 + }, + { + "epoch": 0.33649, + "grad_norm": 0.7582238882941134, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 33649 + }, + { + "epoch": 0.3365, + "grad_norm": 0.7138576101773915, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 33650 + }, + { + "epoch": 0.33651, + "grad_norm": 0.6492023567373897, + "learning_rate": 0.003, + "loss": 3.98, + "step": 33651 + }, + { + "epoch": 0.33652, + "grad_norm": 0.6740390369750205, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 33652 + }, + { + "epoch": 0.33653, + "grad_norm": 0.655246311652235, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 33653 + }, + { + "epoch": 0.33654, + "grad_norm": 0.7914626733814226, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 33654 + }, + { + "epoch": 0.33655, + "grad_norm": 1.0514422089748663, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 33655 + }, + { + "epoch": 0.33656, + "grad_norm": 1.1994407881060334, + "learning_rate": 0.003, + "loss": 4.014, + "step": 33656 + }, + { + "epoch": 0.33657, + "grad_norm": 0.8072012273427926, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 33657 + }, + { + "epoch": 0.33658, + "grad_norm": 0.9008736617431873, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 33658 + }, + { + "epoch": 0.33659, + "grad_norm": 1.0089604379330699, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 33659 + }, + { + "epoch": 0.3366, + "grad_norm": 1.0622500206628893, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 33660 + }, + { + "epoch": 0.33661, + "grad_norm": 0.8550968240166889, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 33661 + }, + { + "epoch": 0.33662, + "grad_norm": 0.7969889900181886, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 33662 + }, + { + "epoch": 0.33663, + "grad_norm": 0.837574958384487, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 33663 + }, + { + "epoch": 0.33664, + "grad_norm": 0.6697154149231108, + "learning_rate": 0.003, + "loss": 4.009, + "step": 33664 + }, + { + "epoch": 0.33665, + "grad_norm": 0.6058527195833822, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 33665 + }, + { + "epoch": 0.33666, + "grad_norm": 0.6528461565751452, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 33666 + }, + { + "epoch": 0.33667, + "grad_norm": 0.6491043901796164, + "learning_rate": 0.003, + "loss": 3.985, + "step": 33667 + }, + { + "epoch": 0.33668, + "grad_norm": 0.5700823431183046, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 33668 + }, + { + "epoch": 0.33669, + "grad_norm": 0.5867825848961103, + "learning_rate": 0.003, + "loss": 3.981, + "step": 33669 + }, + { + "epoch": 0.3367, + "grad_norm": 0.6904838521867018, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 33670 + }, + { + "epoch": 0.33671, + "grad_norm": 0.7825593734572992, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 33671 + }, + { + "epoch": 0.33672, + "grad_norm": 0.7370403200233678, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 33672 + }, + { + "epoch": 0.33673, + "grad_norm": 0.6152914856063478, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 33673 + }, + { + "epoch": 0.33674, + "grad_norm": 0.6537333149812922, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 33674 + }, + { + "epoch": 0.33675, + "grad_norm": 0.669270983719719, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 33675 + }, + { + "epoch": 0.33676, + "grad_norm": 0.6278013998552558, + "learning_rate": 0.003, + "loss": 4.003, + "step": 33676 + }, + { + "epoch": 0.33677, + "grad_norm": 0.701187534472829, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 33677 + }, + { + "epoch": 0.33678, + "grad_norm": 0.7864948133461899, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 33678 + }, + { + "epoch": 0.33679, + "grad_norm": 0.839131146639385, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 33679 + }, + { + "epoch": 0.3368, + "grad_norm": 0.8445773006706427, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 33680 + }, + { + "epoch": 0.33681, + "grad_norm": 0.9717897236563401, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 33681 + }, + { + "epoch": 0.33682, + "grad_norm": 1.111655361344383, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 33682 + }, + { + "epoch": 0.33683, + "grad_norm": 0.9176433690170097, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 33683 + }, + { + "epoch": 0.33684, + "grad_norm": 0.7160407964682486, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 33684 + }, + { + "epoch": 0.33685, + "grad_norm": 0.7715351042585386, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 33685 + }, + { + "epoch": 0.33686, + "grad_norm": 0.801054394904306, + "learning_rate": 0.003, + "loss": 3.978, + "step": 33686 + }, + { + "epoch": 0.33687, + "grad_norm": 0.9057875937031179, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 33687 + }, + { + "epoch": 0.33688, + "grad_norm": 1.0717549980510173, + "learning_rate": 0.003, + "loss": 4.025, + "step": 33688 + }, + { + "epoch": 0.33689, + "grad_norm": 0.9392762695736924, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 33689 + }, + { + "epoch": 0.3369, + "grad_norm": 0.9125034478134512, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 33690 + }, + { + "epoch": 0.33691, + "grad_norm": 0.8265689073516037, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 33691 + }, + { + "epoch": 0.33692, + "grad_norm": 0.696358803445847, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 33692 + }, + { + "epoch": 0.33693, + "grad_norm": 0.6860439730278369, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 33693 + }, + { + "epoch": 0.33694, + "grad_norm": 0.6389955718796352, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 33694 + }, + { + "epoch": 0.33695, + "grad_norm": 0.6501881929809441, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 33695 + }, + { + "epoch": 0.33696, + "grad_norm": 0.5967772531668006, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 33696 + }, + { + "epoch": 0.33697, + "grad_norm": 0.5854704675344893, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 33697 + }, + { + "epoch": 0.33698, + "grad_norm": 0.4688294182656902, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 33698 + }, + { + "epoch": 0.33699, + "grad_norm": 0.536944519936617, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 33699 + }, + { + "epoch": 0.337, + "grad_norm": 0.5026662774551542, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 33700 + }, + { + "epoch": 0.33701, + "grad_norm": 0.46443585862007136, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 33701 + }, + { + "epoch": 0.33702, + "grad_norm": 0.5074664227684133, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 33702 + }, + { + "epoch": 0.33703, + "grad_norm": 0.49605907512194286, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 33703 + }, + { + "epoch": 0.33704, + "grad_norm": 0.4475592515047952, + "learning_rate": 0.003, + "loss": 3.954, + "step": 33704 + }, + { + "epoch": 0.33705, + "grad_norm": 0.4307549518258386, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 33705 + }, + { + "epoch": 0.33706, + "grad_norm": 0.39993904431047916, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 33706 + }, + { + "epoch": 0.33707, + "grad_norm": 0.4550521738791303, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 33707 + }, + { + "epoch": 0.33708, + "grad_norm": 0.5512765419001475, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 33708 + }, + { + "epoch": 0.33709, + "grad_norm": 0.7034138192480129, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 33709 + }, + { + "epoch": 0.3371, + "grad_norm": 0.853152766486652, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 33710 + }, + { + "epoch": 0.33711, + "grad_norm": 1.081165106250889, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 33711 + }, + { + "epoch": 0.33712, + "grad_norm": 0.9878299117469601, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 33712 + }, + { + "epoch": 0.33713, + "grad_norm": 0.9152672325105637, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 33713 + }, + { + "epoch": 0.33714, + "grad_norm": 0.8349065570762332, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 33714 + }, + { + "epoch": 0.33715, + "grad_norm": 0.9019628036885479, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 33715 + }, + { + "epoch": 0.33716, + "grad_norm": 0.8467274475415633, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 33716 + }, + { + "epoch": 0.33717, + "grad_norm": 0.8625476099415053, + "learning_rate": 0.003, + "loss": 3.983, + "step": 33717 + }, + { + "epoch": 0.33718, + "grad_norm": 0.9003665493752877, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 33718 + }, + { + "epoch": 0.33719, + "grad_norm": 1.056294043826762, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 33719 + }, + { + "epoch": 0.3372, + "grad_norm": 1.0325579228171828, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 33720 + }, + { + "epoch": 0.33721, + "grad_norm": 0.9758610324186762, + "learning_rate": 0.003, + "loss": 4.007, + "step": 33721 + }, + { + "epoch": 0.33722, + "grad_norm": 1.0695892212940863, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 33722 + }, + { + "epoch": 0.33723, + "grad_norm": 0.9311748367592814, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 33723 + }, + { + "epoch": 0.33724, + "grad_norm": 1.0755307037871362, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 33724 + }, + { + "epoch": 0.33725, + "grad_norm": 0.9892019185513795, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 33725 + }, + { + "epoch": 0.33726, + "grad_norm": 0.9904527949465693, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 33726 + }, + { + "epoch": 0.33727, + "grad_norm": 0.9454186793220298, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 33727 + }, + { + "epoch": 0.33728, + "grad_norm": 0.8989455180959215, + "learning_rate": 0.003, + "loss": 4.004, + "step": 33728 + }, + { + "epoch": 0.33729, + "grad_norm": 0.897983889293674, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 33729 + }, + { + "epoch": 0.3373, + "grad_norm": 1.0318353154515307, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 33730 + }, + { + "epoch": 0.33731, + "grad_norm": 0.9942336253023509, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 33731 + }, + { + "epoch": 0.33732, + "grad_norm": 0.8119332107309263, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 33732 + }, + { + "epoch": 0.33733, + "grad_norm": 0.8658151865437123, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 33733 + }, + { + "epoch": 0.33734, + "grad_norm": 0.8873919559153474, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 33734 + }, + { + "epoch": 0.33735, + "grad_norm": 0.9371384042211308, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 33735 + }, + { + "epoch": 0.33736, + "grad_norm": 0.8850480554648705, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 33736 + }, + { + "epoch": 0.33737, + "grad_norm": 0.8948052947969544, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 33737 + }, + { + "epoch": 0.33738, + "grad_norm": 0.857813963507418, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 33738 + }, + { + "epoch": 0.33739, + "grad_norm": 0.844246131167956, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 33739 + }, + { + "epoch": 0.3374, + "grad_norm": 0.7994906258716141, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 33740 + }, + { + "epoch": 0.33741, + "grad_norm": 0.788304452222258, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 33741 + }, + { + "epoch": 0.33742, + "grad_norm": 0.7554575077762605, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 33742 + }, + { + "epoch": 0.33743, + "grad_norm": 0.8005322973636292, + "learning_rate": 0.003, + "loss": 3.963, + "step": 33743 + }, + { + "epoch": 0.33744, + "grad_norm": 0.8095971452485761, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 33744 + }, + { + "epoch": 0.33745, + "grad_norm": 0.8527709608372662, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 33745 + }, + { + "epoch": 0.33746, + "grad_norm": 0.819302998100733, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 33746 + }, + { + "epoch": 0.33747, + "grad_norm": 0.8046928275081865, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 33747 + }, + { + "epoch": 0.33748, + "grad_norm": 0.7150855298876738, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 33748 + }, + { + "epoch": 0.33749, + "grad_norm": 0.6537729556129859, + "learning_rate": 0.003, + "loss": 3.9483, + "step": 33749 + }, + { + "epoch": 0.3375, + "grad_norm": 0.6640002960893957, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 33750 + }, + { + "epoch": 0.33751, + "grad_norm": 0.5539755533106645, + "learning_rate": 0.003, + "loss": 4.0324, + "step": 33751 + }, + { + "epoch": 0.33752, + "grad_norm": 0.6148065439303024, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 33752 + }, + { + "epoch": 0.33753, + "grad_norm": 0.6289609148285268, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 33753 + }, + { + "epoch": 0.33754, + "grad_norm": 0.7533437155213003, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 33754 + }, + { + "epoch": 0.33755, + "grad_norm": 0.8461683838129879, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 33755 + }, + { + "epoch": 0.33756, + "grad_norm": 0.8811682259265571, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 33756 + }, + { + "epoch": 0.33757, + "grad_norm": 0.9052923835599851, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 33757 + }, + { + "epoch": 0.33758, + "grad_norm": 0.8851760987002658, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 33758 + }, + { + "epoch": 0.33759, + "grad_norm": 0.8174233748547785, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 33759 + }, + { + "epoch": 0.3376, + "grad_norm": 0.7357912144361363, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 33760 + }, + { + "epoch": 0.33761, + "grad_norm": 0.5402505952140684, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 33761 + }, + { + "epoch": 0.33762, + "grad_norm": 0.5425972774785475, + "learning_rate": 0.003, + "loss": 4.006, + "step": 33762 + }, + { + "epoch": 0.33763, + "grad_norm": 0.5326410438951421, + "learning_rate": 0.003, + "loss": 3.9483, + "step": 33763 + }, + { + "epoch": 0.33764, + "grad_norm": 0.4396934013111538, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 33764 + }, + { + "epoch": 0.33765, + "grad_norm": 0.452668417325524, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 33765 + }, + { + "epoch": 0.33766, + "grad_norm": 0.4112616093445322, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 33766 + }, + { + "epoch": 0.33767, + "grad_norm": 0.4626897979233109, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 33767 + }, + { + "epoch": 0.33768, + "grad_norm": 0.5518594579313814, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 33768 + }, + { + "epoch": 0.33769, + "grad_norm": 0.7362303786333685, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 33769 + }, + { + "epoch": 0.3377, + "grad_norm": 0.8865461784019608, + "learning_rate": 0.003, + "loss": 3.9437, + "step": 33770 + }, + { + "epoch": 0.33771, + "grad_norm": 0.9495710543181587, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 33771 + }, + { + "epoch": 0.33772, + "grad_norm": 0.807434944357693, + "learning_rate": 0.003, + "loss": 3.969, + "step": 33772 + }, + { + "epoch": 0.33773, + "grad_norm": 0.6044350727395845, + "learning_rate": 0.003, + "loss": 3.976, + "step": 33773 + }, + { + "epoch": 0.33774, + "grad_norm": 0.5073771696177949, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 33774 + }, + { + "epoch": 0.33775, + "grad_norm": 0.5816245139984405, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 33775 + }, + { + "epoch": 0.33776, + "grad_norm": 0.6432706163967697, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 33776 + }, + { + "epoch": 0.33777, + "grad_norm": 0.7150454924413755, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 33777 + }, + { + "epoch": 0.33778, + "grad_norm": 0.7702507324059991, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 33778 + }, + { + "epoch": 0.33779, + "grad_norm": 0.5940479542148058, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 33779 + }, + { + "epoch": 0.3378, + "grad_norm": 0.47336708278327605, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 33780 + }, + { + "epoch": 0.33781, + "grad_norm": 0.5701899816776752, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 33781 + }, + { + "epoch": 0.33782, + "grad_norm": 0.6287105018465148, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 33782 + }, + { + "epoch": 0.33783, + "grad_norm": 0.6352852786431046, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 33783 + }, + { + "epoch": 0.33784, + "grad_norm": 0.6396954171778082, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 33784 + }, + { + "epoch": 0.33785, + "grad_norm": 0.8188979997005079, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 33785 + }, + { + "epoch": 0.33786, + "grad_norm": 0.997778138094196, + "learning_rate": 0.003, + "loss": 3.986, + "step": 33786 + }, + { + "epoch": 0.33787, + "grad_norm": 1.1828186471206676, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 33787 + }, + { + "epoch": 0.33788, + "grad_norm": 0.7127352165601428, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 33788 + }, + { + "epoch": 0.33789, + "grad_norm": 0.8518914978612416, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 33789 + }, + { + "epoch": 0.3379, + "grad_norm": 0.9265293019022051, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 33790 + }, + { + "epoch": 0.33791, + "grad_norm": 1.0802159427753875, + "learning_rate": 0.003, + "loss": 3.977, + "step": 33791 + }, + { + "epoch": 0.33792, + "grad_norm": 0.9271033379535782, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 33792 + }, + { + "epoch": 0.33793, + "grad_norm": 0.9903029698322556, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 33793 + }, + { + "epoch": 0.33794, + "grad_norm": 0.998686627886744, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 33794 + }, + { + "epoch": 0.33795, + "grad_norm": 0.9377868112354748, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 33795 + }, + { + "epoch": 0.33796, + "grad_norm": 0.8354991269290906, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 33796 + }, + { + "epoch": 0.33797, + "grad_norm": 0.7735752248038108, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 33797 + }, + { + "epoch": 0.33798, + "grad_norm": 0.756687492806736, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 33798 + }, + { + "epoch": 0.33799, + "grad_norm": 0.7519771067968118, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 33799 + }, + { + "epoch": 0.338, + "grad_norm": 0.7449395810409749, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 33800 + }, + { + "epoch": 0.33801, + "grad_norm": 0.5941772702423047, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 33801 + }, + { + "epoch": 0.33802, + "grad_norm": 0.6518803029611363, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 33802 + }, + { + "epoch": 0.33803, + "grad_norm": 0.7024553417604132, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 33803 + }, + { + "epoch": 0.33804, + "grad_norm": 0.8585138696745663, + "learning_rate": 0.003, + "loss": 3.937, + "step": 33804 + }, + { + "epoch": 0.33805, + "grad_norm": 0.9948291271227013, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 33805 + }, + { + "epoch": 0.33806, + "grad_norm": 1.0145927924211022, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 33806 + }, + { + "epoch": 0.33807, + "grad_norm": 0.8743851205348834, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 33807 + }, + { + "epoch": 0.33808, + "grad_norm": 0.7729314304117211, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 33808 + }, + { + "epoch": 0.33809, + "grad_norm": 0.7904624459490709, + "learning_rate": 0.003, + "loss": 3.972, + "step": 33809 + }, + { + "epoch": 0.3381, + "grad_norm": 0.8204932358310844, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 33810 + }, + { + "epoch": 0.33811, + "grad_norm": 0.8034286426368792, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 33811 + }, + { + "epoch": 0.33812, + "grad_norm": 0.784796717205867, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 33812 + }, + { + "epoch": 0.33813, + "grad_norm": 0.7469500257529667, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 33813 + }, + { + "epoch": 0.33814, + "grad_norm": 0.7198236293992658, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 33814 + }, + { + "epoch": 0.33815, + "grad_norm": 0.6787592813350848, + "learning_rate": 0.003, + "loss": 3.973, + "step": 33815 + }, + { + "epoch": 0.33816, + "grad_norm": 0.6187827581132074, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 33816 + }, + { + "epoch": 0.33817, + "grad_norm": 0.6638207126006309, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 33817 + }, + { + "epoch": 0.33818, + "grad_norm": 0.6857866267195424, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 33818 + }, + { + "epoch": 0.33819, + "grad_norm": 0.7221925934969443, + "learning_rate": 0.003, + "loss": 4.004, + "step": 33819 + }, + { + "epoch": 0.3382, + "grad_norm": 0.7821663783996308, + "learning_rate": 0.003, + "loss": 3.997, + "step": 33820 + }, + { + "epoch": 0.33821, + "grad_norm": 0.7671053432998737, + "learning_rate": 0.003, + "loss": 4.0, + "step": 33821 + }, + { + "epoch": 0.33822, + "grad_norm": 0.7902810737696403, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 33822 + }, + { + "epoch": 0.33823, + "grad_norm": 0.9058798514206302, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 33823 + }, + { + "epoch": 0.33824, + "grad_norm": 0.9353736752576906, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 33824 + }, + { + "epoch": 0.33825, + "grad_norm": 0.8060900007290764, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 33825 + }, + { + "epoch": 0.33826, + "grad_norm": 0.6413665541497153, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 33826 + }, + { + "epoch": 0.33827, + "grad_norm": 0.5741256665814242, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 33827 + }, + { + "epoch": 0.33828, + "grad_norm": 0.6043577593457192, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 33828 + }, + { + "epoch": 0.33829, + "grad_norm": 0.6443490703425467, + "learning_rate": 0.003, + "loss": 3.978, + "step": 33829 + }, + { + "epoch": 0.3383, + "grad_norm": 0.7127051821311174, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 33830 + }, + { + "epoch": 0.33831, + "grad_norm": 0.9128964054999652, + "learning_rate": 0.003, + "loss": 3.981, + "step": 33831 + }, + { + "epoch": 0.33832, + "grad_norm": 1.136159471052604, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 33832 + }, + { + "epoch": 0.33833, + "grad_norm": 0.875698259429525, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 33833 + }, + { + "epoch": 0.33834, + "grad_norm": 0.814937818850221, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 33834 + }, + { + "epoch": 0.33835, + "grad_norm": 0.8849646337423716, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 33835 + }, + { + "epoch": 0.33836, + "grad_norm": 0.8940306778867901, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 33836 + }, + { + "epoch": 0.33837, + "grad_norm": 0.8412964533683042, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 33837 + }, + { + "epoch": 0.33838, + "grad_norm": 0.77664524799901, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 33838 + }, + { + "epoch": 0.33839, + "grad_norm": 0.6925925686234258, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 33839 + }, + { + "epoch": 0.3384, + "grad_norm": 0.5342611543937601, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 33840 + }, + { + "epoch": 0.33841, + "grad_norm": 0.5360644380864846, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 33841 + }, + { + "epoch": 0.33842, + "grad_norm": 0.5221805349862988, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 33842 + }, + { + "epoch": 0.33843, + "grad_norm": 0.5962302666391536, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 33843 + }, + { + "epoch": 0.33844, + "grad_norm": 0.6194389764391284, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 33844 + }, + { + "epoch": 0.33845, + "grad_norm": 0.7442924787272216, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 33845 + }, + { + "epoch": 0.33846, + "grad_norm": 0.9182019387644941, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 33846 + }, + { + "epoch": 0.33847, + "grad_norm": 1.088559304206429, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 33847 + }, + { + "epoch": 0.33848, + "grad_norm": 0.777291079923187, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 33848 + }, + { + "epoch": 0.33849, + "grad_norm": 0.5903017471715641, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 33849 + }, + { + "epoch": 0.3385, + "grad_norm": 0.6655980239844811, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 33850 + }, + { + "epoch": 0.33851, + "grad_norm": 0.806019588149564, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 33851 + }, + { + "epoch": 0.33852, + "grad_norm": 0.8392665629547084, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 33852 + }, + { + "epoch": 0.33853, + "grad_norm": 0.7662842070587722, + "learning_rate": 0.003, + "loss": 3.991, + "step": 33853 + }, + { + "epoch": 0.33854, + "grad_norm": 0.7730562072777611, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 33854 + }, + { + "epoch": 0.33855, + "grad_norm": 0.8055094632866456, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 33855 + }, + { + "epoch": 0.33856, + "grad_norm": 0.7870318196466084, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 33856 + }, + { + "epoch": 0.33857, + "grad_norm": 0.8074138718190667, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 33857 + }, + { + "epoch": 0.33858, + "grad_norm": 0.7550485589901808, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 33858 + }, + { + "epoch": 0.33859, + "grad_norm": 0.8242757940701076, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 33859 + }, + { + "epoch": 0.3386, + "grad_norm": 0.7920181673259781, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 33860 + }, + { + "epoch": 0.33861, + "grad_norm": 0.8438629640302419, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 33861 + }, + { + "epoch": 0.33862, + "grad_norm": 0.9819822404972544, + "learning_rate": 0.003, + "loss": 4.0364, + "step": 33862 + }, + { + "epoch": 0.33863, + "grad_norm": 1.0265520740915846, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 33863 + }, + { + "epoch": 0.33864, + "grad_norm": 0.9311568094728455, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 33864 + }, + { + "epoch": 0.33865, + "grad_norm": 0.8255302927873313, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 33865 + }, + { + "epoch": 0.33866, + "grad_norm": 0.5962228269833322, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 33866 + }, + { + "epoch": 0.33867, + "grad_norm": 0.6784630444345298, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 33867 + }, + { + "epoch": 0.33868, + "grad_norm": 0.6849611718046962, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 33868 + }, + { + "epoch": 0.33869, + "grad_norm": 0.7071316485639915, + "learning_rate": 0.003, + "loss": 4.003, + "step": 33869 + }, + { + "epoch": 0.3387, + "grad_norm": 0.8807793157607177, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 33870 + }, + { + "epoch": 0.33871, + "grad_norm": 1.1445002857794253, + "learning_rate": 0.003, + "loss": 4.026, + "step": 33871 + }, + { + "epoch": 0.33872, + "grad_norm": 0.8905900901129686, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 33872 + }, + { + "epoch": 0.33873, + "grad_norm": 0.6758288027230905, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 33873 + }, + { + "epoch": 0.33874, + "grad_norm": 0.6540384431713129, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 33874 + }, + { + "epoch": 0.33875, + "grad_norm": 0.8345188754257213, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 33875 + }, + { + "epoch": 0.33876, + "grad_norm": 1.0103674167997934, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 33876 + }, + { + "epoch": 0.33877, + "grad_norm": 1.0444744317354409, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 33877 + }, + { + "epoch": 0.33878, + "grad_norm": 0.8461842725199641, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 33878 + }, + { + "epoch": 0.33879, + "grad_norm": 0.9051501725139476, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 33879 + }, + { + "epoch": 0.3388, + "grad_norm": 0.7907826091359991, + "learning_rate": 0.003, + "loss": 3.997, + "step": 33880 + }, + { + "epoch": 0.33881, + "grad_norm": 0.7143929935135009, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 33881 + }, + { + "epoch": 0.33882, + "grad_norm": 0.6827816018125588, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 33882 + }, + { + "epoch": 0.33883, + "grad_norm": 0.7673110728446694, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 33883 + }, + { + "epoch": 0.33884, + "grad_norm": 0.871196189963555, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 33884 + }, + { + "epoch": 0.33885, + "grad_norm": 0.8183392274059955, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 33885 + }, + { + "epoch": 0.33886, + "grad_norm": 0.8132636082962279, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 33886 + }, + { + "epoch": 0.33887, + "grad_norm": 0.8539616665971689, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 33887 + }, + { + "epoch": 0.33888, + "grad_norm": 1.0365057556609962, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 33888 + }, + { + "epoch": 0.33889, + "grad_norm": 0.8550971529914325, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 33889 + }, + { + "epoch": 0.3389, + "grad_norm": 0.7541740348900634, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 33890 + }, + { + "epoch": 0.33891, + "grad_norm": 0.8177741378551586, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 33891 + }, + { + "epoch": 0.33892, + "grad_norm": 0.9146945962347115, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 33892 + }, + { + "epoch": 0.33893, + "grad_norm": 0.7975269713387072, + "learning_rate": 0.003, + "loss": 4.003, + "step": 33893 + }, + { + "epoch": 0.33894, + "grad_norm": 0.6707734126089103, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 33894 + }, + { + "epoch": 0.33895, + "grad_norm": 0.576945574399668, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 33895 + }, + { + "epoch": 0.33896, + "grad_norm": 0.58936720264363, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 33896 + }, + { + "epoch": 0.33897, + "grad_norm": 0.6533248194259971, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 33897 + }, + { + "epoch": 0.33898, + "grad_norm": 0.6782892496580255, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 33898 + }, + { + "epoch": 0.33899, + "grad_norm": 0.6947743778842246, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 33899 + }, + { + "epoch": 0.339, + "grad_norm": 0.6950312766331888, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 33900 + }, + { + "epoch": 0.33901, + "grad_norm": 0.5752207599440468, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 33901 + }, + { + "epoch": 0.33902, + "grad_norm": 0.623355010179186, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 33902 + }, + { + "epoch": 0.33903, + "grad_norm": 0.6452872102596818, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 33903 + }, + { + "epoch": 0.33904, + "grad_norm": 0.6394548803131666, + "learning_rate": 0.003, + "loss": 3.966, + "step": 33904 + }, + { + "epoch": 0.33905, + "grad_norm": 0.5303434526551674, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 33905 + }, + { + "epoch": 0.33906, + "grad_norm": 0.4881100743836828, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 33906 + }, + { + "epoch": 0.33907, + "grad_norm": 0.46435181599234454, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 33907 + }, + { + "epoch": 0.33908, + "grad_norm": 0.46909789058993406, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 33908 + }, + { + "epoch": 0.33909, + "grad_norm": 0.5345275224751789, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 33909 + }, + { + "epoch": 0.3391, + "grad_norm": 0.583556178600899, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 33910 + }, + { + "epoch": 0.33911, + "grad_norm": 0.5917239985383286, + "learning_rate": 0.003, + "loss": 3.986, + "step": 33911 + }, + { + "epoch": 0.33912, + "grad_norm": 0.6087183927731455, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 33912 + }, + { + "epoch": 0.33913, + "grad_norm": 0.6667246657614693, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 33913 + }, + { + "epoch": 0.33914, + "grad_norm": 0.771826494488798, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 33914 + }, + { + "epoch": 0.33915, + "grad_norm": 0.888026047011596, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 33915 + }, + { + "epoch": 0.33916, + "grad_norm": 1.0132052431550171, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 33916 + }, + { + "epoch": 0.33917, + "grad_norm": 1.148455743450329, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 33917 + }, + { + "epoch": 0.33918, + "grad_norm": 0.919557283535974, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 33918 + }, + { + "epoch": 0.33919, + "grad_norm": 1.1292406287151588, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 33919 + }, + { + "epoch": 0.3392, + "grad_norm": 1.1540006552731514, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 33920 + }, + { + "epoch": 0.33921, + "grad_norm": 0.8559043093656644, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 33921 + }, + { + "epoch": 0.33922, + "grad_norm": 0.8101141449108049, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 33922 + }, + { + "epoch": 0.33923, + "grad_norm": 0.7926667986140079, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 33923 + }, + { + "epoch": 0.33924, + "grad_norm": 0.5805152148750147, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 33924 + }, + { + "epoch": 0.33925, + "grad_norm": 0.6001703271313961, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 33925 + }, + { + "epoch": 0.33926, + "grad_norm": 0.5426448065798505, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 33926 + }, + { + "epoch": 0.33927, + "grad_norm": 0.6117911199792728, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 33927 + }, + { + "epoch": 0.33928, + "grad_norm": 0.7558883611278402, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 33928 + }, + { + "epoch": 0.33929, + "grad_norm": 0.9011462746572283, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 33929 + }, + { + "epoch": 0.3393, + "grad_norm": 0.8539639843201267, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 33930 + }, + { + "epoch": 0.33931, + "grad_norm": 0.8435664340592358, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 33931 + }, + { + "epoch": 0.33932, + "grad_norm": 0.8855353909088137, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 33932 + }, + { + "epoch": 0.33933, + "grad_norm": 0.867631857933132, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 33933 + }, + { + "epoch": 0.33934, + "grad_norm": 0.7946012452839009, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 33934 + }, + { + "epoch": 0.33935, + "grad_norm": 0.8549923434006909, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 33935 + }, + { + "epoch": 0.33936, + "grad_norm": 0.8312682484475415, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 33936 + }, + { + "epoch": 0.33937, + "grad_norm": 0.8704291699798327, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 33937 + }, + { + "epoch": 0.33938, + "grad_norm": 0.9141077813030195, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 33938 + }, + { + "epoch": 0.33939, + "grad_norm": 1.0061185144414218, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 33939 + }, + { + "epoch": 0.3394, + "grad_norm": 1.0481387683172465, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 33940 + }, + { + "epoch": 0.33941, + "grad_norm": 0.8979368599900782, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 33941 + }, + { + "epoch": 0.33942, + "grad_norm": 0.9352780057336881, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 33942 + }, + { + "epoch": 0.33943, + "grad_norm": 0.7936715425748535, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 33943 + }, + { + "epoch": 0.33944, + "grad_norm": 0.7841868585348767, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 33944 + }, + { + "epoch": 0.33945, + "grad_norm": 0.772363540762368, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 33945 + }, + { + "epoch": 0.33946, + "grad_norm": 0.8852411191036482, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 33946 + }, + { + "epoch": 0.33947, + "grad_norm": 1.005777874232702, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 33947 + }, + { + "epoch": 0.33948, + "grad_norm": 1.0406631689703458, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 33948 + }, + { + "epoch": 0.33949, + "grad_norm": 0.9601258602364137, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 33949 + }, + { + "epoch": 0.3395, + "grad_norm": 0.843812966294435, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 33950 + }, + { + "epoch": 0.33951, + "grad_norm": 0.8277699252931541, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 33951 + }, + { + "epoch": 0.33952, + "grad_norm": 0.7347836122776936, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 33952 + }, + { + "epoch": 0.33953, + "grad_norm": 0.7791474576531647, + "learning_rate": 0.003, + "loss": 3.97, + "step": 33953 + }, + { + "epoch": 0.33954, + "grad_norm": 0.7037566768814358, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 33954 + }, + { + "epoch": 0.33955, + "grad_norm": 0.7173962920393533, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 33955 + }, + { + "epoch": 0.33956, + "grad_norm": 0.6642327703337485, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 33956 + }, + { + "epoch": 0.33957, + "grad_norm": 0.6504872854346899, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 33957 + }, + { + "epoch": 0.33958, + "grad_norm": 0.6621875708553915, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 33958 + }, + { + "epoch": 0.33959, + "grad_norm": 0.621496999350215, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 33959 + }, + { + "epoch": 0.3396, + "grad_norm": 0.6110756568014146, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 33960 + }, + { + "epoch": 0.33961, + "grad_norm": 0.6446526957471146, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 33961 + }, + { + "epoch": 0.33962, + "grad_norm": 0.6783543083856779, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 33962 + }, + { + "epoch": 0.33963, + "grad_norm": 0.8107845126900755, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 33963 + }, + { + "epoch": 0.33964, + "grad_norm": 0.8503519275653216, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 33964 + }, + { + "epoch": 0.33965, + "grad_norm": 0.7750236973617601, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 33965 + }, + { + "epoch": 0.33966, + "grad_norm": 0.7235634458530678, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 33966 + }, + { + "epoch": 0.33967, + "grad_norm": 0.7353824725113232, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 33967 + }, + { + "epoch": 0.33968, + "grad_norm": 0.7978205059582537, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 33968 + }, + { + "epoch": 0.33969, + "grad_norm": 0.9860313966484218, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 33969 + }, + { + "epoch": 0.3397, + "grad_norm": 0.9067529633569802, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 33970 + }, + { + "epoch": 0.33971, + "grad_norm": 0.8314881606207352, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 33971 + }, + { + "epoch": 0.33972, + "grad_norm": 0.8184777995959399, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 33972 + }, + { + "epoch": 0.33973, + "grad_norm": 0.7363308401512432, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 33973 + }, + { + "epoch": 0.33974, + "grad_norm": 0.7201299966088686, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 33974 + }, + { + "epoch": 0.33975, + "grad_norm": 0.7814289401405623, + "learning_rate": 0.003, + "loss": 3.9465, + "step": 33975 + }, + { + "epoch": 0.33976, + "grad_norm": 0.8479786412788433, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 33976 + }, + { + "epoch": 0.33977, + "grad_norm": 0.8885357914128639, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 33977 + }, + { + "epoch": 0.33978, + "grad_norm": 0.8972888705979684, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 33978 + }, + { + "epoch": 0.33979, + "grad_norm": 0.7142806749605035, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 33979 + }, + { + "epoch": 0.3398, + "grad_norm": 0.7192173596260215, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 33980 + }, + { + "epoch": 0.33981, + "grad_norm": 0.646411714401557, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 33981 + }, + { + "epoch": 0.33982, + "grad_norm": 0.6439829938105855, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 33982 + }, + { + "epoch": 0.33983, + "grad_norm": 0.7979240922680498, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 33983 + }, + { + "epoch": 0.33984, + "grad_norm": 0.8599666144429616, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 33984 + }, + { + "epoch": 0.33985, + "grad_norm": 0.9098911661611266, + "learning_rate": 0.003, + "loss": 3.991, + "step": 33985 + }, + { + "epoch": 0.33986, + "grad_norm": 0.91268843723064, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 33986 + }, + { + "epoch": 0.33987, + "grad_norm": 0.8142616804361923, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 33987 + }, + { + "epoch": 0.33988, + "grad_norm": 0.7100503364210313, + "learning_rate": 0.003, + "loss": 4.015, + "step": 33988 + }, + { + "epoch": 0.33989, + "grad_norm": 0.7287992074719206, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 33989 + }, + { + "epoch": 0.3399, + "grad_norm": 0.6404771087443316, + "learning_rate": 0.003, + "loss": 3.97, + "step": 33990 + }, + { + "epoch": 0.33991, + "grad_norm": 0.7126036475436127, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 33991 + }, + { + "epoch": 0.33992, + "grad_norm": 0.6546890698109173, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 33992 + }, + { + "epoch": 0.33993, + "grad_norm": 0.6951665125271383, + "learning_rate": 0.003, + "loss": 3.977, + "step": 33993 + }, + { + "epoch": 0.33994, + "grad_norm": 0.7150625714265, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 33994 + }, + { + "epoch": 0.33995, + "grad_norm": 0.6321622097502146, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 33995 + }, + { + "epoch": 0.33996, + "grad_norm": 0.6626215255959491, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 33996 + }, + { + "epoch": 0.33997, + "grad_norm": 0.6315485514859263, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 33997 + }, + { + "epoch": 0.33998, + "grad_norm": 0.6654807152652539, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 33998 + }, + { + "epoch": 0.33999, + "grad_norm": 0.6825252481997703, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 33999 + }, + { + "epoch": 0.34, + "grad_norm": 0.7442772999794048, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 34000 + }, + { + "epoch": 0.34001, + "grad_norm": 0.9236596970270415, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 34001 + }, + { + "epoch": 0.34002, + "grad_norm": 1.1340259468909606, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 34002 + }, + { + "epoch": 0.34003, + "grad_norm": 0.9121433807890176, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 34003 + }, + { + "epoch": 0.34004, + "grad_norm": 0.8313584968848139, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 34004 + }, + { + "epoch": 0.34005, + "grad_norm": 0.8837512816954232, + "learning_rate": 0.003, + "loss": 3.989, + "step": 34005 + }, + { + "epoch": 0.34006, + "grad_norm": 0.9808491156532761, + "learning_rate": 0.003, + "loss": 3.972, + "step": 34006 + }, + { + "epoch": 0.34007, + "grad_norm": 1.0863874588398077, + "learning_rate": 0.003, + "loss": 4.002, + "step": 34007 + }, + { + "epoch": 0.34008, + "grad_norm": 0.9490993978948171, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 34008 + }, + { + "epoch": 0.34009, + "grad_norm": 0.844620160841228, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 34009 + }, + { + "epoch": 0.3401, + "grad_norm": 0.7407250276443776, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 34010 + }, + { + "epoch": 0.34011, + "grad_norm": 0.7319268305768843, + "learning_rate": 0.003, + "loss": 3.982, + "step": 34011 + }, + { + "epoch": 0.34012, + "grad_norm": 0.696160052838333, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 34012 + }, + { + "epoch": 0.34013, + "grad_norm": 0.6655859693784034, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 34013 + }, + { + "epoch": 0.34014, + "grad_norm": 0.658263227181382, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 34014 + }, + { + "epoch": 0.34015, + "grad_norm": 0.6451252925895015, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 34015 + }, + { + "epoch": 0.34016, + "grad_norm": 0.6632671605499009, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 34016 + }, + { + "epoch": 0.34017, + "grad_norm": 0.6516017220628093, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 34017 + }, + { + "epoch": 0.34018, + "grad_norm": 0.6384734353893379, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 34018 + }, + { + "epoch": 0.34019, + "grad_norm": 0.6359628728071917, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 34019 + }, + { + "epoch": 0.3402, + "grad_norm": 0.6766660998590348, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 34020 + }, + { + "epoch": 0.34021, + "grad_norm": 0.7225279413529492, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 34021 + }, + { + "epoch": 0.34022, + "grad_norm": 0.7462018733644171, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 34022 + }, + { + "epoch": 0.34023, + "grad_norm": 0.8467479397601448, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 34023 + }, + { + "epoch": 0.34024, + "grad_norm": 0.9291951596864996, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 34024 + }, + { + "epoch": 0.34025, + "grad_norm": 1.0174412795995211, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 34025 + }, + { + "epoch": 0.34026, + "grad_norm": 1.0438870447371482, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 34026 + }, + { + "epoch": 0.34027, + "grad_norm": 0.9734000300034621, + "learning_rate": 0.003, + "loss": 4.006, + "step": 34027 + }, + { + "epoch": 0.34028, + "grad_norm": 0.9128345812633337, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 34028 + }, + { + "epoch": 0.34029, + "grad_norm": 0.8560337435989438, + "learning_rate": 0.003, + "loss": 3.972, + "step": 34029 + }, + { + "epoch": 0.3403, + "grad_norm": 0.9027919555578886, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 34030 + }, + { + "epoch": 0.34031, + "grad_norm": 0.9237039109158754, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 34031 + }, + { + "epoch": 0.34032, + "grad_norm": 1.0410206539309006, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 34032 + }, + { + "epoch": 0.34033, + "grad_norm": 0.9391476469880128, + "learning_rate": 0.003, + "loss": 4.007, + "step": 34033 + }, + { + "epoch": 0.34034, + "grad_norm": 0.937333602707067, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 34034 + }, + { + "epoch": 0.34035, + "grad_norm": 0.9543520493817418, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 34035 + }, + { + "epoch": 0.34036, + "grad_norm": 0.9637772136852486, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 34036 + }, + { + "epoch": 0.34037, + "grad_norm": 0.9106338127063414, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 34037 + }, + { + "epoch": 0.34038, + "grad_norm": 0.9375530157298707, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 34038 + }, + { + "epoch": 0.34039, + "grad_norm": 0.8571702985794173, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 34039 + }, + { + "epoch": 0.3404, + "grad_norm": 0.7735179521702212, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 34040 + }, + { + "epoch": 0.34041, + "grad_norm": 0.7241018483937822, + "learning_rate": 0.003, + "loss": 3.923, + "step": 34041 + }, + { + "epoch": 0.34042, + "grad_norm": 0.783313942402669, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 34042 + }, + { + "epoch": 0.34043, + "grad_norm": 0.8721808915499221, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 34043 + }, + { + "epoch": 0.34044, + "grad_norm": 0.9139131098619426, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 34044 + }, + { + "epoch": 0.34045, + "grad_norm": 0.764787011745297, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 34045 + }, + { + "epoch": 0.34046, + "grad_norm": 0.5859957926034419, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 34046 + }, + { + "epoch": 0.34047, + "grad_norm": 0.6561668964066263, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 34047 + }, + { + "epoch": 0.34048, + "grad_norm": 0.6694284888928305, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 34048 + }, + { + "epoch": 0.34049, + "grad_norm": 0.6793731073102937, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 34049 + }, + { + "epoch": 0.3405, + "grad_norm": 0.594148492178686, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 34050 + }, + { + "epoch": 0.34051, + "grad_norm": 0.6685718643328741, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 34051 + }, + { + "epoch": 0.34052, + "grad_norm": 0.8214336595877086, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 34052 + }, + { + "epoch": 0.34053, + "grad_norm": 0.7737827897430797, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 34053 + }, + { + "epoch": 0.34054, + "grad_norm": 0.6688785289039708, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 34054 + }, + { + "epoch": 0.34055, + "grad_norm": 0.6680879522728009, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 34055 + }, + { + "epoch": 0.34056, + "grad_norm": 0.6657189297305418, + "learning_rate": 0.003, + "loss": 3.997, + "step": 34056 + }, + { + "epoch": 0.34057, + "grad_norm": 0.7071259298708669, + "learning_rate": 0.003, + "loss": 3.989, + "step": 34057 + }, + { + "epoch": 0.34058, + "grad_norm": 0.7253999830871587, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 34058 + }, + { + "epoch": 0.34059, + "grad_norm": 0.7266495010642249, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 34059 + }, + { + "epoch": 0.3406, + "grad_norm": 0.7899866043194206, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 34060 + }, + { + "epoch": 0.34061, + "grad_norm": 0.7462327707361636, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 34061 + }, + { + "epoch": 0.34062, + "grad_norm": 0.7325398124180481, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 34062 + }, + { + "epoch": 0.34063, + "grad_norm": 0.6628461302329818, + "learning_rate": 0.003, + "loss": 3.997, + "step": 34063 + }, + { + "epoch": 0.34064, + "grad_norm": 0.5600962410225888, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 34064 + }, + { + "epoch": 0.34065, + "grad_norm": 0.6322801730658237, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 34065 + }, + { + "epoch": 0.34066, + "grad_norm": 0.630435444449193, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 34066 + }, + { + "epoch": 0.34067, + "grad_norm": 0.6868067956560588, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 34067 + }, + { + "epoch": 0.34068, + "grad_norm": 0.7109895992871491, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 34068 + }, + { + "epoch": 0.34069, + "grad_norm": 0.8666441150573179, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 34069 + }, + { + "epoch": 0.3407, + "grad_norm": 0.988101135879782, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 34070 + }, + { + "epoch": 0.34071, + "grad_norm": 1.0122803088638044, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 34071 + }, + { + "epoch": 0.34072, + "grad_norm": 1.0711064391099379, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 34072 + }, + { + "epoch": 0.34073, + "grad_norm": 0.7699481050282878, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 34073 + }, + { + "epoch": 0.34074, + "grad_norm": 0.7079231812260203, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 34074 + }, + { + "epoch": 0.34075, + "grad_norm": 0.7528700161363434, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 34075 + }, + { + "epoch": 0.34076, + "grad_norm": 0.8019339516234248, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 34076 + }, + { + "epoch": 0.34077, + "grad_norm": 0.8451491718920846, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 34077 + }, + { + "epoch": 0.34078, + "grad_norm": 0.8392697654996586, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 34078 + }, + { + "epoch": 0.34079, + "grad_norm": 0.8390248818023525, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 34079 + }, + { + "epoch": 0.3408, + "grad_norm": 0.8871029673351983, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 34080 + }, + { + "epoch": 0.34081, + "grad_norm": 0.7572587145674868, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 34081 + }, + { + "epoch": 0.34082, + "grad_norm": 0.6488092896296196, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 34082 + }, + { + "epoch": 0.34083, + "grad_norm": 0.6229059957401921, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 34083 + }, + { + "epoch": 0.34084, + "grad_norm": 0.6495079672240276, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 34084 + }, + { + "epoch": 0.34085, + "grad_norm": 0.6104577632228138, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 34085 + }, + { + "epoch": 0.34086, + "grad_norm": 0.6699795047024822, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 34086 + }, + { + "epoch": 0.34087, + "grad_norm": 0.7729083745150179, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 34087 + }, + { + "epoch": 0.34088, + "grad_norm": 0.8955454815589542, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 34088 + }, + { + "epoch": 0.34089, + "grad_norm": 1.0042292373177997, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 34089 + }, + { + "epoch": 0.3409, + "grad_norm": 1.0917886540525135, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 34090 + }, + { + "epoch": 0.34091, + "grad_norm": 0.9866814420617198, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 34091 + }, + { + "epoch": 0.34092, + "grad_norm": 0.9270956556955812, + "learning_rate": 0.003, + "loss": 3.995, + "step": 34092 + }, + { + "epoch": 0.34093, + "grad_norm": 0.8427081162417729, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 34093 + }, + { + "epoch": 0.34094, + "grad_norm": 0.7485746720049199, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34094 + }, + { + "epoch": 0.34095, + "grad_norm": 0.740312487946092, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 34095 + }, + { + "epoch": 0.34096, + "grad_norm": 0.693305201641169, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 34096 + }, + { + "epoch": 0.34097, + "grad_norm": 0.7758725347389314, + "learning_rate": 0.003, + "loss": 3.998, + "step": 34097 + }, + { + "epoch": 0.34098, + "grad_norm": 0.7622459186138834, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 34098 + }, + { + "epoch": 0.34099, + "grad_norm": 0.771580963097879, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 34099 + }, + { + "epoch": 0.341, + "grad_norm": 0.7876810572780948, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 34100 + }, + { + "epoch": 0.34101, + "grad_norm": 0.7811564393679472, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 34101 + }, + { + "epoch": 0.34102, + "grad_norm": 0.7240828676483808, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 34102 + }, + { + "epoch": 0.34103, + "grad_norm": 0.8760825793279595, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 34103 + }, + { + "epoch": 0.34104, + "grad_norm": 1.080738087206659, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 34104 + }, + { + "epoch": 0.34105, + "grad_norm": 0.968977731530353, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 34105 + }, + { + "epoch": 0.34106, + "grad_norm": 0.7806081190313565, + "learning_rate": 0.003, + "loss": 3.977, + "step": 34106 + }, + { + "epoch": 0.34107, + "grad_norm": 0.8039704415960179, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 34107 + }, + { + "epoch": 0.34108, + "grad_norm": 0.7356328605265812, + "learning_rate": 0.003, + "loss": 3.992, + "step": 34108 + }, + { + "epoch": 0.34109, + "grad_norm": 0.6936041808834816, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 34109 + }, + { + "epoch": 0.3411, + "grad_norm": 0.6565988325570049, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 34110 + }, + { + "epoch": 0.34111, + "grad_norm": 0.6281422485772418, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 34111 + }, + { + "epoch": 0.34112, + "grad_norm": 0.5670742496721844, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 34112 + }, + { + "epoch": 0.34113, + "grad_norm": 0.6475256923907644, + "learning_rate": 0.003, + "loss": 3.996, + "step": 34113 + }, + { + "epoch": 0.34114, + "grad_norm": 0.6305012230987764, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 34114 + }, + { + "epoch": 0.34115, + "grad_norm": 0.7010230649278287, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 34115 + }, + { + "epoch": 0.34116, + "grad_norm": 0.8148912832766142, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 34116 + }, + { + "epoch": 0.34117, + "grad_norm": 0.8399189378689852, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 34117 + }, + { + "epoch": 0.34118, + "grad_norm": 0.7906737712101282, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 34118 + }, + { + "epoch": 0.34119, + "grad_norm": 0.790984460831878, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 34119 + }, + { + "epoch": 0.3412, + "grad_norm": 0.6536406724562132, + "learning_rate": 0.003, + "loss": 3.9463, + "step": 34120 + }, + { + "epoch": 0.34121, + "grad_norm": 0.6566249172630813, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 34121 + }, + { + "epoch": 0.34122, + "grad_norm": 0.6887237440441302, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 34122 + }, + { + "epoch": 0.34123, + "grad_norm": 0.7706799573975522, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 34123 + }, + { + "epoch": 0.34124, + "grad_norm": 0.7514589850347897, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 34124 + }, + { + "epoch": 0.34125, + "grad_norm": 0.7303442426116119, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 34125 + }, + { + "epoch": 0.34126, + "grad_norm": 0.8043077876907663, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 34126 + }, + { + "epoch": 0.34127, + "grad_norm": 0.9120454334806409, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 34127 + }, + { + "epoch": 0.34128, + "grad_norm": 1.077695668627912, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 34128 + }, + { + "epoch": 0.34129, + "grad_norm": 1.0305172548844692, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 34129 + }, + { + "epoch": 0.3413, + "grad_norm": 1.048469978489923, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 34130 + }, + { + "epoch": 0.34131, + "grad_norm": 0.7680145601710543, + "learning_rate": 0.003, + "loss": 4.004, + "step": 34131 + }, + { + "epoch": 0.34132, + "grad_norm": 0.6023996016884281, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 34132 + }, + { + "epoch": 0.34133, + "grad_norm": 0.620588513412957, + "learning_rate": 0.003, + "loss": 3.957, + "step": 34133 + }, + { + "epoch": 0.34134, + "grad_norm": 0.6460329175553975, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 34134 + }, + { + "epoch": 0.34135, + "grad_norm": 0.7425345829119853, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 34135 + }, + { + "epoch": 0.34136, + "grad_norm": 0.9347772568165781, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 34136 + }, + { + "epoch": 0.34137, + "grad_norm": 1.0804681635450377, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34137 + }, + { + "epoch": 0.34138, + "grad_norm": 0.931651725232988, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 34138 + }, + { + "epoch": 0.34139, + "grad_norm": 0.9311856651553089, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 34139 + }, + { + "epoch": 0.3414, + "grad_norm": 0.9049117134714305, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 34140 + }, + { + "epoch": 0.34141, + "grad_norm": 0.7695934328492324, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 34141 + }, + { + "epoch": 0.34142, + "grad_norm": 0.7026410717360018, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 34142 + }, + { + "epoch": 0.34143, + "grad_norm": 0.6928523094331916, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 34143 + }, + { + "epoch": 0.34144, + "grad_norm": 0.7046565733811669, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 34144 + }, + { + "epoch": 0.34145, + "grad_norm": 0.7706187169788519, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 34145 + }, + { + "epoch": 0.34146, + "grad_norm": 0.6180689477638579, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 34146 + }, + { + "epoch": 0.34147, + "grad_norm": 0.7266762925948904, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 34147 + }, + { + "epoch": 0.34148, + "grad_norm": 0.6941367965132705, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 34148 + }, + { + "epoch": 0.34149, + "grad_norm": 0.6717992905040658, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 34149 + }, + { + "epoch": 0.3415, + "grad_norm": 0.7527612734525229, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 34150 + }, + { + "epoch": 0.34151, + "grad_norm": 0.783856872475952, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 34151 + }, + { + "epoch": 0.34152, + "grad_norm": 0.8436419390267204, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 34152 + }, + { + "epoch": 0.34153, + "grad_norm": 0.8570504242138842, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 34153 + }, + { + "epoch": 0.34154, + "grad_norm": 0.7952388226111111, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 34154 + }, + { + "epoch": 0.34155, + "grad_norm": 0.74307505589083, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 34155 + }, + { + "epoch": 0.34156, + "grad_norm": 0.6829967298132319, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 34156 + }, + { + "epoch": 0.34157, + "grad_norm": 0.7320120497875062, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 34157 + }, + { + "epoch": 0.34158, + "grad_norm": 0.7355155572141154, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 34158 + }, + { + "epoch": 0.34159, + "grad_norm": 0.7199857222016598, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 34159 + }, + { + "epoch": 0.3416, + "grad_norm": 0.6546788126395119, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 34160 + }, + { + "epoch": 0.34161, + "grad_norm": 0.6968366143763279, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 34161 + }, + { + "epoch": 0.34162, + "grad_norm": 0.6596356028913388, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 34162 + }, + { + "epoch": 0.34163, + "grad_norm": 0.6811908293052156, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 34163 + }, + { + "epoch": 0.34164, + "grad_norm": 0.7098191030534313, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 34164 + }, + { + "epoch": 0.34165, + "grad_norm": 0.8149810281060333, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 34165 + }, + { + "epoch": 0.34166, + "grad_norm": 0.7774715057119678, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 34166 + }, + { + "epoch": 0.34167, + "grad_norm": 0.7212345126300245, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 34167 + }, + { + "epoch": 0.34168, + "grad_norm": 0.7886691340041806, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 34168 + }, + { + "epoch": 0.34169, + "grad_norm": 0.7780699563920273, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 34169 + }, + { + "epoch": 0.3417, + "grad_norm": 0.730536446685556, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 34170 + }, + { + "epoch": 0.34171, + "grad_norm": 0.7758035398089002, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 34171 + }, + { + "epoch": 0.34172, + "grad_norm": 0.7344207287143286, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 34172 + }, + { + "epoch": 0.34173, + "grad_norm": 0.771808462560549, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 34173 + }, + { + "epoch": 0.34174, + "grad_norm": 0.9571969701295114, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 34174 + }, + { + "epoch": 0.34175, + "grad_norm": 0.8876550354744493, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 34175 + }, + { + "epoch": 0.34176, + "grad_norm": 0.7138572562236265, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 34176 + }, + { + "epoch": 0.34177, + "grad_norm": 0.7662561241957739, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 34177 + }, + { + "epoch": 0.34178, + "grad_norm": 0.8819921311748844, + "learning_rate": 0.003, + "loss": 3.981, + "step": 34178 + }, + { + "epoch": 0.34179, + "grad_norm": 1.0418238619782996, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 34179 + }, + { + "epoch": 0.3418, + "grad_norm": 1.2507542179347413, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 34180 + }, + { + "epoch": 0.34181, + "grad_norm": 0.696595973077881, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 34181 + }, + { + "epoch": 0.34182, + "grad_norm": 0.7885455463093872, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 34182 + }, + { + "epoch": 0.34183, + "grad_norm": 0.8724579190978734, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 34183 + }, + { + "epoch": 0.34184, + "grad_norm": 0.8924629860656539, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 34184 + }, + { + "epoch": 0.34185, + "grad_norm": 0.9276214546785733, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 34185 + }, + { + "epoch": 0.34186, + "grad_norm": 1.066627239752835, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 34186 + }, + { + "epoch": 0.34187, + "grad_norm": 0.8723113486410413, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 34187 + }, + { + "epoch": 0.34188, + "grad_norm": 0.7467045858122325, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 34188 + }, + { + "epoch": 0.34189, + "grad_norm": 0.745099617983941, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 34189 + }, + { + "epoch": 0.3419, + "grad_norm": 0.8562379608087892, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 34190 + }, + { + "epoch": 0.34191, + "grad_norm": 0.8709930055288989, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 34191 + }, + { + "epoch": 0.34192, + "grad_norm": 0.800464079144911, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 34192 + }, + { + "epoch": 0.34193, + "grad_norm": 0.8909142344093872, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 34193 + }, + { + "epoch": 0.34194, + "grad_norm": 1.030910285957417, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 34194 + }, + { + "epoch": 0.34195, + "grad_norm": 1.0306319526969436, + "learning_rate": 0.003, + "loss": 4.036, + "step": 34195 + }, + { + "epoch": 0.34196, + "grad_norm": 0.9903236570369595, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 34196 + }, + { + "epoch": 0.34197, + "grad_norm": 1.0440420952693295, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 34197 + }, + { + "epoch": 0.34198, + "grad_norm": 0.9507077168284387, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 34198 + }, + { + "epoch": 0.34199, + "grad_norm": 0.8822981817953165, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 34199 + }, + { + "epoch": 0.342, + "grad_norm": 0.8179117960428819, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 34200 + }, + { + "epoch": 0.34201, + "grad_norm": 0.7224533972540612, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 34201 + }, + { + "epoch": 0.34202, + "grad_norm": 0.6246079652137763, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 34202 + }, + { + "epoch": 0.34203, + "grad_norm": 0.6645475947743167, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 34203 + }, + { + "epoch": 0.34204, + "grad_norm": 0.7433078254714365, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 34204 + }, + { + "epoch": 0.34205, + "grad_norm": 0.8893821032160696, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 34205 + }, + { + "epoch": 0.34206, + "grad_norm": 1.1419749802052226, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 34206 + }, + { + "epoch": 0.34207, + "grad_norm": 0.7603553499382077, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 34207 + }, + { + "epoch": 0.34208, + "grad_norm": 0.5972277388814297, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 34208 + }, + { + "epoch": 0.34209, + "grad_norm": 0.6259691698884732, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 34209 + }, + { + "epoch": 0.3421, + "grad_norm": 0.6165850084213889, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 34210 + }, + { + "epoch": 0.34211, + "grad_norm": 0.6053920426070034, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 34211 + }, + { + "epoch": 0.34212, + "grad_norm": 0.5339199914313103, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 34212 + }, + { + "epoch": 0.34213, + "grad_norm": 0.5604458210431593, + "learning_rate": 0.003, + "loss": 3.959, + "step": 34213 + }, + { + "epoch": 0.34214, + "grad_norm": 0.5857652674016881, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 34214 + }, + { + "epoch": 0.34215, + "grad_norm": 0.552176113155831, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 34215 + }, + { + "epoch": 0.34216, + "grad_norm": 0.46812979757740725, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 34216 + }, + { + "epoch": 0.34217, + "grad_norm": 0.49962087802435756, + "learning_rate": 0.003, + "loss": 3.989, + "step": 34217 + }, + { + "epoch": 0.34218, + "grad_norm": 0.549129228264793, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 34218 + }, + { + "epoch": 0.34219, + "grad_norm": 0.5674567231708654, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 34219 + }, + { + "epoch": 0.3422, + "grad_norm": 0.6734342165151923, + "learning_rate": 0.003, + "loss": 3.984, + "step": 34220 + }, + { + "epoch": 0.34221, + "grad_norm": 0.8404273970432529, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 34221 + }, + { + "epoch": 0.34222, + "grad_norm": 1.007988840934782, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 34222 + }, + { + "epoch": 0.34223, + "grad_norm": 1.173850657612207, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 34223 + }, + { + "epoch": 0.34224, + "grad_norm": 0.728049047080327, + "learning_rate": 0.003, + "loss": 4.004, + "step": 34224 + }, + { + "epoch": 0.34225, + "grad_norm": 0.682792431375473, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 34225 + }, + { + "epoch": 0.34226, + "grad_norm": 0.7196447452755005, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 34226 + }, + { + "epoch": 0.34227, + "grad_norm": 0.6282260234589351, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 34227 + }, + { + "epoch": 0.34228, + "grad_norm": 0.5403766270330679, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 34228 + }, + { + "epoch": 0.34229, + "grad_norm": 0.532228480528133, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 34229 + }, + { + "epoch": 0.3423, + "grad_norm": 0.5175980454465258, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 34230 + }, + { + "epoch": 0.34231, + "grad_norm": 0.5771159704504015, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 34231 + }, + { + "epoch": 0.34232, + "grad_norm": 0.6023478392605335, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 34232 + }, + { + "epoch": 0.34233, + "grad_norm": 0.6436612513642449, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 34233 + }, + { + "epoch": 0.34234, + "grad_norm": 0.7006502428694751, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 34234 + }, + { + "epoch": 0.34235, + "grad_norm": 0.7680529701868932, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 34235 + }, + { + "epoch": 0.34236, + "grad_norm": 0.9919467019321411, + "learning_rate": 0.003, + "loss": 3.993, + "step": 34236 + }, + { + "epoch": 0.34237, + "grad_norm": 1.1167191462397303, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 34237 + }, + { + "epoch": 0.34238, + "grad_norm": 0.7057957639435094, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 34238 + }, + { + "epoch": 0.34239, + "grad_norm": 0.7025023230432705, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 34239 + }, + { + "epoch": 0.3424, + "grad_norm": 0.9161892637529727, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 34240 + }, + { + "epoch": 0.34241, + "grad_norm": 1.2739607548714447, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 34241 + }, + { + "epoch": 0.34242, + "grad_norm": 0.9177205519985403, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 34242 + }, + { + "epoch": 0.34243, + "grad_norm": 0.843721821013009, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 34243 + }, + { + "epoch": 0.34244, + "grad_norm": 0.8432339851005155, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 34244 + }, + { + "epoch": 0.34245, + "grad_norm": 0.9087819156816456, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 34245 + }, + { + "epoch": 0.34246, + "grad_norm": 0.9212086308828045, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 34246 + }, + { + "epoch": 0.34247, + "grad_norm": 0.8503778960846038, + "learning_rate": 0.003, + "loss": 4.02, + "step": 34247 + }, + { + "epoch": 0.34248, + "grad_norm": 0.7342685243962305, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 34248 + }, + { + "epoch": 0.34249, + "grad_norm": 0.6891243985571656, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 34249 + }, + { + "epoch": 0.3425, + "grad_norm": 0.5941112896015796, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 34250 + }, + { + "epoch": 0.34251, + "grad_norm": 0.6511420134364976, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 34251 + }, + { + "epoch": 0.34252, + "grad_norm": 0.7455485131248929, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 34252 + }, + { + "epoch": 0.34253, + "grad_norm": 0.8507788079342047, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 34253 + }, + { + "epoch": 0.34254, + "grad_norm": 0.931127093595612, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 34254 + }, + { + "epoch": 0.34255, + "grad_norm": 0.9912002799948626, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 34255 + }, + { + "epoch": 0.34256, + "grad_norm": 1.1550393757724586, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 34256 + }, + { + "epoch": 0.34257, + "grad_norm": 0.688946107951795, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 34257 + }, + { + "epoch": 0.34258, + "grad_norm": 0.7507950726009835, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 34258 + }, + { + "epoch": 0.34259, + "grad_norm": 0.8337373913308778, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 34259 + }, + { + "epoch": 0.3426, + "grad_norm": 0.8006695572936795, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 34260 + }, + { + "epoch": 0.34261, + "grad_norm": 0.8063320674823166, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 34261 + }, + { + "epoch": 0.34262, + "grad_norm": 0.7711896146905228, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 34262 + }, + { + "epoch": 0.34263, + "grad_norm": 0.7000411193285051, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 34263 + }, + { + "epoch": 0.34264, + "grad_norm": 0.774983673605465, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 34264 + }, + { + "epoch": 0.34265, + "grad_norm": 0.7720309071470038, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 34265 + }, + { + "epoch": 0.34266, + "grad_norm": 0.7243118348548883, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 34266 + }, + { + "epoch": 0.34267, + "grad_norm": 0.7196413607483494, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 34267 + }, + { + "epoch": 0.34268, + "grad_norm": 0.7088260119730678, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 34268 + }, + { + "epoch": 0.34269, + "grad_norm": 0.7884149888598119, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34269 + }, + { + "epoch": 0.3427, + "grad_norm": 0.826980979704376, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 34270 + }, + { + "epoch": 0.34271, + "grad_norm": 0.8599176994797756, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 34271 + }, + { + "epoch": 0.34272, + "grad_norm": 0.8667469718271907, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 34272 + }, + { + "epoch": 0.34273, + "grad_norm": 0.7857936797402967, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 34273 + }, + { + "epoch": 0.34274, + "grad_norm": 0.8714233221961698, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 34274 + }, + { + "epoch": 0.34275, + "grad_norm": 0.9284039734351917, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 34275 + }, + { + "epoch": 0.34276, + "grad_norm": 0.8997231080054487, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 34276 + }, + { + "epoch": 0.34277, + "grad_norm": 0.8032784054046223, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 34277 + }, + { + "epoch": 0.34278, + "grad_norm": 0.7529210766772331, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 34278 + }, + { + "epoch": 0.34279, + "grad_norm": 0.7960302304999772, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 34279 + }, + { + "epoch": 0.3428, + "grad_norm": 0.9001400422158872, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 34280 + }, + { + "epoch": 0.34281, + "grad_norm": 0.9517569908554305, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 34281 + }, + { + "epoch": 0.34282, + "grad_norm": 0.9020094300153533, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 34282 + }, + { + "epoch": 0.34283, + "grad_norm": 0.9205960041257212, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 34283 + }, + { + "epoch": 0.34284, + "grad_norm": 0.8613460548663237, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 34284 + }, + { + "epoch": 0.34285, + "grad_norm": 1.0080321294330588, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 34285 + }, + { + "epoch": 0.34286, + "grad_norm": 1.1956423928300688, + "learning_rate": 0.003, + "loss": 4.05, + "step": 34286 + }, + { + "epoch": 0.34287, + "grad_norm": 0.9560525776772165, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 34287 + }, + { + "epoch": 0.34288, + "grad_norm": 0.8677950944060172, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 34288 + }, + { + "epoch": 0.34289, + "grad_norm": 0.8501794178279558, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 34289 + }, + { + "epoch": 0.3429, + "grad_norm": 0.9152124769360128, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 34290 + }, + { + "epoch": 0.34291, + "grad_norm": 0.8910993011540973, + "learning_rate": 0.003, + "loss": 4.0, + "step": 34291 + }, + { + "epoch": 0.34292, + "grad_norm": 0.9134687677737385, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 34292 + }, + { + "epoch": 0.34293, + "grad_norm": 0.7142497021794452, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 34293 + }, + { + "epoch": 0.34294, + "grad_norm": 0.6951779774602732, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 34294 + }, + { + "epoch": 0.34295, + "grad_norm": 0.7025288252191041, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 34295 + }, + { + "epoch": 0.34296, + "grad_norm": 0.7636060345463019, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 34296 + }, + { + "epoch": 0.34297, + "grad_norm": 0.7575775884947915, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 34297 + }, + { + "epoch": 0.34298, + "grad_norm": 0.847501404164838, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 34298 + }, + { + "epoch": 0.34299, + "grad_norm": 0.8362222119390472, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 34299 + }, + { + "epoch": 0.343, + "grad_norm": 0.7670942923460796, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 34300 + }, + { + "epoch": 0.34301, + "grad_norm": 0.6992083117814367, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 34301 + }, + { + "epoch": 0.34302, + "grad_norm": 0.6955929290017533, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 34302 + }, + { + "epoch": 0.34303, + "grad_norm": 0.6828620585748527, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 34303 + }, + { + "epoch": 0.34304, + "grad_norm": 0.6136337343878535, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 34304 + }, + { + "epoch": 0.34305, + "grad_norm": 0.6219439566169794, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 34305 + }, + { + "epoch": 0.34306, + "grad_norm": 0.6956071867528374, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 34306 + }, + { + "epoch": 0.34307, + "grad_norm": 0.787899191566734, + "learning_rate": 0.003, + "loss": 3.974, + "step": 34307 + }, + { + "epoch": 0.34308, + "grad_norm": 0.7925978516343322, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 34308 + }, + { + "epoch": 0.34309, + "grad_norm": 0.800141134305652, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 34309 + }, + { + "epoch": 0.3431, + "grad_norm": 0.8141320567876819, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 34310 + }, + { + "epoch": 0.34311, + "grad_norm": 0.8994166160115219, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 34311 + }, + { + "epoch": 0.34312, + "grad_norm": 0.9392852675130462, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 34312 + }, + { + "epoch": 0.34313, + "grad_norm": 0.833158706443411, + "learning_rate": 0.003, + "loss": 4.002, + "step": 34313 + }, + { + "epoch": 0.34314, + "grad_norm": 0.7816591814111882, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 34314 + }, + { + "epoch": 0.34315, + "grad_norm": 0.7297378418948767, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 34315 + }, + { + "epoch": 0.34316, + "grad_norm": 0.6939001506157642, + "learning_rate": 0.003, + "loss": 3.997, + "step": 34316 + }, + { + "epoch": 0.34317, + "grad_norm": 0.6783665826705796, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 34317 + }, + { + "epoch": 0.34318, + "grad_norm": 0.718398348020722, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 34318 + }, + { + "epoch": 0.34319, + "grad_norm": 0.6936627599948662, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 34319 + }, + { + "epoch": 0.3432, + "grad_norm": 0.6774474804801417, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 34320 + }, + { + "epoch": 0.34321, + "grad_norm": 0.6950770669056517, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 34321 + }, + { + "epoch": 0.34322, + "grad_norm": 0.6903481813142425, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 34322 + }, + { + "epoch": 0.34323, + "grad_norm": 0.6431031047456699, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 34323 + }, + { + "epoch": 0.34324, + "grad_norm": 0.6571822011939831, + "learning_rate": 0.003, + "loss": 3.98, + "step": 34324 + }, + { + "epoch": 0.34325, + "grad_norm": 0.686250445124119, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34325 + }, + { + "epoch": 0.34326, + "grad_norm": 0.8292355957350662, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 34326 + }, + { + "epoch": 0.34327, + "grad_norm": 1.0249894987650983, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 34327 + }, + { + "epoch": 0.34328, + "grad_norm": 1.020394407028967, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 34328 + }, + { + "epoch": 0.34329, + "grad_norm": 0.8464807738973757, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 34329 + }, + { + "epoch": 0.3433, + "grad_norm": 0.6984741911452185, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 34330 + }, + { + "epoch": 0.34331, + "grad_norm": 0.8446893655445561, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 34331 + }, + { + "epoch": 0.34332, + "grad_norm": 0.9678562288621534, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 34332 + }, + { + "epoch": 0.34333, + "grad_norm": 0.9506226598525508, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 34333 + }, + { + "epoch": 0.34334, + "grad_norm": 0.8626189440400881, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 34334 + }, + { + "epoch": 0.34335, + "grad_norm": 0.6973146427059723, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 34335 + }, + { + "epoch": 0.34336, + "grad_norm": 0.7379602166075333, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 34336 + }, + { + "epoch": 0.34337, + "grad_norm": 0.6823886245801383, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 34337 + }, + { + "epoch": 0.34338, + "grad_norm": 0.579783518272865, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 34338 + }, + { + "epoch": 0.34339, + "grad_norm": 0.6514280387188477, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 34339 + }, + { + "epoch": 0.3434, + "grad_norm": 0.643053386872874, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 34340 + }, + { + "epoch": 0.34341, + "grad_norm": 0.6178443174189074, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 34341 + }, + { + "epoch": 0.34342, + "grad_norm": 0.5787678749301375, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 34342 + }, + { + "epoch": 0.34343, + "grad_norm": 0.6876933635073121, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 34343 + }, + { + "epoch": 0.34344, + "grad_norm": 0.7862954180710621, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 34344 + }, + { + "epoch": 0.34345, + "grad_norm": 0.8698480431875983, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 34345 + }, + { + "epoch": 0.34346, + "grad_norm": 0.8167436638515403, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 34346 + }, + { + "epoch": 0.34347, + "grad_norm": 0.7749517924631217, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 34347 + }, + { + "epoch": 0.34348, + "grad_norm": 0.7844729074622154, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 34348 + }, + { + "epoch": 0.34349, + "grad_norm": 0.8492893341335653, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 34349 + }, + { + "epoch": 0.3435, + "grad_norm": 0.8541545910040689, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 34350 + }, + { + "epoch": 0.34351, + "grad_norm": 0.8120315054551888, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 34351 + }, + { + "epoch": 0.34352, + "grad_norm": 0.7861995188953623, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 34352 + }, + { + "epoch": 0.34353, + "grad_norm": 0.7794960131481653, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 34353 + }, + { + "epoch": 0.34354, + "grad_norm": 0.7802409472214958, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 34354 + }, + { + "epoch": 0.34355, + "grad_norm": 0.8748354786445356, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 34355 + }, + { + "epoch": 0.34356, + "grad_norm": 0.9734393708642434, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 34356 + }, + { + "epoch": 0.34357, + "grad_norm": 0.9493519300169171, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 34357 + }, + { + "epoch": 0.34358, + "grad_norm": 0.8268623309380495, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 34358 + }, + { + "epoch": 0.34359, + "grad_norm": 0.717999547709811, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 34359 + }, + { + "epoch": 0.3436, + "grad_norm": 0.6869967138455889, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 34360 + }, + { + "epoch": 0.34361, + "grad_norm": 0.7386739294921034, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 34361 + }, + { + "epoch": 0.34362, + "grad_norm": 0.6811990518542272, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 34362 + }, + { + "epoch": 0.34363, + "grad_norm": 0.636446582673226, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 34363 + }, + { + "epoch": 0.34364, + "grad_norm": 0.705650860892254, + "learning_rate": 0.003, + "loss": 4.0415, + "step": 34364 + }, + { + "epoch": 0.34365, + "grad_norm": 0.7367847559438475, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34365 + }, + { + "epoch": 0.34366, + "grad_norm": 0.6957785358038686, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 34366 + }, + { + "epoch": 0.34367, + "grad_norm": 0.6960687539223067, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 34367 + }, + { + "epoch": 0.34368, + "grad_norm": 0.7468589484773068, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 34368 + }, + { + "epoch": 0.34369, + "grad_norm": 0.8674290175648025, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 34369 + }, + { + "epoch": 0.3437, + "grad_norm": 0.9400816557980674, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 34370 + }, + { + "epoch": 0.34371, + "grad_norm": 0.9924826932561196, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 34371 + }, + { + "epoch": 0.34372, + "grad_norm": 0.9561256722978039, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 34372 + }, + { + "epoch": 0.34373, + "grad_norm": 0.7934883740672555, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 34373 + }, + { + "epoch": 0.34374, + "grad_norm": 0.661182008398807, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 34374 + }, + { + "epoch": 0.34375, + "grad_norm": 0.686807439028785, + "learning_rate": 0.003, + "loss": 3.989, + "step": 34375 + }, + { + "epoch": 0.34376, + "grad_norm": 0.6507138621591945, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 34376 + }, + { + "epoch": 0.34377, + "grad_norm": 0.6302191657474336, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 34377 + }, + { + "epoch": 0.34378, + "grad_norm": 0.5971197954694174, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 34378 + }, + { + "epoch": 0.34379, + "grad_norm": 0.5745584487493781, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 34379 + }, + { + "epoch": 0.3438, + "grad_norm": 0.6267464108696496, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 34380 + }, + { + "epoch": 0.34381, + "grad_norm": 0.6905280978365872, + "learning_rate": 0.003, + "loss": 3.973, + "step": 34381 + }, + { + "epoch": 0.34382, + "grad_norm": 0.6951548757567476, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 34382 + }, + { + "epoch": 0.34383, + "grad_norm": 0.7003085625784101, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 34383 + }, + { + "epoch": 0.34384, + "grad_norm": 0.8026112543004142, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 34384 + }, + { + "epoch": 0.34385, + "grad_norm": 0.8911938931233754, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 34385 + }, + { + "epoch": 0.34386, + "grad_norm": 0.8375346535691529, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 34386 + }, + { + "epoch": 0.34387, + "grad_norm": 0.8668679671609235, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 34387 + }, + { + "epoch": 0.34388, + "grad_norm": 1.0085811221838734, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 34388 + }, + { + "epoch": 0.34389, + "grad_norm": 1.1352456491384189, + "learning_rate": 0.003, + "loss": 3.993, + "step": 34389 + }, + { + "epoch": 0.3439, + "grad_norm": 0.9501289525418795, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 34390 + }, + { + "epoch": 0.34391, + "grad_norm": 1.041271495100212, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 34391 + }, + { + "epoch": 0.34392, + "grad_norm": 1.088581484168248, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 34392 + }, + { + "epoch": 0.34393, + "grad_norm": 0.8694901225284282, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 34393 + }, + { + "epoch": 0.34394, + "grad_norm": 0.9152789186217929, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 34394 + }, + { + "epoch": 0.34395, + "grad_norm": 0.8845172836749723, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 34395 + }, + { + "epoch": 0.34396, + "grad_norm": 0.8001873826054913, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 34396 + }, + { + "epoch": 0.34397, + "grad_norm": 0.7554529148787172, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 34397 + }, + { + "epoch": 0.34398, + "grad_norm": 0.8541399245262953, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 34398 + }, + { + "epoch": 0.34399, + "grad_norm": 0.9004870948844522, + "learning_rate": 0.003, + "loss": 4.016, + "step": 34399 + }, + { + "epoch": 0.344, + "grad_norm": 0.8170968023764155, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 34400 + }, + { + "epoch": 0.34401, + "grad_norm": 0.6642331179857544, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 34401 + }, + { + "epoch": 0.34402, + "grad_norm": 0.6865245808517993, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 34402 + }, + { + "epoch": 0.34403, + "grad_norm": 0.6583626494722635, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 34403 + }, + { + "epoch": 0.34404, + "grad_norm": 0.6168369481463026, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 34404 + }, + { + "epoch": 0.34405, + "grad_norm": 0.5379214691275847, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 34405 + }, + { + "epoch": 0.34406, + "grad_norm": 0.5180581982182425, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 34406 + }, + { + "epoch": 0.34407, + "grad_norm": 0.5651341569437723, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 34407 + }, + { + "epoch": 0.34408, + "grad_norm": 0.5678229778884978, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 34408 + }, + { + "epoch": 0.34409, + "grad_norm": 0.581813754698933, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 34409 + }, + { + "epoch": 0.3441, + "grad_norm": 0.555067626027503, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 34410 + }, + { + "epoch": 0.34411, + "grad_norm": 0.546957691895348, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 34411 + }, + { + "epoch": 0.34412, + "grad_norm": 0.585161901668965, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 34412 + }, + { + "epoch": 0.34413, + "grad_norm": 0.6651308181883108, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 34413 + }, + { + "epoch": 0.34414, + "grad_norm": 0.722009806483119, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 34414 + }, + { + "epoch": 0.34415, + "grad_norm": 0.7491723660838507, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 34415 + }, + { + "epoch": 0.34416, + "grad_norm": 0.7877962417464688, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 34416 + }, + { + "epoch": 0.34417, + "grad_norm": 0.7957601398483674, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 34417 + }, + { + "epoch": 0.34418, + "grad_norm": 0.8218827836215163, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 34418 + }, + { + "epoch": 0.34419, + "grad_norm": 0.8067820311061543, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 34419 + }, + { + "epoch": 0.3442, + "grad_norm": 0.8716093895144513, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 34420 + }, + { + "epoch": 0.34421, + "grad_norm": 0.826700696343803, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 34421 + }, + { + "epoch": 0.34422, + "grad_norm": 0.8135144286451343, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 34422 + }, + { + "epoch": 0.34423, + "grad_norm": 1.1604325120600698, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 34423 + }, + { + "epoch": 0.34424, + "grad_norm": 1.0428853131043916, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 34424 + }, + { + "epoch": 0.34425, + "grad_norm": 0.8427223929391611, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 34425 + }, + { + "epoch": 0.34426, + "grad_norm": 0.8483868685133468, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 34426 + }, + { + "epoch": 0.34427, + "grad_norm": 0.8491919314231303, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 34427 + }, + { + "epoch": 0.34428, + "grad_norm": 0.6497276969616298, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 34428 + }, + { + "epoch": 0.34429, + "grad_norm": 0.5785020859338538, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34429 + }, + { + "epoch": 0.3443, + "grad_norm": 0.5659920183115081, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 34430 + }, + { + "epoch": 0.34431, + "grad_norm": 0.5439249978486551, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 34431 + }, + { + "epoch": 0.34432, + "grad_norm": 0.6160793646767538, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 34432 + }, + { + "epoch": 0.34433, + "grad_norm": 0.7227641366402295, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 34433 + }, + { + "epoch": 0.34434, + "grad_norm": 0.749181808280106, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 34434 + }, + { + "epoch": 0.34435, + "grad_norm": 0.7510781918845321, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 34435 + }, + { + "epoch": 0.34436, + "grad_norm": 0.8300116923102817, + "learning_rate": 0.003, + "loss": 3.975, + "step": 34436 + }, + { + "epoch": 0.34437, + "grad_norm": 0.8404942638614588, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 34437 + }, + { + "epoch": 0.34438, + "grad_norm": 0.7006036913846607, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 34438 + }, + { + "epoch": 0.34439, + "grad_norm": 0.6337848987712085, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 34439 + }, + { + "epoch": 0.3444, + "grad_norm": 0.6127128786400186, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 34440 + }, + { + "epoch": 0.34441, + "grad_norm": 0.6842152549713936, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 34441 + }, + { + "epoch": 0.34442, + "grad_norm": 0.7160717389359447, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 34442 + }, + { + "epoch": 0.34443, + "grad_norm": 0.7324850472276581, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 34443 + }, + { + "epoch": 0.34444, + "grad_norm": 0.7932722299026898, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 34444 + }, + { + "epoch": 0.34445, + "grad_norm": 1.0316740742875843, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 34445 + }, + { + "epoch": 0.34446, + "grad_norm": 1.2382171272030698, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 34446 + }, + { + "epoch": 0.34447, + "grad_norm": 0.7250398839753655, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 34447 + }, + { + "epoch": 0.34448, + "grad_norm": 0.71092021784933, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 34448 + }, + { + "epoch": 0.34449, + "grad_norm": 0.7706665923799967, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 34449 + }, + { + "epoch": 0.3445, + "grad_norm": 0.7503483296031918, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 34450 + }, + { + "epoch": 0.34451, + "grad_norm": 0.7672122375107989, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 34451 + }, + { + "epoch": 0.34452, + "grad_norm": 0.8031701556388688, + "learning_rate": 0.003, + "loss": 3.967, + "step": 34452 + }, + { + "epoch": 0.34453, + "grad_norm": 0.9692810319172789, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 34453 + }, + { + "epoch": 0.34454, + "grad_norm": 1.0392886022462924, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 34454 + }, + { + "epoch": 0.34455, + "grad_norm": 0.9958783943664848, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 34455 + }, + { + "epoch": 0.34456, + "grad_norm": 0.875550248602773, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34456 + }, + { + "epoch": 0.34457, + "grad_norm": 0.9053378356082498, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 34457 + }, + { + "epoch": 0.34458, + "grad_norm": 1.0068877443567483, + "learning_rate": 0.003, + "loss": 4.0328, + "step": 34458 + }, + { + "epoch": 0.34459, + "grad_norm": 0.9937027136142078, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 34459 + }, + { + "epoch": 0.3446, + "grad_norm": 0.9520899820182426, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 34460 + }, + { + "epoch": 0.34461, + "grad_norm": 1.0922684765861639, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 34461 + }, + { + "epoch": 0.34462, + "grad_norm": 0.9865554431366947, + "learning_rate": 0.003, + "loss": 4.0293, + "step": 34462 + }, + { + "epoch": 0.34463, + "grad_norm": 1.0297750826216998, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 34463 + }, + { + "epoch": 0.34464, + "grad_norm": 0.8795428780591752, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 34464 + }, + { + "epoch": 0.34465, + "grad_norm": 0.8349389065872335, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 34465 + }, + { + "epoch": 0.34466, + "grad_norm": 0.9397572882384254, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 34466 + }, + { + "epoch": 0.34467, + "grad_norm": 1.1768890394076927, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 34467 + }, + { + "epoch": 0.34468, + "grad_norm": 0.938927968291535, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 34468 + }, + { + "epoch": 0.34469, + "grad_norm": 1.0302233267092193, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 34469 + }, + { + "epoch": 0.3447, + "grad_norm": 1.0185881997834225, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 34470 + }, + { + "epoch": 0.34471, + "grad_norm": 0.8930925698968641, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 34471 + }, + { + "epoch": 0.34472, + "grad_norm": 0.7835140455906806, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 34472 + }, + { + "epoch": 0.34473, + "grad_norm": 0.7868847083814344, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 34473 + }, + { + "epoch": 0.34474, + "grad_norm": 0.7825371061580894, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 34474 + }, + { + "epoch": 0.34475, + "grad_norm": 0.8237612068293588, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 34475 + }, + { + "epoch": 0.34476, + "grad_norm": 0.7796207171374915, + "learning_rate": 0.003, + "loss": 3.981, + "step": 34476 + }, + { + "epoch": 0.34477, + "grad_norm": 0.8393607978044355, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 34477 + }, + { + "epoch": 0.34478, + "grad_norm": 0.8052795999342226, + "learning_rate": 0.003, + "loss": 3.974, + "step": 34478 + }, + { + "epoch": 0.34479, + "grad_norm": 0.7894998468904938, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 34479 + }, + { + "epoch": 0.3448, + "grad_norm": 0.7924835408871443, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 34480 + }, + { + "epoch": 0.34481, + "grad_norm": 0.7780118159179037, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 34481 + }, + { + "epoch": 0.34482, + "grad_norm": 0.7907170144281953, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 34482 + }, + { + "epoch": 0.34483, + "grad_norm": 0.7196671200995218, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 34483 + }, + { + "epoch": 0.34484, + "grad_norm": 0.6340468216698141, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 34484 + }, + { + "epoch": 0.34485, + "grad_norm": 0.6305314025819071, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 34485 + }, + { + "epoch": 0.34486, + "grad_norm": 0.5881699806501539, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 34486 + }, + { + "epoch": 0.34487, + "grad_norm": 0.5808970995523858, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 34487 + }, + { + "epoch": 0.34488, + "grad_norm": 0.5592833038249967, + "learning_rate": 0.003, + "loss": 3.937, + "step": 34488 + }, + { + "epoch": 0.34489, + "grad_norm": 0.5483514107090999, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 34489 + }, + { + "epoch": 0.3449, + "grad_norm": 0.503529622760941, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 34490 + }, + { + "epoch": 0.34491, + "grad_norm": 0.5176995985989031, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 34491 + }, + { + "epoch": 0.34492, + "grad_norm": 0.44848237861708223, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 34492 + }, + { + "epoch": 0.34493, + "grad_norm": 0.46984692786983256, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 34493 + }, + { + "epoch": 0.34494, + "grad_norm": 0.4550536660642751, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 34494 + }, + { + "epoch": 0.34495, + "grad_norm": 0.4791321045316757, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 34495 + }, + { + "epoch": 0.34496, + "grad_norm": 0.6640989097563673, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 34496 + }, + { + "epoch": 0.34497, + "grad_norm": 0.9908783274680155, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 34497 + }, + { + "epoch": 0.34498, + "grad_norm": 1.2812928862470419, + "learning_rate": 0.003, + "loss": 4.002, + "step": 34498 + }, + { + "epoch": 0.34499, + "grad_norm": 0.5437035495476265, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 34499 + }, + { + "epoch": 0.345, + "grad_norm": 0.7747810776032181, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 34500 + }, + { + "epoch": 0.34501, + "grad_norm": 0.9490212679015411, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 34501 + }, + { + "epoch": 0.34502, + "grad_norm": 0.9796345042097999, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 34502 + }, + { + "epoch": 0.34503, + "grad_norm": 0.7607087912007325, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 34503 + }, + { + "epoch": 0.34504, + "grad_norm": 0.7234427149524187, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 34504 + }, + { + "epoch": 0.34505, + "grad_norm": 0.7671696085559419, + "learning_rate": 0.003, + "loss": 3.99, + "step": 34505 + }, + { + "epoch": 0.34506, + "grad_norm": 0.709173828315792, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 34506 + }, + { + "epoch": 0.34507, + "grad_norm": 0.6670944637897445, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 34507 + }, + { + "epoch": 0.34508, + "grad_norm": 0.6842367145729621, + "learning_rate": 0.003, + "loss": 4.024, + "step": 34508 + }, + { + "epoch": 0.34509, + "grad_norm": 0.6982122072202317, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 34509 + }, + { + "epoch": 0.3451, + "grad_norm": 0.8258147190767848, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 34510 + }, + { + "epoch": 0.34511, + "grad_norm": 0.8396099146743888, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 34511 + }, + { + "epoch": 0.34512, + "grad_norm": 0.8195827906714173, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 34512 + }, + { + "epoch": 0.34513, + "grad_norm": 0.7769346008775107, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 34513 + }, + { + "epoch": 0.34514, + "grad_norm": 0.7084838779942048, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 34514 + }, + { + "epoch": 0.34515, + "grad_norm": 0.7197592448557495, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 34515 + }, + { + "epoch": 0.34516, + "grad_norm": 0.6244433923759531, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 34516 + }, + { + "epoch": 0.34517, + "grad_norm": 0.6668239855530365, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 34517 + }, + { + "epoch": 0.34518, + "grad_norm": 0.7593360507839759, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 34518 + }, + { + "epoch": 0.34519, + "grad_norm": 0.9608536761496398, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 34519 + }, + { + "epoch": 0.3452, + "grad_norm": 1.0136391958412074, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 34520 + }, + { + "epoch": 0.34521, + "grad_norm": 0.9564687279585071, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 34521 + }, + { + "epoch": 0.34522, + "grad_norm": 0.810377909284908, + "learning_rate": 0.003, + "loss": 3.981, + "step": 34522 + }, + { + "epoch": 0.34523, + "grad_norm": 0.8735687193771736, + "learning_rate": 0.003, + "loss": 4.008, + "step": 34523 + }, + { + "epoch": 0.34524, + "grad_norm": 0.9427412967022402, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 34524 + }, + { + "epoch": 0.34525, + "grad_norm": 1.007208628091444, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 34525 + }, + { + "epoch": 0.34526, + "grad_norm": 1.0310327573878832, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 34526 + }, + { + "epoch": 0.34527, + "grad_norm": 1.026177287130178, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 34527 + }, + { + "epoch": 0.34528, + "grad_norm": 1.0431458821104167, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 34528 + }, + { + "epoch": 0.34529, + "grad_norm": 0.9416645909090376, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 34529 + }, + { + "epoch": 0.3453, + "grad_norm": 1.0021685535740805, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 34530 + }, + { + "epoch": 0.34531, + "grad_norm": 1.132864085671578, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 34531 + }, + { + "epoch": 0.34532, + "grad_norm": 0.92681873059181, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 34532 + }, + { + "epoch": 0.34533, + "grad_norm": 0.9258261050132089, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 34533 + }, + { + "epoch": 0.34534, + "grad_norm": 0.9170806545205432, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 34534 + }, + { + "epoch": 0.34535, + "grad_norm": 0.939374492706547, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 34535 + }, + { + "epoch": 0.34536, + "grad_norm": 0.9756903668023905, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 34536 + }, + { + "epoch": 0.34537, + "grad_norm": 0.9542404611672902, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 34537 + }, + { + "epoch": 0.34538, + "grad_norm": 0.882730782414391, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 34538 + }, + { + "epoch": 0.34539, + "grad_norm": 0.8466372056866195, + "learning_rate": 0.003, + "loss": 4.028, + "step": 34539 + }, + { + "epoch": 0.3454, + "grad_norm": 0.7696125758444318, + "learning_rate": 0.003, + "loss": 4.008, + "step": 34540 + }, + { + "epoch": 0.34541, + "grad_norm": 0.8251755349004091, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 34541 + }, + { + "epoch": 0.34542, + "grad_norm": 0.7251112502998093, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 34542 + }, + { + "epoch": 0.34543, + "grad_norm": 0.5803812827936994, + "learning_rate": 0.003, + "loss": 3.985, + "step": 34543 + }, + { + "epoch": 0.34544, + "grad_norm": 0.4845834584711425, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 34544 + }, + { + "epoch": 0.34545, + "grad_norm": 0.5068845089352839, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 34545 + }, + { + "epoch": 0.34546, + "grad_norm": 0.4607125169150372, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 34546 + }, + { + "epoch": 0.34547, + "grad_norm": 0.4966716166997889, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 34547 + }, + { + "epoch": 0.34548, + "grad_norm": 0.5538213377057469, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 34548 + }, + { + "epoch": 0.34549, + "grad_norm": 0.6084624236353674, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 34549 + }, + { + "epoch": 0.3455, + "grad_norm": 0.6000823689404557, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 34550 + }, + { + "epoch": 0.34551, + "grad_norm": 0.6480095930240921, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 34551 + }, + { + "epoch": 0.34552, + "grad_norm": 0.8538185879383512, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 34552 + }, + { + "epoch": 0.34553, + "grad_norm": 1.1379734965727955, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 34553 + }, + { + "epoch": 0.34554, + "grad_norm": 1.07424048576428, + "learning_rate": 0.003, + "loss": 3.99, + "step": 34554 + }, + { + "epoch": 0.34555, + "grad_norm": 0.7165122699632196, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 34555 + }, + { + "epoch": 0.34556, + "grad_norm": 0.5694744284347003, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 34556 + }, + { + "epoch": 0.34557, + "grad_norm": 0.664790142568161, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 34557 + }, + { + "epoch": 0.34558, + "grad_norm": 0.6504256983495245, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 34558 + }, + { + "epoch": 0.34559, + "grad_norm": 0.5784586379423562, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 34559 + }, + { + "epoch": 0.3456, + "grad_norm": 0.5325582243328215, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 34560 + }, + { + "epoch": 0.34561, + "grad_norm": 0.5481267620895507, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 34561 + }, + { + "epoch": 0.34562, + "grad_norm": 0.5409836515008456, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 34562 + }, + { + "epoch": 0.34563, + "grad_norm": 0.6542312215908354, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 34563 + }, + { + "epoch": 0.34564, + "grad_norm": 0.7175422399496296, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 34564 + }, + { + "epoch": 0.34565, + "grad_norm": 0.6732953937933944, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 34565 + }, + { + "epoch": 0.34566, + "grad_norm": 0.5908369020829956, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 34566 + }, + { + "epoch": 0.34567, + "grad_norm": 0.5590434373363438, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 34567 + }, + { + "epoch": 0.34568, + "grad_norm": 0.5275611372887329, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 34568 + }, + { + "epoch": 0.34569, + "grad_norm": 0.5111403071856525, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 34569 + }, + { + "epoch": 0.3457, + "grad_norm": 0.4968980147657393, + "learning_rate": 0.003, + "loss": 3.973, + "step": 34570 + }, + { + "epoch": 0.34571, + "grad_norm": 0.5056690331795851, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 34571 + }, + { + "epoch": 0.34572, + "grad_norm": 0.5546844750491257, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 34572 + }, + { + "epoch": 0.34573, + "grad_norm": 0.6944108730792272, + "learning_rate": 0.003, + "loss": 3.9449, + "step": 34573 + }, + { + "epoch": 0.34574, + "grad_norm": 0.9794585390168075, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 34574 + }, + { + "epoch": 0.34575, + "grad_norm": 1.002590645093892, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 34575 + }, + { + "epoch": 0.34576, + "grad_norm": 0.9687691021896264, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 34576 + }, + { + "epoch": 0.34577, + "grad_norm": 1.019115431166376, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 34577 + }, + { + "epoch": 0.34578, + "grad_norm": 0.9621622319789566, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 34578 + }, + { + "epoch": 0.34579, + "grad_norm": 0.8258501985366085, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 34579 + }, + { + "epoch": 0.3458, + "grad_norm": 0.7881174194768098, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 34580 + }, + { + "epoch": 0.34581, + "grad_norm": 0.8458914550261059, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 34581 + }, + { + "epoch": 0.34582, + "grad_norm": 0.9672984632422915, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 34582 + }, + { + "epoch": 0.34583, + "grad_norm": 0.9015372276722142, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 34583 + }, + { + "epoch": 0.34584, + "grad_norm": 0.9593454635588653, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 34584 + }, + { + "epoch": 0.34585, + "grad_norm": 0.9391287666546809, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 34585 + }, + { + "epoch": 0.34586, + "grad_norm": 1.0092373168702657, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 34586 + }, + { + "epoch": 0.34587, + "grad_norm": 1.0469239416757412, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 34587 + }, + { + "epoch": 0.34588, + "grad_norm": 0.9195732253845567, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 34588 + }, + { + "epoch": 0.34589, + "grad_norm": 0.9082517870837965, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 34589 + }, + { + "epoch": 0.3459, + "grad_norm": 0.9980365177706227, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 34590 + }, + { + "epoch": 0.34591, + "grad_norm": 1.1329062623216646, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 34591 + }, + { + "epoch": 0.34592, + "grad_norm": 1.0155391736633022, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 34592 + }, + { + "epoch": 0.34593, + "grad_norm": 0.8553549132876743, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 34593 + }, + { + "epoch": 0.34594, + "grad_norm": 0.8001810821579916, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 34594 + }, + { + "epoch": 0.34595, + "grad_norm": 0.8447749023894181, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 34595 + }, + { + "epoch": 0.34596, + "grad_norm": 0.8967148969466506, + "learning_rate": 0.003, + "loss": 4.0408, + "step": 34596 + }, + { + "epoch": 0.34597, + "grad_norm": 0.9978529922362006, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 34597 + }, + { + "epoch": 0.34598, + "grad_norm": 1.175622254262496, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 34598 + }, + { + "epoch": 0.34599, + "grad_norm": 0.9397884230658138, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 34599 + }, + { + "epoch": 0.346, + "grad_norm": 0.8438618297572745, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 34600 + }, + { + "epoch": 0.34601, + "grad_norm": 0.7936188362788934, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 34601 + }, + { + "epoch": 0.34602, + "grad_norm": 0.7098911338970787, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 34602 + }, + { + "epoch": 0.34603, + "grad_norm": 0.5583592132553611, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 34603 + }, + { + "epoch": 0.34604, + "grad_norm": 0.5200392862275083, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 34604 + }, + { + "epoch": 0.34605, + "grad_norm": 0.53696836895974, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 34605 + }, + { + "epoch": 0.34606, + "grad_norm": 0.5869826588347887, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 34606 + }, + { + "epoch": 0.34607, + "grad_norm": 0.5170766746214915, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 34607 + }, + { + "epoch": 0.34608, + "grad_norm": 0.48573851675358215, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 34608 + }, + { + "epoch": 0.34609, + "grad_norm": 0.5346644098975338, + "learning_rate": 0.003, + "loss": 3.972, + "step": 34609 + }, + { + "epoch": 0.3461, + "grad_norm": 0.5182571135901357, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 34610 + }, + { + "epoch": 0.34611, + "grad_norm": 0.5282295456351086, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 34611 + }, + { + "epoch": 0.34612, + "grad_norm": 0.5019839458660303, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 34612 + }, + { + "epoch": 0.34613, + "grad_norm": 0.5309834289550097, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 34613 + }, + { + "epoch": 0.34614, + "grad_norm": 0.6388356007346614, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 34614 + }, + { + "epoch": 0.34615, + "grad_norm": 0.7597654846883329, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 34615 + }, + { + "epoch": 0.34616, + "grad_norm": 0.9990072856989818, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 34616 + }, + { + "epoch": 0.34617, + "grad_norm": 1.2551084868312412, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 34617 + }, + { + "epoch": 0.34618, + "grad_norm": 0.6114798369928837, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 34618 + }, + { + "epoch": 0.34619, + "grad_norm": 0.5385284484394199, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 34619 + }, + { + "epoch": 0.3462, + "grad_norm": 0.6621780808204997, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 34620 + }, + { + "epoch": 0.34621, + "grad_norm": 0.6100341158230358, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 34621 + }, + { + "epoch": 0.34622, + "grad_norm": 0.592664980810067, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 34622 + }, + { + "epoch": 0.34623, + "grad_norm": 0.5548220934231071, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 34623 + }, + { + "epoch": 0.34624, + "grad_norm": 0.6647147455609169, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 34624 + }, + { + "epoch": 0.34625, + "grad_norm": 0.7904620716135136, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 34625 + }, + { + "epoch": 0.34626, + "grad_norm": 0.7831916475618694, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 34626 + }, + { + "epoch": 0.34627, + "grad_norm": 0.78211284512424, + "learning_rate": 0.003, + "loss": 3.9453, + "step": 34627 + }, + { + "epoch": 0.34628, + "grad_norm": 0.8457176252714174, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 34628 + }, + { + "epoch": 0.34629, + "grad_norm": 0.7616564382157593, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 34629 + }, + { + "epoch": 0.3463, + "grad_norm": 0.6526555749512406, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 34630 + }, + { + "epoch": 0.34631, + "grad_norm": 0.7481217768674719, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 34631 + }, + { + "epoch": 0.34632, + "grad_norm": 0.8417226653408912, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 34632 + }, + { + "epoch": 0.34633, + "grad_norm": 0.8658079312471826, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34633 + }, + { + "epoch": 0.34634, + "grad_norm": 0.8520376638617471, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 34634 + }, + { + "epoch": 0.34635, + "grad_norm": 0.8250253970408346, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 34635 + }, + { + "epoch": 0.34636, + "grad_norm": 0.8408471045223515, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 34636 + }, + { + "epoch": 0.34637, + "grad_norm": 0.8990150564144304, + "learning_rate": 0.003, + "loss": 3.985, + "step": 34637 + }, + { + "epoch": 0.34638, + "grad_norm": 0.9115502145309292, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 34638 + }, + { + "epoch": 0.34639, + "grad_norm": 0.9334406873433861, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 34639 + }, + { + "epoch": 0.3464, + "grad_norm": 1.126674073522983, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 34640 + }, + { + "epoch": 0.34641, + "grad_norm": 0.8046750393689268, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 34641 + }, + { + "epoch": 0.34642, + "grad_norm": 0.7745059369759463, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 34642 + }, + { + "epoch": 0.34643, + "grad_norm": 0.8382640545643129, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 34643 + }, + { + "epoch": 0.34644, + "grad_norm": 0.8070672162370429, + "learning_rate": 0.003, + "loss": 3.982, + "step": 34644 + }, + { + "epoch": 0.34645, + "grad_norm": 0.8367862228853232, + "learning_rate": 0.003, + "loss": 4.0259, + "step": 34645 + }, + { + "epoch": 0.34646, + "grad_norm": 0.7701081948073749, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 34646 + }, + { + "epoch": 0.34647, + "grad_norm": 0.8239879841937084, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 34647 + }, + { + "epoch": 0.34648, + "grad_norm": 0.8098424536890371, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 34648 + }, + { + "epoch": 0.34649, + "grad_norm": 0.9959230435488614, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 34649 + }, + { + "epoch": 0.3465, + "grad_norm": 1.2160148593277025, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 34650 + }, + { + "epoch": 0.34651, + "grad_norm": 0.7776315332405761, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 34651 + }, + { + "epoch": 0.34652, + "grad_norm": 0.6341095223292658, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 34652 + }, + { + "epoch": 0.34653, + "grad_norm": 0.6112243666975048, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 34653 + }, + { + "epoch": 0.34654, + "grad_norm": 0.7387452391927023, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 34654 + }, + { + "epoch": 0.34655, + "grad_norm": 0.8094147201365236, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 34655 + }, + { + "epoch": 0.34656, + "grad_norm": 0.7801102660163671, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 34656 + }, + { + "epoch": 0.34657, + "grad_norm": 0.7432563011477636, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 34657 + }, + { + "epoch": 0.34658, + "grad_norm": 0.7859122698512347, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 34658 + }, + { + "epoch": 0.34659, + "grad_norm": 0.8070116610199709, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 34659 + }, + { + "epoch": 0.3466, + "grad_norm": 0.7977389312000231, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 34660 + }, + { + "epoch": 0.34661, + "grad_norm": 0.8007804427659698, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 34661 + }, + { + "epoch": 0.34662, + "grad_norm": 0.7903615759458127, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 34662 + }, + { + "epoch": 0.34663, + "grad_norm": 0.7729956453224861, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 34663 + }, + { + "epoch": 0.34664, + "grad_norm": 0.7966773764877998, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 34664 + }, + { + "epoch": 0.34665, + "grad_norm": 0.7211621390596913, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34665 + }, + { + "epoch": 0.34666, + "grad_norm": 0.7200871173327317, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 34666 + }, + { + "epoch": 0.34667, + "grad_norm": 0.6852685208385848, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 34667 + }, + { + "epoch": 0.34668, + "grad_norm": 0.635009275580486, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 34668 + }, + { + "epoch": 0.34669, + "grad_norm": 0.7023122918053438, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 34669 + }, + { + "epoch": 0.3467, + "grad_norm": 0.8247822111416547, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 34670 + }, + { + "epoch": 0.34671, + "grad_norm": 0.9965812153043034, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 34671 + }, + { + "epoch": 0.34672, + "grad_norm": 0.9894376884739792, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 34672 + }, + { + "epoch": 0.34673, + "grad_norm": 0.9739588979174483, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 34673 + }, + { + "epoch": 0.34674, + "grad_norm": 0.9362891334777395, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 34674 + }, + { + "epoch": 0.34675, + "grad_norm": 0.8724826425819007, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 34675 + }, + { + "epoch": 0.34676, + "grad_norm": 0.819311499423285, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 34676 + }, + { + "epoch": 0.34677, + "grad_norm": 0.8240424792063369, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 34677 + }, + { + "epoch": 0.34678, + "grad_norm": 0.8377866900051366, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 34678 + }, + { + "epoch": 0.34679, + "grad_norm": 0.8555958704291454, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 34679 + }, + { + "epoch": 0.3468, + "grad_norm": 0.9252602925181008, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 34680 + }, + { + "epoch": 0.34681, + "grad_norm": 0.8460714607057139, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 34681 + }, + { + "epoch": 0.34682, + "grad_norm": 0.7742024939513834, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 34682 + }, + { + "epoch": 0.34683, + "grad_norm": 0.624366473774761, + "learning_rate": 0.003, + "loss": 3.975, + "step": 34683 + }, + { + "epoch": 0.34684, + "grad_norm": 0.5855164892985883, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 34684 + }, + { + "epoch": 0.34685, + "grad_norm": 0.7687331729918104, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 34685 + }, + { + "epoch": 0.34686, + "grad_norm": 0.819951455966055, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 34686 + }, + { + "epoch": 0.34687, + "grad_norm": 0.7620065319798353, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 34687 + }, + { + "epoch": 0.34688, + "grad_norm": 0.6556910565119801, + "learning_rate": 0.003, + "loss": 3.996, + "step": 34688 + }, + { + "epoch": 0.34689, + "grad_norm": 0.7189438713972905, + "learning_rate": 0.003, + "loss": 3.964, + "step": 34689 + }, + { + "epoch": 0.3469, + "grad_norm": 0.829622048014883, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 34690 + }, + { + "epoch": 0.34691, + "grad_norm": 0.7950725310777959, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 34691 + }, + { + "epoch": 0.34692, + "grad_norm": 0.699311681581213, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 34692 + }, + { + "epoch": 0.34693, + "grad_norm": 0.656605049696138, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 34693 + }, + { + "epoch": 0.34694, + "grad_norm": 0.6693555895646092, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 34694 + }, + { + "epoch": 0.34695, + "grad_norm": 0.6556688300141402, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 34695 + }, + { + "epoch": 0.34696, + "grad_norm": 0.7365836170373734, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 34696 + }, + { + "epoch": 0.34697, + "grad_norm": 0.724361653995037, + "learning_rate": 0.003, + "loss": 3.9276, + "step": 34697 + }, + { + "epoch": 0.34698, + "grad_norm": 0.6538749444052013, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 34698 + }, + { + "epoch": 0.34699, + "grad_norm": 0.7079182681996693, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 34699 + }, + { + "epoch": 0.347, + "grad_norm": 0.7280296884476207, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 34700 + }, + { + "epoch": 0.34701, + "grad_norm": 0.9033795588446615, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 34701 + }, + { + "epoch": 0.34702, + "grad_norm": 0.9136377720032001, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 34702 + }, + { + "epoch": 0.34703, + "grad_norm": 0.915194824991687, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 34703 + }, + { + "epoch": 0.34704, + "grad_norm": 0.680537988038325, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 34704 + }, + { + "epoch": 0.34705, + "grad_norm": 0.6396663161180235, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 34705 + }, + { + "epoch": 0.34706, + "grad_norm": 0.691172232126016, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 34706 + }, + { + "epoch": 0.34707, + "grad_norm": 0.8186051489516243, + "learning_rate": 0.003, + "loss": 3.988, + "step": 34707 + }, + { + "epoch": 0.34708, + "grad_norm": 0.8129340999906731, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 34708 + }, + { + "epoch": 0.34709, + "grad_norm": 0.7594302558698203, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 34709 + }, + { + "epoch": 0.3471, + "grad_norm": 0.8106629807780971, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 34710 + }, + { + "epoch": 0.34711, + "grad_norm": 0.8343175999816773, + "learning_rate": 0.003, + "loss": 3.986, + "step": 34711 + }, + { + "epoch": 0.34712, + "grad_norm": 0.8450255105603796, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 34712 + }, + { + "epoch": 0.34713, + "grad_norm": 0.9044736831045991, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 34713 + }, + { + "epoch": 0.34714, + "grad_norm": 0.8809976272411, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 34714 + }, + { + "epoch": 0.34715, + "grad_norm": 0.8442440192337273, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 34715 + }, + { + "epoch": 0.34716, + "grad_norm": 0.8339173518882728, + "learning_rate": 0.003, + "loss": 3.977, + "step": 34716 + }, + { + "epoch": 0.34717, + "grad_norm": 0.6820207248176045, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 34717 + }, + { + "epoch": 0.34718, + "grad_norm": 0.620289236214015, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 34718 + }, + { + "epoch": 0.34719, + "grad_norm": 0.5227546815818903, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 34719 + }, + { + "epoch": 0.3472, + "grad_norm": 0.5492611655632008, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 34720 + }, + { + "epoch": 0.34721, + "grad_norm": 0.6009578546656819, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 34721 + }, + { + "epoch": 0.34722, + "grad_norm": 0.6755611829741777, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 34722 + }, + { + "epoch": 0.34723, + "grad_norm": 0.7337097273201606, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 34723 + }, + { + "epoch": 0.34724, + "grad_norm": 0.7540082188696164, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 34724 + }, + { + "epoch": 0.34725, + "grad_norm": 0.7454064736116462, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 34725 + }, + { + "epoch": 0.34726, + "grad_norm": 0.8048717049632155, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 34726 + }, + { + "epoch": 0.34727, + "grad_norm": 0.7793098790216872, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 34727 + }, + { + "epoch": 0.34728, + "grad_norm": 0.9989755415596152, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 34728 + }, + { + "epoch": 0.34729, + "grad_norm": 1.2448839525861957, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 34729 + }, + { + "epoch": 0.3473, + "grad_norm": 0.8877017039243186, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 34730 + }, + { + "epoch": 0.34731, + "grad_norm": 0.8678108714902464, + "learning_rate": 0.003, + "loss": 4.0397, + "step": 34731 + }, + { + "epoch": 0.34732, + "grad_norm": 0.8690359315516637, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 34732 + }, + { + "epoch": 0.34733, + "grad_norm": 0.9213913149294074, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 34733 + }, + { + "epoch": 0.34734, + "grad_norm": 1.1069378134181354, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 34734 + }, + { + "epoch": 0.34735, + "grad_norm": 0.9443415741154899, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 34735 + }, + { + "epoch": 0.34736, + "grad_norm": 0.7828455928771378, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 34736 + }, + { + "epoch": 0.34737, + "grad_norm": 0.6717054710398868, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 34737 + }, + { + "epoch": 0.34738, + "grad_norm": 0.7475437922892384, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 34738 + }, + { + "epoch": 0.34739, + "grad_norm": 0.6848070015592534, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 34739 + }, + { + "epoch": 0.3474, + "grad_norm": 0.6846255545688111, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 34740 + }, + { + "epoch": 0.34741, + "grad_norm": 0.7210109610785396, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 34741 + }, + { + "epoch": 0.34742, + "grad_norm": 0.8103432277662995, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 34742 + }, + { + "epoch": 0.34743, + "grad_norm": 0.8211843547071781, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 34743 + }, + { + "epoch": 0.34744, + "grad_norm": 0.7835445707413108, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 34744 + }, + { + "epoch": 0.34745, + "grad_norm": 0.821283941868032, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 34745 + }, + { + "epoch": 0.34746, + "grad_norm": 0.7861208421382275, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 34746 + }, + { + "epoch": 0.34747, + "grad_norm": 0.6561479510836793, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 34747 + }, + { + "epoch": 0.34748, + "grad_norm": 0.660203258282951, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 34748 + }, + { + "epoch": 0.34749, + "grad_norm": 0.7586598097061281, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 34749 + }, + { + "epoch": 0.3475, + "grad_norm": 0.8056602835028877, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 34750 + }, + { + "epoch": 0.34751, + "grad_norm": 0.8459198990602361, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 34751 + }, + { + "epoch": 0.34752, + "grad_norm": 0.8484531079827956, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 34752 + }, + { + "epoch": 0.34753, + "grad_norm": 0.7679974207330568, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 34753 + }, + { + "epoch": 0.34754, + "grad_norm": 0.8077501971941643, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 34754 + }, + { + "epoch": 0.34755, + "grad_norm": 0.7233764720825072, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 34755 + }, + { + "epoch": 0.34756, + "grad_norm": 0.7567001049656905, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 34756 + }, + { + "epoch": 0.34757, + "grad_norm": 0.7657638906722419, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 34757 + }, + { + "epoch": 0.34758, + "grad_norm": 0.7449352130389405, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 34758 + }, + { + "epoch": 0.34759, + "grad_norm": 0.6557797772070024, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 34759 + }, + { + "epoch": 0.3476, + "grad_norm": 0.5819467699773107, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 34760 + }, + { + "epoch": 0.34761, + "grad_norm": 0.5765275891562186, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 34761 + }, + { + "epoch": 0.34762, + "grad_norm": 0.6616828049455267, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 34762 + }, + { + "epoch": 0.34763, + "grad_norm": 0.6827133343520579, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 34763 + }, + { + "epoch": 0.34764, + "grad_norm": 0.8043983995374219, + "learning_rate": 0.003, + "loss": 4.0, + "step": 34764 + }, + { + "epoch": 0.34765, + "grad_norm": 0.8923760933646483, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 34765 + }, + { + "epoch": 0.34766, + "grad_norm": 0.8653225028872373, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 34766 + }, + { + "epoch": 0.34767, + "grad_norm": 0.9647998080570787, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 34767 + }, + { + "epoch": 0.34768, + "grad_norm": 1.0493466268473426, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 34768 + }, + { + "epoch": 0.34769, + "grad_norm": 0.8453743503736192, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34769 + }, + { + "epoch": 0.3477, + "grad_norm": 0.7549599465822927, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 34770 + }, + { + "epoch": 0.34771, + "grad_norm": 0.7153450760614578, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 34771 + }, + { + "epoch": 0.34772, + "grad_norm": 0.7244867573422911, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 34772 + }, + { + "epoch": 0.34773, + "grad_norm": 0.8470916137335528, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 34773 + }, + { + "epoch": 0.34774, + "grad_norm": 0.8719409740134055, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34774 + }, + { + "epoch": 0.34775, + "grad_norm": 0.8566415559327685, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 34775 + }, + { + "epoch": 0.34776, + "grad_norm": 0.85893528254898, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 34776 + }, + { + "epoch": 0.34777, + "grad_norm": 0.7772446178855904, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 34777 + }, + { + "epoch": 0.34778, + "grad_norm": 0.6996523603852993, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 34778 + }, + { + "epoch": 0.34779, + "grad_norm": 0.662244736886022, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 34779 + }, + { + "epoch": 0.3478, + "grad_norm": 0.6575645224755892, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 34780 + }, + { + "epoch": 0.34781, + "grad_norm": 0.5913568309982737, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 34781 + }, + { + "epoch": 0.34782, + "grad_norm": 0.6428770913057542, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 34782 + }, + { + "epoch": 0.34783, + "grad_norm": 0.6378261365245445, + "learning_rate": 0.003, + "loss": 3.985, + "step": 34783 + }, + { + "epoch": 0.34784, + "grad_norm": 0.6618810103894059, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 34784 + }, + { + "epoch": 0.34785, + "grad_norm": 0.6992695041117191, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 34785 + }, + { + "epoch": 0.34786, + "grad_norm": 0.7512130297389273, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 34786 + }, + { + "epoch": 0.34787, + "grad_norm": 0.9009041522769371, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 34787 + }, + { + "epoch": 0.34788, + "grad_norm": 1.2707287043753892, + "learning_rate": 0.003, + "loss": 3.999, + "step": 34788 + }, + { + "epoch": 0.34789, + "grad_norm": 0.7689946173584695, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 34789 + }, + { + "epoch": 0.3479, + "grad_norm": 0.6925535311467651, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 34790 + }, + { + "epoch": 0.34791, + "grad_norm": 0.6364443389101093, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 34791 + }, + { + "epoch": 0.34792, + "grad_norm": 0.6948719523420448, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 34792 + }, + { + "epoch": 0.34793, + "grad_norm": 0.6171114820128942, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 34793 + }, + { + "epoch": 0.34794, + "grad_norm": 0.6380976208785433, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 34794 + }, + { + "epoch": 0.34795, + "grad_norm": 0.6084077629015404, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 34795 + }, + { + "epoch": 0.34796, + "grad_norm": 0.6052211695401223, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 34796 + }, + { + "epoch": 0.34797, + "grad_norm": 0.5805930419195066, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 34797 + }, + { + "epoch": 0.34798, + "grad_norm": 0.6354468762580573, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 34798 + }, + { + "epoch": 0.34799, + "grad_norm": 0.7144199247902889, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34799 + }, + { + "epoch": 0.348, + "grad_norm": 0.8002086233229098, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 34800 + }, + { + "epoch": 0.34801, + "grad_norm": 0.8553898335756376, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 34801 + }, + { + "epoch": 0.34802, + "grad_norm": 0.9106482538602209, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 34802 + }, + { + "epoch": 0.34803, + "grad_norm": 1.0384914903912503, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 34803 + }, + { + "epoch": 0.34804, + "grad_norm": 0.9488554597468459, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 34804 + }, + { + "epoch": 0.34805, + "grad_norm": 1.0260820187225599, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 34805 + }, + { + "epoch": 0.34806, + "grad_norm": 0.9772167324308508, + "learning_rate": 0.003, + "loss": 4.004, + "step": 34806 + }, + { + "epoch": 0.34807, + "grad_norm": 1.025702090300723, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 34807 + }, + { + "epoch": 0.34808, + "grad_norm": 1.1095976750473637, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 34808 + }, + { + "epoch": 0.34809, + "grad_norm": 1.1062575917234205, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 34809 + }, + { + "epoch": 0.3481, + "grad_norm": 1.2954168766784075, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 34810 + }, + { + "epoch": 0.34811, + "grad_norm": 0.9025590529755357, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 34811 + }, + { + "epoch": 0.34812, + "grad_norm": 1.026494217462317, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 34812 + }, + { + "epoch": 0.34813, + "grad_norm": 1.1527096739003915, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 34813 + }, + { + "epoch": 0.34814, + "grad_norm": 0.8372433846987508, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 34814 + }, + { + "epoch": 0.34815, + "grad_norm": 0.682095061229316, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 34815 + }, + { + "epoch": 0.34816, + "grad_norm": 0.6703571690590994, + "learning_rate": 0.003, + "loss": 3.989, + "step": 34816 + }, + { + "epoch": 0.34817, + "grad_norm": 0.6730043328795021, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 34817 + }, + { + "epoch": 0.34818, + "grad_norm": 0.671475078335034, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 34818 + }, + { + "epoch": 0.34819, + "grad_norm": 0.6695956658526793, + "learning_rate": 0.003, + "loss": 3.98, + "step": 34819 + }, + { + "epoch": 0.3482, + "grad_norm": 0.6922150063197128, + "learning_rate": 0.003, + "loss": 3.993, + "step": 34820 + }, + { + "epoch": 0.34821, + "grad_norm": 0.8743970059129051, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 34821 + }, + { + "epoch": 0.34822, + "grad_norm": 1.0963144374463702, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 34822 + }, + { + "epoch": 0.34823, + "grad_norm": 0.9850599105343288, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 34823 + }, + { + "epoch": 0.34824, + "grad_norm": 0.7156938918725613, + "learning_rate": 0.003, + "loss": 3.998, + "step": 34824 + }, + { + "epoch": 0.34825, + "grad_norm": 0.7162198694187574, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 34825 + }, + { + "epoch": 0.34826, + "grad_norm": 0.8814086018742212, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 34826 + }, + { + "epoch": 0.34827, + "grad_norm": 0.891234414060706, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 34827 + }, + { + "epoch": 0.34828, + "grad_norm": 0.7854833917666104, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 34828 + }, + { + "epoch": 0.34829, + "grad_norm": 0.7301226235574744, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 34829 + }, + { + "epoch": 0.3483, + "grad_norm": 0.7134479808155871, + "learning_rate": 0.003, + "loss": 4.004, + "step": 34830 + }, + { + "epoch": 0.34831, + "grad_norm": 0.7724489115712424, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 34831 + }, + { + "epoch": 0.34832, + "grad_norm": 0.7946600337734799, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 34832 + }, + { + "epoch": 0.34833, + "grad_norm": 0.7088625266449013, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 34833 + }, + { + "epoch": 0.34834, + "grad_norm": 0.6903589495532986, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 34834 + }, + { + "epoch": 0.34835, + "grad_norm": 0.6725882741769397, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 34835 + }, + { + "epoch": 0.34836, + "grad_norm": 0.6157664628574653, + "learning_rate": 0.003, + "loss": 3.995, + "step": 34836 + }, + { + "epoch": 0.34837, + "grad_norm": 0.6151187848315215, + "learning_rate": 0.003, + "loss": 3.991, + "step": 34837 + }, + { + "epoch": 0.34838, + "grad_norm": 0.7259745479824297, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 34838 + }, + { + "epoch": 0.34839, + "grad_norm": 0.7213490838636404, + "learning_rate": 0.003, + "loss": 3.973, + "step": 34839 + }, + { + "epoch": 0.3484, + "grad_norm": 0.5583626533204766, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 34840 + }, + { + "epoch": 0.34841, + "grad_norm": 0.5142916573814807, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 34841 + }, + { + "epoch": 0.34842, + "grad_norm": 0.4946004177478286, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 34842 + }, + { + "epoch": 0.34843, + "grad_norm": 0.5154670927189977, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 34843 + }, + { + "epoch": 0.34844, + "grad_norm": 0.5327309231690557, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 34844 + }, + { + "epoch": 0.34845, + "grad_norm": 0.5049392674637726, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 34845 + }, + { + "epoch": 0.34846, + "grad_norm": 0.5519135384800032, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 34846 + }, + { + "epoch": 0.34847, + "grad_norm": 0.595556080954557, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 34847 + }, + { + "epoch": 0.34848, + "grad_norm": 0.6192786588276801, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 34848 + }, + { + "epoch": 0.34849, + "grad_norm": 0.6652680995648584, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 34849 + }, + { + "epoch": 0.3485, + "grad_norm": 0.7553724736719389, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 34850 + }, + { + "epoch": 0.34851, + "grad_norm": 0.9194708904656893, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 34851 + }, + { + "epoch": 0.34852, + "grad_norm": 1.00087449710663, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 34852 + }, + { + "epoch": 0.34853, + "grad_norm": 0.9980826757203882, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 34853 + }, + { + "epoch": 0.34854, + "grad_norm": 0.9014084125761085, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 34854 + }, + { + "epoch": 0.34855, + "grad_norm": 0.8772342347028064, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 34855 + }, + { + "epoch": 0.34856, + "grad_norm": 1.0934094752168593, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 34856 + }, + { + "epoch": 0.34857, + "grad_norm": 0.9564382611918101, + "learning_rate": 0.003, + "loss": 4.002, + "step": 34857 + }, + { + "epoch": 0.34858, + "grad_norm": 1.0853146865566317, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 34858 + }, + { + "epoch": 0.34859, + "grad_norm": 0.9193272056849373, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 34859 + }, + { + "epoch": 0.3486, + "grad_norm": 0.8619350654449217, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 34860 + }, + { + "epoch": 0.34861, + "grad_norm": 0.8316557693878881, + "learning_rate": 0.003, + "loss": 4.009, + "step": 34861 + }, + { + "epoch": 0.34862, + "grad_norm": 0.6899030950624614, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 34862 + }, + { + "epoch": 0.34863, + "grad_norm": 0.6887807106229537, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 34863 + }, + { + "epoch": 0.34864, + "grad_norm": 0.6951464086601844, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 34864 + }, + { + "epoch": 0.34865, + "grad_norm": 0.6964070950565204, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 34865 + }, + { + "epoch": 0.34866, + "grad_norm": 0.70937951487294, + "learning_rate": 0.003, + "loss": 3.971, + "step": 34866 + }, + { + "epoch": 0.34867, + "grad_norm": 0.749331723160959, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 34867 + }, + { + "epoch": 0.34868, + "grad_norm": 0.9395002240979768, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 34868 + }, + { + "epoch": 0.34869, + "grad_norm": 0.985432239903274, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 34869 + }, + { + "epoch": 0.3487, + "grad_norm": 0.9376388493571642, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 34870 + }, + { + "epoch": 0.34871, + "grad_norm": 0.8335096708119997, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 34871 + }, + { + "epoch": 0.34872, + "grad_norm": 0.7815127663640674, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 34872 + }, + { + "epoch": 0.34873, + "grad_norm": 0.7997969610582267, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 34873 + }, + { + "epoch": 0.34874, + "grad_norm": 0.9363751093029384, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 34874 + }, + { + "epoch": 0.34875, + "grad_norm": 1.0662326496337782, + "learning_rate": 0.003, + "loss": 3.995, + "step": 34875 + }, + { + "epoch": 0.34876, + "grad_norm": 0.7878634944400105, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 34876 + }, + { + "epoch": 0.34877, + "grad_norm": 0.6905699171191518, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 34877 + }, + { + "epoch": 0.34878, + "grad_norm": 0.730477727478146, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 34878 + }, + { + "epoch": 0.34879, + "grad_norm": 0.7171647510898287, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 34879 + }, + { + "epoch": 0.3488, + "grad_norm": 0.8504580433187792, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 34880 + }, + { + "epoch": 0.34881, + "grad_norm": 1.0201051234381622, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 34881 + }, + { + "epoch": 0.34882, + "grad_norm": 1.108986795078144, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 34882 + }, + { + "epoch": 0.34883, + "grad_norm": 0.8938185069371197, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 34883 + }, + { + "epoch": 0.34884, + "grad_norm": 0.7810584768529342, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 34884 + }, + { + "epoch": 0.34885, + "grad_norm": 0.7825333489923283, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 34885 + }, + { + "epoch": 0.34886, + "grad_norm": 0.7186567972866363, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 34886 + }, + { + "epoch": 0.34887, + "grad_norm": 0.7020353682584405, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 34887 + }, + { + "epoch": 0.34888, + "grad_norm": 0.672881836040391, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 34888 + }, + { + "epoch": 0.34889, + "grad_norm": 0.6819822839110674, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 34889 + }, + { + "epoch": 0.3489, + "grad_norm": 0.716559641546245, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 34890 + }, + { + "epoch": 0.34891, + "grad_norm": 0.7548037530325272, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 34891 + }, + { + "epoch": 0.34892, + "grad_norm": 0.6640473841258083, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 34892 + }, + { + "epoch": 0.34893, + "grad_norm": 0.5962443467796702, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 34893 + }, + { + "epoch": 0.34894, + "grad_norm": 0.6815205996332551, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 34894 + }, + { + "epoch": 0.34895, + "grad_norm": 0.7608527986178597, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 34895 + }, + { + "epoch": 0.34896, + "grad_norm": 0.8438215962867064, + "learning_rate": 0.003, + "loss": 3.983, + "step": 34896 + }, + { + "epoch": 0.34897, + "grad_norm": 0.8583109347959037, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 34897 + }, + { + "epoch": 0.34898, + "grad_norm": 0.9170897261124352, + "learning_rate": 0.003, + "loss": 3.997, + "step": 34898 + }, + { + "epoch": 0.34899, + "grad_norm": 0.8733940994871391, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 34899 + }, + { + "epoch": 0.349, + "grad_norm": 0.8949830342897834, + "learning_rate": 0.003, + "loss": 4.016, + "step": 34900 + }, + { + "epoch": 0.34901, + "grad_norm": 1.1004939071118347, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 34901 + }, + { + "epoch": 0.34902, + "grad_norm": 1.0045098292915389, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 34902 + }, + { + "epoch": 0.34903, + "grad_norm": 1.0896020580065526, + "learning_rate": 0.003, + "loss": 3.971, + "step": 34903 + }, + { + "epoch": 0.34904, + "grad_norm": 0.7872851731505157, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 34904 + }, + { + "epoch": 0.34905, + "grad_norm": 0.6258893240115019, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 34905 + }, + { + "epoch": 0.34906, + "grad_norm": 0.6743179987158667, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 34906 + }, + { + "epoch": 0.34907, + "grad_norm": 0.7025350454772048, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 34907 + }, + { + "epoch": 0.34908, + "grad_norm": 0.6869677613853151, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 34908 + }, + { + "epoch": 0.34909, + "grad_norm": 0.7171373883460457, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 34909 + }, + { + "epoch": 0.3491, + "grad_norm": 0.7002738018242388, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 34910 + }, + { + "epoch": 0.34911, + "grad_norm": 0.6995804987506483, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 34911 + }, + { + "epoch": 0.34912, + "grad_norm": 0.7852672618305858, + "learning_rate": 0.003, + "loss": 3.991, + "step": 34912 + }, + { + "epoch": 0.34913, + "grad_norm": 0.7325468601286192, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 34913 + }, + { + "epoch": 0.34914, + "grad_norm": 0.5935506522050412, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 34914 + }, + { + "epoch": 0.34915, + "grad_norm": 0.5259398959729304, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 34915 + }, + { + "epoch": 0.34916, + "grad_norm": 0.5611704584104195, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 34916 + }, + { + "epoch": 0.34917, + "grad_norm": 0.5146460418414344, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 34917 + }, + { + "epoch": 0.34918, + "grad_norm": 0.522172352519563, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 34918 + }, + { + "epoch": 0.34919, + "grad_norm": 0.5264779334327766, + "learning_rate": 0.003, + "loss": 3.983, + "step": 34919 + }, + { + "epoch": 0.3492, + "grad_norm": 0.6147385353675904, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 34920 + }, + { + "epoch": 0.34921, + "grad_norm": 0.71627567845427, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 34921 + }, + { + "epoch": 0.34922, + "grad_norm": 0.8150120525294959, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 34922 + }, + { + "epoch": 0.34923, + "grad_norm": 0.891923467198432, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 34923 + }, + { + "epoch": 0.34924, + "grad_norm": 0.8761160500064893, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 34924 + }, + { + "epoch": 0.34925, + "grad_norm": 0.7411065742579024, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 34925 + }, + { + "epoch": 0.34926, + "grad_norm": 0.7236688496945832, + "learning_rate": 0.003, + "loss": 3.9354, + "step": 34926 + }, + { + "epoch": 0.34927, + "grad_norm": 0.6707780053195398, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 34927 + }, + { + "epoch": 0.34928, + "grad_norm": 0.6959398484568103, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 34928 + }, + { + "epoch": 0.34929, + "grad_norm": 0.7833966256806991, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 34929 + }, + { + "epoch": 0.3493, + "grad_norm": 0.9665748314207764, + "learning_rate": 0.003, + "loss": 3.984, + "step": 34930 + }, + { + "epoch": 0.34931, + "grad_norm": 1.1244952594603117, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 34931 + }, + { + "epoch": 0.34932, + "grad_norm": 0.9321426121778862, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 34932 + }, + { + "epoch": 0.34933, + "grad_norm": 0.7925134932897813, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 34933 + }, + { + "epoch": 0.34934, + "grad_norm": 0.7696141354411105, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 34934 + }, + { + "epoch": 0.34935, + "grad_norm": 0.8426098202650187, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 34935 + }, + { + "epoch": 0.34936, + "grad_norm": 0.8714452970946213, + "learning_rate": 0.003, + "loss": 3.987, + "step": 34936 + }, + { + "epoch": 0.34937, + "grad_norm": 0.7699213046492595, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 34937 + }, + { + "epoch": 0.34938, + "grad_norm": 0.7415665254050225, + "learning_rate": 0.003, + "loss": 4.019, + "step": 34938 + }, + { + "epoch": 0.34939, + "grad_norm": 0.8177922231145608, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 34939 + }, + { + "epoch": 0.3494, + "grad_norm": 0.7932179323798088, + "learning_rate": 0.003, + "loss": 3.987, + "step": 34940 + }, + { + "epoch": 0.34941, + "grad_norm": 0.7236928175204871, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 34941 + }, + { + "epoch": 0.34942, + "grad_norm": 0.7608091034751483, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 34942 + }, + { + "epoch": 0.34943, + "grad_norm": 0.7047211063405152, + "learning_rate": 0.003, + "loss": 3.985, + "step": 34943 + }, + { + "epoch": 0.34944, + "grad_norm": 0.7390424759007496, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 34944 + }, + { + "epoch": 0.34945, + "grad_norm": 0.7168632100802169, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 34945 + }, + { + "epoch": 0.34946, + "grad_norm": 0.846261146418119, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 34946 + }, + { + "epoch": 0.34947, + "grad_norm": 0.9721342646185456, + "learning_rate": 0.003, + "loss": 3.973, + "step": 34947 + }, + { + "epoch": 0.34948, + "grad_norm": 1.0523282202181654, + "learning_rate": 0.003, + "loss": 3.999, + "step": 34948 + }, + { + "epoch": 0.34949, + "grad_norm": 0.8139669917200718, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 34949 + }, + { + "epoch": 0.3495, + "grad_norm": 0.7440541250522961, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 34950 + }, + { + "epoch": 0.34951, + "grad_norm": 0.8160252047626679, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 34951 + }, + { + "epoch": 0.34952, + "grad_norm": 0.9264501290608168, + "learning_rate": 0.003, + "loss": 3.988, + "step": 34952 + }, + { + "epoch": 0.34953, + "grad_norm": 0.9877580521004224, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 34953 + }, + { + "epoch": 0.34954, + "grad_norm": 1.1355718327178046, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 34954 + }, + { + "epoch": 0.34955, + "grad_norm": 1.0502939098994337, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 34955 + }, + { + "epoch": 0.34956, + "grad_norm": 1.0316460529174125, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 34956 + }, + { + "epoch": 0.34957, + "grad_norm": 0.9167504810347322, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 34957 + }, + { + "epoch": 0.34958, + "grad_norm": 0.8158931155468964, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 34958 + }, + { + "epoch": 0.34959, + "grad_norm": 0.7471166952504412, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 34959 + }, + { + "epoch": 0.3496, + "grad_norm": 0.7397051659317718, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 34960 + }, + { + "epoch": 0.34961, + "grad_norm": 0.6240997288549556, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 34961 + }, + { + "epoch": 0.34962, + "grad_norm": 0.648866947481713, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 34962 + }, + { + "epoch": 0.34963, + "grad_norm": 0.6859521832254701, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 34963 + }, + { + "epoch": 0.34964, + "grad_norm": 0.7457879275176661, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 34964 + }, + { + "epoch": 0.34965, + "grad_norm": 0.7892925363678451, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 34965 + }, + { + "epoch": 0.34966, + "grad_norm": 0.7305890970566543, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 34966 + }, + { + "epoch": 0.34967, + "grad_norm": 0.7312311396361528, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 34967 + }, + { + "epoch": 0.34968, + "grad_norm": 0.7246137194791026, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 34968 + }, + { + "epoch": 0.34969, + "grad_norm": 0.7896342637609052, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 34969 + }, + { + "epoch": 0.3497, + "grad_norm": 0.8538151555538105, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 34970 + }, + { + "epoch": 0.34971, + "grad_norm": 0.7782920358368681, + "learning_rate": 0.003, + "loss": 3.991, + "step": 34971 + }, + { + "epoch": 0.34972, + "grad_norm": 0.7005896917163198, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 34972 + }, + { + "epoch": 0.34973, + "grad_norm": 0.7462691860561799, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 34973 + }, + { + "epoch": 0.34974, + "grad_norm": 0.6386143778308095, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 34974 + }, + { + "epoch": 0.34975, + "grad_norm": 0.6252927607752565, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 34975 + }, + { + "epoch": 0.34976, + "grad_norm": 0.6003515504007964, + "learning_rate": 0.003, + "loss": 3.9473, + "step": 34976 + }, + { + "epoch": 0.34977, + "grad_norm": 0.6468601071449286, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 34977 + }, + { + "epoch": 0.34978, + "grad_norm": 0.6647687392629823, + "learning_rate": 0.003, + "loss": 3.987, + "step": 34978 + }, + { + "epoch": 0.34979, + "grad_norm": 0.7090299812630432, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 34979 + }, + { + "epoch": 0.3498, + "grad_norm": 0.7321197527836898, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 34980 + }, + { + "epoch": 0.34981, + "grad_norm": 0.6823029943678487, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 34981 + }, + { + "epoch": 0.34982, + "grad_norm": 0.6741927516315572, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 34982 + }, + { + "epoch": 0.34983, + "grad_norm": 0.6166635066763855, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 34983 + }, + { + "epoch": 0.34984, + "grad_norm": 0.7529707435109018, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 34984 + }, + { + "epoch": 0.34985, + "grad_norm": 0.9589522344527159, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 34985 + }, + { + "epoch": 0.34986, + "grad_norm": 1.164820497539043, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 34986 + }, + { + "epoch": 0.34987, + "grad_norm": 0.834943595211956, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 34987 + }, + { + "epoch": 0.34988, + "grad_norm": 0.6693457699782669, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 34988 + }, + { + "epoch": 0.34989, + "grad_norm": 0.6636696614974225, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 34989 + }, + { + "epoch": 0.3499, + "grad_norm": 0.6883432429246298, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 34990 + }, + { + "epoch": 0.34991, + "grad_norm": 0.7103128399369376, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 34991 + }, + { + "epoch": 0.34992, + "grad_norm": 0.8787813667866676, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 34992 + }, + { + "epoch": 0.34993, + "grad_norm": 0.878507183911766, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 34993 + }, + { + "epoch": 0.34994, + "grad_norm": 0.9183475705297266, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 34994 + }, + { + "epoch": 0.34995, + "grad_norm": 0.8892380127845354, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 34995 + }, + { + "epoch": 0.34996, + "grad_norm": 0.9385377261959528, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 34996 + }, + { + "epoch": 0.34997, + "grad_norm": 0.9584931269843664, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 34997 + }, + { + "epoch": 0.34998, + "grad_norm": 0.9328470931891171, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 34998 + }, + { + "epoch": 0.34999, + "grad_norm": 0.8202726030903011, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 34999 + }, + { + "epoch": 0.35, + "grad_norm": 0.8903243744528959, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 35000 + }, + { + "epoch": 0.35001, + "grad_norm": 0.9477854191517203, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 35001 + }, + { + "epoch": 0.35002, + "grad_norm": 0.9549116858070917, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 35002 + }, + { + "epoch": 0.35003, + "grad_norm": 1.01647574654741, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 35003 + }, + { + "epoch": 0.35004, + "grad_norm": 0.8335099682831303, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 35004 + }, + { + "epoch": 0.35005, + "grad_norm": 0.7840278846758676, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 35005 + }, + { + "epoch": 0.35006, + "grad_norm": 0.916423630364647, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 35006 + }, + { + "epoch": 0.35007, + "grad_norm": 1.012921792578281, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 35007 + }, + { + "epoch": 0.35008, + "grad_norm": 0.9219209633474328, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 35008 + }, + { + "epoch": 0.35009, + "grad_norm": 0.7734580602653918, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 35009 + }, + { + "epoch": 0.3501, + "grad_norm": 0.6960206330777283, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 35010 + }, + { + "epoch": 0.35011, + "grad_norm": 0.8531423625334243, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 35011 + }, + { + "epoch": 0.35012, + "grad_norm": 0.9854711422202582, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 35012 + }, + { + "epoch": 0.35013, + "grad_norm": 1.0283811344833726, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 35013 + }, + { + "epoch": 0.35014, + "grad_norm": 0.951277378194737, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 35014 + }, + { + "epoch": 0.35015, + "grad_norm": 0.9050200902366206, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 35015 + }, + { + "epoch": 0.35016, + "grad_norm": 0.8874169111961365, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 35016 + }, + { + "epoch": 0.35017, + "grad_norm": 0.7562673087132956, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 35017 + }, + { + "epoch": 0.35018, + "grad_norm": 0.8287649317403567, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 35018 + }, + { + "epoch": 0.35019, + "grad_norm": 0.9046991283513127, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 35019 + }, + { + "epoch": 0.3502, + "grad_norm": 0.9467269876955546, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 35020 + }, + { + "epoch": 0.35021, + "grad_norm": 0.9986473486826952, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 35021 + }, + { + "epoch": 0.35022, + "grad_norm": 0.989224853950532, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 35022 + }, + { + "epoch": 0.35023, + "grad_norm": 0.8254232678885789, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 35023 + }, + { + "epoch": 0.35024, + "grad_norm": 0.7303390135013608, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 35024 + }, + { + "epoch": 0.35025, + "grad_norm": 0.564434230050381, + "learning_rate": 0.003, + "loss": 4.003, + "step": 35025 + }, + { + "epoch": 0.35026, + "grad_norm": 0.6041621560865433, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 35026 + }, + { + "epoch": 0.35027, + "grad_norm": 0.6499832607491719, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 35027 + }, + { + "epoch": 0.35028, + "grad_norm": 0.6049573465337056, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 35028 + }, + { + "epoch": 0.35029, + "grad_norm": 0.6234600577853535, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 35029 + }, + { + "epoch": 0.3503, + "grad_norm": 0.6025425751028075, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 35030 + }, + { + "epoch": 0.35031, + "grad_norm": 0.6100972063280379, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 35031 + }, + { + "epoch": 0.35032, + "grad_norm": 0.548225190116448, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 35032 + }, + { + "epoch": 0.35033, + "grad_norm": 0.5087518184104968, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 35033 + }, + { + "epoch": 0.35034, + "grad_norm": 0.623335146408655, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 35034 + }, + { + "epoch": 0.35035, + "grad_norm": 0.7162103084828718, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 35035 + }, + { + "epoch": 0.35036, + "grad_norm": 0.7517559684126996, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 35036 + }, + { + "epoch": 0.35037, + "grad_norm": 0.7879957583587381, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 35037 + }, + { + "epoch": 0.35038, + "grad_norm": 0.8025234190597722, + "learning_rate": 0.003, + "loss": 3.961, + "step": 35038 + }, + { + "epoch": 0.35039, + "grad_norm": 0.8840568411015236, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 35039 + }, + { + "epoch": 0.3504, + "grad_norm": 0.9279160180484846, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 35040 + }, + { + "epoch": 0.35041, + "grad_norm": 0.8242234966864597, + "learning_rate": 0.003, + "loss": 3.944, + "step": 35041 + }, + { + "epoch": 0.35042, + "grad_norm": 0.7312225533554098, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 35042 + }, + { + "epoch": 0.35043, + "grad_norm": 0.7694184131714776, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 35043 + }, + { + "epoch": 0.35044, + "grad_norm": 0.6770767499299863, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 35044 + }, + { + "epoch": 0.35045, + "grad_norm": 0.6445965484015593, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 35045 + }, + { + "epoch": 0.35046, + "grad_norm": 0.6840605057450249, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 35046 + }, + { + "epoch": 0.35047, + "grad_norm": 0.6985177419221055, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 35047 + }, + { + "epoch": 0.35048, + "grad_norm": 0.6768566690579317, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 35048 + }, + { + "epoch": 0.35049, + "grad_norm": 0.5892881223792524, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35049 + }, + { + "epoch": 0.3505, + "grad_norm": 0.537733083939155, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 35050 + }, + { + "epoch": 0.35051, + "grad_norm": 0.54713433117366, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 35051 + }, + { + "epoch": 0.35052, + "grad_norm": 0.6138418374085444, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 35052 + }, + { + "epoch": 0.35053, + "grad_norm": 0.7458247018007961, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 35053 + }, + { + "epoch": 0.35054, + "grad_norm": 0.8055148793400309, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 35054 + }, + { + "epoch": 0.35055, + "grad_norm": 0.8851610004528415, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 35055 + }, + { + "epoch": 0.35056, + "grad_norm": 1.017630959447587, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 35056 + }, + { + "epoch": 0.35057, + "grad_norm": 1.0120082161355133, + "learning_rate": 0.003, + "loss": 3.999, + "step": 35057 + }, + { + "epoch": 0.35058, + "grad_norm": 0.8814747291592374, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 35058 + }, + { + "epoch": 0.35059, + "grad_norm": 0.6573241032427828, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 35059 + }, + { + "epoch": 0.3506, + "grad_norm": 0.624707094216453, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 35060 + }, + { + "epoch": 0.35061, + "grad_norm": 0.6184225690232736, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 35061 + }, + { + "epoch": 0.35062, + "grad_norm": 0.6719856484747473, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 35062 + }, + { + "epoch": 0.35063, + "grad_norm": 0.594242364989012, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 35063 + }, + { + "epoch": 0.35064, + "grad_norm": 0.6491531127268477, + "learning_rate": 0.003, + "loss": 3.967, + "step": 35064 + }, + { + "epoch": 0.35065, + "grad_norm": 0.7257110845527307, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 35065 + }, + { + "epoch": 0.35066, + "grad_norm": 0.7753149663514951, + "learning_rate": 0.003, + "loss": 3.99, + "step": 35066 + }, + { + "epoch": 0.35067, + "grad_norm": 0.8293315496637955, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35067 + }, + { + "epoch": 0.35068, + "grad_norm": 0.9615270468168079, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 35068 + }, + { + "epoch": 0.35069, + "grad_norm": 1.0887101297263946, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 35069 + }, + { + "epoch": 0.3507, + "grad_norm": 0.8585028633748609, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 35070 + }, + { + "epoch": 0.35071, + "grad_norm": 0.8214550740095159, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 35071 + }, + { + "epoch": 0.35072, + "grad_norm": 0.9130696834551836, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 35072 + }, + { + "epoch": 0.35073, + "grad_norm": 1.0155004576371147, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 35073 + }, + { + "epoch": 0.35074, + "grad_norm": 0.9346440365220319, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 35074 + }, + { + "epoch": 0.35075, + "grad_norm": 0.7732329377713347, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 35075 + }, + { + "epoch": 0.35076, + "grad_norm": 0.8336219842501223, + "learning_rate": 0.003, + "loss": 3.978, + "step": 35076 + }, + { + "epoch": 0.35077, + "grad_norm": 0.9274469034705262, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 35077 + }, + { + "epoch": 0.35078, + "grad_norm": 1.0300810940039358, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 35078 + }, + { + "epoch": 0.35079, + "grad_norm": 1.0425140954568024, + "learning_rate": 0.003, + "loss": 4.003, + "step": 35079 + }, + { + "epoch": 0.3508, + "grad_norm": 1.0297300698836482, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 35080 + }, + { + "epoch": 0.35081, + "grad_norm": 1.0289567397271395, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 35081 + }, + { + "epoch": 0.35082, + "grad_norm": 0.9613487394841475, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 35082 + }, + { + "epoch": 0.35083, + "grad_norm": 0.8035050389916052, + "learning_rate": 0.003, + "loss": 4.008, + "step": 35083 + }, + { + "epoch": 0.35084, + "grad_norm": 0.7943572160899416, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 35084 + }, + { + "epoch": 0.35085, + "grad_norm": 0.8946852874232641, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 35085 + }, + { + "epoch": 0.35086, + "grad_norm": 0.8029527997816318, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 35086 + }, + { + "epoch": 0.35087, + "grad_norm": 0.7447963780711048, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 35087 + }, + { + "epoch": 0.35088, + "grad_norm": 0.6851610864436829, + "learning_rate": 0.003, + "loss": 3.991, + "step": 35088 + }, + { + "epoch": 0.35089, + "grad_norm": 0.6879005590461745, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 35089 + }, + { + "epoch": 0.3509, + "grad_norm": 0.6343287703499273, + "learning_rate": 0.003, + "loss": 4.005, + "step": 35090 + }, + { + "epoch": 0.35091, + "grad_norm": 0.6602731963548358, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 35091 + }, + { + "epoch": 0.35092, + "grad_norm": 0.6598876810148144, + "learning_rate": 0.003, + "loss": 3.966, + "step": 35092 + }, + { + "epoch": 0.35093, + "grad_norm": 0.7142391374865125, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 35093 + }, + { + "epoch": 0.35094, + "grad_norm": 0.7475678330962675, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 35094 + }, + { + "epoch": 0.35095, + "grad_norm": 0.7440067508522922, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 35095 + }, + { + "epoch": 0.35096, + "grad_norm": 0.7052572391033733, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 35096 + }, + { + "epoch": 0.35097, + "grad_norm": 0.7629255120142312, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 35097 + }, + { + "epoch": 0.35098, + "grad_norm": 1.0334632678293982, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 35098 + }, + { + "epoch": 0.35099, + "grad_norm": 1.1132930274738875, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 35099 + }, + { + "epoch": 0.351, + "grad_norm": 0.7448186400272139, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 35100 + }, + { + "epoch": 0.35101, + "grad_norm": 0.8678161975817981, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 35101 + }, + { + "epoch": 0.35102, + "grad_norm": 0.7934258684025436, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 35102 + }, + { + "epoch": 0.35103, + "grad_norm": 0.6724590671774107, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 35103 + }, + { + "epoch": 0.35104, + "grad_norm": 0.7546859314228401, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 35104 + }, + { + "epoch": 0.35105, + "grad_norm": 0.6938287434186544, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 35105 + }, + { + "epoch": 0.35106, + "grad_norm": 0.6374417641551684, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 35106 + }, + { + "epoch": 0.35107, + "grad_norm": 0.6373218700094393, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 35107 + }, + { + "epoch": 0.35108, + "grad_norm": 0.6295073040691214, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 35108 + }, + { + "epoch": 0.35109, + "grad_norm": 0.5979779634973402, + "learning_rate": 0.003, + "loss": 3.979, + "step": 35109 + }, + { + "epoch": 0.3511, + "grad_norm": 0.5689224620379114, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 35110 + }, + { + "epoch": 0.35111, + "grad_norm": 0.5610890369152016, + "learning_rate": 0.003, + "loss": 3.956, + "step": 35111 + }, + { + "epoch": 0.35112, + "grad_norm": 0.5179876674299987, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 35112 + }, + { + "epoch": 0.35113, + "grad_norm": 0.5035975502021163, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 35113 + }, + { + "epoch": 0.35114, + "grad_norm": 0.5758561488739063, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 35114 + }, + { + "epoch": 0.35115, + "grad_norm": 0.6607359570437864, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 35115 + }, + { + "epoch": 0.35116, + "grad_norm": 0.8759649107296588, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 35116 + }, + { + "epoch": 0.35117, + "grad_norm": 1.0891199268255978, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 35117 + }, + { + "epoch": 0.35118, + "grad_norm": 0.8116326767219566, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 35118 + }, + { + "epoch": 0.35119, + "grad_norm": 0.6810762878741473, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 35119 + }, + { + "epoch": 0.3512, + "grad_norm": 0.6977362215588522, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 35120 + }, + { + "epoch": 0.35121, + "grad_norm": 0.7727524636230622, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 35121 + }, + { + "epoch": 0.35122, + "grad_norm": 0.9239070731953622, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 35122 + }, + { + "epoch": 0.35123, + "grad_norm": 0.9222506043664969, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 35123 + }, + { + "epoch": 0.35124, + "grad_norm": 0.8619686437488844, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 35124 + }, + { + "epoch": 0.35125, + "grad_norm": 0.7741397771255982, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 35125 + }, + { + "epoch": 0.35126, + "grad_norm": 0.7224352225612325, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 35126 + }, + { + "epoch": 0.35127, + "grad_norm": 0.669266316226832, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 35127 + }, + { + "epoch": 0.35128, + "grad_norm": 0.5754756562909285, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 35128 + }, + { + "epoch": 0.35129, + "grad_norm": 0.6324758698682023, + "learning_rate": 0.003, + "loss": 3.985, + "step": 35129 + }, + { + "epoch": 0.3513, + "grad_norm": 0.6655890386968716, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 35130 + }, + { + "epoch": 0.35131, + "grad_norm": 0.7843200352875617, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 35131 + }, + { + "epoch": 0.35132, + "grad_norm": 0.8464845294394847, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 35132 + }, + { + "epoch": 0.35133, + "grad_norm": 0.7755388393054115, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 35133 + }, + { + "epoch": 0.35134, + "grad_norm": 0.8334009045128419, + "learning_rate": 0.003, + "loss": 4.011, + "step": 35134 + }, + { + "epoch": 0.35135, + "grad_norm": 1.0232518556162327, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 35135 + }, + { + "epoch": 0.35136, + "grad_norm": 0.9802164625943152, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 35136 + }, + { + "epoch": 0.35137, + "grad_norm": 0.8924184189948352, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 35137 + }, + { + "epoch": 0.35138, + "grad_norm": 0.9105617446674067, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 35138 + }, + { + "epoch": 0.35139, + "grad_norm": 0.8436137525302351, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 35139 + }, + { + "epoch": 0.3514, + "grad_norm": 0.8734596146230033, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 35140 + }, + { + "epoch": 0.35141, + "grad_norm": 0.8857274884195293, + "learning_rate": 0.003, + "loss": 3.995, + "step": 35141 + }, + { + "epoch": 0.35142, + "grad_norm": 0.9144658612633819, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 35142 + }, + { + "epoch": 0.35143, + "grad_norm": 0.8975560381947459, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 35143 + }, + { + "epoch": 0.35144, + "grad_norm": 0.8981248437987156, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 35144 + }, + { + "epoch": 0.35145, + "grad_norm": 1.0268332595532654, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 35145 + }, + { + "epoch": 0.35146, + "grad_norm": 1.0174618277613792, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 35146 + }, + { + "epoch": 0.35147, + "grad_norm": 1.121862078965585, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 35147 + }, + { + "epoch": 0.35148, + "grad_norm": 0.8995116413271744, + "learning_rate": 0.003, + "loss": 3.981, + "step": 35148 + }, + { + "epoch": 0.35149, + "grad_norm": 0.7429667520291701, + "learning_rate": 0.003, + "loss": 3.989, + "step": 35149 + }, + { + "epoch": 0.3515, + "grad_norm": 0.7282824358185572, + "learning_rate": 0.003, + "loss": 4.005, + "step": 35150 + }, + { + "epoch": 0.35151, + "grad_norm": 0.7053576781245685, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 35151 + }, + { + "epoch": 0.35152, + "grad_norm": 0.8220769252277198, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 35152 + }, + { + "epoch": 0.35153, + "grad_norm": 0.8344512933731711, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 35153 + }, + { + "epoch": 0.35154, + "grad_norm": 0.9148446902403905, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 35154 + }, + { + "epoch": 0.35155, + "grad_norm": 0.9955761841015612, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 35155 + }, + { + "epoch": 0.35156, + "grad_norm": 0.9452381127925114, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 35156 + }, + { + "epoch": 0.35157, + "grad_norm": 0.7671657362693468, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 35157 + }, + { + "epoch": 0.35158, + "grad_norm": 0.662010880348974, + "learning_rate": 0.003, + "loss": 3.976, + "step": 35158 + }, + { + "epoch": 0.35159, + "grad_norm": 0.7236614370856348, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 35159 + }, + { + "epoch": 0.3516, + "grad_norm": 0.7110713980001004, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 35160 + }, + { + "epoch": 0.35161, + "grad_norm": 0.6291071434977072, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 35161 + }, + { + "epoch": 0.35162, + "grad_norm": 0.5972494892166744, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 35162 + }, + { + "epoch": 0.35163, + "grad_norm": 0.5357582347102519, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 35163 + }, + { + "epoch": 0.35164, + "grad_norm": 0.6265319495583049, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 35164 + }, + { + "epoch": 0.35165, + "grad_norm": 0.6743322248574881, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 35165 + }, + { + "epoch": 0.35166, + "grad_norm": 0.7418459202792177, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 35166 + }, + { + "epoch": 0.35167, + "grad_norm": 0.7740442235757397, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 35167 + }, + { + "epoch": 0.35168, + "grad_norm": 0.8454981434427356, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 35168 + }, + { + "epoch": 0.35169, + "grad_norm": 0.9702589972623438, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 35169 + }, + { + "epoch": 0.3517, + "grad_norm": 0.9742477827643687, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 35170 + }, + { + "epoch": 0.35171, + "grad_norm": 0.8869742660846764, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 35171 + }, + { + "epoch": 0.35172, + "grad_norm": 0.8426087164461468, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 35172 + }, + { + "epoch": 0.35173, + "grad_norm": 0.8424108854303376, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 35173 + }, + { + "epoch": 0.35174, + "grad_norm": 0.7871921844200074, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 35174 + }, + { + "epoch": 0.35175, + "grad_norm": 0.8270452130336224, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 35175 + }, + { + "epoch": 0.35176, + "grad_norm": 0.8515773006091213, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 35176 + }, + { + "epoch": 0.35177, + "grad_norm": 0.7140423504702215, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 35177 + }, + { + "epoch": 0.35178, + "grad_norm": 0.738435955794416, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 35178 + }, + { + "epoch": 0.35179, + "grad_norm": 0.767314248334635, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 35179 + }, + { + "epoch": 0.3518, + "grad_norm": 0.769200695134979, + "learning_rate": 0.003, + "loss": 3.9427, + "step": 35180 + }, + { + "epoch": 0.35181, + "grad_norm": 0.7684038637087931, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 35181 + }, + { + "epoch": 0.35182, + "grad_norm": 0.8070336696762399, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 35182 + }, + { + "epoch": 0.35183, + "grad_norm": 0.7482207580627204, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 35183 + }, + { + "epoch": 0.35184, + "grad_norm": 0.6140092716217912, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 35184 + }, + { + "epoch": 0.35185, + "grad_norm": 0.5689069250694937, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 35185 + }, + { + "epoch": 0.35186, + "grad_norm": 0.559871162641255, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 35186 + }, + { + "epoch": 0.35187, + "grad_norm": 0.5506445269935085, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 35187 + }, + { + "epoch": 0.35188, + "grad_norm": 0.6286951472296207, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 35188 + }, + { + "epoch": 0.35189, + "grad_norm": 0.7194752864159302, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 35189 + }, + { + "epoch": 0.3519, + "grad_norm": 0.8962278696207676, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 35190 + }, + { + "epoch": 0.35191, + "grad_norm": 1.00867124152483, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 35191 + }, + { + "epoch": 0.35192, + "grad_norm": 0.8977019835489463, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 35192 + }, + { + "epoch": 0.35193, + "grad_norm": 0.679983884430616, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 35193 + }, + { + "epoch": 0.35194, + "grad_norm": 0.8427790652963326, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 35194 + }, + { + "epoch": 0.35195, + "grad_norm": 0.9518559753465087, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 35195 + }, + { + "epoch": 0.35196, + "grad_norm": 0.8937617799508619, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 35196 + }, + { + "epoch": 0.35197, + "grad_norm": 0.9416804118513958, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 35197 + }, + { + "epoch": 0.35198, + "grad_norm": 0.9110627953727788, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 35198 + }, + { + "epoch": 0.35199, + "grad_norm": 0.8127341029618395, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 35199 + }, + { + "epoch": 0.352, + "grad_norm": 0.7955520312251638, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 35200 + }, + { + "epoch": 0.35201, + "grad_norm": 0.8016691171288524, + "learning_rate": 0.003, + "loss": 3.986, + "step": 35201 + }, + { + "epoch": 0.35202, + "grad_norm": 0.833070131428296, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 35202 + }, + { + "epoch": 0.35203, + "grad_norm": 0.8555877302525411, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 35203 + }, + { + "epoch": 0.35204, + "grad_norm": 0.9443454039293888, + "learning_rate": 0.003, + "loss": 4.0373, + "step": 35204 + }, + { + "epoch": 0.35205, + "grad_norm": 0.826545514395122, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 35205 + }, + { + "epoch": 0.35206, + "grad_norm": 0.6685075592995955, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 35206 + }, + { + "epoch": 0.35207, + "grad_norm": 0.6046034431663432, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 35207 + }, + { + "epoch": 0.35208, + "grad_norm": 0.5652551448356513, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 35208 + }, + { + "epoch": 0.35209, + "grad_norm": 0.5255939782724987, + "learning_rate": 0.003, + "loss": 3.989, + "step": 35209 + }, + { + "epoch": 0.3521, + "grad_norm": 0.5859489712149528, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 35210 + }, + { + "epoch": 0.35211, + "grad_norm": 0.6496397601736231, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 35211 + }, + { + "epoch": 0.35212, + "grad_norm": 0.7210172431194988, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 35212 + }, + { + "epoch": 0.35213, + "grad_norm": 0.7851000773457087, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 35213 + }, + { + "epoch": 0.35214, + "grad_norm": 0.6935982580580488, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 35214 + }, + { + "epoch": 0.35215, + "grad_norm": 0.6611877375071818, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 35215 + }, + { + "epoch": 0.35216, + "grad_norm": 0.6765110237489163, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 35216 + }, + { + "epoch": 0.35217, + "grad_norm": 0.6800866785224909, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 35217 + }, + { + "epoch": 0.35218, + "grad_norm": 0.7013645545877565, + "learning_rate": 0.003, + "loss": 3.9475, + "step": 35218 + }, + { + "epoch": 0.35219, + "grad_norm": 0.5871371147922856, + "learning_rate": 0.003, + "loss": 3.98, + "step": 35219 + }, + { + "epoch": 0.3522, + "grad_norm": 0.6374804766529026, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 35220 + }, + { + "epoch": 0.35221, + "grad_norm": 0.6978138190783271, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 35221 + }, + { + "epoch": 0.35222, + "grad_norm": 0.8135848588015081, + "learning_rate": 0.003, + "loss": 3.984, + "step": 35222 + }, + { + "epoch": 0.35223, + "grad_norm": 0.9929816904861578, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 35223 + }, + { + "epoch": 0.35224, + "grad_norm": 0.9477150057534572, + "learning_rate": 0.003, + "loss": 3.99, + "step": 35224 + }, + { + "epoch": 0.35225, + "grad_norm": 0.8043812557527764, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 35225 + }, + { + "epoch": 0.35226, + "grad_norm": 0.7855470363138349, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 35226 + }, + { + "epoch": 0.35227, + "grad_norm": 0.7401263237992022, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 35227 + }, + { + "epoch": 0.35228, + "grad_norm": 0.8025055163385245, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 35228 + }, + { + "epoch": 0.35229, + "grad_norm": 0.8383577734060044, + "learning_rate": 0.003, + "loss": 3.981, + "step": 35229 + }, + { + "epoch": 0.3523, + "grad_norm": 0.8536493171211973, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 35230 + }, + { + "epoch": 0.35231, + "grad_norm": 0.7900765629836491, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 35231 + }, + { + "epoch": 0.35232, + "grad_norm": 0.7080075213208787, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 35232 + }, + { + "epoch": 0.35233, + "grad_norm": 0.6469292228720794, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 35233 + }, + { + "epoch": 0.35234, + "grad_norm": 0.6556292703811335, + "learning_rate": 0.003, + "loss": 4.005, + "step": 35234 + }, + { + "epoch": 0.35235, + "grad_norm": 0.7404177897723406, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 35235 + }, + { + "epoch": 0.35236, + "grad_norm": 0.6793857923895756, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 35236 + }, + { + "epoch": 0.35237, + "grad_norm": 0.6138529857320534, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 35237 + }, + { + "epoch": 0.35238, + "grad_norm": 0.6722639353197672, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 35238 + }, + { + "epoch": 0.35239, + "grad_norm": 0.7582881780308157, + "learning_rate": 0.003, + "loss": 3.981, + "step": 35239 + }, + { + "epoch": 0.3524, + "grad_norm": 0.791549376904117, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 35240 + }, + { + "epoch": 0.35241, + "grad_norm": 0.901753799193598, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 35241 + }, + { + "epoch": 0.35242, + "grad_norm": 0.991489664119746, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 35242 + }, + { + "epoch": 0.35243, + "grad_norm": 1.0093711362805355, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 35243 + }, + { + "epoch": 0.35244, + "grad_norm": 1.0020777907724505, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 35244 + }, + { + "epoch": 0.35245, + "grad_norm": 1.07323308559677, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 35245 + }, + { + "epoch": 0.35246, + "grad_norm": 0.8800190148498481, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 35246 + }, + { + "epoch": 0.35247, + "grad_norm": 0.7627134367475871, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 35247 + }, + { + "epoch": 0.35248, + "grad_norm": 0.774222584467208, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 35248 + }, + { + "epoch": 0.35249, + "grad_norm": 0.9653373458131779, + "learning_rate": 0.003, + "loss": 3.988, + "step": 35249 + }, + { + "epoch": 0.3525, + "grad_norm": 1.2378987659027665, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 35250 + }, + { + "epoch": 0.35251, + "grad_norm": 0.8500120165842283, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 35251 + }, + { + "epoch": 0.35252, + "grad_norm": 0.8585525999689816, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 35252 + }, + { + "epoch": 0.35253, + "grad_norm": 0.8879791916568326, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 35253 + }, + { + "epoch": 0.35254, + "grad_norm": 0.8374846895529252, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 35254 + }, + { + "epoch": 0.35255, + "grad_norm": 0.7075078760412431, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 35255 + }, + { + "epoch": 0.35256, + "grad_norm": 0.7507031586803892, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 35256 + }, + { + "epoch": 0.35257, + "grad_norm": 0.8376909781331675, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 35257 + }, + { + "epoch": 0.35258, + "grad_norm": 0.8451598286288334, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 35258 + }, + { + "epoch": 0.35259, + "grad_norm": 0.9174692882565547, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 35259 + }, + { + "epoch": 0.3526, + "grad_norm": 1.075277085701998, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 35260 + }, + { + "epoch": 0.35261, + "grad_norm": 1.0238760196221062, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 35261 + }, + { + "epoch": 0.35262, + "grad_norm": 0.8674753199408173, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 35262 + }, + { + "epoch": 0.35263, + "grad_norm": 0.783784549788334, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 35263 + }, + { + "epoch": 0.35264, + "grad_norm": 0.7450255842886812, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 35264 + }, + { + "epoch": 0.35265, + "grad_norm": 0.751685297666641, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 35265 + }, + { + "epoch": 0.35266, + "grad_norm": 0.8513294260194353, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 35266 + }, + { + "epoch": 0.35267, + "grad_norm": 0.7757042583816063, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 35267 + }, + { + "epoch": 0.35268, + "grad_norm": 0.8073837184475262, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 35268 + }, + { + "epoch": 0.35269, + "grad_norm": 0.7662141099567918, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 35269 + }, + { + "epoch": 0.3527, + "grad_norm": 0.6257798345038158, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 35270 + }, + { + "epoch": 0.35271, + "grad_norm": 0.6521659643672066, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 35271 + }, + { + "epoch": 0.35272, + "grad_norm": 0.6399775884409636, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 35272 + }, + { + "epoch": 0.35273, + "grad_norm": 0.6590651615190849, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 35273 + }, + { + "epoch": 0.35274, + "grad_norm": 0.7319256212304084, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 35274 + }, + { + "epoch": 0.35275, + "grad_norm": 0.7587240881286365, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 35275 + }, + { + "epoch": 0.35276, + "grad_norm": 0.7172734974469288, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 35276 + }, + { + "epoch": 0.35277, + "grad_norm": 0.6888767011710143, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 35277 + }, + { + "epoch": 0.35278, + "grad_norm": 0.7745569494033533, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 35278 + }, + { + "epoch": 0.35279, + "grad_norm": 0.7550392873285757, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 35279 + }, + { + "epoch": 0.3528, + "grad_norm": 0.8065571400988523, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 35280 + }, + { + "epoch": 0.35281, + "grad_norm": 0.8252553336788109, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 35281 + }, + { + "epoch": 0.35282, + "grad_norm": 0.8724733816900653, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 35282 + }, + { + "epoch": 0.35283, + "grad_norm": 0.8941463363140301, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 35283 + }, + { + "epoch": 0.35284, + "grad_norm": 0.8303597010834197, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 35284 + }, + { + "epoch": 0.35285, + "grad_norm": 0.7679349461104323, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 35285 + }, + { + "epoch": 0.35286, + "grad_norm": 0.7504020331599559, + "learning_rate": 0.003, + "loss": 3.977, + "step": 35286 + }, + { + "epoch": 0.35287, + "grad_norm": 0.878899667727674, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 35287 + }, + { + "epoch": 0.35288, + "grad_norm": 1.0579269040099897, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 35288 + }, + { + "epoch": 0.35289, + "grad_norm": 0.9301903753278686, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 35289 + }, + { + "epoch": 0.3529, + "grad_norm": 0.9295024308101909, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 35290 + }, + { + "epoch": 0.35291, + "grad_norm": 0.8089410130505252, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 35291 + }, + { + "epoch": 0.35292, + "grad_norm": 0.7976166077256627, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 35292 + }, + { + "epoch": 0.35293, + "grad_norm": 0.7010141532842858, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 35293 + }, + { + "epoch": 0.35294, + "grad_norm": 0.7042748128224153, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 35294 + }, + { + "epoch": 0.35295, + "grad_norm": 0.7449660372128635, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 35295 + }, + { + "epoch": 0.35296, + "grad_norm": 0.6937216951608792, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 35296 + }, + { + "epoch": 0.35297, + "grad_norm": 0.6339029867544108, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 35297 + }, + { + "epoch": 0.35298, + "grad_norm": 0.737661132091014, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 35298 + }, + { + "epoch": 0.35299, + "grad_norm": 0.6770717808242984, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35299 + }, + { + "epoch": 0.353, + "grad_norm": 0.5671894010464263, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 35300 + }, + { + "epoch": 0.35301, + "grad_norm": 0.5566463715130063, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 35301 + }, + { + "epoch": 0.35302, + "grad_norm": 0.5710859500014515, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 35302 + }, + { + "epoch": 0.35303, + "grad_norm": 0.5952302868102828, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 35303 + }, + { + "epoch": 0.35304, + "grad_norm": 0.6423405841447043, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 35304 + }, + { + "epoch": 0.35305, + "grad_norm": 0.7278582448692761, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 35305 + }, + { + "epoch": 0.35306, + "grad_norm": 0.8066906934622549, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 35306 + }, + { + "epoch": 0.35307, + "grad_norm": 0.9906365601208028, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 35307 + }, + { + "epoch": 0.35308, + "grad_norm": 1.087057577200313, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 35308 + }, + { + "epoch": 0.35309, + "grad_norm": 0.7834622354440436, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 35309 + }, + { + "epoch": 0.3531, + "grad_norm": 0.7773220356804946, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 35310 + }, + { + "epoch": 0.35311, + "grad_norm": 0.829647728524844, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 35311 + }, + { + "epoch": 0.35312, + "grad_norm": 0.7425137657991339, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 35312 + }, + { + "epoch": 0.35313, + "grad_norm": 0.6397365687957433, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 35313 + }, + { + "epoch": 0.35314, + "grad_norm": 0.6512249588297362, + "learning_rate": 0.003, + "loss": 3.938, + "step": 35314 + }, + { + "epoch": 0.35315, + "grad_norm": 0.7365156800327416, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 35315 + }, + { + "epoch": 0.35316, + "grad_norm": 0.6995845735554747, + "learning_rate": 0.003, + "loss": 3.9342, + "step": 35316 + }, + { + "epoch": 0.35317, + "grad_norm": 0.6332932514735726, + "learning_rate": 0.003, + "loss": 3.962, + "step": 35317 + }, + { + "epoch": 0.35318, + "grad_norm": 0.6527199045452428, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 35318 + }, + { + "epoch": 0.35319, + "grad_norm": 0.7438875187116398, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 35319 + }, + { + "epoch": 0.3532, + "grad_norm": 0.7472085368456367, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 35320 + }, + { + "epoch": 0.35321, + "grad_norm": 0.7672965790854946, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 35321 + }, + { + "epoch": 0.35322, + "grad_norm": 0.8547191076375731, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 35322 + }, + { + "epoch": 0.35323, + "grad_norm": 1.0241030858036058, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 35323 + }, + { + "epoch": 0.35324, + "grad_norm": 1.0149791382047681, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 35324 + }, + { + "epoch": 0.35325, + "grad_norm": 0.8343823837849879, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 35325 + }, + { + "epoch": 0.35326, + "grad_norm": 0.8316248466992976, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 35326 + }, + { + "epoch": 0.35327, + "grad_norm": 0.8478828370683472, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 35327 + }, + { + "epoch": 0.35328, + "grad_norm": 0.7731423165003826, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 35328 + }, + { + "epoch": 0.35329, + "grad_norm": 0.7546894009842994, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 35329 + }, + { + "epoch": 0.3533, + "grad_norm": 0.7977254930525408, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 35330 + }, + { + "epoch": 0.35331, + "grad_norm": 0.966666589668533, + "learning_rate": 0.003, + "loss": 4.02, + "step": 35331 + }, + { + "epoch": 0.35332, + "grad_norm": 0.9771901375656873, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 35332 + }, + { + "epoch": 0.35333, + "grad_norm": 0.943440079891669, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 35333 + }, + { + "epoch": 0.35334, + "grad_norm": 0.9424518069985606, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 35334 + }, + { + "epoch": 0.35335, + "grad_norm": 0.8990084117567437, + "learning_rate": 0.003, + "loss": 4.002, + "step": 35335 + }, + { + "epoch": 0.35336, + "grad_norm": 0.8439029472535591, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 35336 + }, + { + "epoch": 0.35337, + "grad_norm": 0.8298530709915036, + "learning_rate": 0.003, + "loss": 3.967, + "step": 35337 + }, + { + "epoch": 0.35338, + "grad_norm": 0.690701438290764, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 35338 + }, + { + "epoch": 0.35339, + "grad_norm": 0.6109033228136536, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 35339 + }, + { + "epoch": 0.3534, + "grad_norm": 0.6639082079847702, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 35340 + }, + { + "epoch": 0.35341, + "grad_norm": 0.6581995887243517, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 35341 + }, + { + "epoch": 0.35342, + "grad_norm": 0.60621573979601, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 35342 + }, + { + "epoch": 0.35343, + "grad_norm": 0.6872363244080597, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 35343 + }, + { + "epoch": 0.35344, + "grad_norm": 0.8684656301435701, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 35344 + }, + { + "epoch": 0.35345, + "grad_norm": 0.8857147402083894, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 35345 + }, + { + "epoch": 0.35346, + "grad_norm": 0.8702393873908972, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 35346 + }, + { + "epoch": 0.35347, + "grad_norm": 0.8998186068701604, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 35347 + }, + { + "epoch": 0.35348, + "grad_norm": 0.8575327844605694, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 35348 + }, + { + "epoch": 0.35349, + "grad_norm": 0.9938412200285683, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 35349 + }, + { + "epoch": 0.3535, + "grad_norm": 1.1834877528468193, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 35350 + }, + { + "epoch": 0.35351, + "grad_norm": 0.9301656107782601, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 35351 + }, + { + "epoch": 0.35352, + "grad_norm": 0.9480241386718654, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 35352 + }, + { + "epoch": 0.35353, + "grad_norm": 0.9213326682161619, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 35353 + }, + { + "epoch": 0.35354, + "grad_norm": 0.8198951799304933, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 35354 + }, + { + "epoch": 0.35355, + "grad_norm": 0.716703467114866, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 35355 + }, + { + "epoch": 0.35356, + "grad_norm": 0.6658115853992342, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 35356 + }, + { + "epoch": 0.35357, + "grad_norm": 0.7402204409889601, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 35357 + }, + { + "epoch": 0.35358, + "grad_norm": 0.7664145133437602, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 35358 + }, + { + "epoch": 0.35359, + "grad_norm": 0.7593524610610143, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 35359 + }, + { + "epoch": 0.3536, + "grad_norm": 0.6804061154354848, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 35360 + }, + { + "epoch": 0.35361, + "grad_norm": 0.6921187105839803, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 35361 + }, + { + "epoch": 0.35362, + "grad_norm": 0.6461842285943236, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 35362 + }, + { + "epoch": 0.35363, + "grad_norm": 0.5921667816239918, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 35363 + }, + { + "epoch": 0.35364, + "grad_norm": 0.5600367916283101, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 35364 + }, + { + "epoch": 0.35365, + "grad_norm": 0.6162865868289445, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 35365 + }, + { + "epoch": 0.35366, + "grad_norm": 0.6112337377028648, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 35366 + }, + { + "epoch": 0.35367, + "grad_norm": 0.6622452673573525, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 35367 + }, + { + "epoch": 0.35368, + "grad_norm": 0.703366022303163, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 35368 + }, + { + "epoch": 0.35369, + "grad_norm": 0.70885890577448, + "learning_rate": 0.003, + "loss": 3.9377, + "step": 35369 + }, + { + "epoch": 0.3537, + "grad_norm": 0.9163031852809645, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 35370 + }, + { + "epoch": 0.35371, + "grad_norm": 1.20715296372092, + "learning_rate": 0.003, + "loss": 4.014, + "step": 35371 + }, + { + "epoch": 0.35372, + "grad_norm": 0.6981331590545259, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 35372 + }, + { + "epoch": 0.35373, + "grad_norm": 0.7568055672475779, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 35373 + }, + { + "epoch": 0.35374, + "grad_norm": 0.7192813761618958, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 35374 + }, + { + "epoch": 0.35375, + "grad_norm": 0.7550389061708063, + "learning_rate": 0.003, + "loss": 3.98, + "step": 35375 + }, + { + "epoch": 0.35376, + "grad_norm": 0.8335880364599136, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 35376 + }, + { + "epoch": 0.35377, + "grad_norm": 0.8583475330755386, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 35377 + }, + { + "epoch": 0.35378, + "grad_norm": 0.8057018815510162, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 35378 + }, + { + "epoch": 0.35379, + "grad_norm": 0.7752174444157116, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 35379 + }, + { + "epoch": 0.3538, + "grad_norm": 0.8624655750687317, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 35380 + }, + { + "epoch": 0.35381, + "grad_norm": 0.955402963431668, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 35381 + }, + { + "epoch": 0.35382, + "grad_norm": 0.9957578310434146, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 35382 + }, + { + "epoch": 0.35383, + "grad_norm": 1.0189571585450246, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 35383 + }, + { + "epoch": 0.35384, + "grad_norm": 0.9323308860623984, + "learning_rate": 0.003, + "loss": 3.995, + "step": 35384 + }, + { + "epoch": 0.35385, + "grad_norm": 0.8402372447210306, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 35385 + }, + { + "epoch": 0.35386, + "grad_norm": 0.8697156682418397, + "learning_rate": 0.003, + "loss": 3.988, + "step": 35386 + }, + { + "epoch": 0.35387, + "grad_norm": 1.0158592785838885, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 35387 + }, + { + "epoch": 0.35388, + "grad_norm": 1.0320795649371763, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 35388 + }, + { + "epoch": 0.35389, + "grad_norm": 0.9623896895706622, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 35389 + }, + { + "epoch": 0.3539, + "grad_norm": 0.896224396185343, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 35390 + }, + { + "epoch": 0.35391, + "grad_norm": 0.7890050312375666, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 35391 + }, + { + "epoch": 0.35392, + "grad_norm": 0.7368913056103593, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 35392 + }, + { + "epoch": 0.35393, + "grad_norm": 0.6972769721879877, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 35393 + }, + { + "epoch": 0.35394, + "grad_norm": 0.7499531600582074, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 35394 + }, + { + "epoch": 0.35395, + "grad_norm": 0.7732159397508519, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 35395 + }, + { + "epoch": 0.35396, + "grad_norm": 0.8964113762930592, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 35396 + }, + { + "epoch": 0.35397, + "grad_norm": 0.9326188649662972, + "learning_rate": 0.003, + "loss": 4.026, + "step": 35397 + }, + { + "epoch": 0.35398, + "grad_norm": 0.8638472003490948, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 35398 + }, + { + "epoch": 0.35399, + "grad_norm": 0.6933612783807711, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 35399 + }, + { + "epoch": 0.354, + "grad_norm": 0.6943114532196061, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 35400 + }, + { + "epoch": 0.35401, + "grad_norm": 0.6224376818694005, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 35401 + }, + { + "epoch": 0.35402, + "grad_norm": 0.6032439017591434, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 35402 + }, + { + "epoch": 0.35403, + "grad_norm": 0.6425184673449571, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 35403 + }, + { + "epoch": 0.35404, + "grad_norm": 0.6691359939581684, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 35404 + }, + { + "epoch": 0.35405, + "grad_norm": 0.6473182374703839, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 35405 + }, + { + "epoch": 0.35406, + "grad_norm": 0.625147639479095, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 35406 + }, + { + "epoch": 0.35407, + "grad_norm": 0.5819534415704684, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 35407 + }, + { + "epoch": 0.35408, + "grad_norm": 0.6208884444214644, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 35408 + }, + { + "epoch": 0.35409, + "grad_norm": 0.5878224842519727, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 35409 + }, + { + "epoch": 0.3541, + "grad_norm": 0.5858263241656942, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 35410 + }, + { + "epoch": 0.35411, + "grad_norm": 0.5872290947135378, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 35411 + }, + { + "epoch": 0.35412, + "grad_norm": 0.5786141419748668, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 35412 + }, + { + "epoch": 0.35413, + "grad_norm": 0.6268984146055695, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 35413 + }, + { + "epoch": 0.35414, + "grad_norm": 0.6141863080736925, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 35414 + }, + { + "epoch": 0.35415, + "grad_norm": 0.6042097905601251, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 35415 + }, + { + "epoch": 0.35416, + "grad_norm": 0.6340868222027165, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 35416 + }, + { + "epoch": 0.35417, + "grad_norm": 0.7372201932438378, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 35417 + }, + { + "epoch": 0.35418, + "grad_norm": 0.8388552901899043, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 35418 + }, + { + "epoch": 0.35419, + "grad_norm": 1.0660576214363797, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 35419 + }, + { + "epoch": 0.3542, + "grad_norm": 1.1215594237057314, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 35420 + }, + { + "epoch": 0.35421, + "grad_norm": 0.8046068980597274, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 35421 + }, + { + "epoch": 0.35422, + "grad_norm": 0.6686363163673289, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 35422 + }, + { + "epoch": 0.35423, + "grad_norm": 0.6594010537372805, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 35423 + }, + { + "epoch": 0.35424, + "grad_norm": 0.6506910175648156, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 35424 + }, + { + "epoch": 0.35425, + "grad_norm": 0.6963219199314024, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 35425 + }, + { + "epoch": 0.35426, + "grad_norm": 0.805184164139208, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 35426 + }, + { + "epoch": 0.35427, + "grad_norm": 0.7951069943203765, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 35427 + }, + { + "epoch": 0.35428, + "grad_norm": 0.7916595361935131, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 35428 + }, + { + "epoch": 0.35429, + "grad_norm": 0.7075980682994478, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 35429 + }, + { + "epoch": 0.3543, + "grad_norm": 0.5715658867938771, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 35430 + }, + { + "epoch": 0.35431, + "grad_norm": 0.6405299860694649, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 35431 + }, + { + "epoch": 0.35432, + "grad_norm": 0.6313990655352308, + "learning_rate": 0.003, + "loss": 3.967, + "step": 35432 + }, + { + "epoch": 0.35433, + "grad_norm": 0.5898902200610359, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 35433 + }, + { + "epoch": 0.35434, + "grad_norm": 0.5634671326849607, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 35434 + }, + { + "epoch": 0.35435, + "grad_norm": 0.5676651365085336, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 35435 + }, + { + "epoch": 0.35436, + "grad_norm": 0.7220153107509684, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 35436 + }, + { + "epoch": 0.35437, + "grad_norm": 0.7816243211988316, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 35437 + }, + { + "epoch": 0.35438, + "grad_norm": 0.7535474807746095, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 35438 + }, + { + "epoch": 0.35439, + "grad_norm": 0.6959667534988673, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 35439 + }, + { + "epoch": 0.3544, + "grad_norm": 0.543261596956507, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 35440 + }, + { + "epoch": 0.35441, + "grad_norm": 0.5257703219323926, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 35441 + }, + { + "epoch": 0.35442, + "grad_norm": 0.5975411740678588, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 35442 + }, + { + "epoch": 0.35443, + "grad_norm": 0.7418115101185594, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 35443 + }, + { + "epoch": 0.35444, + "grad_norm": 0.9781455272552838, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 35444 + }, + { + "epoch": 0.35445, + "grad_norm": 1.0479754712225766, + "learning_rate": 0.003, + "loss": 4.0379, + "step": 35445 + }, + { + "epoch": 0.35446, + "grad_norm": 0.9481465293348388, + "learning_rate": 0.003, + "loss": 3.961, + "step": 35446 + }, + { + "epoch": 0.35447, + "grad_norm": 1.0126060730576514, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 35447 + }, + { + "epoch": 0.35448, + "grad_norm": 1.0990664892340662, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 35448 + }, + { + "epoch": 0.35449, + "grad_norm": 1.181590748496734, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 35449 + }, + { + "epoch": 0.3545, + "grad_norm": 0.9842358588115236, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 35450 + }, + { + "epoch": 0.35451, + "grad_norm": 1.0750423191803753, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 35451 + }, + { + "epoch": 0.35452, + "grad_norm": 0.9533637730427588, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 35452 + }, + { + "epoch": 0.35453, + "grad_norm": 0.9105120128448195, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 35453 + }, + { + "epoch": 0.35454, + "grad_norm": 0.8787124944004169, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 35454 + }, + { + "epoch": 0.35455, + "grad_norm": 0.8973975032559364, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35455 + }, + { + "epoch": 0.35456, + "grad_norm": 0.918456466021666, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 35456 + }, + { + "epoch": 0.35457, + "grad_norm": 0.8353946878353345, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 35457 + }, + { + "epoch": 0.35458, + "grad_norm": 0.8208056438422113, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 35458 + }, + { + "epoch": 0.35459, + "grad_norm": 0.8023047513496299, + "learning_rate": 0.003, + "loss": 3.991, + "step": 35459 + }, + { + "epoch": 0.3546, + "grad_norm": 0.777971705314103, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 35460 + }, + { + "epoch": 0.35461, + "grad_norm": 0.9120659812160249, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 35461 + }, + { + "epoch": 0.35462, + "grad_norm": 1.1280406406912373, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 35462 + }, + { + "epoch": 0.35463, + "grad_norm": 1.0478836607119713, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 35463 + }, + { + "epoch": 0.35464, + "grad_norm": 1.0544257384838223, + "learning_rate": 0.003, + "loss": 3.979, + "step": 35464 + }, + { + "epoch": 0.35465, + "grad_norm": 0.9552322818645534, + "learning_rate": 0.003, + "loss": 4.014, + "step": 35465 + }, + { + "epoch": 0.35466, + "grad_norm": 0.905697110294471, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 35466 + }, + { + "epoch": 0.35467, + "grad_norm": 0.9057890240811255, + "learning_rate": 0.003, + "loss": 4.002, + "step": 35467 + }, + { + "epoch": 0.35468, + "grad_norm": 0.8118138560458572, + "learning_rate": 0.003, + "loss": 4.014, + "step": 35468 + }, + { + "epoch": 0.35469, + "grad_norm": 0.8021080657121514, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 35469 + }, + { + "epoch": 0.3547, + "grad_norm": 0.8026162570281353, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 35470 + }, + { + "epoch": 0.35471, + "grad_norm": 0.6373634619640578, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 35471 + }, + { + "epoch": 0.35472, + "grad_norm": 0.593469969277902, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 35472 + }, + { + "epoch": 0.35473, + "grad_norm": 0.5995711876921903, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 35473 + }, + { + "epoch": 0.35474, + "grad_norm": 0.7306732411820367, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 35474 + }, + { + "epoch": 0.35475, + "grad_norm": 0.7718470754967452, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 35475 + }, + { + "epoch": 0.35476, + "grad_norm": 0.8395776815361596, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 35476 + }, + { + "epoch": 0.35477, + "grad_norm": 0.780133216651119, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35477 + }, + { + "epoch": 0.35478, + "grad_norm": 0.6931281013987679, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 35478 + }, + { + "epoch": 0.35479, + "grad_norm": 0.6070100783888545, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 35479 + }, + { + "epoch": 0.3548, + "grad_norm": 0.5748912805125825, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 35480 + }, + { + "epoch": 0.35481, + "grad_norm": 0.63961434450767, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 35481 + }, + { + "epoch": 0.35482, + "grad_norm": 0.6583723587310887, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 35482 + }, + { + "epoch": 0.35483, + "grad_norm": 0.5772338886161298, + "learning_rate": 0.003, + "loss": 3.972, + "step": 35483 + }, + { + "epoch": 0.35484, + "grad_norm": 0.4901786479141106, + "learning_rate": 0.003, + "loss": 3.998, + "step": 35484 + }, + { + "epoch": 0.35485, + "grad_norm": 0.5439527037942317, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 35485 + }, + { + "epoch": 0.35486, + "grad_norm": 0.5456204225155848, + "learning_rate": 0.003, + "loss": 3.978, + "step": 35486 + }, + { + "epoch": 0.35487, + "grad_norm": 0.6728408450275941, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 35487 + }, + { + "epoch": 0.35488, + "grad_norm": 0.8423615071145342, + "learning_rate": 0.003, + "loss": 3.972, + "step": 35488 + }, + { + "epoch": 0.35489, + "grad_norm": 0.9652182513109531, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 35489 + }, + { + "epoch": 0.3549, + "grad_norm": 1.0731218341193862, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 35490 + }, + { + "epoch": 0.35491, + "grad_norm": 0.8660301179722975, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 35491 + }, + { + "epoch": 0.35492, + "grad_norm": 0.7963616591012429, + "learning_rate": 0.003, + "loss": 3.993, + "step": 35492 + }, + { + "epoch": 0.35493, + "grad_norm": 0.8234678051785309, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 35493 + }, + { + "epoch": 0.35494, + "grad_norm": 0.8850124065397298, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 35494 + }, + { + "epoch": 0.35495, + "grad_norm": 0.7986985882086995, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 35495 + }, + { + "epoch": 0.35496, + "grad_norm": 0.7984444324313134, + "learning_rate": 0.003, + "loss": 3.9414, + "step": 35496 + }, + { + "epoch": 0.35497, + "grad_norm": 0.8593081170125779, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 35497 + }, + { + "epoch": 0.35498, + "grad_norm": 1.004254572357562, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 35498 + }, + { + "epoch": 0.35499, + "grad_norm": 1.0566522222830226, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 35499 + }, + { + "epoch": 0.355, + "grad_norm": 0.7581062256873788, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 35500 + }, + { + "epoch": 0.35501, + "grad_norm": 0.8161173378336305, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 35501 + }, + { + "epoch": 0.35502, + "grad_norm": 0.917927714027004, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 35502 + }, + { + "epoch": 0.35503, + "grad_norm": 0.912004422578734, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 35503 + }, + { + "epoch": 0.35504, + "grad_norm": 0.9124176243153621, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 35504 + }, + { + "epoch": 0.35505, + "grad_norm": 0.9399490222343301, + "learning_rate": 0.003, + "loss": 4.002, + "step": 35505 + }, + { + "epoch": 0.35506, + "grad_norm": 0.8689357301603601, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 35506 + }, + { + "epoch": 0.35507, + "grad_norm": 0.7379438720987997, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 35507 + }, + { + "epoch": 0.35508, + "grad_norm": 0.8651395756128508, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 35508 + }, + { + "epoch": 0.35509, + "grad_norm": 0.8699836069746909, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 35509 + }, + { + "epoch": 0.3551, + "grad_norm": 0.8649470365078773, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 35510 + }, + { + "epoch": 0.35511, + "grad_norm": 0.7545757795985819, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 35511 + }, + { + "epoch": 0.35512, + "grad_norm": 0.7432286670361977, + "learning_rate": 0.003, + "loss": 4.006, + "step": 35512 + }, + { + "epoch": 0.35513, + "grad_norm": 0.677096599894341, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 35513 + }, + { + "epoch": 0.35514, + "grad_norm": 0.6272643253376298, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 35514 + }, + { + "epoch": 0.35515, + "grad_norm": 0.6014843840918238, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 35515 + }, + { + "epoch": 0.35516, + "grad_norm": 0.5612556635735406, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 35516 + }, + { + "epoch": 0.35517, + "grad_norm": 0.5724811503832979, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 35517 + }, + { + "epoch": 0.35518, + "grad_norm": 0.5786145768796307, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 35518 + }, + { + "epoch": 0.35519, + "grad_norm": 0.5595564027870885, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 35519 + }, + { + "epoch": 0.3552, + "grad_norm": 0.601816196240334, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 35520 + }, + { + "epoch": 0.35521, + "grad_norm": 0.5536244648163662, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 35521 + }, + { + "epoch": 0.35522, + "grad_norm": 0.5352291849357353, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 35522 + }, + { + "epoch": 0.35523, + "grad_norm": 0.55280483979541, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 35523 + }, + { + "epoch": 0.35524, + "grad_norm": 0.6458105821751907, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 35524 + }, + { + "epoch": 0.35525, + "grad_norm": 0.6090896369350148, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 35525 + }, + { + "epoch": 0.35526, + "grad_norm": 0.5623934687627222, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 35526 + }, + { + "epoch": 0.35527, + "grad_norm": 0.6571140113153346, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 35527 + }, + { + "epoch": 0.35528, + "grad_norm": 0.7280954806346054, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 35528 + }, + { + "epoch": 0.35529, + "grad_norm": 0.723718086620164, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 35529 + }, + { + "epoch": 0.3553, + "grad_norm": 0.7169642836699902, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 35530 + }, + { + "epoch": 0.35531, + "grad_norm": 0.8858819236276312, + "learning_rate": 0.003, + "loss": 3.978, + "step": 35531 + }, + { + "epoch": 0.35532, + "grad_norm": 1.076841955723765, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 35532 + }, + { + "epoch": 0.35533, + "grad_norm": 1.1669995394329755, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 35533 + }, + { + "epoch": 0.35534, + "grad_norm": 0.8232884087985999, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 35534 + }, + { + "epoch": 0.35535, + "grad_norm": 0.8045239386422811, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 35535 + }, + { + "epoch": 0.35536, + "grad_norm": 0.9278822592485458, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 35536 + }, + { + "epoch": 0.35537, + "grad_norm": 0.9824429131439949, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 35537 + }, + { + "epoch": 0.35538, + "grad_norm": 1.1233553077453666, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 35538 + }, + { + "epoch": 0.35539, + "grad_norm": 0.843075320030642, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 35539 + }, + { + "epoch": 0.3554, + "grad_norm": 0.8717692094848617, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 35540 + }, + { + "epoch": 0.35541, + "grad_norm": 0.9395928991347757, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 35541 + }, + { + "epoch": 0.35542, + "grad_norm": 0.8718036826088882, + "learning_rate": 0.003, + "loss": 4.008, + "step": 35542 + }, + { + "epoch": 0.35543, + "grad_norm": 0.8297884108427753, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 35543 + }, + { + "epoch": 0.35544, + "grad_norm": 0.8647045134639326, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 35544 + }, + { + "epoch": 0.35545, + "grad_norm": 0.8250815574181756, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 35545 + }, + { + "epoch": 0.35546, + "grad_norm": 0.6587231492313167, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 35546 + }, + { + "epoch": 0.35547, + "grad_norm": 0.6828979212277232, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 35547 + }, + { + "epoch": 0.35548, + "grad_norm": 0.7388782958416166, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 35548 + }, + { + "epoch": 0.35549, + "grad_norm": 0.7187806134602264, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 35549 + }, + { + "epoch": 0.3555, + "grad_norm": 0.6199613539132961, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 35550 + }, + { + "epoch": 0.35551, + "grad_norm": 0.5533077479222785, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 35551 + }, + { + "epoch": 0.35552, + "grad_norm": 0.5649344290878027, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 35552 + }, + { + "epoch": 0.35553, + "grad_norm": 0.5769754110957034, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 35553 + }, + { + "epoch": 0.35554, + "grad_norm": 0.6470108146365483, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 35554 + }, + { + "epoch": 0.35555, + "grad_norm": 0.7564563981013352, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 35555 + }, + { + "epoch": 0.35556, + "grad_norm": 0.8121824002436308, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 35556 + }, + { + "epoch": 0.35557, + "grad_norm": 0.9168901621292134, + "learning_rate": 0.003, + "loss": 3.97, + "step": 35557 + }, + { + "epoch": 0.35558, + "grad_norm": 0.9711942183161434, + "learning_rate": 0.003, + "loss": 4.005, + "step": 35558 + }, + { + "epoch": 0.35559, + "grad_norm": 0.8834719977620541, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 35559 + }, + { + "epoch": 0.3556, + "grad_norm": 0.7524365117523197, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 35560 + }, + { + "epoch": 0.35561, + "grad_norm": 0.8314960242046088, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 35561 + }, + { + "epoch": 0.35562, + "grad_norm": 0.8876328518776356, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 35562 + }, + { + "epoch": 0.35563, + "grad_norm": 0.8289717840262316, + "learning_rate": 0.003, + "loss": 3.976, + "step": 35563 + }, + { + "epoch": 0.35564, + "grad_norm": 0.7439970698402438, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 35564 + }, + { + "epoch": 0.35565, + "grad_norm": 0.7641384106043213, + "learning_rate": 0.003, + "loss": 3.97, + "step": 35565 + }, + { + "epoch": 0.35566, + "grad_norm": 0.6410437540672357, + "learning_rate": 0.003, + "loss": 3.996, + "step": 35566 + }, + { + "epoch": 0.35567, + "grad_norm": 0.6015318614138794, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 35567 + }, + { + "epoch": 0.35568, + "grad_norm": 0.6333416748748014, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 35568 + }, + { + "epoch": 0.35569, + "grad_norm": 0.681366462421066, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 35569 + }, + { + "epoch": 0.3557, + "grad_norm": 0.7622668170765392, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 35570 + }, + { + "epoch": 0.35571, + "grad_norm": 0.7435083969347431, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 35571 + }, + { + "epoch": 0.35572, + "grad_norm": 0.6745272842968276, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 35572 + }, + { + "epoch": 0.35573, + "grad_norm": 0.5989494931027183, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 35573 + }, + { + "epoch": 0.35574, + "grad_norm": 0.62390761063717, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 35574 + }, + { + "epoch": 0.35575, + "grad_norm": 0.6900006247839154, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 35575 + }, + { + "epoch": 0.35576, + "grad_norm": 0.7568292706548242, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 35576 + }, + { + "epoch": 0.35577, + "grad_norm": 0.8403189534910179, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 35577 + }, + { + "epoch": 0.35578, + "grad_norm": 0.89370155538455, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 35578 + }, + { + "epoch": 0.35579, + "grad_norm": 0.8827083993098808, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 35579 + }, + { + "epoch": 0.3558, + "grad_norm": 0.9149628085909068, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 35580 + }, + { + "epoch": 0.35581, + "grad_norm": 0.9177284501249146, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 35581 + }, + { + "epoch": 0.35582, + "grad_norm": 0.9457256449220612, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 35582 + }, + { + "epoch": 0.35583, + "grad_norm": 0.7218561194593853, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 35583 + }, + { + "epoch": 0.35584, + "grad_norm": 0.6589564087573246, + "learning_rate": 0.003, + "loss": 3.985, + "step": 35584 + }, + { + "epoch": 0.35585, + "grad_norm": 0.6916763568729399, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 35585 + }, + { + "epoch": 0.35586, + "grad_norm": 0.7101271924199637, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 35586 + }, + { + "epoch": 0.35587, + "grad_norm": 0.823273082041167, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 35587 + }, + { + "epoch": 0.35588, + "grad_norm": 0.8562902061571648, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 35588 + }, + { + "epoch": 0.35589, + "grad_norm": 1.0956053582056293, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 35589 + }, + { + "epoch": 0.3559, + "grad_norm": 1.1461428739133805, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 35590 + }, + { + "epoch": 0.35591, + "grad_norm": 0.7309767019458201, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 35591 + }, + { + "epoch": 0.35592, + "grad_norm": 0.6943774921116099, + "learning_rate": 0.003, + "loss": 3.991, + "step": 35592 + }, + { + "epoch": 0.35593, + "grad_norm": 0.7878531312165592, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 35593 + }, + { + "epoch": 0.35594, + "grad_norm": 0.8637624963342206, + "learning_rate": 0.003, + "loss": 3.994, + "step": 35594 + }, + { + "epoch": 0.35595, + "grad_norm": 1.0234175810803519, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 35595 + }, + { + "epoch": 0.35596, + "grad_norm": 1.09520565352884, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 35596 + }, + { + "epoch": 0.35597, + "grad_norm": 0.9121888281848621, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 35597 + }, + { + "epoch": 0.35598, + "grad_norm": 0.7712815696167692, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 35598 + }, + { + "epoch": 0.35599, + "grad_norm": 0.5885891338486048, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 35599 + }, + { + "epoch": 0.356, + "grad_norm": 0.5825878752213534, + "learning_rate": 0.003, + "loss": 3.966, + "step": 35600 + }, + { + "epoch": 0.35601, + "grad_norm": 0.5692507786045419, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 35601 + }, + { + "epoch": 0.35602, + "grad_norm": 0.5534363651528622, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 35602 + }, + { + "epoch": 0.35603, + "grad_norm": 0.5404603630740845, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 35603 + }, + { + "epoch": 0.35604, + "grad_norm": 0.5947337105755364, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 35604 + }, + { + "epoch": 0.35605, + "grad_norm": 0.5749640186985604, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 35605 + }, + { + "epoch": 0.35606, + "grad_norm": 0.5308907182091431, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 35606 + }, + { + "epoch": 0.35607, + "grad_norm": 0.5095185507960407, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 35607 + }, + { + "epoch": 0.35608, + "grad_norm": 0.644645767034932, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 35608 + }, + { + "epoch": 0.35609, + "grad_norm": 0.7054157320050506, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 35609 + }, + { + "epoch": 0.3561, + "grad_norm": 0.7789828401951023, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 35610 + }, + { + "epoch": 0.35611, + "grad_norm": 0.8741318269350741, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 35611 + }, + { + "epoch": 0.35612, + "grad_norm": 0.9641926887652761, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 35612 + }, + { + "epoch": 0.35613, + "grad_norm": 0.9801720055251548, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 35613 + }, + { + "epoch": 0.35614, + "grad_norm": 0.8550602603620173, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 35614 + }, + { + "epoch": 0.35615, + "grad_norm": 0.7057557587712554, + "learning_rate": 0.003, + "loss": 3.988, + "step": 35615 + }, + { + "epoch": 0.35616, + "grad_norm": 0.744209497724092, + "learning_rate": 0.003, + "loss": 3.994, + "step": 35616 + }, + { + "epoch": 0.35617, + "grad_norm": 0.8170562285598418, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 35617 + }, + { + "epoch": 0.35618, + "grad_norm": 0.9470509611807868, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 35618 + }, + { + "epoch": 0.35619, + "grad_norm": 0.9362055242154014, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 35619 + }, + { + "epoch": 0.3562, + "grad_norm": 0.8824389232430062, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 35620 + }, + { + "epoch": 0.35621, + "grad_norm": 0.841802641722331, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 35621 + }, + { + "epoch": 0.35622, + "grad_norm": 0.8607279148497157, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 35622 + }, + { + "epoch": 0.35623, + "grad_norm": 0.8510945850341001, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 35623 + }, + { + "epoch": 0.35624, + "grad_norm": 0.9833210114453282, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 35624 + }, + { + "epoch": 0.35625, + "grad_norm": 1.0690656587866112, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 35625 + }, + { + "epoch": 0.35626, + "grad_norm": 0.9071812625772443, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 35626 + }, + { + "epoch": 0.35627, + "grad_norm": 0.9554748595688869, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 35627 + }, + { + "epoch": 0.35628, + "grad_norm": 1.0898787187598966, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 35628 + }, + { + "epoch": 0.35629, + "grad_norm": 0.9342297879642388, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 35629 + }, + { + "epoch": 0.3563, + "grad_norm": 0.8031143981003513, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 35630 + }, + { + "epoch": 0.35631, + "grad_norm": 0.8540663957642496, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 35631 + }, + { + "epoch": 0.35632, + "grad_norm": 0.8873966673924816, + "learning_rate": 0.003, + "loss": 3.9465, + "step": 35632 + }, + { + "epoch": 0.35633, + "grad_norm": 0.9434484867293144, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 35633 + }, + { + "epoch": 0.35634, + "grad_norm": 1.015489179065083, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 35634 + }, + { + "epoch": 0.35635, + "grad_norm": 1.044931921331795, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 35635 + }, + { + "epoch": 0.35636, + "grad_norm": 0.96691484853415, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 35636 + }, + { + "epoch": 0.35637, + "grad_norm": 0.7545565002338911, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 35637 + }, + { + "epoch": 0.35638, + "grad_norm": 0.7420710204434833, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 35638 + }, + { + "epoch": 0.35639, + "grad_norm": 0.684917488531726, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 35639 + }, + { + "epoch": 0.3564, + "grad_norm": 0.7052691460510632, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 35640 + }, + { + "epoch": 0.35641, + "grad_norm": 0.7535268278786265, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 35641 + }, + { + "epoch": 0.35642, + "grad_norm": 0.7025035630339741, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 35642 + }, + { + "epoch": 0.35643, + "grad_norm": 0.6900082063596651, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 35643 + }, + { + "epoch": 0.35644, + "grad_norm": 0.7860098923204106, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 35644 + }, + { + "epoch": 0.35645, + "grad_norm": 0.8013570471905881, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 35645 + }, + { + "epoch": 0.35646, + "grad_norm": 0.7853927973906882, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 35646 + }, + { + "epoch": 0.35647, + "grad_norm": 0.7871818074473261, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 35647 + }, + { + "epoch": 0.35648, + "grad_norm": 0.8472621859890879, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 35648 + }, + { + "epoch": 0.35649, + "grad_norm": 0.7604974037030353, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 35649 + }, + { + "epoch": 0.3565, + "grad_norm": 0.6694805282965562, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 35650 + }, + { + "epoch": 0.35651, + "grad_norm": 0.6819340085486704, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 35651 + }, + { + "epoch": 0.35652, + "grad_norm": 0.5874768747723985, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 35652 + }, + { + "epoch": 0.35653, + "grad_norm": 0.6279044379413871, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 35653 + }, + { + "epoch": 0.35654, + "grad_norm": 0.7005146415259635, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 35654 + }, + { + "epoch": 0.35655, + "grad_norm": 0.7022503612168782, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 35655 + }, + { + "epoch": 0.35656, + "grad_norm": 0.6268980079821952, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 35656 + }, + { + "epoch": 0.35657, + "grad_norm": 0.6533256893908017, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 35657 + }, + { + "epoch": 0.35658, + "grad_norm": 0.7097048504372211, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 35658 + }, + { + "epoch": 0.35659, + "grad_norm": 0.8979301661417088, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 35659 + }, + { + "epoch": 0.3566, + "grad_norm": 1.0544138696026382, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 35660 + }, + { + "epoch": 0.35661, + "grad_norm": 0.8456214837282942, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 35661 + }, + { + "epoch": 0.35662, + "grad_norm": 0.6550123403191308, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 35662 + }, + { + "epoch": 0.35663, + "grad_norm": 0.5652672229634894, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 35663 + }, + { + "epoch": 0.35664, + "grad_norm": 0.636565786209613, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 35664 + }, + { + "epoch": 0.35665, + "grad_norm": 0.6604339167710419, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 35665 + }, + { + "epoch": 0.35666, + "grad_norm": 0.7467309706835823, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 35666 + }, + { + "epoch": 0.35667, + "grad_norm": 0.6994456333149737, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 35667 + }, + { + "epoch": 0.35668, + "grad_norm": 0.657544642167639, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 35668 + }, + { + "epoch": 0.35669, + "grad_norm": 0.6749849074561849, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 35669 + }, + { + "epoch": 0.3567, + "grad_norm": 0.671204564426563, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 35670 + }, + { + "epoch": 0.35671, + "grad_norm": 0.6767795173606838, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 35671 + }, + { + "epoch": 0.35672, + "grad_norm": 0.6516381592998084, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 35672 + }, + { + "epoch": 0.35673, + "grad_norm": 0.7562964564414798, + "learning_rate": 0.003, + "loss": 3.998, + "step": 35673 + }, + { + "epoch": 0.35674, + "grad_norm": 0.8382206186461812, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 35674 + }, + { + "epoch": 0.35675, + "grad_norm": 0.8804019743059012, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 35675 + }, + { + "epoch": 0.35676, + "grad_norm": 0.8895324027867533, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 35676 + }, + { + "epoch": 0.35677, + "grad_norm": 0.9628744503968615, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 35677 + }, + { + "epoch": 0.35678, + "grad_norm": 1.0205595252908182, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 35678 + }, + { + "epoch": 0.35679, + "grad_norm": 0.9033046450846817, + "learning_rate": 0.003, + "loss": 3.997, + "step": 35679 + }, + { + "epoch": 0.3568, + "grad_norm": 0.7408734390185969, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 35680 + }, + { + "epoch": 0.35681, + "grad_norm": 0.8319989392585881, + "learning_rate": 0.003, + "loss": 3.983, + "step": 35681 + }, + { + "epoch": 0.35682, + "grad_norm": 0.8320413125027317, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 35682 + }, + { + "epoch": 0.35683, + "grad_norm": 0.7705244277794188, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 35683 + }, + { + "epoch": 0.35684, + "grad_norm": 0.821006161981525, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 35684 + }, + { + "epoch": 0.35685, + "grad_norm": 0.8038087058018827, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 35685 + }, + { + "epoch": 0.35686, + "grad_norm": 0.6766795518440272, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 35686 + }, + { + "epoch": 0.35687, + "grad_norm": 0.7036393327813463, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 35687 + }, + { + "epoch": 0.35688, + "grad_norm": 0.739834076763171, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 35688 + }, + { + "epoch": 0.35689, + "grad_norm": 0.7463794860920486, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 35689 + }, + { + "epoch": 0.3569, + "grad_norm": 0.6255516563481153, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 35690 + }, + { + "epoch": 0.35691, + "grad_norm": 0.5921467764608879, + "learning_rate": 0.003, + "loss": 4.012, + "step": 35691 + }, + { + "epoch": 0.35692, + "grad_norm": 0.6167579448483422, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 35692 + }, + { + "epoch": 0.35693, + "grad_norm": 0.704909238124695, + "learning_rate": 0.003, + "loss": 3.991, + "step": 35693 + }, + { + "epoch": 0.35694, + "grad_norm": 0.7530781111410215, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 35694 + }, + { + "epoch": 0.35695, + "grad_norm": 0.8601884002564131, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 35695 + }, + { + "epoch": 0.35696, + "grad_norm": 0.916912874615838, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 35696 + }, + { + "epoch": 0.35697, + "grad_norm": 0.867369751905969, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 35697 + }, + { + "epoch": 0.35698, + "grad_norm": 1.003469318070057, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 35698 + }, + { + "epoch": 0.35699, + "grad_norm": 1.0688699504875994, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 35699 + }, + { + "epoch": 0.357, + "grad_norm": 0.9880104082677789, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 35700 + }, + { + "epoch": 0.35701, + "grad_norm": 0.9033686471950637, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 35701 + }, + { + "epoch": 0.35702, + "grad_norm": 0.7681785490149349, + "learning_rate": 0.003, + "loss": 3.978, + "step": 35702 + }, + { + "epoch": 0.35703, + "grad_norm": 0.7721597283534919, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 35703 + }, + { + "epoch": 0.35704, + "grad_norm": 0.9898233803749608, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 35704 + }, + { + "epoch": 0.35705, + "grad_norm": 1.211803742912677, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 35705 + }, + { + "epoch": 0.35706, + "grad_norm": 0.9474642747724861, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 35706 + }, + { + "epoch": 0.35707, + "grad_norm": 0.9259679021844555, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 35707 + }, + { + "epoch": 0.35708, + "grad_norm": 0.9517681322309768, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 35708 + }, + { + "epoch": 0.35709, + "grad_norm": 0.932209004042754, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 35709 + }, + { + "epoch": 0.3571, + "grad_norm": 0.9589680360551965, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 35710 + }, + { + "epoch": 0.35711, + "grad_norm": 0.8675422808868247, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 35711 + }, + { + "epoch": 0.35712, + "grad_norm": 0.8429458322738338, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 35712 + }, + { + "epoch": 0.35713, + "grad_norm": 0.866706566685559, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 35713 + }, + { + "epoch": 0.35714, + "grad_norm": 0.7418975981981816, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 35714 + }, + { + "epoch": 0.35715, + "grad_norm": 0.6009073837164728, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 35715 + }, + { + "epoch": 0.35716, + "grad_norm": 0.661097724923929, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 35716 + }, + { + "epoch": 0.35717, + "grad_norm": 0.7850352538592601, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 35717 + }, + { + "epoch": 0.35718, + "grad_norm": 0.8185655607060424, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 35718 + }, + { + "epoch": 0.35719, + "grad_norm": 0.6895014904966076, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 35719 + }, + { + "epoch": 0.3572, + "grad_norm": 0.5606470394523667, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 35720 + }, + { + "epoch": 0.35721, + "grad_norm": 0.5831121952385722, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 35721 + }, + { + "epoch": 0.35722, + "grad_norm": 0.5695138966607579, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 35722 + }, + { + "epoch": 0.35723, + "grad_norm": 0.5949392924309267, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 35723 + }, + { + "epoch": 0.35724, + "grad_norm": 0.6602072655043134, + "learning_rate": 0.003, + "loss": 3.9454, + "step": 35724 + }, + { + "epoch": 0.35725, + "grad_norm": 0.8626724768330842, + "learning_rate": 0.003, + "loss": 3.975, + "step": 35725 + }, + { + "epoch": 0.35726, + "grad_norm": 0.9525100153099254, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 35726 + }, + { + "epoch": 0.35727, + "grad_norm": 0.9237163304038558, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 35727 + }, + { + "epoch": 0.35728, + "grad_norm": 0.7711154899398577, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 35728 + }, + { + "epoch": 0.35729, + "grad_norm": 0.8342501882233946, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 35729 + }, + { + "epoch": 0.3573, + "grad_norm": 0.8030745901536911, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 35730 + }, + { + "epoch": 0.35731, + "grad_norm": 0.6801180624347943, + "learning_rate": 0.003, + "loss": 4.006, + "step": 35731 + }, + { + "epoch": 0.35732, + "grad_norm": 0.6391966294589374, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 35732 + }, + { + "epoch": 0.35733, + "grad_norm": 0.6928339809525549, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 35733 + }, + { + "epoch": 0.35734, + "grad_norm": 0.7138902258683874, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 35734 + }, + { + "epoch": 0.35735, + "grad_norm": 0.7627811313132818, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 35735 + }, + { + "epoch": 0.35736, + "grad_norm": 0.8552167660662104, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35736 + }, + { + "epoch": 0.35737, + "grad_norm": 0.9987405847323757, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 35737 + }, + { + "epoch": 0.35738, + "grad_norm": 1.0923561817510323, + "learning_rate": 0.003, + "loss": 3.971, + "step": 35738 + }, + { + "epoch": 0.35739, + "grad_norm": 0.9992350154539634, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 35739 + }, + { + "epoch": 0.3574, + "grad_norm": 0.8712254087184549, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 35740 + }, + { + "epoch": 0.35741, + "grad_norm": 0.7192287486007113, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 35741 + }, + { + "epoch": 0.35742, + "grad_norm": 0.732458678954973, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 35742 + }, + { + "epoch": 0.35743, + "grad_norm": 0.7941760700998907, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 35743 + }, + { + "epoch": 0.35744, + "grad_norm": 0.8499518954698266, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 35744 + }, + { + "epoch": 0.35745, + "grad_norm": 0.8284390452852567, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 35745 + }, + { + "epoch": 0.35746, + "grad_norm": 0.849228140086554, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 35746 + }, + { + "epoch": 0.35747, + "grad_norm": 0.8056724694383414, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 35747 + }, + { + "epoch": 0.35748, + "grad_norm": 0.9043188546352657, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 35748 + }, + { + "epoch": 0.35749, + "grad_norm": 0.985422021585614, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 35749 + }, + { + "epoch": 0.3575, + "grad_norm": 1.0438971517648394, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 35750 + }, + { + "epoch": 0.35751, + "grad_norm": 0.9454581836152652, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 35751 + }, + { + "epoch": 0.35752, + "grad_norm": 0.8703790990300649, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 35752 + }, + { + "epoch": 0.35753, + "grad_norm": 0.8081193463316395, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 35753 + }, + { + "epoch": 0.35754, + "grad_norm": 0.7716600883980803, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 35754 + }, + { + "epoch": 0.35755, + "grad_norm": 0.7537087459297771, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 35755 + }, + { + "epoch": 0.35756, + "grad_norm": 0.6738488531343823, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 35756 + }, + { + "epoch": 0.35757, + "grad_norm": 0.5608233873860055, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 35757 + }, + { + "epoch": 0.35758, + "grad_norm": 0.5717701576934606, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 35758 + }, + { + "epoch": 0.35759, + "grad_norm": 0.5126744425752714, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 35759 + }, + { + "epoch": 0.3576, + "grad_norm": 0.5748875918176631, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 35760 + }, + { + "epoch": 0.35761, + "grad_norm": 0.626379024387596, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 35761 + }, + { + "epoch": 0.35762, + "grad_norm": 0.6328409157699102, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 35762 + }, + { + "epoch": 0.35763, + "grad_norm": 0.6752298907996153, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 35763 + }, + { + "epoch": 0.35764, + "grad_norm": 0.7636032801027302, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 35764 + }, + { + "epoch": 0.35765, + "grad_norm": 0.8240652988951083, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 35765 + }, + { + "epoch": 0.35766, + "grad_norm": 0.9868918219542331, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 35766 + }, + { + "epoch": 0.35767, + "grad_norm": 1.1747241599561815, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 35767 + }, + { + "epoch": 0.35768, + "grad_norm": 0.554392850994767, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 35768 + }, + { + "epoch": 0.35769, + "grad_norm": 0.6491674188480715, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 35769 + }, + { + "epoch": 0.3577, + "grad_norm": 0.6705680334775365, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 35770 + }, + { + "epoch": 0.35771, + "grad_norm": 0.7064964108039542, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 35771 + }, + { + "epoch": 0.35772, + "grad_norm": 0.7051534499670435, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 35772 + }, + { + "epoch": 0.35773, + "grad_norm": 0.6441554723587485, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 35773 + }, + { + "epoch": 0.35774, + "grad_norm": 0.6397327015298437, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 35774 + }, + { + "epoch": 0.35775, + "grad_norm": 0.6921915754151878, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 35775 + }, + { + "epoch": 0.35776, + "grad_norm": 0.721019396185895, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 35776 + }, + { + "epoch": 0.35777, + "grad_norm": 0.8526080478245184, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 35777 + }, + { + "epoch": 0.35778, + "grad_norm": 0.8753215563106432, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 35778 + }, + { + "epoch": 0.35779, + "grad_norm": 0.8977746275006611, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 35779 + }, + { + "epoch": 0.3578, + "grad_norm": 0.7931011610769791, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 35780 + }, + { + "epoch": 0.35781, + "grad_norm": 0.6108961386202559, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 35781 + }, + { + "epoch": 0.35782, + "grad_norm": 0.5259186896881958, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 35782 + }, + { + "epoch": 0.35783, + "grad_norm": 0.5074532238918892, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 35783 + }, + { + "epoch": 0.35784, + "grad_norm": 0.44210488261062375, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 35784 + }, + { + "epoch": 0.35785, + "grad_norm": 0.4556578083935324, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 35785 + }, + { + "epoch": 0.35786, + "grad_norm": 0.5025766460767508, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 35786 + }, + { + "epoch": 0.35787, + "grad_norm": 0.6719349004833436, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 35787 + }, + { + "epoch": 0.35788, + "grad_norm": 0.9050876674500101, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 35788 + }, + { + "epoch": 0.35789, + "grad_norm": 1.001988456856351, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 35789 + }, + { + "epoch": 0.3579, + "grad_norm": 1.0951251619671896, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 35790 + }, + { + "epoch": 0.35791, + "grad_norm": 0.8635578168195862, + "learning_rate": 0.003, + "loss": 3.9423, + "step": 35791 + }, + { + "epoch": 0.35792, + "grad_norm": 0.705408360404511, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 35792 + }, + { + "epoch": 0.35793, + "grad_norm": 0.753261013457465, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 35793 + }, + { + "epoch": 0.35794, + "grad_norm": 0.8835777614411533, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 35794 + }, + { + "epoch": 0.35795, + "grad_norm": 0.932224460160577, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 35795 + }, + { + "epoch": 0.35796, + "grad_norm": 1.0180825967980434, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 35796 + }, + { + "epoch": 0.35797, + "grad_norm": 1.0169867702422617, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 35797 + }, + { + "epoch": 0.35798, + "grad_norm": 1.0563077358697268, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 35798 + }, + { + "epoch": 0.35799, + "grad_norm": 0.9721452463341305, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 35799 + }, + { + "epoch": 0.358, + "grad_norm": 0.8384324234548272, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 35800 + }, + { + "epoch": 0.35801, + "grad_norm": 0.7572001329716449, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 35801 + }, + { + "epoch": 0.35802, + "grad_norm": 0.8154875254747772, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 35802 + }, + { + "epoch": 0.35803, + "grad_norm": 0.9498975142120674, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 35803 + }, + { + "epoch": 0.35804, + "grad_norm": 1.0287178390352594, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 35804 + }, + { + "epoch": 0.35805, + "grad_norm": 0.8248715151386454, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 35805 + }, + { + "epoch": 0.35806, + "grad_norm": 0.7300082131335685, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 35806 + }, + { + "epoch": 0.35807, + "grad_norm": 0.8449901715609801, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 35807 + }, + { + "epoch": 0.35808, + "grad_norm": 0.7936739669508888, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 35808 + }, + { + "epoch": 0.35809, + "grad_norm": 0.7665895280178088, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 35809 + }, + { + "epoch": 0.3581, + "grad_norm": 0.8783290976124959, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 35810 + }, + { + "epoch": 0.35811, + "grad_norm": 0.8890628918026522, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 35811 + }, + { + "epoch": 0.35812, + "grad_norm": 0.8559783548750234, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 35812 + }, + { + "epoch": 0.35813, + "grad_norm": 0.7645967733120328, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 35813 + }, + { + "epoch": 0.35814, + "grad_norm": 0.7438292369917675, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35814 + }, + { + "epoch": 0.35815, + "grad_norm": 0.6672870107595, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 35815 + }, + { + "epoch": 0.35816, + "grad_norm": 0.6951134283812441, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 35816 + }, + { + "epoch": 0.35817, + "grad_norm": 0.6595396824076063, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 35817 + }, + { + "epoch": 0.35818, + "grad_norm": 0.6620953276093257, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 35818 + }, + { + "epoch": 0.35819, + "grad_norm": 0.5797160016966358, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 35819 + }, + { + "epoch": 0.3582, + "grad_norm": 0.6193370191180632, + "learning_rate": 0.003, + "loss": 3.973, + "step": 35820 + }, + { + "epoch": 0.35821, + "grad_norm": 0.7388298058667735, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 35821 + }, + { + "epoch": 0.35822, + "grad_norm": 0.9552044838735626, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 35822 + }, + { + "epoch": 0.35823, + "grad_norm": 1.0793734471188745, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 35823 + }, + { + "epoch": 0.35824, + "grad_norm": 0.7216005416579208, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 35824 + }, + { + "epoch": 0.35825, + "grad_norm": 0.7101027767914012, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 35825 + }, + { + "epoch": 0.35826, + "grad_norm": 0.8048543038786217, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 35826 + }, + { + "epoch": 0.35827, + "grad_norm": 0.966387174489259, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 35827 + }, + { + "epoch": 0.35828, + "grad_norm": 1.1591466764982146, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 35828 + }, + { + "epoch": 0.35829, + "grad_norm": 0.7040415335567463, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 35829 + }, + { + "epoch": 0.3583, + "grad_norm": 0.7620541365534974, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 35830 + }, + { + "epoch": 0.35831, + "grad_norm": 0.7515081085209993, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 35831 + }, + { + "epoch": 0.35832, + "grad_norm": 0.7541461835048767, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 35832 + }, + { + "epoch": 0.35833, + "grad_norm": 0.6828892150876766, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 35833 + }, + { + "epoch": 0.35834, + "grad_norm": 0.6164462218442378, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 35834 + }, + { + "epoch": 0.35835, + "grad_norm": 0.6413150247140695, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 35835 + }, + { + "epoch": 0.35836, + "grad_norm": 0.7126619204374977, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 35836 + }, + { + "epoch": 0.35837, + "grad_norm": 0.6274452988118366, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 35837 + }, + { + "epoch": 0.35838, + "grad_norm": 0.6764414925652403, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 35838 + }, + { + "epoch": 0.35839, + "grad_norm": 0.7938622762306647, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 35839 + }, + { + "epoch": 0.3584, + "grad_norm": 0.8496768649109193, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 35840 + }, + { + "epoch": 0.35841, + "grad_norm": 0.7867627789521375, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 35841 + }, + { + "epoch": 0.35842, + "grad_norm": 0.7985981170311826, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 35842 + }, + { + "epoch": 0.35843, + "grad_norm": 0.8743281453091141, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 35843 + }, + { + "epoch": 0.35844, + "grad_norm": 0.9833779005719335, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 35844 + }, + { + "epoch": 0.35845, + "grad_norm": 0.8237065319434711, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 35845 + }, + { + "epoch": 0.35846, + "grad_norm": 0.7663193832415807, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 35846 + }, + { + "epoch": 0.35847, + "grad_norm": 0.8202028191604948, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 35847 + }, + { + "epoch": 0.35848, + "grad_norm": 0.8751387279615032, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 35848 + }, + { + "epoch": 0.35849, + "grad_norm": 0.9232704532891879, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 35849 + }, + { + "epoch": 0.3585, + "grad_norm": 1.1094206988695654, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 35850 + }, + { + "epoch": 0.35851, + "grad_norm": 0.944365428894095, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 35851 + }, + { + "epoch": 0.35852, + "grad_norm": 0.8765305942110068, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 35852 + }, + { + "epoch": 0.35853, + "grad_norm": 0.788173212190267, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 35853 + }, + { + "epoch": 0.35854, + "grad_norm": 0.7353692348246313, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 35854 + }, + { + "epoch": 0.35855, + "grad_norm": 0.7238038232842504, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 35855 + }, + { + "epoch": 0.35856, + "grad_norm": 0.7434031392199845, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 35856 + }, + { + "epoch": 0.35857, + "grad_norm": 0.6794402408249819, + "learning_rate": 0.003, + "loss": 4.002, + "step": 35857 + }, + { + "epoch": 0.35858, + "grad_norm": 0.6864104958177653, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 35858 + }, + { + "epoch": 0.35859, + "grad_norm": 0.6864921643187413, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 35859 + }, + { + "epoch": 0.3586, + "grad_norm": 0.7411231258476269, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 35860 + }, + { + "epoch": 0.35861, + "grad_norm": 0.7877957214710982, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 35861 + }, + { + "epoch": 0.35862, + "grad_norm": 0.8076326440155489, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 35862 + }, + { + "epoch": 0.35863, + "grad_norm": 0.7894793061823305, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 35863 + }, + { + "epoch": 0.35864, + "grad_norm": 0.7384657512393856, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 35864 + }, + { + "epoch": 0.35865, + "grad_norm": 0.5953471045375031, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 35865 + }, + { + "epoch": 0.35866, + "grad_norm": 0.6892919461323321, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 35866 + }, + { + "epoch": 0.35867, + "grad_norm": 0.7069185035176755, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 35867 + }, + { + "epoch": 0.35868, + "grad_norm": 0.6875933266758232, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 35868 + }, + { + "epoch": 0.35869, + "grad_norm": 0.6628947014703047, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 35869 + }, + { + "epoch": 0.3587, + "grad_norm": 0.6466119085011243, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 35870 + }, + { + "epoch": 0.35871, + "grad_norm": 0.6651769255342617, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 35871 + }, + { + "epoch": 0.35872, + "grad_norm": 0.6513905598052389, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 35872 + }, + { + "epoch": 0.35873, + "grad_norm": 0.6746586128578045, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 35873 + }, + { + "epoch": 0.35874, + "grad_norm": 0.7476866759532522, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 35874 + }, + { + "epoch": 0.35875, + "grad_norm": 0.7493762332775447, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 35875 + }, + { + "epoch": 0.35876, + "grad_norm": 0.8398092843030316, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 35876 + }, + { + "epoch": 0.35877, + "grad_norm": 0.927569240022322, + "learning_rate": 0.003, + "loss": 3.988, + "step": 35877 + }, + { + "epoch": 0.35878, + "grad_norm": 1.0155128058537501, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 35878 + }, + { + "epoch": 0.35879, + "grad_norm": 0.9906976976478516, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 35879 + }, + { + "epoch": 0.3588, + "grad_norm": 1.074235239861476, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 35880 + }, + { + "epoch": 0.35881, + "grad_norm": 1.0837518472720062, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 35881 + }, + { + "epoch": 0.35882, + "grad_norm": 1.0041878035342175, + "learning_rate": 0.003, + "loss": 4.0329, + "step": 35882 + }, + { + "epoch": 0.35883, + "grad_norm": 1.133841514671948, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 35883 + }, + { + "epoch": 0.35884, + "grad_norm": 0.827888802811209, + "learning_rate": 0.003, + "loss": 4.0301, + "step": 35884 + }, + { + "epoch": 0.35885, + "grad_norm": 0.837107194484381, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 35885 + }, + { + "epoch": 0.35886, + "grad_norm": 1.0004084955911192, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 35886 + }, + { + "epoch": 0.35887, + "grad_norm": 1.0507352272163033, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 35887 + }, + { + "epoch": 0.35888, + "grad_norm": 0.9824994056509556, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 35888 + }, + { + "epoch": 0.35889, + "grad_norm": 0.9770458405099605, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 35889 + }, + { + "epoch": 0.3589, + "grad_norm": 0.9354704301721345, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 35890 + }, + { + "epoch": 0.35891, + "grad_norm": 0.9082382148310509, + "learning_rate": 0.003, + "loss": 3.994, + "step": 35891 + }, + { + "epoch": 0.35892, + "grad_norm": 0.9551312321238026, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 35892 + }, + { + "epoch": 0.35893, + "grad_norm": 0.8412296381735355, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 35893 + }, + { + "epoch": 0.35894, + "grad_norm": 0.9400355388889873, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 35894 + }, + { + "epoch": 0.35895, + "grad_norm": 0.9236564060571277, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 35895 + }, + { + "epoch": 0.35896, + "grad_norm": 0.7987528741388954, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 35896 + }, + { + "epoch": 0.35897, + "grad_norm": 0.8560154251826639, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 35897 + }, + { + "epoch": 0.35898, + "grad_norm": 0.8233913867302424, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 35898 + }, + { + "epoch": 0.35899, + "grad_norm": 0.8353080937143263, + "learning_rate": 0.003, + "loss": 4.0, + "step": 35899 + }, + { + "epoch": 0.359, + "grad_norm": 0.9456370308563062, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 35900 + }, + { + "epoch": 0.35901, + "grad_norm": 1.0857278535707697, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 35901 + }, + { + "epoch": 0.35902, + "grad_norm": 0.8293843216572335, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 35902 + }, + { + "epoch": 0.35903, + "grad_norm": 0.718710243856656, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 35903 + }, + { + "epoch": 0.35904, + "grad_norm": 0.6112159832202221, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 35904 + }, + { + "epoch": 0.35905, + "grad_norm": 0.5010456202652428, + "learning_rate": 0.003, + "loss": 3.978, + "step": 35905 + }, + { + "epoch": 0.35906, + "grad_norm": 0.462646678263429, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 35906 + }, + { + "epoch": 0.35907, + "grad_norm": 0.46118999105754777, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 35907 + }, + { + "epoch": 0.35908, + "grad_norm": 0.4696268557330063, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 35908 + }, + { + "epoch": 0.35909, + "grad_norm": 0.44731485767008683, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 35909 + }, + { + "epoch": 0.3591, + "grad_norm": 0.43125591774260774, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 35910 + }, + { + "epoch": 0.35911, + "grad_norm": 0.4566114479908388, + "learning_rate": 0.003, + "loss": 3.989, + "step": 35911 + }, + { + "epoch": 0.35912, + "grad_norm": 0.5067599994600465, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 35912 + }, + { + "epoch": 0.35913, + "grad_norm": 0.5845619185127312, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 35913 + }, + { + "epoch": 0.35914, + "grad_norm": 0.7035398963188703, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 35914 + }, + { + "epoch": 0.35915, + "grad_norm": 0.8619939006739915, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 35915 + }, + { + "epoch": 0.35916, + "grad_norm": 0.9996160569097772, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 35916 + }, + { + "epoch": 0.35917, + "grad_norm": 0.8979117920434869, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 35917 + }, + { + "epoch": 0.35918, + "grad_norm": 0.7383696406784073, + "learning_rate": 0.003, + "loss": 3.998, + "step": 35918 + }, + { + "epoch": 0.35919, + "grad_norm": 0.8532345591157356, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 35919 + }, + { + "epoch": 0.3592, + "grad_norm": 0.9893490584351287, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 35920 + }, + { + "epoch": 0.35921, + "grad_norm": 1.0578724181661263, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 35921 + }, + { + "epoch": 0.35922, + "grad_norm": 0.7461722609557704, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 35922 + }, + { + "epoch": 0.35923, + "grad_norm": 0.6778347502433888, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 35923 + }, + { + "epoch": 0.35924, + "grad_norm": 0.6672168610185322, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 35924 + }, + { + "epoch": 0.35925, + "grad_norm": 0.6772833571485473, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 35925 + }, + { + "epoch": 0.35926, + "grad_norm": 0.6079093698775991, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 35926 + }, + { + "epoch": 0.35927, + "grad_norm": 0.6767437201026785, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 35927 + }, + { + "epoch": 0.35928, + "grad_norm": 0.6795600245358759, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 35928 + }, + { + "epoch": 0.35929, + "grad_norm": 0.5669589674400884, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 35929 + }, + { + "epoch": 0.3593, + "grad_norm": 0.5696853821816887, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 35930 + }, + { + "epoch": 0.35931, + "grad_norm": 0.6871059423532117, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 35931 + }, + { + "epoch": 0.35932, + "grad_norm": 0.8115332423023776, + "learning_rate": 0.003, + "loss": 3.976, + "step": 35932 + }, + { + "epoch": 0.35933, + "grad_norm": 0.9362112508456097, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 35933 + }, + { + "epoch": 0.35934, + "grad_norm": 0.8536532808742586, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 35934 + }, + { + "epoch": 0.35935, + "grad_norm": 0.775407601864158, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 35935 + }, + { + "epoch": 0.35936, + "grad_norm": 0.7099356347606224, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 35936 + }, + { + "epoch": 0.35937, + "grad_norm": 0.7400251414170733, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 35937 + }, + { + "epoch": 0.35938, + "grad_norm": 0.79077195876092, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 35938 + }, + { + "epoch": 0.35939, + "grad_norm": 0.8355989089102145, + "learning_rate": 0.003, + "loss": 3.997, + "step": 35939 + }, + { + "epoch": 0.3594, + "grad_norm": 0.8375412920580225, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 35940 + }, + { + "epoch": 0.35941, + "grad_norm": 0.8115977535438447, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 35941 + }, + { + "epoch": 0.35942, + "grad_norm": 0.9130730397697093, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 35942 + }, + { + "epoch": 0.35943, + "grad_norm": 1.0072073385918907, + "learning_rate": 0.003, + "loss": 3.991, + "step": 35943 + }, + { + "epoch": 0.35944, + "grad_norm": 0.9789979303720144, + "learning_rate": 0.003, + "loss": 4.0, + "step": 35944 + }, + { + "epoch": 0.35945, + "grad_norm": 0.9681755530937668, + "learning_rate": 0.003, + "loss": 3.989, + "step": 35945 + }, + { + "epoch": 0.35946, + "grad_norm": 0.8937965613967089, + "learning_rate": 0.003, + "loss": 3.975, + "step": 35946 + }, + { + "epoch": 0.35947, + "grad_norm": 0.8498470875206555, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 35947 + }, + { + "epoch": 0.35948, + "grad_norm": 0.7788848723842078, + "learning_rate": 0.003, + "loss": 4.0307, + "step": 35948 + }, + { + "epoch": 0.35949, + "grad_norm": 0.8545678050927871, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 35949 + }, + { + "epoch": 0.3595, + "grad_norm": 0.980745149985835, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 35950 + }, + { + "epoch": 0.35951, + "grad_norm": 1.2314112708668612, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 35951 + }, + { + "epoch": 0.35952, + "grad_norm": 0.9749134981105331, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 35952 + }, + { + "epoch": 0.35953, + "grad_norm": 0.8277242464645637, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 35953 + }, + { + "epoch": 0.35954, + "grad_norm": 0.775692690457987, + "learning_rate": 0.003, + "loss": 3.982, + "step": 35954 + }, + { + "epoch": 0.35955, + "grad_norm": 0.8023147013656485, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 35955 + }, + { + "epoch": 0.35956, + "grad_norm": 0.7316058065883608, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 35956 + }, + { + "epoch": 0.35957, + "grad_norm": 0.7079024702029322, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 35957 + }, + { + "epoch": 0.35958, + "grad_norm": 0.687503342573597, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 35958 + }, + { + "epoch": 0.35959, + "grad_norm": 0.5762249358544459, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 35959 + }, + { + "epoch": 0.3596, + "grad_norm": 0.5106080551013461, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 35960 + }, + { + "epoch": 0.35961, + "grad_norm": 0.5065443834654457, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 35961 + }, + { + "epoch": 0.35962, + "grad_norm": 0.5122084979642646, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 35962 + }, + { + "epoch": 0.35963, + "grad_norm": 0.4610368416788814, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 35963 + }, + { + "epoch": 0.35964, + "grad_norm": 0.4604650582594643, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 35964 + }, + { + "epoch": 0.35965, + "grad_norm": 0.4944139337612654, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 35965 + }, + { + "epoch": 0.35966, + "grad_norm": 0.5299289248452191, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 35966 + }, + { + "epoch": 0.35967, + "grad_norm": 0.6612050121646494, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 35967 + }, + { + "epoch": 0.35968, + "grad_norm": 0.7236174958752429, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 35968 + }, + { + "epoch": 0.35969, + "grad_norm": 0.7714778363748357, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 35969 + }, + { + "epoch": 0.3597, + "grad_norm": 0.8672405086122081, + "learning_rate": 0.003, + "loss": 4.014, + "step": 35970 + }, + { + "epoch": 0.35971, + "grad_norm": 1.077308876854535, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 35971 + }, + { + "epoch": 0.35972, + "grad_norm": 0.9827349512326913, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 35972 + }, + { + "epoch": 0.35973, + "grad_norm": 0.8927807207387957, + "learning_rate": 0.003, + "loss": 3.997, + "step": 35973 + }, + { + "epoch": 0.35974, + "grad_norm": 0.8323613758735025, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 35974 + }, + { + "epoch": 0.35975, + "grad_norm": 1.0075282697610444, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 35975 + }, + { + "epoch": 0.35976, + "grad_norm": 0.9901466531863552, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 35976 + }, + { + "epoch": 0.35977, + "grad_norm": 1.064820106331608, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 35977 + }, + { + "epoch": 0.35978, + "grad_norm": 0.8381722170283149, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 35978 + }, + { + "epoch": 0.35979, + "grad_norm": 0.8178717435817393, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 35979 + }, + { + "epoch": 0.3598, + "grad_norm": 0.9032279648576015, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 35980 + }, + { + "epoch": 0.35981, + "grad_norm": 0.8561435268473161, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 35981 + }, + { + "epoch": 0.35982, + "grad_norm": 0.7783567349155922, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 35982 + }, + { + "epoch": 0.35983, + "grad_norm": 0.7207225388774943, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 35983 + }, + { + "epoch": 0.35984, + "grad_norm": 0.806745828348914, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 35984 + }, + { + "epoch": 0.35985, + "grad_norm": 0.9187459223592492, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 35985 + }, + { + "epoch": 0.35986, + "grad_norm": 1.0866426676413523, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 35986 + }, + { + "epoch": 0.35987, + "grad_norm": 1.0541180209409011, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 35987 + }, + { + "epoch": 0.35988, + "grad_norm": 0.783372325159347, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 35988 + }, + { + "epoch": 0.35989, + "grad_norm": 0.6534995662698321, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 35989 + }, + { + "epoch": 0.3599, + "grad_norm": 0.6898183789309571, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 35990 + }, + { + "epoch": 0.35991, + "grad_norm": 0.6760111632000148, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 35991 + }, + { + "epoch": 0.35992, + "grad_norm": 0.6121175035519294, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 35992 + }, + { + "epoch": 0.35993, + "grad_norm": 0.7029054683328486, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 35993 + }, + { + "epoch": 0.35994, + "grad_norm": 0.8291477729605738, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 35994 + }, + { + "epoch": 0.35995, + "grad_norm": 0.7804782924042898, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 35995 + }, + { + "epoch": 0.35996, + "grad_norm": 0.7667174766933175, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 35996 + }, + { + "epoch": 0.35997, + "grad_norm": 0.8161727759970109, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 35997 + }, + { + "epoch": 0.35998, + "grad_norm": 0.7410935189092508, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 35998 + }, + { + "epoch": 0.35999, + "grad_norm": 0.7319526796783715, + "learning_rate": 0.003, + "loss": 4.041, + "step": 35999 + }, + { + "epoch": 0.36, + "grad_norm": 0.6692092661055209, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 36000 + }, + { + "epoch": 0.36001, + "grad_norm": 0.6301157918501054, + "learning_rate": 0.003, + "loss": 3.977, + "step": 36001 + }, + { + "epoch": 0.36002, + "grad_norm": 0.5696807832029028, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 36002 + }, + { + "epoch": 0.36003, + "grad_norm": 0.6014891690859316, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 36003 + }, + { + "epoch": 0.36004, + "grad_norm": 0.6684813154990447, + "learning_rate": 0.003, + "loss": 3.989, + "step": 36004 + }, + { + "epoch": 0.36005, + "grad_norm": 0.8168699858128405, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 36005 + }, + { + "epoch": 0.36006, + "grad_norm": 1.0185646932608585, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 36006 + }, + { + "epoch": 0.36007, + "grad_norm": 1.1246139335923881, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 36007 + }, + { + "epoch": 0.36008, + "grad_norm": 0.7776764386320302, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 36008 + }, + { + "epoch": 0.36009, + "grad_norm": 0.7050245629477478, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 36009 + }, + { + "epoch": 0.3601, + "grad_norm": 0.7837321002254156, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 36010 + }, + { + "epoch": 0.36011, + "grad_norm": 0.7997096475838853, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 36011 + }, + { + "epoch": 0.36012, + "grad_norm": 0.7663381129397775, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 36012 + }, + { + "epoch": 0.36013, + "grad_norm": 0.7560669226392559, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 36013 + }, + { + "epoch": 0.36014, + "grad_norm": 0.7450029068325517, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 36014 + }, + { + "epoch": 0.36015, + "grad_norm": 0.7747200854834491, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 36015 + }, + { + "epoch": 0.36016, + "grad_norm": 0.8758132065262622, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 36016 + }, + { + "epoch": 0.36017, + "grad_norm": 0.8993631365162593, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 36017 + }, + { + "epoch": 0.36018, + "grad_norm": 0.8735333307025405, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 36018 + }, + { + "epoch": 0.36019, + "grad_norm": 0.7315521385993907, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 36019 + }, + { + "epoch": 0.3602, + "grad_norm": 0.6686035677269178, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 36020 + }, + { + "epoch": 0.36021, + "grad_norm": 0.6484658238078612, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 36021 + }, + { + "epoch": 0.36022, + "grad_norm": 0.6279626851950553, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 36022 + }, + { + "epoch": 0.36023, + "grad_norm": 0.6292820917125551, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 36023 + }, + { + "epoch": 0.36024, + "grad_norm": 0.6358161161501591, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 36024 + }, + { + "epoch": 0.36025, + "grad_norm": 0.6308436739609975, + "learning_rate": 0.003, + "loss": 4.002, + "step": 36025 + }, + { + "epoch": 0.36026, + "grad_norm": 0.6127541802611383, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 36026 + }, + { + "epoch": 0.36027, + "grad_norm": 0.5630844610873664, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 36027 + }, + { + "epoch": 0.36028, + "grad_norm": 0.5823650023546639, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 36028 + }, + { + "epoch": 0.36029, + "grad_norm": 0.6091944699062517, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 36029 + }, + { + "epoch": 0.3603, + "grad_norm": 0.6977212078082738, + "learning_rate": 0.003, + "loss": 4.004, + "step": 36030 + }, + { + "epoch": 0.36031, + "grad_norm": 0.8933501608901504, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 36031 + }, + { + "epoch": 0.36032, + "grad_norm": 0.9990496165464786, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 36032 + }, + { + "epoch": 0.36033, + "grad_norm": 0.9734144852144075, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 36033 + }, + { + "epoch": 0.36034, + "grad_norm": 0.9817955479736831, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 36034 + }, + { + "epoch": 0.36035, + "grad_norm": 0.7011639002837787, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 36035 + }, + { + "epoch": 0.36036, + "grad_norm": 0.59851772209042, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 36036 + }, + { + "epoch": 0.36037, + "grad_norm": 0.6346744516808234, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 36037 + }, + { + "epoch": 0.36038, + "grad_norm": 0.6546885621209354, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 36038 + }, + { + "epoch": 0.36039, + "grad_norm": 0.719521892720225, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 36039 + }, + { + "epoch": 0.3604, + "grad_norm": 0.7733397890323926, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 36040 + }, + { + "epoch": 0.36041, + "grad_norm": 0.8686570142429481, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 36041 + }, + { + "epoch": 0.36042, + "grad_norm": 1.0280729558099198, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 36042 + }, + { + "epoch": 0.36043, + "grad_norm": 1.0446208987281969, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 36043 + }, + { + "epoch": 0.36044, + "grad_norm": 0.7887098251964204, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 36044 + }, + { + "epoch": 0.36045, + "grad_norm": 0.7461006603545812, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 36045 + }, + { + "epoch": 0.36046, + "grad_norm": 0.7810400306770022, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 36046 + }, + { + "epoch": 0.36047, + "grad_norm": 0.680867983159428, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 36047 + }, + { + "epoch": 0.36048, + "grad_norm": 0.6585380695186083, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 36048 + }, + { + "epoch": 0.36049, + "grad_norm": 0.8316458407948295, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 36049 + }, + { + "epoch": 0.3605, + "grad_norm": 0.897746530666105, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 36050 + }, + { + "epoch": 0.36051, + "grad_norm": 1.0963183345447103, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 36051 + }, + { + "epoch": 0.36052, + "grad_norm": 0.9795524084461475, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 36052 + }, + { + "epoch": 0.36053, + "grad_norm": 0.9312520015859714, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 36053 + }, + { + "epoch": 0.36054, + "grad_norm": 0.8353205848518358, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 36054 + }, + { + "epoch": 0.36055, + "grad_norm": 0.6749953820331079, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 36055 + }, + { + "epoch": 0.36056, + "grad_norm": 0.6051635779829546, + "learning_rate": 0.003, + "loss": 4.0338, + "step": 36056 + }, + { + "epoch": 0.36057, + "grad_norm": 0.6180757217326807, + "learning_rate": 0.003, + "loss": 3.998, + "step": 36057 + }, + { + "epoch": 0.36058, + "grad_norm": 0.6586448017331794, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 36058 + }, + { + "epoch": 0.36059, + "grad_norm": 0.6943353908142402, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 36059 + }, + { + "epoch": 0.3606, + "grad_norm": 0.7173543923924625, + "learning_rate": 0.003, + "loss": 3.954, + "step": 36060 + }, + { + "epoch": 0.36061, + "grad_norm": 0.8376146585349856, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 36061 + }, + { + "epoch": 0.36062, + "grad_norm": 0.9476762998799805, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 36062 + }, + { + "epoch": 0.36063, + "grad_norm": 0.9812341609462926, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 36063 + }, + { + "epoch": 0.36064, + "grad_norm": 1.0047966916383098, + "learning_rate": 0.003, + "loss": 4.002, + "step": 36064 + }, + { + "epoch": 0.36065, + "grad_norm": 0.9173469273835857, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 36065 + }, + { + "epoch": 0.36066, + "grad_norm": 0.923083374424454, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 36066 + }, + { + "epoch": 0.36067, + "grad_norm": 1.0230894787180063, + "learning_rate": 0.003, + "loss": 4.0255, + "step": 36067 + }, + { + "epoch": 0.36068, + "grad_norm": 0.9458071103604461, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 36068 + }, + { + "epoch": 0.36069, + "grad_norm": 1.054101787694798, + "learning_rate": 0.003, + "loss": 4.005, + "step": 36069 + }, + { + "epoch": 0.3607, + "grad_norm": 1.0454326082737035, + "learning_rate": 0.003, + "loss": 4.003, + "step": 36070 + }, + { + "epoch": 0.36071, + "grad_norm": 1.1136379321695375, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 36071 + }, + { + "epoch": 0.36072, + "grad_norm": 0.9010406424247928, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 36072 + }, + { + "epoch": 0.36073, + "grad_norm": 0.8280901096764739, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 36073 + }, + { + "epoch": 0.36074, + "grad_norm": 0.7342346002596034, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 36074 + }, + { + "epoch": 0.36075, + "grad_norm": 0.7363844451226509, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 36075 + }, + { + "epoch": 0.36076, + "grad_norm": 0.7857491552412381, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 36076 + }, + { + "epoch": 0.36077, + "grad_norm": 0.8036228195295786, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 36077 + }, + { + "epoch": 0.36078, + "grad_norm": 0.8939801148864389, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 36078 + }, + { + "epoch": 0.36079, + "grad_norm": 0.9124968576449756, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 36079 + }, + { + "epoch": 0.3608, + "grad_norm": 0.8287469942061315, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 36080 + }, + { + "epoch": 0.36081, + "grad_norm": 0.6586861110389302, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 36081 + }, + { + "epoch": 0.36082, + "grad_norm": 0.6626322727035726, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 36082 + }, + { + "epoch": 0.36083, + "grad_norm": 0.7547924947524738, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 36083 + }, + { + "epoch": 0.36084, + "grad_norm": 0.7742030169314982, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 36084 + }, + { + "epoch": 0.36085, + "grad_norm": 0.7278524479570013, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 36085 + }, + { + "epoch": 0.36086, + "grad_norm": 0.7001086333748372, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 36086 + }, + { + "epoch": 0.36087, + "grad_norm": 0.695519021859781, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 36087 + }, + { + "epoch": 0.36088, + "grad_norm": 0.6617583318103488, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 36088 + }, + { + "epoch": 0.36089, + "grad_norm": 0.7075652534976217, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 36089 + }, + { + "epoch": 0.3609, + "grad_norm": 0.7091207062866398, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 36090 + }, + { + "epoch": 0.36091, + "grad_norm": 0.7083564689607653, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 36091 + }, + { + "epoch": 0.36092, + "grad_norm": 0.7979425204171445, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 36092 + }, + { + "epoch": 0.36093, + "grad_norm": 0.8063248210995279, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 36093 + }, + { + "epoch": 0.36094, + "grad_norm": 0.8109293421511266, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 36094 + }, + { + "epoch": 0.36095, + "grad_norm": 0.786724343057865, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 36095 + }, + { + "epoch": 0.36096, + "grad_norm": 0.7021117998738727, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 36096 + }, + { + "epoch": 0.36097, + "grad_norm": 0.7067380150533984, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 36097 + }, + { + "epoch": 0.36098, + "grad_norm": 0.697712817539772, + "learning_rate": 0.003, + "loss": 4.004, + "step": 36098 + }, + { + "epoch": 0.36099, + "grad_norm": 0.6853271737742758, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 36099 + }, + { + "epoch": 0.361, + "grad_norm": 0.6284946913134329, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 36100 + }, + { + "epoch": 0.36101, + "grad_norm": 0.6489469784311828, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 36101 + }, + { + "epoch": 0.36102, + "grad_norm": 0.7160274727000285, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 36102 + }, + { + "epoch": 0.36103, + "grad_norm": 0.7860710631351632, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 36103 + }, + { + "epoch": 0.36104, + "grad_norm": 0.7228592010588855, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 36104 + }, + { + "epoch": 0.36105, + "grad_norm": 0.5671713285124131, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 36105 + }, + { + "epoch": 0.36106, + "grad_norm": 0.6211526187861642, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 36106 + }, + { + "epoch": 0.36107, + "grad_norm": 0.661408057917548, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 36107 + }, + { + "epoch": 0.36108, + "grad_norm": 0.7960472347810169, + "learning_rate": 0.003, + "loss": 4.003, + "step": 36108 + }, + { + "epoch": 0.36109, + "grad_norm": 0.9263309158717443, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 36109 + }, + { + "epoch": 0.3611, + "grad_norm": 1.0743844274921832, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 36110 + }, + { + "epoch": 0.36111, + "grad_norm": 1.0285730741437937, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 36111 + }, + { + "epoch": 0.36112, + "grad_norm": 0.9569124841957083, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 36112 + }, + { + "epoch": 0.36113, + "grad_norm": 1.0066290921013208, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 36113 + }, + { + "epoch": 0.36114, + "grad_norm": 0.9490132528706207, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 36114 + }, + { + "epoch": 0.36115, + "grad_norm": 0.9807333213967783, + "learning_rate": 0.003, + "loss": 3.979, + "step": 36115 + }, + { + "epoch": 0.36116, + "grad_norm": 0.879619576999744, + "learning_rate": 0.003, + "loss": 3.986, + "step": 36116 + }, + { + "epoch": 0.36117, + "grad_norm": 0.7468059794141234, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 36117 + }, + { + "epoch": 0.36118, + "grad_norm": 0.697042429534565, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 36118 + }, + { + "epoch": 0.36119, + "grad_norm": 0.7407812527566059, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 36119 + }, + { + "epoch": 0.3612, + "grad_norm": 0.8226293068687742, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 36120 + }, + { + "epoch": 0.36121, + "grad_norm": 0.789987134440657, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 36121 + }, + { + "epoch": 0.36122, + "grad_norm": 0.6777987611237015, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 36122 + }, + { + "epoch": 0.36123, + "grad_norm": 0.5746812274619743, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 36123 + }, + { + "epoch": 0.36124, + "grad_norm": 0.4999893719298798, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 36124 + }, + { + "epoch": 0.36125, + "grad_norm": 0.6232481327771792, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 36125 + }, + { + "epoch": 0.36126, + "grad_norm": 0.7224404617271906, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 36126 + }, + { + "epoch": 0.36127, + "grad_norm": 0.780598207065133, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 36127 + }, + { + "epoch": 0.36128, + "grad_norm": 0.8468276976759895, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 36128 + }, + { + "epoch": 0.36129, + "grad_norm": 1.0677318234140878, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 36129 + }, + { + "epoch": 0.3613, + "grad_norm": 0.9991863980287546, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 36130 + }, + { + "epoch": 0.36131, + "grad_norm": 0.8630156051760434, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 36131 + }, + { + "epoch": 0.36132, + "grad_norm": 0.7568578431453746, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 36132 + }, + { + "epoch": 0.36133, + "grad_norm": 0.6900021049970169, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 36133 + }, + { + "epoch": 0.36134, + "grad_norm": 0.7364539543554925, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 36134 + }, + { + "epoch": 0.36135, + "grad_norm": 0.7550265363467387, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 36135 + }, + { + "epoch": 0.36136, + "grad_norm": 0.7499857630607109, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 36136 + }, + { + "epoch": 0.36137, + "grad_norm": 0.7073873448455503, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 36137 + }, + { + "epoch": 0.36138, + "grad_norm": 0.7668414045916963, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 36138 + }, + { + "epoch": 0.36139, + "grad_norm": 0.8155854643599342, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 36139 + }, + { + "epoch": 0.3614, + "grad_norm": 0.8477186706631795, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 36140 + }, + { + "epoch": 0.36141, + "grad_norm": 0.8662546517811139, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 36141 + }, + { + "epoch": 0.36142, + "grad_norm": 0.8850973364702642, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 36142 + }, + { + "epoch": 0.36143, + "grad_norm": 0.914780371232853, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 36143 + }, + { + "epoch": 0.36144, + "grad_norm": 0.9849202486746463, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 36144 + }, + { + "epoch": 0.36145, + "grad_norm": 1.0485212121922882, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 36145 + }, + { + "epoch": 0.36146, + "grad_norm": 1.009243240002195, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 36146 + }, + { + "epoch": 0.36147, + "grad_norm": 1.0896479402191688, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 36147 + }, + { + "epoch": 0.36148, + "grad_norm": 0.8304856809656125, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 36148 + }, + { + "epoch": 0.36149, + "grad_norm": 0.8969156072998892, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 36149 + }, + { + "epoch": 0.3615, + "grad_norm": 0.9943963961249569, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 36150 + }, + { + "epoch": 0.36151, + "grad_norm": 1.0408312948676919, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 36151 + }, + { + "epoch": 0.36152, + "grad_norm": 1.0227929195694563, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 36152 + }, + { + "epoch": 0.36153, + "grad_norm": 0.9693773596176812, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 36153 + }, + { + "epoch": 0.36154, + "grad_norm": 0.9614478877695228, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 36154 + }, + { + "epoch": 0.36155, + "grad_norm": 0.8517472463302511, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 36155 + }, + { + "epoch": 0.36156, + "grad_norm": 0.7904002109591328, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 36156 + }, + { + "epoch": 0.36157, + "grad_norm": 0.8652903655222919, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 36157 + }, + { + "epoch": 0.36158, + "grad_norm": 0.8239817437904867, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 36158 + }, + { + "epoch": 0.36159, + "grad_norm": 0.5932671516289173, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 36159 + }, + { + "epoch": 0.3616, + "grad_norm": 0.5912606740176982, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 36160 + }, + { + "epoch": 0.36161, + "grad_norm": 0.4968467487136162, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 36161 + }, + { + "epoch": 0.36162, + "grad_norm": 0.4918950232831561, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 36162 + }, + { + "epoch": 0.36163, + "grad_norm": 0.5626963172210498, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 36163 + }, + { + "epoch": 0.36164, + "grad_norm": 0.5841158758226975, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 36164 + }, + { + "epoch": 0.36165, + "grad_norm": 0.592784309773468, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 36165 + }, + { + "epoch": 0.36166, + "grad_norm": 0.4710473411023664, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 36166 + }, + { + "epoch": 0.36167, + "grad_norm": 0.4537802130839176, + "learning_rate": 0.003, + "loss": 3.976, + "step": 36167 + }, + { + "epoch": 0.36168, + "grad_norm": 0.48782369807177145, + "learning_rate": 0.003, + "loss": 3.97, + "step": 36168 + }, + { + "epoch": 0.36169, + "grad_norm": 0.5451964542225893, + "learning_rate": 0.003, + "loss": 3.9329, + "step": 36169 + }, + { + "epoch": 0.3617, + "grad_norm": 0.6986412129122822, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 36170 + }, + { + "epoch": 0.36171, + "grad_norm": 0.8316623056825552, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 36171 + }, + { + "epoch": 0.36172, + "grad_norm": 0.9675859648593087, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 36172 + }, + { + "epoch": 0.36173, + "grad_norm": 0.9846083314571136, + "learning_rate": 0.003, + "loss": 3.965, + "step": 36173 + }, + { + "epoch": 0.36174, + "grad_norm": 0.957213966650663, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 36174 + }, + { + "epoch": 0.36175, + "grad_norm": 0.7891328077454641, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 36175 + }, + { + "epoch": 0.36176, + "grad_norm": 0.6700335866995166, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 36176 + }, + { + "epoch": 0.36177, + "grad_norm": 0.6386225700198103, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 36177 + }, + { + "epoch": 0.36178, + "grad_norm": 0.6699633690362037, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 36178 + }, + { + "epoch": 0.36179, + "grad_norm": 0.6359091180698718, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 36179 + }, + { + "epoch": 0.3618, + "grad_norm": 0.624304434399466, + "learning_rate": 0.003, + "loss": 3.9287, + "step": 36180 + }, + { + "epoch": 0.36181, + "grad_norm": 0.5701091811367675, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 36181 + }, + { + "epoch": 0.36182, + "grad_norm": 0.5354611505539254, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 36182 + }, + { + "epoch": 0.36183, + "grad_norm": 0.5031213300209743, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 36183 + }, + { + "epoch": 0.36184, + "grad_norm": 0.5818711276651438, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 36184 + }, + { + "epoch": 0.36185, + "grad_norm": 0.613425706660839, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 36185 + }, + { + "epoch": 0.36186, + "grad_norm": 0.7232689693950513, + "learning_rate": 0.003, + "loss": 3.9334, + "step": 36186 + }, + { + "epoch": 0.36187, + "grad_norm": 0.9305522779734635, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 36187 + }, + { + "epoch": 0.36188, + "grad_norm": 1.116841337206311, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 36188 + }, + { + "epoch": 0.36189, + "grad_norm": 0.77945303983206, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 36189 + }, + { + "epoch": 0.3619, + "grad_norm": 0.7151567733725992, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 36190 + }, + { + "epoch": 0.36191, + "grad_norm": 0.785308368406591, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 36191 + }, + { + "epoch": 0.36192, + "grad_norm": 0.7594938319331259, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 36192 + }, + { + "epoch": 0.36193, + "grad_norm": 0.7666523092328063, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 36193 + }, + { + "epoch": 0.36194, + "grad_norm": 0.7645287287638741, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 36194 + }, + { + "epoch": 0.36195, + "grad_norm": 0.868889406245545, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 36195 + }, + { + "epoch": 0.36196, + "grad_norm": 0.9142188401409324, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 36196 + }, + { + "epoch": 0.36197, + "grad_norm": 0.9550798126288419, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 36197 + }, + { + "epoch": 0.36198, + "grad_norm": 0.7999637921138538, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 36198 + }, + { + "epoch": 0.36199, + "grad_norm": 0.7225475373932329, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 36199 + }, + { + "epoch": 0.362, + "grad_norm": 0.7206472835603995, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 36200 + }, + { + "epoch": 0.36201, + "grad_norm": 0.8159342676940927, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 36201 + }, + { + "epoch": 0.36202, + "grad_norm": 0.8922828753711015, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 36202 + }, + { + "epoch": 0.36203, + "grad_norm": 0.8256317278591391, + "learning_rate": 0.003, + "loss": 4.014, + "step": 36203 + }, + { + "epoch": 0.36204, + "grad_norm": 0.7201036413925909, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 36204 + }, + { + "epoch": 0.36205, + "grad_norm": 0.779051875607736, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 36205 + }, + { + "epoch": 0.36206, + "grad_norm": 0.7800220425337698, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 36206 + }, + { + "epoch": 0.36207, + "grad_norm": 0.8677631633596163, + "learning_rate": 0.003, + "loss": 3.963, + "step": 36207 + }, + { + "epoch": 0.36208, + "grad_norm": 0.8528526988368768, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 36208 + }, + { + "epoch": 0.36209, + "grad_norm": 0.9085856894337434, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 36209 + }, + { + "epoch": 0.3621, + "grad_norm": 1.033876169024203, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 36210 + }, + { + "epoch": 0.36211, + "grad_norm": 1.3447915969710076, + "learning_rate": 0.003, + "loss": 4.0354, + "step": 36211 + }, + { + "epoch": 0.36212, + "grad_norm": 0.8145268593937279, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 36212 + }, + { + "epoch": 0.36213, + "grad_norm": 0.7409921646877149, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 36213 + }, + { + "epoch": 0.36214, + "grad_norm": 0.7284330352140229, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 36214 + }, + { + "epoch": 0.36215, + "grad_norm": 0.7792571589086127, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 36215 + }, + { + "epoch": 0.36216, + "grad_norm": 0.7652715244535283, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 36216 + }, + { + "epoch": 0.36217, + "grad_norm": 0.8592419961034332, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 36217 + }, + { + "epoch": 0.36218, + "grad_norm": 0.8570522124347945, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 36218 + }, + { + "epoch": 0.36219, + "grad_norm": 0.7635638207551402, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 36219 + }, + { + "epoch": 0.3622, + "grad_norm": 0.7811766788614256, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 36220 + }, + { + "epoch": 0.36221, + "grad_norm": 0.6483087479095254, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 36221 + }, + { + "epoch": 0.36222, + "grad_norm": 0.5930314729584273, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 36222 + }, + { + "epoch": 0.36223, + "grad_norm": 0.652734516830896, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 36223 + }, + { + "epoch": 0.36224, + "grad_norm": 0.6279785066313875, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 36224 + }, + { + "epoch": 0.36225, + "grad_norm": 0.6558457592051072, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 36225 + }, + { + "epoch": 0.36226, + "grad_norm": 0.6410750477793979, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 36226 + }, + { + "epoch": 0.36227, + "grad_norm": 0.558799324607615, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 36227 + }, + { + "epoch": 0.36228, + "grad_norm": 0.5129939034198119, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 36228 + }, + { + "epoch": 0.36229, + "grad_norm": 0.5725539025366239, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 36229 + }, + { + "epoch": 0.3623, + "grad_norm": 0.6263980302798658, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 36230 + }, + { + "epoch": 0.36231, + "grad_norm": 0.6885122301880153, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 36231 + }, + { + "epoch": 0.36232, + "grad_norm": 0.8269490223464139, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 36232 + }, + { + "epoch": 0.36233, + "grad_norm": 0.870790110335388, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 36233 + }, + { + "epoch": 0.36234, + "grad_norm": 1.0115428085851865, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 36234 + }, + { + "epoch": 0.36235, + "grad_norm": 1.238534462880614, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 36235 + }, + { + "epoch": 0.36236, + "grad_norm": 0.808453663457491, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 36236 + }, + { + "epoch": 0.36237, + "grad_norm": 1.0662532097441015, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 36237 + }, + { + "epoch": 0.36238, + "grad_norm": 1.033184502216387, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 36238 + }, + { + "epoch": 0.36239, + "grad_norm": 0.8480444224753354, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 36239 + }, + { + "epoch": 0.3624, + "grad_norm": 0.7441267655041379, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 36240 + }, + { + "epoch": 0.36241, + "grad_norm": 0.7781566701916047, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 36241 + }, + { + "epoch": 0.36242, + "grad_norm": 0.7653282603814147, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 36242 + }, + { + "epoch": 0.36243, + "grad_norm": 0.6463929636472617, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 36243 + }, + { + "epoch": 0.36244, + "grad_norm": 0.5726258639971762, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 36244 + }, + { + "epoch": 0.36245, + "grad_norm": 0.6308843790358027, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 36245 + }, + { + "epoch": 0.36246, + "grad_norm": 0.5920624486065669, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 36246 + }, + { + "epoch": 0.36247, + "grad_norm": 0.5824353568796482, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 36247 + }, + { + "epoch": 0.36248, + "grad_norm": 0.5780140037660252, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 36248 + }, + { + "epoch": 0.36249, + "grad_norm": 0.5920148721636336, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 36249 + }, + { + "epoch": 0.3625, + "grad_norm": 0.613055684128019, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 36250 + }, + { + "epoch": 0.36251, + "grad_norm": 0.6286641708832837, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 36251 + }, + { + "epoch": 0.36252, + "grad_norm": 0.5145855157213327, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 36252 + }, + { + "epoch": 0.36253, + "grad_norm": 0.4986966112855566, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 36253 + }, + { + "epoch": 0.36254, + "grad_norm": 0.5890673073922651, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 36254 + }, + { + "epoch": 0.36255, + "grad_norm": 0.707847769170496, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 36255 + }, + { + "epoch": 0.36256, + "grad_norm": 0.8648264866928199, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 36256 + }, + { + "epoch": 0.36257, + "grad_norm": 0.940609986377098, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 36257 + }, + { + "epoch": 0.36258, + "grad_norm": 0.8548030895068668, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 36258 + }, + { + "epoch": 0.36259, + "grad_norm": 0.7164017211839672, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 36259 + }, + { + "epoch": 0.3626, + "grad_norm": 0.7751835959021768, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 36260 + }, + { + "epoch": 0.36261, + "grad_norm": 0.7860634139591716, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 36261 + }, + { + "epoch": 0.36262, + "grad_norm": 0.8577983103264119, + "learning_rate": 0.003, + "loss": 3.971, + "step": 36262 + }, + { + "epoch": 0.36263, + "grad_norm": 0.9019761052438237, + "learning_rate": 0.003, + "loss": 3.982, + "step": 36263 + }, + { + "epoch": 0.36264, + "grad_norm": 0.8841938608624463, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 36264 + }, + { + "epoch": 0.36265, + "grad_norm": 0.9039861948167854, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 36265 + }, + { + "epoch": 0.36266, + "grad_norm": 0.8850853495279126, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 36266 + }, + { + "epoch": 0.36267, + "grad_norm": 0.8293487913292814, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 36267 + }, + { + "epoch": 0.36268, + "grad_norm": 0.7644397218612676, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 36268 + }, + { + "epoch": 0.36269, + "grad_norm": 0.7883437207485574, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 36269 + }, + { + "epoch": 0.3627, + "grad_norm": 0.7364036907169972, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 36270 + }, + { + "epoch": 0.36271, + "grad_norm": 0.7364746828437313, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 36271 + }, + { + "epoch": 0.36272, + "grad_norm": 0.7258138082109775, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 36272 + }, + { + "epoch": 0.36273, + "grad_norm": 0.7962398924242226, + "learning_rate": 0.003, + "loss": 3.992, + "step": 36273 + }, + { + "epoch": 0.36274, + "grad_norm": 1.018651758176802, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 36274 + }, + { + "epoch": 0.36275, + "grad_norm": 1.134135404959161, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 36275 + }, + { + "epoch": 0.36276, + "grad_norm": 0.7993240896699674, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 36276 + }, + { + "epoch": 0.36277, + "grad_norm": 0.9287928182188089, + "learning_rate": 0.003, + "loss": 3.994, + "step": 36277 + }, + { + "epoch": 0.36278, + "grad_norm": 1.0021589148227992, + "learning_rate": 0.003, + "loss": 4.005, + "step": 36278 + }, + { + "epoch": 0.36279, + "grad_norm": 0.9284685205123999, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 36279 + }, + { + "epoch": 0.3628, + "grad_norm": 0.7827482046274662, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 36280 + }, + { + "epoch": 0.36281, + "grad_norm": 0.8181639606729773, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 36281 + }, + { + "epoch": 0.36282, + "grad_norm": 0.7971308686976185, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 36282 + }, + { + "epoch": 0.36283, + "grad_norm": 0.8891523478602905, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 36283 + }, + { + "epoch": 0.36284, + "grad_norm": 0.859369604839796, + "learning_rate": 0.003, + "loss": 4.018, + "step": 36284 + }, + { + "epoch": 0.36285, + "grad_norm": 0.7023653253039751, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 36285 + }, + { + "epoch": 0.36286, + "grad_norm": 0.733892508545896, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 36286 + }, + { + "epoch": 0.36287, + "grad_norm": 0.8530456752242771, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 36287 + }, + { + "epoch": 0.36288, + "grad_norm": 1.0920509217181644, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 36288 + }, + { + "epoch": 0.36289, + "grad_norm": 0.9045436047921612, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 36289 + }, + { + "epoch": 0.3629, + "grad_norm": 0.8598622723946289, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 36290 + }, + { + "epoch": 0.36291, + "grad_norm": 0.7423595385106935, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 36291 + }, + { + "epoch": 0.36292, + "grad_norm": 0.6420676359757923, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 36292 + }, + { + "epoch": 0.36293, + "grad_norm": 0.6129034745428672, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 36293 + }, + { + "epoch": 0.36294, + "grad_norm": 0.6219828825134838, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 36294 + }, + { + "epoch": 0.36295, + "grad_norm": 0.6508252642463691, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 36295 + }, + { + "epoch": 0.36296, + "grad_norm": 0.6950363034590706, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 36296 + }, + { + "epoch": 0.36297, + "grad_norm": 0.7253601154060485, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 36297 + }, + { + "epoch": 0.36298, + "grad_norm": 0.8268671952921252, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 36298 + }, + { + "epoch": 0.36299, + "grad_norm": 0.8725008206786862, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 36299 + }, + { + "epoch": 0.363, + "grad_norm": 1.0108678763024967, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 36300 + }, + { + "epoch": 0.36301, + "grad_norm": 0.9509883857215937, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 36301 + }, + { + "epoch": 0.36302, + "grad_norm": 0.7287646956550231, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 36302 + }, + { + "epoch": 0.36303, + "grad_norm": 0.6717916118138987, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 36303 + }, + { + "epoch": 0.36304, + "grad_norm": 0.6489544009355827, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 36304 + }, + { + "epoch": 0.36305, + "grad_norm": 0.7136544825972334, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 36305 + }, + { + "epoch": 0.36306, + "grad_norm": 0.9309668520898553, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 36306 + }, + { + "epoch": 0.36307, + "grad_norm": 0.9763277607922167, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 36307 + }, + { + "epoch": 0.36308, + "grad_norm": 0.755475411755065, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 36308 + }, + { + "epoch": 0.36309, + "grad_norm": 0.6711555167491022, + "learning_rate": 0.003, + "loss": 3.978, + "step": 36309 + }, + { + "epoch": 0.3631, + "grad_norm": 0.8850471186854472, + "learning_rate": 0.003, + "loss": 3.991, + "step": 36310 + }, + { + "epoch": 0.36311, + "grad_norm": 1.0665868991290146, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 36311 + }, + { + "epoch": 0.36312, + "grad_norm": 0.8393012406992445, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 36312 + }, + { + "epoch": 0.36313, + "grad_norm": 0.6886756079648054, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 36313 + }, + { + "epoch": 0.36314, + "grad_norm": 0.6827456075947519, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 36314 + }, + { + "epoch": 0.36315, + "grad_norm": 0.7061388240672337, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 36315 + }, + { + "epoch": 0.36316, + "grad_norm": 0.7161301908161544, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 36316 + }, + { + "epoch": 0.36317, + "grad_norm": 0.5727185298577121, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 36317 + }, + { + "epoch": 0.36318, + "grad_norm": 0.5321704563697585, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 36318 + }, + { + "epoch": 0.36319, + "grad_norm": 0.4675395080155422, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 36319 + }, + { + "epoch": 0.3632, + "grad_norm": 0.48667542985390755, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 36320 + }, + { + "epoch": 0.36321, + "grad_norm": 0.4929185018769133, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 36321 + }, + { + "epoch": 0.36322, + "grad_norm": 0.5688620044785493, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 36322 + }, + { + "epoch": 0.36323, + "grad_norm": 0.6079889907109771, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 36323 + }, + { + "epoch": 0.36324, + "grad_norm": 0.5508567847000859, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 36324 + }, + { + "epoch": 0.36325, + "grad_norm": 0.6521908166591298, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 36325 + }, + { + "epoch": 0.36326, + "grad_norm": 0.7040494701614646, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 36326 + }, + { + "epoch": 0.36327, + "grad_norm": 0.8075551284529706, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 36327 + }, + { + "epoch": 0.36328, + "grad_norm": 0.8823415662304881, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 36328 + }, + { + "epoch": 0.36329, + "grad_norm": 0.9542502458860478, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 36329 + }, + { + "epoch": 0.3633, + "grad_norm": 1.0065729772131948, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 36330 + }, + { + "epoch": 0.36331, + "grad_norm": 0.9678688439649723, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 36331 + }, + { + "epoch": 0.36332, + "grad_norm": 1.0079125763717773, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 36332 + }, + { + "epoch": 0.36333, + "grad_norm": 0.8690867652763333, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 36333 + }, + { + "epoch": 0.36334, + "grad_norm": 0.9684563617213806, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 36334 + }, + { + "epoch": 0.36335, + "grad_norm": 0.9084695092155991, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 36335 + }, + { + "epoch": 0.36336, + "grad_norm": 0.9143512437686235, + "learning_rate": 0.003, + "loss": 3.975, + "step": 36336 + }, + { + "epoch": 0.36337, + "grad_norm": 1.0997736908262026, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 36337 + }, + { + "epoch": 0.36338, + "grad_norm": 1.0014719636915599, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 36338 + }, + { + "epoch": 0.36339, + "grad_norm": 0.8925842622148618, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 36339 + }, + { + "epoch": 0.3634, + "grad_norm": 0.9627681680834757, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 36340 + }, + { + "epoch": 0.36341, + "grad_norm": 0.9732242487335764, + "learning_rate": 0.003, + "loss": 4.0, + "step": 36341 + }, + { + "epoch": 0.36342, + "grad_norm": 0.9754453534409858, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 36342 + }, + { + "epoch": 0.36343, + "grad_norm": 0.9588318864719246, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 36343 + }, + { + "epoch": 0.36344, + "grad_norm": 0.9549806521885167, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 36344 + }, + { + "epoch": 0.36345, + "grad_norm": 0.9597297882459908, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 36345 + }, + { + "epoch": 0.36346, + "grad_norm": 1.020035954218298, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 36346 + }, + { + "epoch": 0.36347, + "grad_norm": 0.846508632154279, + "learning_rate": 0.003, + "loss": 4.0358, + "step": 36347 + }, + { + "epoch": 0.36348, + "grad_norm": 0.8564428803575749, + "learning_rate": 0.003, + "loss": 4.0304, + "step": 36348 + }, + { + "epoch": 0.36349, + "grad_norm": 1.0137823879526178, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 36349 + }, + { + "epoch": 0.3635, + "grad_norm": 0.984261899785057, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 36350 + }, + { + "epoch": 0.36351, + "grad_norm": 0.7659268516973846, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 36351 + }, + { + "epoch": 0.36352, + "grad_norm": 0.6656251804699231, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 36352 + }, + { + "epoch": 0.36353, + "grad_norm": 0.6411060806793136, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 36353 + }, + { + "epoch": 0.36354, + "grad_norm": 0.6814828026217363, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 36354 + }, + { + "epoch": 0.36355, + "grad_norm": 0.679428827904559, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 36355 + }, + { + "epoch": 0.36356, + "grad_norm": 0.6615916681456189, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 36356 + }, + { + "epoch": 0.36357, + "grad_norm": 0.6298355695532912, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 36357 + }, + { + "epoch": 0.36358, + "grad_norm": 0.5554683322347722, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 36358 + }, + { + "epoch": 0.36359, + "grad_norm": 0.5079388559716683, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 36359 + }, + { + "epoch": 0.3636, + "grad_norm": 0.547934441831116, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 36360 + }, + { + "epoch": 0.36361, + "grad_norm": 0.6084453029960716, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 36361 + }, + { + "epoch": 0.36362, + "grad_norm": 0.6100475708717723, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 36362 + }, + { + "epoch": 0.36363, + "grad_norm": 0.691517874340762, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 36363 + }, + { + "epoch": 0.36364, + "grad_norm": 0.7509498935184743, + "learning_rate": 0.003, + "loss": 3.998, + "step": 36364 + }, + { + "epoch": 0.36365, + "grad_norm": 0.8476065129551122, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 36365 + }, + { + "epoch": 0.36366, + "grad_norm": 0.9043272068278216, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 36366 + }, + { + "epoch": 0.36367, + "grad_norm": 0.8046358560447459, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 36367 + }, + { + "epoch": 0.36368, + "grad_norm": 0.7826573926663385, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 36368 + }, + { + "epoch": 0.36369, + "grad_norm": 0.7535940977530147, + "learning_rate": 0.003, + "loss": 3.9454, + "step": 36369 + }, + { + "epoch": 0.3637, + "grad_norm": 0.8096230985542527, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 36370 + }, + { + "epoch": 0.36371, + "grad_norm": 0.9569741641888482, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 36371 + }, + { + "epoch": 0.36372, + "grad_norm": 1.088017915913959, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 36372 + }, + { + "epoch": 0.36373, + "grad_norm": 0.8812432105843406, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 36373 + }, + { + "epoch": 0.36374, + "grad_norm": 0.8076630682397237, + "learning_rate": 0.003, + "loss": 3.993, + "step": 36374 + }, + { + "epoch": 0.36375, + "grad_norm": 0.7646689698018122, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 36375 + }, + { + "epoch": 0.36376, + "grad_norm": 0.6814571392101529, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 36376 + }, + { + "epoch": 0.36377, + "grad_norm": 0.7288502917272492, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 36377 + }, + { + "epoch": 0.36378, + "grad_norm": 0.6917144596411257, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 36378 + }, + { + "epoch": 0.36379, + "grad_norm": 0.6498343565302744, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 36379 + }, + { + "epoch": 0.3638, + "grad_norm": 0.7761716969475662, + "learning_rate": 0.003, + "loss": 3.978, + "step": 36380 + }, + { + "epoch": 0.36381, + "grad_norm": 0.8005279776252305, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 36381 + }, + { + "epoch": 0.36382, + "grad_norm": 0.7781482868730656, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 36382 + }, + { + "epoch": 0.36383, + "grad_norm": 0.6966445776085768, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 36383 + }, + { + "epoch": 0.36384, + "grad_norm": 0.7061010436228707, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 36384 + }, + { + "epoch": 0.36385, + "grad_norm": 0.6530432869760708, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 36385 + }, + { + "epoch": 0.36386, + "grad_norm": 0.5908095096855353, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 36386 + }, + { + "epoch": 0.36387, + "grad_norm": 0.5406587001308878, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 36387 + }, + { + "epoch": 0.36388, + "grad_norm": 0.5557483232192022, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 36388 + }, + { + "epoch": 0.36389, + "grad_norm": 0.6126122493969585, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36389 + }, + { + "epoch": 0.3639, + "grad_norm": 0.6328610893474367, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 36390 + }, + { + "epoch": 0.36391, + "grad_norm": 0.5508262782813212, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 36391 + }, + { + "epoch": 0.36392, + "grad_norm": 0.579521624444295, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 36392 + }, + { + "epoch": 0.36393, + "grad_norm": 0.6975345898773575, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 36393 + }, + { + "epoch": 0.36394, + "grad_norm": 0.75748148454774, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 36394 + }, + { + "epoch": 0.36395, + "grad_norm": 0.7309218713850152, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 36395 + }, + { + "epoch": 0.36396, + "grad_norm": 0.7697580937132107, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 36396 + }, + { + "epoch": 0.36397, + "grad_norm": 0.7853320108191362, + "learning_rate": 0.003, + "loss": 3.938, + "step": 36397 + }, + { + "epoch": 0.36398, + "grad_norm": 0.8679341769721415, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 36398 + }, + { + "epoch": 0.36399, + "grad_norm": 1.1302741982333508, + "learning_rate": 0.003, + "loss": 4.003, + "step": 36399 + }, + { + "epoch": 0.364, + "grad_norm": 1.00328866680757, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 36400 + }, + { + "epoch": 0.36401, + "grad_norm": 0.9754924339610815, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 36401 + }, + { + "epoch": 0.36402, + "grad_norm": 0.9042263500067316, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 36402 + }, + { + "epoch": 0.36403, + "grad_norm": 0.8129236575681639, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 36403 + }, + { + "epoch": 0.36404, + "grad_norm": 0.9516584888313604, + "learning_rate": 0.003, + "loss": 3.984, + "step": 36404 + }, + { + "epoch": 0.36405, + "grad_norm": 1.0122376710251237, + "learning_rate": 0.003, + "loss": 4.016, + "step": 36405 + }, + { + "epoch": 0.36406, + "grad_norm": 0.863533424375761, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 36406 + }, + { + "epoch": 0.36407, + "grad_norm": 0.9020355147762316, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 36407 + }, + { + "epoch": 0.36408, + "grad_norm": 0.8232621602561458, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 36408 + }, + { + "epoch": 0.36409, + "grad_norm": 0.8831346008185105, + "learning_rate": 0.003, + "loss": 3.998, + "step": 36409 + }, + { + "epoch": 0.3641, + "grad_norm": 0.8293316996910209, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 36410 + }, + { + "epoch": 0.36411, + "grad_norm": 0.7876158337198457, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 36411 + }, + { + "epoch": 0.36412, + "grad_norm": 0.8503085267108014, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 36412 + }, + { + "epoch": 0.36413, + "grad_norm": 0.894398530228946, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 36413 + }, + { + "epoch": 0.36414, + "grad_norm": 0.9160557303451616, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 36414 + }, + { + "epoch": 0.36415, + "grad_norm": 0.8513771026101011, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 36415 + }, + { + "epoch": 0.36416, + "grad_norm": 0.8659268841837822, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 36416 + }, + { + "epoch": 0.36417, + "grad_norm": 0.8812519198347109, + "learning_rate": 0.003, + "loss": 4.005, + "step": 36417 + }, + { + "epoch": 0.36418, + "grad_norm": 0.8015562812859589, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 36418 + }, + { + "epoch": 0.36419, + "grad_norm": 0.8230243453862077, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 36419 + }, + { + "epoch": 0.3642, + "grad_norm": 0.870685206789036, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 36420 + }, + { + "epoch": 0.36421, + "grad_norm": 0.9046558139606039, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 36421 + }, + { + "epoch": 0.36422, + "grad_norm": 0.7566797845989844, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 36422 + }, + { + "epoch": 0.36423, + "grad_norm": 0.6835171495411866, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 36423 + }, + { + "epoch": 0.36424, + "grad_norm": 0.6702561050220979, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 36424 + }, + { + "epoch": 0.36425, + "grad_norm": 0.6256891010524971, + "learning_rate": 0.003, + "loss": 3.986, + "step": 36425 + }, + { + "epoch": 0.36426, + "grad_norm": 0.6125634742791394, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 36426 + }, + { + "epoch": 0.36427, + "grad_norm": 0.7213510065553119, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 36427 + }, + { + "epoch": 0.36428, + "grad_norm": 0.8583188290898323, + "learning_rate": 0.003, + "loss": 3.971, + "step": 36428 + }, + { + "epoch": 0.36429, + "grad_norm": 0.8247569012876984, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 36429 + }, + { + "epoch": 0.3643, + "grad_norm": 0.726352959345181, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 36430 + }, + { + "epoch": 0.36431, + "grad_norm": 0.6135426537748868, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 36431 + }, + { + "epoch": 0.36432, + "grad_norm": 0.5580471820864797, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 36432 + }, + { + "epoch": 0.36433, + "grad_norm": 0.6433030352353047, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 36433 + }, + { + "epoch": 0.36434, + "grad_norm": 0.711702028896053, + "learning_rate": 0.003, + "loss": 4.0, + "step": 36434 + }, + { + "epoch": 0.36435, + "grad_norm": 0.6752019850882239, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 36435 + }, + { + "epoch": 0.36436, + "grad_norm": 0.622210132333093, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 36436 + }, + { + "epoch": 0.36437, + "grad_norm": 0.5950398929134019, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 36437 + }, + { + "epoch": 0.36438, + "grad_norm": 0.6232443098454412, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 36438 + }, + { + "epoch": 0.36439, + "grad_norm": 0.6219431564346625, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 36439 + }, + { + "epoch": 0.3644, + "grad_norm": 0.6264437682556371, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 36440 + }, + { + "epoch": 0.36441, + "grad_norm": 0.672437268529299, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 36441 + }, + { + "epoch": 0.36442, + "grad_norm": 0.6979751092780355, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 36442 + }, + { + "epoch": 0.36443, + "grad_norm": 0.7565896417815902, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 36443 + }, + { + "epoch": 0.36444, + "grad_norm": 0.6594125047083171, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 36444 + }, + { + "epoch": 0.36445, + "grad_norm": 0.6509848845964069, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 36445 + }, + { + "epoch": 0.36446, + "grad_norm": 0.6219681313494759, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 36446 + }, + { + "epoch": 0.36447, + "grad_norm": 0.5503742483735655, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 36447 + }, + { + "epoch": 0.36448, + "grad_norm": 0.6471253539626863, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 36448 + }, + { + "epoch": 0.36449, + "grad_norm": 0.804966097186655, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 36449 + }, + { + "epoch": 0.3645, + "grad_norm": 0.9520611714850674, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 36450 + }, + { + "epoch": 0.36451, + "grad_norm": 1.0511931713171898, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 36451 + }, + { + "epoch": 0.36452, + "grad_norm": 0.8559264815148131, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 36452 + }, + { + "epoch": 0.36453, + "grad_norm": 0.6486395142513832, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 36453 + }, + { + "epoch": 0.36454, + "grad_norm": 0.7225379317914379, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 36454 + }, + { + "epoch": 0.36455, + "grad_norm": 0.6726762969687481, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 36455 + }, + { + "epoch": 0.36456, + "grad_norm": 0.7187134895679185, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 36456 + }, + { + "epoch": 0.36457, + "grad_norm": 0.727517463223111, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 36457 + }, + { + "epoch": 0.36458, + "grad_norm": 0.8022750654399886, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 36458 + }, + { + "epoch": 0.36459, + "grad_norm": 0.9449342832903538, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 36459 + }, + { + "epoch": 0.3646, + "grad_norm": 1.0565360926361254, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 36460 + }, + { + "epoch": 0.36461, + "grad_norm": 0.7857884406611055, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 36461 + }, + { + "epoch": 0.36462, + "grad_norm": 0.8045280680465176, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 36462 + }, + { + "epoch": 0.36463, + "grad_norm": 0.8606718722503058, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 36463 + }, + { + "epoch": 0.36464, + "grad_norm": 0.9405234628110184, + "learning_rate": 0.003, + "loss": 4.006, + "step": 36464 + }, + { + "epoch": 0.36465, + "grad_norm": 0.9316085033298432, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 36465 + }, + { + "epoch": 0.36466, + "grad_norm": 0.9129986917813837, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 36466 + }, + { + "epoch": 0.36467, + "grad_norm": 1.0125013602147455, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 36467 + }, + { + "epoch": 0.36468, + "grad_norm": 1.0213943006137907, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 36468 + }, + { + "epoch": 0.36469, + "grad_norm": 1.0432772740191423, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 36469 + }, + { + "epoch": 0.3647, + "grad_norm": 1.1750697668650778, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 36470 + }, + { + "epoch": 0.36471, + "grad_norm": 1.0645450061884707, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 36471 + }, + { + "epoch": 0.36472, + "grad_norm": 0.9487928689782212, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 36472 + }, + { + "epoch": 0.36473, + "grad_norm": 0.8630701467384645, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 36473 + }, + { + "epoch": 0.36474, + "grad_norm": 0.9065084016704901, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 36474 + }, + { + "epoch": 0.36475, + "grad_norm": 0.9040049734226634, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 36475 + }, + { + "epoch": 0.36476, + "grad_norm": 0.8506776852201933, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 36476 + }, + { + "epoch": 0.36477, + "grad_norm": 0.8725130606571769, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 36477 + }, + { + "epoch": 0.36478, + "grad_norm": 0.8503189304809415, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 36478 + }, + { + "epoch": 0.36479, + "grad_norm": 0.9084417210365341, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 36479 + }, + { + "epoch": 0.3648, + "grad_norm": 0.8783605408247995, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 36480 + }, + { + "epoch": 0.36481, + "grad_norm": 0.8113472962262411, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 36481 + }, + { + "epoch": 0.36482, + "grad_norm": 0.7539945357892502, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 36482 + }, + { + "epoch": 0.36483, + "grad_norm": 0.6960902644503565, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 36483 + }, + { + "epoch": 0.36484, + "grad_norm": 0.6567414189671503, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 36484 + }, + { + "epoch": 0.36485, + "grad_norm": 0.729166877297516, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 36485 + }, + { + "epoch": 0.36486, + "grad_norm": 0.7292828092848257, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 36486 + }, + { + "epoch": 0.36487, + "grad_norm": 0.7826876832667998, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 36487 + }, + { + "epoch": 0.36488, + "grad_norm": 0.8234910306500641, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 36488 + }, + { + "epoch": 0.36489, + "grad_norm": 0.8930310966123485, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 36489 + }, + { + "epoch": 0.3649, + "grad_norm": 0.8444346450607454, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 36490 + }, + { + "epoch": 0.36491, + "grad_norm": 0.7643920212543397, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 36491 + }, + { + "epoch": 0.36492, + "grad_norm": 0.7184899339719553, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 36492 + }, + { + "epoch": 0.36493, + "grad_norm": 0.7505727072432233, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36493 + }, + { + "epoch": 0.36494, + "grad_norm": 0.816931491720464, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 36494 + }, + { + "epoch": 0.36495, + "grad_norm": 0.8777577600629848, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 36495 + }, + { + "epoch": 0.36496, + "grad_norm": 0.9627116168196922, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 36496 + }, + { + "epoch": 0.36497, + "grad_norm": 1.0358658940533205, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 36497 + }, + { + "epoch": 0.36498, + "grad_norm": 0.9039088574463211, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 36498 + }, + { + "epoch": 0.36499, + "grad_norm": 0.6859836783453321, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 36499 + }, + { + "epoch": 0.365, + "grad_norm": 0.650974978167186, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 36500 + }, + { + "epoch": 0.36501, + "grad_norm": 0.7318770729360985, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 36501 + }, + { + "epoch": 0.36502, + "grad_norm": 0.6712318636050507, + "learning_rate": 0.003, + "loss": 3.978, + "step": 36502 + }, + { + "epoch": 0.36503, + "grad_norm": 0.5959146452786983, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 36503 + }, + { + "epoch": 0.36504, + "grad_norm": 0.5395620621328415, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 36504 + }, + { + "epoch": 0.36505, + "grad_norm": 0.5079848182618162, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 36505 + }, + { + "epoch": 0.36506, + "grad_norm": 0.5044373211643174, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 36506 + }, + { + "epoch": 0.36507, + "grad_norm": 0.5370273639335376, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 36507 + }, + { + "epoch": 0.36508, + "grad_norm": 0.5580896246875724, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 36508 + }, + { + "epoch": 0.36509, + "grad_norm": 0.6008192218999251, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 36509 + }, + { + "epoch": 0.3651, + "grad_norm": 0.695299979768008, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 36510 + }, + { + "epoch": 0.36511, + "grad_norm": 0.8206955249345628, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 36511 + }, + { + "epoch": 0.36512, + "grad_norm": 0.8240795036188814, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 36512 + }, + { + "epoch": 0.36513, + "grad_norm": 0.667464898552451, + "learning_rate": 0.003, + "loss": 3.9402, + "step": 36513 + }, + { + "epoch": 0.36514, + "grad_norm": 0.7397868592678551, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 36514 + }, + { + "epoch": 0.36515, + "grad_norm": 0.6952032936714039, + "learning_rate": 0.003, + "loss": 3.938, + "step": 36515 + }, + { + "epoch": 0.36516, + "grad_norm": 0.7352347403794479, + "learning_rate": 0.003, + "loss": 3.98, + "step": 36516 + }, + { + "epoch": 0.36517, + "grad_norm": 0.7326317583288867, + "learning_rate": 0.003, + "loss": 3.983, + "step": 36517 + }, + { + "epoch": 0.36518, + "grad_norm": 0.8889462531838546, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 36518 + }, + { + "epoch": 0.36519, + "grad_norm": 1.0635765306792058, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 36519 + }, + { + "epoch": 0.3652, + "grad_norm": 0.9104386088291746, + "learning_rate": 0.003, + "loss": 3.984, + "step": 36520 + }, + { + "epoch": 0.36521, + "grad_norm": 0.8568621629868485, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 36521 + }, + { + "epoch": 0.36522, + "grad_norm": 0.8279148062547059, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 36522 + }, + { + "epoch": 0.36523, + "grad_norm": 0.8010409322239815, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 36523 + }, + { + "epoch": 0.36524, + "grad_norm": 0.8704941777087926, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 36524 + }, + { + "epoch": 0.36525, + "grad_norm": 0.8589914641684682, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 36525 + }, + { + "epoch": 0.36526, + "grad_norm": 0.7826732301520795, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 36526 + }, + { + "epoch": 0.36527, + "grad_norm": 0.752759359991943, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 36527 + }, + { + "epoch": 0.36528, + "grad_norm": 0.7821290253943436, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 36528 + }, + { + "epoch": 0.36529, + "grad_norm": 0.756228419902481, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 36529 + }, + { + "epoch": 0.3653, + "grad_norm": 0.8339728816963186, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 36530 + }, + { + "epoch": 0.36531, + "grad_norm": 0.8290465065397985, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 36531 + }, + { + "epoch": 0.36532, + "grad_norm": 0.8401294651310593, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 36532 + }, + { + "epoch": 0.36533, + "grad_norm": 0.7919983184111117, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 36533 + }, + { + "epoch": 0.36534, + "grad_norm": 0.8054676382133769, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 36534 + }, + { + "epoch": 0.36535, + "grad_norm": 0.9126552280978925, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 36535 + }, + { + "epoch": 0.36536, + "grad_norm": 0.8281584171975176, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 36536 + }, + { + "epoch": 0.36537, + "grad_norm": 0.5717045274358314, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 36537 + }, + { + "epoch": 0.36538, + "grad_norm": 0.5808053252423248, + "learning_rate": 0.003, + "loss": 3.968, + "step": 36538 + }, + { + "epoch": 0.36539, + "grad_norm": 0.6065237191239646, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 36539 + }, + { + "epoch": 0.3654, + "grad_norm": 0.6916622371541866, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 36540 + }, + { + "epoch": 0.36541, + "grad_norm": 0.7592324232754798, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 36541 + }, + { + "epoch": 0.36542, + "grad_norm": 0.7882725235441298, + "learning_rate": 0.003, + "loss": 3.969, + "step": 36542 + }, + { + "epoch": 0.36543, + "grad_norm": 0.8601571472012766, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 36543 + }, + { + "epoch": 0.36544, + "grad_norm": 1.0154605324102728, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 36544 + }, + { + "epoch": 0.36545, + "grad_norm": 0.924193259557064, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 36545 + }, + { + "epoch": 0.36546, + "grad_norm": 0.9427492407780395, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 36546 + }, + { + "epoch": 0.36547, + "grad_norm": 0.9360159024768909, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 36547 + }, + { + "epoch": 0.36548, + "grad_norm": 0.9564938972569625, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 36548 + }, + { + "epoch": 0.36549, + "grad_norm": 1.0621771196806504, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 36549 + }, + { + "epoch": 0.3655, + "grad_norm": 1.0956546909752618, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 36550 + }, + { + "epoch": 0.36551, + "grad_norm": 1.069815563038974, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 36551 + }, + { + "epoch": 0.36552, + "grad_norm": 0.9194828970302611, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 36552 + }, + { + "epoch": 0.36553, + "grad_norm": 1.056007779234898, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 36553 + }, + { + "epoch": 0.36554, + "grad_norm": 1.0283004980538513, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 36554 + }, + { + "epoch": 0.36555, + "grad_norm": 0.9050375569374107, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 36555 + }, + { + "epoch": 0.36556, + "grad_norm": 0.8111779947577006, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 36556 + }, + { + "epoch": 0.36557, + "grad_norm": 0.7540851357215896, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 36557 + }, + { + "epoch": 0.36558, + "grad_norm": 0.8864533199531875, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 36558 + }, + { + "epoch": 0.36559, + "grad_norm": 1.0049344875250195, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 36559 + }, + { + "epoch": 0.3656, + "grad_norm": 0.9563817625060002, + "learning_rate": 0.003, + "loss": 4.0275, + "step": 36560 + }, + { + "epoch": 0.36561, + "grad_norm": 0.8378024414286237, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 36561 + }, + { + "epoch": 0.36562, + "grad_norm": 0.7114196372247082, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 36562 + }, + { + "epoch": 0.36563, + "grad_norm": 0.6750364328068509, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 36563 + }, + { + "epoch": 0.36564, + "grad_norm": 0.5772793794237784, + "learning_rate": 0.003, + "loss": 3.985, + "step": 36564 + }, + { + "epoch": 0.36565, + "grad_norm": 0.580624160368338, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 36565 + }, + { + "epoch": 0.36566, + "grad_norm": 0.5833360210568341, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 36566 + }, + { + "epoch": 0.36567, + "grad_norm": 0.569066649085021, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 36567 + }, + { + "epoch": 0.36568, + "grad_norm": 0.6096190057079838, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 36568 + }, + { + "epoch": 0.36569, + "grad_norm": 0.7202288821924665, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 36569 + }, + { + "epoch": 0.3657, + "grad_norm": 0.8242870625829475, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 36570 + }, + { + "epoch": 0.36571, + "grad_norm": 0.8490991527544308, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 36571 + }, + { + "epoch": 0.36572, + "grad_norm": 0.745322052547895, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 36572 + }, + { + "epoch": 0.36573, + "grad_norm": 0.6976641480009595, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 36573 + }, + { + "epoch": 0.36574, + "grad_norm": 0.6731384886388053, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 36574 + }, + { + "epoch": 0.36575, + "grad_norm": 0.5914496652130036, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 36575 + }, + { + "epoch": 0.36576, + "grad_norm": 0.5929432150261653, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 36576 + }, + { + "epoch": 0.36577, + "grad_norm": 0.5803010404936081, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 36577 + }, + { + "epoch": 0.36578, + "grad_norm": 0.5712863075751028, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 36578 + }, + { + "epoch": 0.36579, + "grad_norm": 0.5876361793084276, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 36579 + }, + { + "epoch": 0.3658, + "grad_norm": 0.5583274390186888, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 36580 + }, + { + "epoch": 0.36581, + "grad_norm": 0.5798722072318466, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 36581 + }, + { + "epoch": 0.36582, + "grad_norm": 0.6992060223690082, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 36582 + }, + { + "epoch": 0.36583, + "grad_norm": 0.7407506670843421, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 36583 + }, + { + "epoch": 0.36584, + "grad_norm": 0.8700286805781223, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 36584 + }, + { + "epoch": 0.36585, + "grad_norm": 1.1371587374300438, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 36585 + }, + { + "epoch": 0.36586, + "grad_norm": 0.8523114272615536, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 36586 + }, + { + "epoch": 0.36587, + "grad_norm": 0.7150729373319576, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 36587 + }, + { + "epoch": 0.36588, + "grad_norm": 0.7459360800145945, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 36588 + }, + { + "epoch": 0.36589, + "grad_norm": 0.7282332812795802, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 36589 + }, + { + "epoch": 0.3659, + "grad_norm": 0.6648619965964683, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 36590 + }, + { + "epoch": 0.36591, + "grad_norm": 0.595290101790867, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 36591 + }, + { + "epoch": 0.36592, + "grad_norm": 0.5955634514491756, + "learning_rate": 0.003, + "loss": 3.983, + "step": 36592 + }, + { + "epoch": 0.36593, + "grad_norm": 0.6687724552671622, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 36593 + }, + { + "epoch": 0.36594, + "grad_norm": 0.7796166418662583, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 36594 + }, + { + "epoch": 0.36595, + "grad_norm": 0.755064642999783, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 36595 + }, + { + "epoch": 0.36596, + "grad_norm": 0.8963287187129717, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 36596 + }, + { + "epoch": 0.36597, + "grad_norm": 0.9184367836976441, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 36597 + }, + { + "epoch": 0.36598, + "grad_norm": 0.9313132353755758, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 36598 + }, + { + "epoch": 0.36599, + "grad_norm": 0.8799923178580925, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 36599 + }, + { + "epoch": 0.366, + "grad_norm": 0.9359780337869117, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 36600 + }, + { + "epoch": 0.36601, + "grad_norm": 0.8902304603545317, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 36601 + }, + { + "epoch": 0.36602, + "grad_norm": 0.969045308722191, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 36602 + }, + { + "epoch": 0.36603, + "grad_norm": 0.988486063752951, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 36603 + }, + { + "epoch": 0.36604, + "grad_norm": 1.0029344427454199, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 36604 + }, + { + "epoch": 0.36605, + "grad_norm": 0.9299608441129362, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 36605 + }, + { + "epoch": 0.36606, + "grad_norm": 0.9075924326813823, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 36606 + }, + { + "epoch": 0.36607, + "grad_norm": 0.9929430391358673, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 36607 + }, + { + "epoch": 0.36608, + "grad_norm": 1.0491859952387532, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 36608 + }, + { + "epoch": 0.36609, + "grad_norm": 0.8971086252928457, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 36609 + }, + { + "epoch": 0.3661, + "grad_norm": 0.8517745445119901, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 36610 + }, + { + "epoch": 0.36611, + "grad_norm": 0.8120838684220478, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 36611 + }, + { + "epoch": 0.36612, + "grad_norm": 0.760266677565241, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 36612 + }, + { + "epoch": 0.36613, + "grad_norm": 0.7576092860395672, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 36613 + }, + { + "epoch": 0.36614, + "grad_norm": 0.7296466456330121, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 36614 + }, + { + "epoch": 0.36615, + "grad_norm": 0.74245782556296, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 36615 + }, + { + "epoch": 0.36616, + "grad_norm": 0.7478094209896418, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 36616 + }, + { + "epoch": 0.36617, + "grad_norm": 0.7465488321331035, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 36617 + }, + { + "epoch": 0.36618, + "grad_norm": 0.750025716536755, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 36618 + }, + { + "epoch": 0.36619, + "grad_norm": 0.6389882828235359, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 36619 + }, + { + "epoch": 0.3662, + "grad_norm": 0.6121634590149976, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 36620 + }, + { + "epoch": 0.36621, + "grad_norm": 0.5982710287371437, + "learning_rate": 0.003, + "loss": 3.975, + "step": 36621 + }, + { + "epoch": 0.36622, + "grad_norm": 0.6311251181030891, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 36622 + }, + { + "epoch": 0.36623, + "grad_norm": 0.7140803342114469, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 36623 + }, + { + "epoch": 0.36624, + "grad_norm": 0.770775076946149, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 36624 + }, + { + "epoch": 0.36625, + "grad_norm": 0.8133261875593062, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 36625 + }, + { + "epoch": 0.36626, + "grad_norm": 0.8408349825921693, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 36626 + }, + { + "epoch": 0.36627, + "grad_norm": 0.9082280207652945, + "learning_rate": 0.003, + "loss": 3.988, + "step": 36627 + }, + { + "epoch": 0.36628, + "grad_norm": 0.9784937939154772, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 36628 + }, + { + "epoch": 0.36629, + "grad_norm": 0.8383378789292012, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 36629 + }, + { + "epoch": 0.3663, + "grad_norm": 0.7891903013778868, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 36630 + }, + { + "epoch": 0.36631, + "grad_norm": 0.7613358138888812, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 36631 + }, + { + "epoch": 0.36632, + "grad_norm": 0.7235538050975857, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 36632 + }, + { + "epoch": 0.36633, + "grad_norm": 0.6630571508589391, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 36633 + }, + { + "epoch": 0.36634, + "grad_norm": 0.7545593441390857, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 36634 + }, + { + "epoch": 0.36635, + "grad_norm": 0.7240824442807149, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 36635 + }, + { + "epoch": 0.36636, + "grad_norm": 0.6690036897781736, + "learning_rate": 0.003, + "loss": 3.975, + "step": 36636 + }, + { + "epoch": 0.36637, + "grad_norm": 0.6785413776577067, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 36637 + }, + { + "epoch": 0.36638, + "grad_norm": 0.7928897883678462, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 36638 + }, + { + "epoch": 0.36639, + "grad_norm": 1.01450283085952, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 36639 + }, + { + "epoch": 0.3664, + "grad_norm": 1.0958723242289041, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 36640 + }, + { + "epoch": 0.36641, + "grad_norm": 0.8756922264237154, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 36641 + }, + { + "epoch": 0.36642, + "grad_norm": 0.7308443991530171, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 36642 + }, + { + "epoch": 0.36643, + "grad_norm": 0.6635540940999773, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 36643 + }, + { + "epoch": 0.36644, + "grad_norm": 0.6012675849102996, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 36644 + }, + { + "epoch": 0.36645, + "grad_norm": 0.4728204345195631, + "learning_rate": 0.003, + "loss": 3.99, + "step": 36645 + }, + { + "epoch": 0.36646, + "grad_norm": 0.5061219105371965, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 36646 + }, + { + "epoch": 0.36647, + "grad_norm": 0.5036090251389372, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 36647 + }, + { + "epoch": 0.36648, + "grad_norm": 0.6040095132290975, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 36648 + }, + { + "epoch": 0.36649, + "grad_norm": 0.6480643913069484, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 36649 + }, + { + "epoch": 0.3665, + "grad_norm": 0.7338653819999622, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 36650 + }, + { + "epoch": 0.36651, + "grad_norm": 0.7672456141172134, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 36651 + }, + { + "epoch": 0.36652, + "grad_norm": 0.7964967239817559, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 36652 + }, + { + "epoch": 0.36653, + "grad_norm": 0.9439042726770607, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 36653 + }, + { + "epoch": 0.36654, + "grad_norm": 1.0629933854051423, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 36654 + }, + { + "epoch": 0.36655, + "grad_norm": 0.8838836472618666, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 36655 + }, + { + "epoch": 0.36656, + "grad_norm": 1.0493231230482447, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 36656 + }, + { + "epoch": 0.36657, + "grad_norm": 1.0564194476987965, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 36657 + }, + { + "epoch": 0.36658, + "grad_norm": 0.8571509589580724, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 36658 + }, + { + "epoch": 0.36659, + "grad_norm": 0.8112456191232403, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 36659 + }, + { + "epoch": 0.3666, + "grad_norm": 0.83451961293313, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 36660 + }, + { + "epoch": 0.36661, + "grad_norm": 0.8873826908526802, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 36661 + }, + { + "epoch": 0.36662, + "grad_norm": 0.9010470078847619, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 36662 + }, + { + "epoch": 0.36663, + "grad_norm": 0.7980362269950625, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 36663 + }, + { + "epoch": 0.36664, + "grad_norm": 0.7554554922476345, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 36664 + }, + { + "epoch": 0.36665, + "grad_norm": 0.6730203624043076, + "learning_rate": 0.003, + "loss": 3.991, + "step": 36665 + }, + { + "epoch": 0.36666, + "grad_norm": 0.6647355554798835, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 36666 + }, + { + "epoch": 0.36667, + "grad_norm": 0.687488555347747, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 36667 + }, + { + "epoch": 0.36668, + "grad_norm": 0.6592149923138939, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 36668 + }, + { + "epoch": 0.36669, + "grad_norm": 0.7876354562027733, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 36669 + }, + { + "epoch": 0.3667, + "grad_norm": 0.8064635971602272, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 36670 + }, + { + "epoch": 0.36671, + "grad_norm": 0.7296160942739872, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 36671 + }, + { + "epoch": 0.36672, + "grad_norm": 0.6910493760518371, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 36672 + }, + { + "epoch": 0.36673, + "grad_norm": 0.7092029080442195, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 36673 + }, + { + "epoch": 0.36674, + "grad_norm": 0.6352768190024655, + "learning_rate": 0.003, + "loss": 3.987, + "step": 36674 + }, + { + "epoch": 0.36675, + "grad_norm": 0.5124794357149711, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 36675 + }, + { + "epoch": 0.36676, + "grad_norm": 0.48730073793101586, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 36676 + }, + { + "epoch": 0.36677, + "grad_norm": 0.5284623267342856, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 36677 + }, + { + "epoch": 0.36678, + "grad_norm": 0.5809823925869414, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 36678 + }, + { + "epoch": 0.36679, + "grad_norm": 0.743104912554208, + "learning_rate": 0.003, + "loss": 4.0, + "step": 36679 + }, + { + "epoch": 0.3668, + "grad_norm": 0.890161967454526, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 36680 + }, + { + "epoch": 0.36681, + "grad_norm": 0.8950283815787294, + "learning_rate": 0.003, + "loss": 3.956, + "step": 36681 + }, + { + "epoch": 0.36682, + "grad_norm": 0.8822984475949327, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 36682 + }, + { + "epoch": 0.36683, + "grad_norm": 1.0650035293768012, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 36683 + }, + { + "epoch": 0.36684, + "grad_norm": 0.9130383208501259, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 36684 + }, + { + "epoch": 0.36685, + "grad_norm": 0.813490295952091, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 36685 + }, + { + "epoch": 0.36686, + "grad_norm": 0.7449217302110597, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 36686 + }, + { + "epoch": 0.36687, + "grad_norm": 0.8357639142818909, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 36687 + }, + { + "epoch": 0.36688, + "grad_norm": 0.8321301251351992, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 36688 + }, + { + "epoch": 0.36689, + "grad_norm": 0.649511657248995, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 36689 + }, + { + "epoch": 0.3669, + "grad_norm": 0.6902630255457113, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 36690 + }, + { + "epoch": 0.36691, + "grad_norm": 0.8201015692198961, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 36691 + }, + { + "epoch": 0.36692, + "grad_norm": 0.9492473518370697, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 36692 + }, + { + "epoch": 0.36693, + "grad_norm": 0.8657519456325855, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 36693 + }, + { + "epoch": 0.36694, + "grad_norm": 0.7374171471675195, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 36694 + }, + { + "epoch": 0.36695, + "grad_norm": 0.7314465320700528, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 36695 + }, + { + "epoch": 0.36696, + "grad_norm": 0.8054952932552014, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 36696 + }, + { + "epoch": 0.36697, + "grad_norm": 0.8547170621368034, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 36697 + }, + { + "epoch": 0.36698, + "grad_norm": 0.8497228014760826, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 36698 + }, + { + "epoch": 0.36699, + "grad_norm": 0.8826155656113099, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 36699 + }, + { + "epoch": 0.367, + "grad_norm": 0.9713356146294647, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 36700 + }, + { + "epoch": 0.36701, + "grad_norm": 1.1565177872335002, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 36701 + }, + { + "epoch": 0.36702, + "grad_norm": 0.9077710633315056, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 36702 + }, + { + "epoch": 0.36703, + "grad_norm": 0.9330165217558347, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 36703 + }, + { + "epoch": 0.36704, + "grad_norm": 0.8735943096837244, + "learning_rate": 0.003, + "loss": 3.983, + "step": 36704 + }, + { + "epoch": 0.36705, + "grad_norm": 0.9018500351057833, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 36705 + }, + { + "epoch": 0.36706, + "grad_norm": 0.9387447553386123, + "learning_rate": 0.003, + "loss": 4.017, + "step": 36706 + }, + { + "epoch": 0.36707, + "grad_norm": 0.8588841863079374, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 36707 + }, + { + "epoch": 0.36708, + "grad_norm": 0.9982411395338049, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 36708 + }, + { + "epoch": 0.36709, + "grad_norm": 1.0080035544932517, + "learning_rate": 0.003, + "loss": 4.016, + "step": 36709 + }, + { + "epoch": 0.3671, + "grad_norm": 0.97488670150338, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 36710 + }, + { + "epoch": 0.36711, + "grad_norm": 0.9565547755006996, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 36711 + }, + { + "epoch": 0.36712, + "grad_norm": 1.0798256316936827, + "learning_rate": 0.003, + "loss": 4.0493, + "step": 36712 + }, + { + "epoch": 0.36713, + "grad_norm": 0.8934202190960626, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 36713 + }, + { + "epoch": 0.36714, + "grad_norm": 0.9231644561739469, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 36714 + }, + { + "epoch": 0.36715, + "grad_norm": 0.8448814530945316, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 36715 + }, + { + "epoch": 0.36716, + "grad_norm": 0.7908164987290391, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 36716 + }, + { + "epoch": 0.36717, + "grad_norm": 0.7578114771834652, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 36717 + }, + { + "epoch": 0.36718, + "grad_norm": 0.7019333372034835, + "learning_rate": 0.003, + "loss": 3.976, + "step": 36718 + }, + { + "epoch": 0.36719, + "grad_norm": 0.5878680296267951, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 36719 + }, + { + "epoch": 0.3672, + "grad_norm": 0.6034555144372111, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 36720 + }, + { + "epoch": 0.36721, + "grad_norm": 0.5651717040072418, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 36721 + }, + { + "epoch": 0.36722, + "grad_norm": 0.48821159090674565, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 36722 + }, + { + "epoch": 0.36723, + "grad_norm": 0.5241644576509481, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 36723 + }, + { + "epoch": 0.36724, + "grad_norm": 0.4643848476195242, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 36724 + }, + { + "epoch": 0.36725, + "grad_norm": 0.5224332369848534, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 36725 + }, + { + "epoch": 0.36726, + "grad_norm": 0.6570265976472003, + "learning_rate": 0.003, + "loss": 3.981, + "step": 36726 + }, + { + "epoch": 0.36727, + "grad_norm": 0.8952215737020999, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 36727 + }, + { + "epoch": 0.36728, + "grad_norm": 1.165872165385745, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 36728 + }, + { + "epoch": 0.36729, + "grad_norm": 0.7136519452201425, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 36729 + }, + { + "epoch": 0.3673, + "grad_norm": 0.5132310778118832, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 36730 + }, + { + "epoch": 0.36731, + "grad_norm": 0.6968356205806551, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 36731 + }, + { + "epoch": 0.36732, + "grad_norm": 0.7740247766068676, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 36732 + }, + { + "epoch": 0.36733, + "grad_norm": 0.758770302315536, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 36733 + }, + { + "epoch": 0.36734, + "grad_norm": 0.7303595880184756, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 36734 + }, + { + "epoch": 0.36735, + "grad_norm": 0.7565974929464548, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 36735 + }, + { + "epoch": 0.36736, + "grad_norm": 0.727419912396365, + "learning_rate": 0.003, + "loss": 4.009, + "step": 36736 + }, + { + "epoch": 0.36737, + "grad_norm": 0.6838570555246697, + "learning_rate": 0.003, + "loss": 3.975, + "step": 36737 + }, + { + "epoch": 0.36738, + "grad_norm": 0.6633362677955154, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 36738 + }, + { + "epoch": 0.36739, + "grad_norm": 0.6066016118214885, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 36739 + }, + { + "epoch": 0.3674, + "grad_norm": 0.6427146022048235, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 36740 + }, + { + "epoch": 0.36741, + "grad_norm": 0.5468530294196627, + "learning_rate": 0.003, + "loss": 3.988, + "step": 36741 + }, + { + "epoch": 0.36742, + "grad_norm": 0.5328898241980566, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 36742 + }, + { + "epoch": 0.36743, + "grad_norm": 0.512958244359255, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 36743 + }, + { + "epoch": 0.36744, + "grad_norm": 0.5000268419310534, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 36744 + }, + { + "epoch": 0.36745, + "grad_norm": 0.6346907767043465, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 36745 + }, + { + "epoch": 0.36746, + "grad_norm": 0.7697324084857694, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 36746 + }, + { + "epoch": 0.36747, + "grad_norm": 0.803569741698267, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 36747 + }, + { + "epoch": 0.36748, + "grad_norm": 0.7730905574806664, + "learning_rate": 0.003, + "loss": 3.9336, + "step": 36748 + }, + { + "epoch": 0.36749, + "grad_norm": 0.806058429169171, + "learning_rate": 0.003, + "loss": 3.984, + "step": 36749 + }, + { + "epoch": 0.3675, + "grad_norm": 0.8857687811846118, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 36750 + }, + { + "epoch": 0.36751, + "grad_norm": 0.88460335592034, + "learning_rate": 0.003, + "loss": 3.982, + "step": 36751 + }, + { + "epoch": 0.36752, + "grad_norm": 0.8318803167742935, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 36752 + }, + { + "epoch": 0.36753, + "grad_norm": 0.7420070785701572, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 36753 + }, + { + "epoch": 0.36754, + "grad_norm": 0.7183185675436161, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 36754 + }, + { + "epoch": 0.36755, + "grad_norm": 0.7865387430395729, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 36755 + }, + { + "epoch": 0.36756, + "grad_norm": 0.8583440445540913, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 36756 + }, + { + "epoch": 0.36757, + "grad_norm": 0.9321603847836587, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 36757 + }, + { + "epoch": 0.36758, + "grad_norm": 0.9505920332604899, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 36758 + }, + { + "epoch": 0.36759, + "grad_norm": 0.9540813295855194, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 36759 + }, + { + "epoch": 0.3676, + "grad_norm": 0.8042661771793124, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 36760 + }, + { + "epoch": 0.36761, + "grad_norm": 0.7500818527889741, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36761 + }, + { + "epoch": 0.36762, + "grad_norm": 0.7297542190338021, + "learning_rate": 0.003, + "loss": 3.978, + "step": 36762 + }, + { + "epoch": 0.36763, + "grad_norm": 0.700280831123055, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 36763 + }, + { + "epoch": 0.36764, + "grad_norm": 0.7515712701788532, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 36764 + }, + { + "epoch": 0.36765, + "grad_norm": 0.7048225675735083, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 36765 + }, + { + "epoch": 0.36766, + "grad_norm": 0.6994293737538612, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 36766 + }, + { + "epoch": 0.36767, + "grad_norm": 0.7444349413868009, + "learning_rate": 0.003, + "loss": 3.994, + "step": 36767 + }, + { + "epoch": 0.36768, + "grad_norm": 0.7320321166529749, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 36768 + }, + { + "epoch": 0.36769, + "grad_norm": 0.7084109838801546, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 36769 + }, + { + "epoch": 0.3677, + "grad_norm": 0.7042467480335615, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 36770 + }, + { + "epoch": 0.36771, + "grad_norm": 0.7034333124290487, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 36771 + }, + { + "epoch": 0.36772, + "grad_norm": 0.6186720931466059, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 36772 + }, + { + "epoch": 0.36773, + "grad_norm": 0.6574789708969766, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 36773 + }, + { + "epoch": 0.36774, + "grad_norm": 0.6616363542997532, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 36774 + }, + { + "epoch": 0.36775, + "grad_norm": 0.8151270770255226, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 36775 + }, + { + "epoch": 0.36776, + "grad_norm": 1.0836860658464669, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 36776 + }, + { + "epoch": 0.36777, + "grad_norm": 1.0458082395913735, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 36777 + }, + { + "epoch": 0.36778, + "grad_norm": 0.9597307059889016, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 36778 + }, + { + "epoch": 0.36779, + "grad_norm": 0.8341841297546259, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 36779 + }, + { + "epoch": 0.3678, + "grad_norm": 0.930327215007682, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 36780 + }, + { + "epoch": 0.36781, + "grad_norm": 0.9175820364143762, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 36781 + }, + { + "epoch": 0.36782, + "grad_norm": 0.9806900909289787, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 36782 + }, + { + "epoch": 0.36783, + "grad_norm": 0.9009323143339094, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 36783 + }, + { + "epoch": 0.36784, + "grad_norm": 0.7953445844581649, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 36784 + }, + { + "epoch": 0.36785, + "grad_norm": 0.835508063094093, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 36785 + }, + { + "epoch": 0.36786, + "grad_norm": 0.8954350195620854, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 36786 + }, + { + "epoch": 0.36787, + "grad_norm": 0.8384699618421654, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 36787 + }, + { + "epoch": 0.36788, + "grad_norm": 0.9685653526382091, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 36788 + }, + { + "epoch": 0.36789, + "grad_norm": 0.8589749793326066, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 36789 + }, + { + "epoch": 0.3679, + "grad_norm": 0.7352537832394329, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 36790 + }, + { + "epoch": 0.36791, + "grad_norm": 0.7315975344018022, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 36791 + }, + { + "epoch": 0.36792, + "grad_norm": 0.7280053023564182, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 36792 + }, + { + "epoch": 0.36793, + "grad_norm": 0.6877596948678619, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 36793 + }, + { + "epoch": 0.36794, + "grad_norm": 0.8044244080360451, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 36794 + }, + { + "epoch": 0.36795, + "grad_norm": 0.8862061806807291, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 36795 + }, + { + "epoch": 0.36796, + "grad_norm": 0.9178621911873787, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 36796 + }, + { + "epoch": 0.36797, + "grad_norm": 0.9439025709504001, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 36797 + }, + { + "epoch": 0.36798, + "grad_norm": 0.876479648238965, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 36798 + }, + { + "epoch": 0.36799, + "grad_norm": 0.8043055425769262, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 36799 + }, + { + "epoch": 0.368, + "grad_norm": 0.6478456821012558, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 36800 + }, + { + "epoch": 0.36801, + "grad_norm": 0.5833862154868115, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 36801 + }, + { + "epoch": 0.36802, + "grad_norm": 0.6264415616191165, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 36802 + }, + { + "epoch": 0.36803, + "grad_norm": 0.6304963333779449, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 36803 + }, + { + "epoch": 0.36804, + "grad_norm": 0.6513889970876493, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 36804 + }, + { + "epoch": 0.36805, + "grad_norm": 0.5805310114702331, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 36805 + }, + { + "epoch": 0.36806, + "grad_norm": 0.6409691815588783, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 36806 + }, + { + "epoch": 0.36807, + "grad_norm": 0.6990710100185158, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 36807 + }, + { + "epoch": 0.36808, + "grad_norm": 0.8082948621258396, + "learning_rate": 0.003, + "loss": 3.978, + "step": 36808 + }, + { + "epoch": 0.36809, + "grad_norm": 0.8037630133472067, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 36809 + }, + { + "epoch": 0.3681, + "grad_norm": 0.7043028441572212, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 36810 + }, + { + "epoch": 0.36811, + "grad_norm": 0.6622664845829583, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 36811 + }, + { + "epoch": 0.36812, + "grad_norm": 0.6794831880229524, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 36812 + }, + { + "epoch": 0.36813, + "grad_norm": 0.6913493016850376, + "learning_rate": 0.003, + "loss": 3.961, + "step": 36813 + }, + { + "epoch": 0.36814, + "grad_norm": 0.7880748015635188, + "learning_rate": 0.003, + "loss": 3.9302, + "step": 36814 + }, + { + "epoch": 0.36815, + "grad_norm": 0.8664682077050049, + "learning_rate": 0.003, + "loss": 3.995, + "step": 36815 + }, + { + "epoch": 0.36816, + "grad_norm": 1.0588295989480563, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 36816 + }, + { + "epoch": 0.36817, + "grad_norm": 0.993943267143678, + "learning_rate": 0.003, + "loss": 4.0, + "step": 36817 + }, + { + "epoch": 0.36818, + "grad_norm": 1.0259050268600591, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 36818 + }, + { + "epoch": 0.36819, + "grad_norm": 0.852987889511621, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 36819 + }, + { + "epoch": 0.3682, + "grad_norm": 0.6840558448786802, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 36820 + }, + { + "epoch": 0.36821, + "grad_norm": 0.6928685425902865, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 36821 + }, + { + "epoch": 0.36822, + "grad_norm": 0.7420301459881085, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 36822 + }, + { + "epoch": 0.36823, + "grad_norm": 0.7394452957633791, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 36823 + }, + { + "epoch": 0.36824, + "grad_norm": 0.728041679126796, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 36824 + }, + { + "epoch": 0.36825, + "grad_norm": 0.8284058528528253, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 36825 + }, + { + "epoch": 0.36826, + "grad_norm": 0.8672040487311543, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 36826 + }, + { + "epoch": 0.36827, + "grad_norm": 0.9548721565753933, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 36827 + }, + { + "epoch": 0.36828, + "grad_norm": 1.0434189762823995, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 36828 + }, + { + "epoch": 0.36829, + "grad_norm": 0.9982518588747656, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 36829 + }, + { + "epoch": 0.3683, + "grad_norm": 1.1248944663781815, + "learning_rate": 0.003, + "loss": 4.003, + "step": 36830 + }, + { + "epoch": 0.36831, + "grad_norm": 0.9482136129396177, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 36831 + }, + { + "epoch": 0.36832, + "grad_norm": 0.9323344206302258, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 36832 + }, + { + "epoch": 0.36833, + "grad_norm": 0.9026755496719738, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 36833 + }, + { + "epoch": 0.36834, + "grad_norm": 0.8962223451122073, + "learning_rate": 0.003, + "loss": 4.014, + "step": 36834 + }, + { + "epoch": 0.36835, + "grad_norm": 0.8847466976077204, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 36835 + }, + { + "epoch": 0.36836, + "grad_norm": 0.9159646902333552, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 36836 + }, + { + "epoch": 0.36837, + "grad_norm": 0.8491242298110344, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 36837 + }, + { + "epoch": 0.36838, + "grad_norm": 0.7405390981979904, + "learning_rate": 0.003, + "loss": 3.99, + "step": 36838 + }, + { + "epoch": 0.36839, + "grad_norm": 0.7059866700232902, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 36839 + }, + { + "epoch": 0.3684, + "grad_norm": 0.6666789891045962, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 36840 + }, + { + "epoch": 0.36841, + "grad_norm": 0.5473876954046222, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 36841 + }, + { + "epoch": 0.36842, + "grad_norm": 0.5197749966731725, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 36842 + }, + { + "epoch": 0.36843, + "grad_norm": 0.5154050076063562, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 36843 + }, + { + "epoch": 0.36844, + "grad_norm": 0.5528584322497028, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 36844 + }, + { + "epoch": 0.36845, + "grad_norm": 0.6046060717452796, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 36845 + }, + { + "epoch": 0.36846, + "grad_norm": 0.7073127800638201, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 36846 + }, + { + "epoch": 0.36847, + "grad_norm": 0.6582298529516827, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 36847 + }, + { + "epoch": 0.36848, + "grad_norm": 0.5476303648500351, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 36848 + }, + { + "epoch": 0.36849, + "grad_norm": 0.6528243469527998, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 36849 + }, + { + "epoch": 0.3685, + "grad_norm": 0.6207788851593611, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 36850 + }, + { + "epoch": 0.36851, + "grad_norm": 0.6784995818252155, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 36851 + }, + { + "epoch": 0.36852, + "grad_norm": 0.6735973083287318, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 36852 + }, + { + "epoch": 0.36853, + "grad_norm": 0.651195843250067, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 36853 + }, + { + "epoch": 0.36854, + "grad_norm": 0.7405796549271432, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 36854 + }, + { + "epoch": 0.36855, + "grad_norm": 0.7904051747597831, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 36855 + }, + { + "epoch": 0.36856, + "grad_norm": 0.8084181814653927, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 36856 + }, + { + "epoch": 0.36857, + "grad_norm": 0.7986147101614982, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 36857 + }, + { + "epoch": 0.36858, + "grad_norm": 0.9174398177630305, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 36858 + }, + { + "epoch": 0.36859, + "grad_norm": 1.2017540823953239, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 36859 + }, + { + "epoch": 0.3686, + "grad_norm": 0.8433315224476371, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 36860 + }, + { + "epoch": 0.36861, + "grad_norm": 0.7936445568071417, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 36861 + }, + { + "epoch": 0.36862, + "grad_norm": 0.8300084027332935, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 36862 + }, + { + "epoch": 0.36863, + "grad_norm": 0.8109196938670066, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 36863 + }, + { + "epoch": 0.36864, + "grad_norm": 0.6874967895950934, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 36864 + }, + { + "epoch": 0.36865, + "grad_norm": 0.6627902957736722, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 36865 + }, + { + "epoch": 0.36866, + "grad_norm": 0.7635471668117677, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 36866 + }, + { + "epoch": 0.36867, + "grad_norm": 0.7987099760304872, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 36867 + }, + { + "epoch": 0.36868, + "grad_norm": 0.7766784789005824, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 36868 + }, + { + "epoch": 0.36869, + "grad_norm": 0.9365308424156874, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 36869 + }, + { + "epoch": 0.3687, + "grad_norm": 1.040526356687193, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 36870 + }, + { + "epoch": 0.36871, + "grad_norm": 0.859175619559244, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 36871 + }, + { + "epoch": 0.36872, + "grad_norm": 0.7652687943389839, + "learning_rate": 0.003, + "loss": 3.938, + "step": 36872 + }, + { + "epoch": 0.36873, + "grad_norm": 0.8104197888838913, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 36873 + }, + { + "epoch": 0.36874, + "grad_norm": 0.8724119971935161, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 36874 + }, + { + "epoch": 0.36875, + "grad_norm": 0.8796739191296173, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 36875 + }, + { + "epoch": 0.36876, + "grad_norm": 0.725057085587044, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 36876 + }, + { + "epoch": 0.36877, + "grad_norm": 0.7951912784401938, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 36877 + }, + { + "epoch": 0.36878, + "grad_norm": 0.8002579853459023, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 36878 + }, + { + "epoch": 0.36879, + "grad_norm": 0.9519893955500935, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 36879 + }, + { + "epoch": 0.3688, + "grad_norm": 1.0157335714361393, + "learning_rate": 0.003, + "loss": 4.007, + "step": 36880 + }, + { + "epoch": 0.36881, + "grad_norm": 1.072223250451355, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 36881 + }, + { + "epoch": 0.36882, + "grad_norm": 1.0935861147861905, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 36882 + }, + { + "epoch": 0.36883, + "grad_norm": 1.0135158213152158, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 36883 + }, + { + "epoch": 0.36884, + "grad_norm": 0.93627420568201, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 36884 + }, + { + "epoch": 0.36885, + "grad_norm": 0.8370816826143768, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 36885 + }, + { + "epoch": 0.36886, + "grad_norm": 0.7432129724750612, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 36886 + }, + { + "epoch": 0.36887, + "grad_norm": 0.6981789379292613, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 36887 + }, + { + "epoch": 0.36888, + "grad_norm": 0.6973623298321245, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 36888 + }, + { + "epoch": 0.36889, + "grad_norm": 0.70122048745508, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 36889 + }, + { + "epoch": 0.3689, + "grad_norm": 0.7489332811327148, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 36890 + }, + { + "epoch": 0.36891, + "grad_norm": 0.7236174479568431, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 36891 + }, + { + "epoch": 0.36892, + "grad_norm": 0.7605564789121838, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 36892 + }, + { + "epoch": 0.36893, + "grad_norm": 0.7792455157984755, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 36893 + }, + { + "epoch": 0.36894, + "grad_norm": 0.7458130611546974, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 36894 + }, + { + "epoch": 0.36895, + "grad_norm": 0.6854374230057354, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 36895 + }, + { + "epoch": 0.36896, + "grad_norm": 0.5561340720062599, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 36896 + }, + { + "epoch": 0.36897, + "grad_norm": 0.6238849438098152, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 36897 + }, + { + "epoch": 0.36898, + "grad_norm": 0.6683451072238483, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 36898 + }, + { + "epoch": 0.36899, + "grad_norm": 0.6453931364723493, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 36899 + }, + { + "epoch": 0.369, + "grad_norm": 0.6428518131015422, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 36900 + }, + { + "epoch": 0.36901, + "grad_norm": 0.6281485747221129, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 36901 + }, + { + "epoch": 0.36902, + "grad_norm": 0.728495687877359, + "learning_rate": 0.003, + "loss": 3.959, + "step": 36902 + }, + { + "epoch": 0.36903, + "grad_norm": 0.6809800112615426, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 36903 + }, + { + "epoch": 0.36904, + "grad_norm": 0.6186306718065144, + "learning_rate": 0.003, + "loss": 3.964, + "step": 36904 + }, + { + "epoch": 0.36905, + "grad_norm": 0.6993521691740291, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 36905 + }, + { + "epoch": 0.36906, + "grad_norm": 0.8455826709927962, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 36906 + }, + { + "epoch": 0.36907, + "grad_norm": 0.9838519835890862, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 36907 + }, + { + "epoch": 0.36908, + "grad_norm": 1.0719786716657789, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 36908 + }, + { + "epoch": 0.36909, + "grad_norm": 0.775630886906731, + "learning_rate": 0.003, + "loss": 3.957, + "step": 36909 + }, + { + "epoch": 0.3691, + "grad_norm": 0.7761594035068623, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 36910 + }, + { + "epoch": 0.36911, + "grad_norm": 0.848066111509584, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 36911 + }, + { + "epoch": 0.36912, + "grad_norm": 0.9249452209662727, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 36912 + }, + { + "epoch": 0.36913, + "grad_norm": 0.9104835844773724, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 36913 + }, + { + "epoch": 0.36914, + "grad_norm": 0.7537042679640136, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 36914 + }, + { + "epoch": 0.36915, + "grad_norm": 0.6377679298549084, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 36915 + }, + { + "epoch": 0.36916, + "grad_norm": 0.5896385070284279, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 36916 + }, + { + "epoch": 0.36917, + "grad_norm": 0.6413565532489058, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 36917 + }, + { + "epoch": 0.36918, + "grad_norm": 0.6288484697992883, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 36918 + }, + { + "epoch": 0.36919, + "grad_norm": 0.6310033917769823, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 36919 + }, + { + "epoch": 0.3692, + "grad_norm": 0.6317417347016704, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 36920 + }, + { + "epoch": 0.36921, + "grad_norm": 0.6767298702269952, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 36921 + }, + { + "epoch": 0.36922, + "grad_norm": 0.6556424037562636, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 36922 + }, + { + "epoch": 0.36923, + "grad_norm": 0.7002957874610136, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 36923 + }, + { + "epoch": 0.36924, + "grad_norm": 0.7229191712717631, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 36924 + }, + { + "epoch": 0.36925, + "grad_norm": 0.6451809155058355, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 36925 + }, + { + "epoch": 0.36926, + "grad_norm": 0.5683852889270148, + "learning_rate": 0.003, + "loss": 4.006, + "step": 36926 + }, + { + "epoch": 0.36927, + "grad_norm": 0.5755777111942447, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 36927 + }, + { + "epoch": 0.36928, + "grad_norm": 0.7171235101142572, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 36928 + }, + { + "epoch": 0.36929, + "grad_norm": 0.8984138171542475, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 36929 + }, + { + "epoch": 0.3693, + "grad_norm": 0.9906787097250834, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 36930 + }, + { + "epoch": 0.36931, + "grad_norm": 0.8980321325069063, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 36931 + }, + { + "epoch": 0.36932, + "grad_norm": 0.9154549397677498, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 36932 + }, + { + "epoch": 0.36933, + "grad_norm": 0.9373097724321264, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 36933 + }, + { + "epoch": 0.36934, + "grad_norm": 0.9632883282587453, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 36934 + }, + { + "epoch": 0.36935, + "grad_norm": 0.9921720075427997, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 36935 + }, + { + "epoch": 0.36936, + "grad_norm": 1.0553265562487317, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 36936 + }, + { + "epoch": 0.36937, + "grad_norm": 1.042146029038658, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 36937 + }, + { + "epoch": 0.36938, + "grad_norm": 0.7983986957006199, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 36938 + }, + { + "epoch": 0.36939, + "grad_norm": 0.7332030228205679, + "learning_rate": 0.003, + "loss": 4.007, + "step": 36939 + }, + { + "epoch": 0.3694, + "grad_norm": 0.6973605048390037, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 36940 + }, + { + "epoch": 0.36941, + "grad_norm": 0.780709062960757, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 36941 + }, + { + "epoch": 0.36942, + "grad_norm": 0.8739200585885918, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 36942 + }, + { + "epoch": 0.36943, + "grad_norm": 0.9847559525732065, + "learning_rate": 0.003, + "loss": 4.013, + "step": 36943 + }, + { + "epoch": 0.36944, + "grad_norm": 0.9027253436641521, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 36944 + }, + { + "epoch": 0.36945, + "grad_norm": 0.864307005480972, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 36945 + }, + { + "epoch": 0.36946, + "grad_norm": 0.8777833027361881, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 36946 + }, + { + "epoch": 0.36947, + "grad_norm": 0.9969908404845462, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 36947 + }, + { + "epoch": 0.36948, + "grad_norm": 0.9858844423524106, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 36948 + }, + { + "epoch": 0.36949, + "grad_norm": 0.9052412924955523, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 36949 + }, + { + "epoch": 0.3695, + "grad_norm": 1.0248513497198597, + "learning_rate": 0.003, + "loss": 4.026, + "step": 36950 + }, + { + "epoch": 0.36951, + "grad_norm": 1.049712240560135, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 36951 + }, + { + "epoch": 0.36952, + "grad_norm": 0.9345595086055839, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 36952 + }, + { + "epoch": 0.36953, + "grad_norm": 0.7860200394237998, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 36953 + }, + { + "epoch": 0.36954, + "grad_norm": 0.7767636311140087, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 36954 + }, + { + "epoch": 0.36955, + "grad_norm": 0.8328683288831727, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 36955 + }, + { + "epoch": 0.36956, + "grad_norm": 0.7745529866841242, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 36956 + }, + { + "epoch": 0.36957, + "grad_norm": 0.835818307038926, + "learning_rate": 0.003, + "loss": 4.035, + "step": 36957 + }, + { + "epoch": 0.36958, + "grad_norm": 0.8891049168153082, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 36958 + }, + { + "epoch": 0.36959, + "grad_norm": 0.907854712105767, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 36959 + }, + { + "epoch": 0.3696, + "grad_norm": 0.8220336138123563, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 36960 + }, + { + "epoch": 0.36961, + "grad_norm": 0.8118611587162053, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 36961 + }, + { + "epoch": 0.36962, + "grad_norm": 0.7748669906595038, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 36962 + }, + { + "epoch": 0.36963, + "grad_norm": 0.8313602597009305, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 36963 + }, + { + "epoch": 0.36964, + "grad_norm": 0.8094472783482518, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 36964 + }, + { + "epoch": 0.36965, + "grad_norm": 0.7094728038776272, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 36965 + }, + { + "epoch": 0.36966, + "grad_norm": 0.5925839083311275, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 36966 + }, + { + "epoch": 0.36967, + "grad_norm": 0.573582128309732, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 36967 + }, + { + "epoch": 0.36968, + "grad_norm": 0.6786146338226023, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 36968 + }, + { + "epoch": 0.36969, + "grad_norm": 0.6853384027535945, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 36969 + }, + { + "epoch": 0.3697, + "grad_norm": 0.7158397675799661, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 36970 + }, + { + "epoch": 0.36971, + "grad_norm": 0.5816206706319269, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 36971 + }, + { + "epoch": 0.36972, + "grad_norm": 0.5343566547859242, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 36972 + }, + { + "epoch": 0.36973, + "grad_norm": 0.6051295598144106, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 36973 + }, + { + "epoch": 0.36974, + "grad_norm": 0.6700470078060317, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 36974 + }, + { + "epoch": 0.36975, + "grad_norm": 0.6813168884601492, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 36975 + }, + { + "epoch": 0.36976, + "grad_norm": 0.5827085519342126, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 36976 + }, + { + "epoch": 0.36977, + "grad_norm": 0.5900355872850798, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 36977 + }, + { + "epoch": 0.36978, + "grad_norm": 0.5626341327289569, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 36978 + }, + { + "epoch": 0.36979, + "grad_norm": 0.6049755774156844, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 36979 + }, + { + "epoch": 0.3698, + "grad_norm": 0.7255387450775939, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 36980 + }, + { + "epoch": 0.36981, + "grad_norm": 0.9531177708687024, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 36981 + }, + { + "epoch": 0.36982, + "grad_norm": 0.9951709244742789, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 36982 + }, + { + "epoch": 0.36983, + "grad_norm": 0.9721265147735424, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 36983 + }, + { + "epoch": 0.36984, + "grad_norm": 0.9435246955293533, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 36984 + }, + { + "epoch": 0.36985, + "grad_norm": 0.8243096424563624, + "learning_rate": 0.003, + "loss": 3.999, + "step": 36985 + }, + { + "epoch": 0.36986, + "grad_norm": 0.7376177270973142, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 36986 + }, + { + "epoch": 0.36987, + "grad_norm": 0.7718312358229713, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 36987 + }, + { + "epoch": 0.36988, + "grad_norm": 0.9446686286813485, + "learning_rate": 0.003, + "loss": 3.988, + "step": 36988 + }, + { + "epoch": 0.36989, + "grad_norm": 1.009640509380454, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 36989 + }, + { + "epoch": 0.3699, + "grad_norm": 0.8944236288545628, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 36990 + }, + { + "epoch": 0.36991, + "grad_norm": 0.8361413851686201, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 36991 + }, + { + "epoch": 0.36992, + "grad_norm": 0.8947330079739885, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 36992 + }, + { + "epoch": 0.36993, + "grad_norm": 0.8870598285688848, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 36993 + }, + { + "epoch": 0.36994, + "grad_norm": 0.8133186780934007, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 36994 + }, + { + "epoch": 0.36995, + "grad_norm": 0.9150847374049611, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 36995 + }, + { + "epoch": 0.36996, + "grad_norm": 0.8694836454885212, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 36996 + }, + { + "epoch": 0.36997, + "grad_norm": 0.770481810332417, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 36997 + }, + { + "epoch": 0.36998, + "grad_norm": 0.6413527669338748, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 36998 + }, + { + "epoch": 0.36999, + "grad_norm": 0.5970089696897923, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 36999 + }, + { + "epoch": 0.37, + "grad_norm": 0.6046513413489216, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 37000 + }, + { + "epoch": 0.37001, + "grad_norm": 0.5871087176139108, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 37001 + }, + { + "epoch": 0.37002, + "grad_norm": 0.5418606389108463, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 37002 + }, + { + "epoch": 0.37003, + "grad_norm": 0.5239013979650224, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 37003 + }, + { + "epoch": 0.37004, + "grad_norm": 0.6645425324234486, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 37004 + }, + { + "epoch": 0.37005, + "grad_norm": 0.81576904508461, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 37005 + }, + { + "epoch": 0.37006, + "grad_norm": 0.9470779231172817, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 37006 + }, + { + "epoch": 0.37007, + "grad_norm": 0.9484681468513808, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 37007 + }, + { + "epoch": 0.37008, + "grad_norm": 0.8799586258022284, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 37008 + }, + { + "epoch": 0.37009, + "grad_norm": 0.8654282767124146, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 37009 + }, + { + "epoch": 0.3701, + "grad_norm": 0.8652839826355533, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 37010 + }, + { + "epoch": 0.37011, + "grad_norm": 0.8463599141080453, + "learning_rate": 0.003, + "loss": 3.972, + "step": 37011 + }, + { + "epoch": 0.37012, + "grad_norm": 0.7980199488711711, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 37012 + }, + { + "epoch": 0.37013, + "grad_norm": 0.7158542801693939, + "learning_rate": 0.003, + "loss": 3.989, + "step": 37013 + }, + { + "epoch": 0.37014, + "grad_norm": 0.5965018327721767, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 37014 + }, + { + "epoch": 0.37015, + "grad_norm": 0.6390780369783482, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 37015 + }, + { + "epoch": 0.37016, + "grad_norm": 0.5437555371350369, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 37016 + }, + { + "epoch": 0.37017, + "grad_norm": 0.5659930800202299, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 37017 + }, + { + "epoch": 0.37018, + "grad_norm": 0.5772379328332059, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 37018 + }, + { + "epoch": 0.37019, + "grad_norm": 0.6861309408187116, + "learning_rate": 0.003, + "loss": 3.985, + "step": 37019 + }, + { + "epoch": 0.3702, + "grad_norm": 0.6937738544079338, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 37020 + }, + { + "epoch": 0.37021, + "grad_norm": 0.5872068286165487, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 37021 + }, + { + "epoch": 0.37022, + "grad_norm": 0.5651182081288284, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 37022 + }, + { + "epoch": 0.37023, + "grad_norm": 0.5643603291195928, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 37023 + }, + { + "epoch": 0.37024, + "grad_norm": 0.6969416180283732, + "learning_rate": 0.003, + "loss": 3.994, + "step": 37024 + }, + { + "epoch": 0.37025, + "grad_norm": 0.8172091585665674, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 37025 + }, + { + "epoch": 0.37026, + "grad_norm": 0.8920095619873384, + "learning_rate": 0.003, + "loss": 3.96, + "step": 37026 + }, + { + "epoch": 0.37027, + "grad_norm": 0.9425276004521971, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 37027 + }, + { + "epoch": 0.37028, + "grad_norm": 1.0346923611353112, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 37028 + }, + { + "epoch": 0.37029, + "grad_norm": 1.075242009063626, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 37029 + }, + { + "epoch": 0.3703, + "grad_norm": 1.0312121511353278, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 37030 + }, + { + "epoch": 0.37031, + "grad_norm": 0.8818420161213226, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 37031 + }, + { + "epoch": 0.37032, + "grad_norm": 0.9188192402475487, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 37032 + }, + { + "epoch": 0.37033, + "grad_norm": 0.9131204337316766, + "learning_rate": 0.003, + "loss": 4.031, + "step": 37033 + }, + { + "epoch": 0.37034, + "grad_norm": 0.8420728400611438, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 37034 + }, + { + "epoch": 0.37035, + "grad_norm": 0.782006418838829, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 37035 + }, + { + "epoch": 0.37036, + "grad_norm": 0.8612753198780135, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 37036 + }, + { + "epoch": 0.37037, + "grad_norm": 0.9009210740828836, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 37037 + }, + { + "epoch": 0.37038, + "grad_norm": 0.9902838983469436, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 37038 + }, + { + "epoch": 0.37039, + "grad_norm": 1.0755563926001785, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 37039 + }, + { + "epoch": 0.3704, + "grad_norm": 0.8422075612567221, + "learning_rate": 0.003, + "loss": 3.957, + "step": 37040 + }, + { + "epoch": 0.37041, + "grad_norm": 0.7734004092529391, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 37041 + }, + { + "epoch": 0.37042, + "grad_norm": 0.8636882685823631, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 37042 + }, + { + "epoch": 0.37043, + "grad_norm": 0.8269818668371682, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 37043 + }, + { + "epoch": 0.37044, + "grad_norm": 0.830954751584064, + "learning_rate": 0.003, + "loss": 3.993, + "step": 37044 + }, + { + "epoch": 0.37045, + "grad_norm": 0.8684246858525275, + "learning_rate": 0.003, + "loss": 3.996, + "step": 37045 + }, + { + "epoch": 0.37046, + "grad_norm": 0.8226610146930954, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 37046 + }, + { + "epoch": 0.37047, + "grad_norm": 0.7426601898281332, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 37047 + }, + { + "epoch": 0.37048, + "grad_norm": 0.8355036616718894, + "learning_rate": 0.003, + "loss": 4.003, + "step": 37048 + }, + { + "epoch": 0.37049, + "grad_norm": 0.8693309799446544, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 37049 + }, + { + "epoch": 0.3705, + "grad_norm": 0.8843262268651375, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 37050 + }, + { + "epoch": 0.37051, + "grad_norm": 0.8609199809210211, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 37051 + }, + { + "epoch": 0.37052, + "grad_norm": 0.8447174211570523, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 37052 + }, + { + "epoch": 0.37053, + "grad_norm": 0.8173925919485506, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 37053 + }, + { + "epoch": 0.37054, + "grad_norm": 0.8154996905493187, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 37054 + }, + { + "epoch": 0.37055, + "grad_norm": 0.754399202843464, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 37055 + }, + { + "epoch": 0.37056, + "grad_norm": 0.7524090598890231, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 37056 + }, + { + "epoch": 0.37057, + "grad_norm": 0.7980911595090846, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 37057 + }, + { + "epoch": 0.37058, + "grad_norm": 0.8913758437772371, + "learning_rate": 0.003, + "loss": 3.99, + "step": 37058 + }, + { + "epoch": 0.37059, + "grad_norm": 1.0131014823559727, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 37059 + }, + { + "epoch": 0.3706, + "grad_norm": 1.03077737128049, + "learning_rate": 0.003, + "loss": 4.008, + "step": 37060 + }, + { + "epoch": 0.37061, + "grad_norm": 0.9081176118570083, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 37061 + }, + { + "epoch": 0.37062, + "grad_norm": 0.6947936301769699, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 37062 + }, + { + "epoch": 0.37063, + "grad_norm": 0.6413319935657784, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 37063 + }, + { + "epoch": 0.37064, + "grad_norm": 0.6489791894277663, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 37064 + }, + { + "epoch": 0.37065, + "grad_norm": 0.5962145949566855, + "learning_rate": 0.003, + "loss": 3.971, + "step": 37065 + }, + { + "epoch": 0.37066, + "grad_norm": 0.5656525767584589, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 37066 + }, + { + "epoch": 0.37067, + "grad_norm": 0.536655719292133, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 37067 + }, + { + "epoch": 0.37068, + "grad_norm": 0.592102827424373, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 37068 + }, + { + "epoch": 0.37069, + "grad_norm": 0.592613821074836, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 37069 + }, + { + "epoch": 0.3707, + "grad_norm": 0.5875806856154263, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 37070 + }, + { + "epoch": 0.37071, + "grad_norm": 0.592038769990577, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 37071 + }, + { + "epoch": 0.37072, + "grad_norm": 0.6048780056084585, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 37072 + }, + { + "epoch": 0.37073, + "grad_norm": 0.5901150249352332, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 37073 + }, + { + "epoch": 0.37074, + "grad_norm": 0.5910294360062618, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 37074 + }, + { + "epoch": 0.37075, + "grad_norm": 0.5836739748743442, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 37075 + }, + { + "epoch": 0.37076, + "grad_norm": 0.6614751575677835, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 37076 + }, + { + "epoch": 0.37077, + "grad_norm": 0.6322915785046115, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37077 + }, + { + "epoch": 0.37078, + "grad_norm": 0.7236412090185833, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 37078 + }, + { + "epoch": 0.37079, + "grad_norm": 0.871139637373613, + "learning_rate": 0.003, + "loss": 3.982, + "step": 37079 + }, + { + "epoch": 0.3708, + "grad_norm": 0.9100498143155387, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 37080 + }, + { + "epoch": 0.37081, + "grad_norm": 0.909938479673984, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 37081 + }, + { + "epoch": 0.37082, + "grad_norm": 0.9676034831676362, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 37082 + }, + { + "epoch": 0.37083, + "grad_norm": 1.0091815193349083, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 37083 + }, + { + "epoch": 0.37084, + "grad_norm": 0.997740850147672, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 37084 + }, + { + "epoch": 0.37085, + "grad_norm": 1.1243284361244905, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 37085 + }, + { + "epoch": 0.37086, + "grad_norm": 1.0475085892619196, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 37086 + }, + { + "epoch": 0.37087, + "grad_norm": 0.9001374431193538, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 37087 + }, + { + "epoch": 0.37088, + "grad_norm": 0.8833422567282079, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 37088 + }, + { + "epoch": 0.37089, + "grad_norm": 0.9203817276731415, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 37089 + }, + { + "epoch": 0.3709, + "grad_norm": 0.7749032886421156, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 37090 + }, + { + "epoch": 0.37091, + "grad_norm": 0.7879160968523721, + "learning_rate": 0.003, + "loss": 3.986, + "step": 37091 + }, + { + "epoch": 0.37092, + "grad_norm": 0.9224078368866951, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 37092 + }, + { + "epoch": 0.37093, + "grad_norm": 0.9204160420804396, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 37093 + }, + { + "epoch": 0.37094, + "grad_norm": 0.8994987658366566, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 37094 + }, + { + "epoch": 0.37095, + "grad_norm": 0.9488897613911068, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 37095 + }, + { + "epoch": 0.37096, + "grad_norm": 0.8808143886213018, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 37096 + }, + { + "epoch": 0.37097, + "grad_norm": 0.7351558687303575, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 37097 + }, + { + "epoch": 0.37098, + "grad_norm": 0.8132685024291842, + "learning_rate": 0.003, + "loss": 4.0421, + "step": 37098 + }, + { + "epoch": 0.37099, + "grad_norm": 0.9518536930709695, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 37099 + }, + { + "epoch": 0.371, + "grad_norm": 1.044757399113619, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 37100 + }, + { + "epoch": 0.37101, + "grad_norm": 0.986437406355914, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 37101 + }, + { + "epoch": 0.37102, + "grad_norm": 1.0802946888028093, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 37102 + }, + { + "epoch": 0.37103, + "grad_norm": 0.9163085237033993, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 37103 + }, + { + "epoch": 0.37104, + "grad_norm": 0.8323314846962786, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 37104 + }, + { + "epoch": 0.37105, + "grad_norm": 0.8845771137649099, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 37105 + }, + { + "epoch": 0.37106, + "grad_norm": 0.7423775428648068, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 37106 + }, + { + "epoch": 0.37107, + "grad_norm": 0.7920010956123799, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 37107 + }, + { + "epoch": 0.37108, + "grad_norm": 0.7234598731497192, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 37108 + }, + { + "epoch": 0.37109, + "grad_norm": 0.6665864061205697, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 37109 + }, + { + "epoch": 0.3711, + "grad_norm": 0.5982736992406668, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 37110 + }, + { + "epoch": 0.37111, + "grad_norm": 0.5637212241065972, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 37111 + }, + { + "epoch": 0.37112, + "grad_norm": 0.5625637656338435, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 37112 + }, + { + "epoch": 0.37113, + "grad_norm": 0.5895708310491067, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 37113 + }, + { + "epoch": 0.37114, + "grad_norm": 0.6238387267280556, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 37114 + }, + { + "epoch": 0.37115, + "grad_norm": 0.593965750014018, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 37115 + }, + { + "epoch": 0.37116, + "grad_norm": 0.6244342693522098, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 37116 + }, + { + "epoch": 0.37117, + "grad_norm": 0.621851110907642, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 37117 + }, + { + "epoch": 0.37118, + "grad_norm": 0.5957379394053167, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 37118 + }, + { + "epoch": 0.37119, + "grad_norm": 0.6479141945950675, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 37119 + }, + { + "epoch": 0.3712, + "grad_norm": 0.7360109976656449, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 37120 + }, + { + "epoch": 0.37121, + "grad_norm": 0.7487045455667308, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 37121 + }, + { + "epoch": 0.37122, + "grad_norm": 0.7422012529095823, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 37122 + }, + { + "epoch": 0.37123, + "grad_norm": 0.6570970836089868, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 37123 + }, + { + "epoch": 0.37124, + "grad_norm": 0.5849180131238513, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 37124 + }, + { + "epoch": 0.37125, + "grad_norm": 0.6086230730686318, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 37125 + }, + { + "epoch": 0.37126, + "grad_norm": 0.6488145730489912, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 37126 + }, + { + "epoch": 0.37127, + "grad_norm": 0.6966867955809996, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 37127 + }, + { + "epoch": 0.37128, + "grad_norm": 0.7027000494137495, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 37128 + }, + { + "epoch": 0.37129, + "grad_norm": 0.88819357730216, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 37129 + }, + { + "epoch": 0.3713, + "grad_norm": 0.9746349825935089, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 37130 + }, + { + "epoch": 0.37131, + "grad_norm": 0.8941448619802829, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 37131 + }, + { + "epoch": 0.37132, + "grad_norm": 0.8464880970242015, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 37132 + }, + { + "epoch": 0.37133, + "grad_norm": 0.8052984565967549, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 37133 + }, + { + "epoch": 0.37134, + "grad_norm": 0.7874647816667323, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 37134 + }, + { + "epoch": 0.37135, + "grad_norm": 0.7336037699050414, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 37135 + }, + { + "epoch": 0.37136, + "grad_norm": 0.7906538835871904, + "learning_rate": 0.003, + "loss": 3.965, + "step": 37136 + }, + { + "epoch": 0.37137, + "grad_norm": 0.7300820660209405, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 37137 + }, + { + "epoch": 0.37138, + "grad_norm": 0.7769789298422859, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 37138 + }, + { + "epoch": 0.37139, + "grad_norm": 0.7903795980865495, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 37139 + }, + { + "epoch": 0.3714, + "grad_norm": 0.9330935169289987, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 37140 + }, + { + "epoch": 0.37141, + "grad_norm": 1.0220363214985388, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 37141 + }, + { + "epoch": 0.37142, + "grad_norm": 0.9660906326775309, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 37142 + }, + { + "epoch": 0.37143, + "grad_norm": 1.0045335628425591, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 37143 + }, + { + "epoch": 0.37144, + "grad_norm": 1.0615233666904913, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 37144 + }, + { + "epoch": 0.37145, + "grad_norm": 0.9689769049601, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 37145 + }, + { + "epoch": 0.37146, + "grad_norm": 0.9839087027845753, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 37146 + }, + { + "epoch": 0.37147, + "grad_norm": 0.8874316393084082, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 37147 + }, + { + "epoch": 0.37148, + "grad_norm": 0.8425393941116597, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 37148 + }, + { + "epoch": 0.37149, + "grad_norm": 0.8093148312393753, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 37149 + }, + { + "epoch": 0.3715, + "grad_norm": 0.845420178567355, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 37150 + }, + { + "epoch": 0.37151, + "grad_norm": 0.8878316928841249, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 37151 + }, + { + "epoch": 0.37152, + "grad_norm": 0.8320166186909694, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 37152 + }, + { + "epoch": 0.37153, + "grad_norm": 0.6953863843488153, + "learning_rate": 0.003, + "loss": 3.986, + "step": 37153 + }, + { + "epoch": 0.37154, + "grad_norm": 0.7318676244470558, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 37154 + }, + { + "epoch": 0.37155, + "grad_norm": 0.7824701241633045, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 37155 + }, + { + "epoch": 0.37156, + "grad_norm": 0.780386781598146, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 37156 + }, + { + "epoch": 0.37157, + "grad_norm": 0.7902625043044152, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 37157 + }, + { + "epoch": 0.37158, + "grad_norm": 0.7323664682750896, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 37158 + }, + { + "epoch": 0.37159, + "grad_norm": 0.6726726597941559, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 37159 + }, + { + "epoch": 0.3716, + "grad_norm": 0.672618289540419, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 37160 + }, + { + "epoch": 0.37161, + "grad_norm": 0.6360585715026064, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 37161 + }, + { + "epoch": 0.37162, + "grad_norm": 0.5863427927754924, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 37162 + }, + { + "epoch": 0.37163, + "grad_norm": 0.6003194783604408, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 37163 + }, + { + "epoch": 0.37164, + "grad_norm": 0.5948581951399624, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 37164 + }, + { + "epoch": 0.37165, + "grad_norm": 0.6784701776083961, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 37165 + }, + { + "epoch": 0.37166, + "grad_norm": 0.8042231795132508, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 37166 + }, + { + "epoch": 0.37167, + "grad_norm": 0.9672135668578914, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 37167 + }, + { + "epoch": 0.37168, + "grad_norm": 1.0283170884043642, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 37168 + }, + { + "epoch": 0.37169, + "grad_norm": 0.8016905977389114, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 37169 + }, + { + "epoch": 0.3717, + "grad_norm": 0.5963316612465994, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 37170 + }, + { + "epoch": 0.37171, + "grad_norm": 0.5393884790186553, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 37171 + }, + { + "epoch": 0.37172, + "grad_norm": 0.6376385421204562, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 37172 + }, + { + "epoch": 0.37173, + "grad_norm": 0.5760991616289581, + "learning_rate": 0.003, + "loss": 3.959, + "step": 37173 + }, + { + "epoch": 0.37174, + "grad_norm": 0.48141944677264503, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 37174 + }, + { + "epoch": 0.37175, + "grad_norm": 0.4818663814533445, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 37175 + }, + { + "epoch": 0.37176, + "grad_norm": 0.4812003803128463, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 37176 + }, + { + "epoch": 0.37177, + "grad_norm": 0.581896589380631, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 37177 + }, + { + "epoch": 0.37178, + "grad_norm": 0.6763881571229436, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 37178 + }, + { + "epoch": 0.37179, + "grad_norm": 0.8488296294138494, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 37179 + }, + { + "epoch": 0.3718, + "grad_norm": 0.9526247609788898, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 37180 + }, + { + "epoch": 0.37181, + "grad_norm": 1.0380892955545251, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 37181 + }, + { + "epoch": 0.37182, + "grad_norm": 0.8992856331733916, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 37182 + }, + { + "epoch": 0.37183, + "grad_norm": 0.8006459324297821, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 37183 + }, + { + "epoch": 0.37184, + "grad_norm": 0.9305777883546598, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 37184 + }, + { + "epoch": 0.37185, + "grad_norm": 1.0821452804673537, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 37185 + }, + { + "epoch": 0.37186, + "grad_norm": 0.9399356267466079, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 37186 + }, + { + "epoch": 0.37187, + "grad_norm": 0.8087424366590575, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 37187 + }, + { + "epoch": 0.37188, + "grad_norm": 0.8687608998789332, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 37188 + }, + { + "epoch": 0.37189, + "grad_norm": 0.8073643714922674, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 37189 + }, + { + "epoch": 0.3719, + "grad_norm": 0.7647882119902668, + "learning_rate": 0.003, + "loss": 3.975, + "step": 37190 + }, + { + "epoch": 0.37191, + "grad_norm": 0.6769708419098248, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 37191 + }, + { + "epoch": 0.37192, + "grad_norm": 0.7280861504892978, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 37192 + }, + { + "epoch": 0.37193, + "grad_norm": 0.7874179808240448, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 37193 + }, + { + "epoch": 0.37194, + "grad_norm": 0.9389743650253026, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 37194 + }, + { + "epoch": 0.37195, + "grad_norm": 1.1367622108251265, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 37195 + }, + { + "epoch": 0.37196, + "grad_norm": 0.9404557488524555, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 37196 + }, + { + "epoch": 0.37197, + "grad_norm": 0.873804978683216, + "learning_rate": 0.003, + "loss": 3.999, + "step": 37197 + }, + { + "epoch": 0.37198, + "grad_norm": 0.829548135329442, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 37198 + }, + { + "epoch": 0.37199, + "grad_norm": 0.7241579270388676, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 37199 + }, + { + "epoch": 0.372, + "grad_norm": 0.7275713447152184, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 37200 + }, + { + "epoch": 0.37201, + "grad_norm": 0.7156159659543586, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 37201 + }, + { + "epoch": 0.37202, + "grad_norm": 0.6464873554657404, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 37202 + }, + { + "epoch": 0.37203, + "grad_norm": 0.6768505067654226, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 37203 + }, + { + "epoch": 0.37204, + "grad_norm": 0.628267751822524, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37204 + }, + { + "epoch": 0.37205, + "grad_norm": 0.6640279704280923, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 37205 + }, + { + "epoch": 0.37206, + "grad_norm": 0.6250692532161229, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 37206 + }, + { + "epoch": 0.37207, + "grad_norm": 0.6076377652447936, + "learning_rate": 0.003, + "loss": 3.962, + "step": 37207 + }, + { + "epoch": 0.37208, + "grad_norm": 0.6813228386826617, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 37208 + }, + { + "epoch": 0.37209, + "grad_norm": 0.7391558444650779, + "learning_rate": 0.003, + "loss": 3.981, + "step": 37209 + }, + { + "epoch": 0.3721, + "grad_norm": 0.9205772762615536, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 37210 + }, + { + "epoch": 0.37211, + "grad_norm": 0.9950743891078002, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 37211 + }, + { + "epoch": 0.37212, + "grad_norm": 0.8967418268212312, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 37212 + }, + { + "epoch": 0.37213, + "grad_norm": 0.736905254894296, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 37213 + }, + { + "epoch": 0.37214, + "grad_norm": 0.8064734759711605, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 37214 + }, + { + "epoch": 0.37215, + "grad_norm": 0.7042788721470031, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 37215 + }, + { + "epoch": 0.37216, + "grad_norm": 0.6518703779280443, + "learning_rate": 0.003, + "loss": 3.97, + "step": 37216 + }, + { + "epoch": 0.37217, + "grad_norm": 0.6782165864673351, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 37217 + }, + { + "epoch": 0.37218, + "grad_norm": 0.7838969253749593, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 37218 + }, + { + "epoch": 0.37219, + "grad_norm": 0.8597934704384081, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 37219 + }, + { + "epoch": 0.3722, + "grad_norm": 0.8768711140658839, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 37220 + }, + { + "epoch": 0.37221, + "grad_norm": 0.8251475256670476, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 37221 + }, + { + "epoch": 0.37222, + "grad_norm": 0.766336013950482, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 37222 + }, + { + "epoch": 0.37223, + "grad_norm": 0.7066934681796341, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 37223 + }, + { + "epoch": 0.37224, + "grad_norm": 0.6825465552535525, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 37224 + }, + { + "epoch": 0.37225, + "grad_norm": 0.6791773282331581, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 37225 + }, + { + "epoch": 0.37226, + "grad_norm": 0.8298344444219002, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 37226 + }, + { + "epoch": 0.37227, + "grad_norm": 0.9438437974948402, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 37227 + }, + { + "epoch": 0.37228, + "grad_norm": 0.894773751563642, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 37228 + }, + { + "epoch": 0.37229, + "grad_norm": 0.7410985047325207, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 37229 + }, + { + "epoch": 0.3723, + "grad_norm": 0.7870518334238757, + "learning_rate": 0.003, + "loss": 3.995, + "step": 37230 + }, + { + "epoch": 0.37231, + "grad_norm": 0.8191013345577083, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 37231 + }, + { + "epoch": 0.37232, + "grad_norm": 0.9016382935634136, + "learning_rate": 0.003, + "loss": 3.996, + "step": 37232 + }, + { + "epoch": 0.37233, + "grad_norm": 0.9448828153045367, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 37233 + }, + { + "epoch": 0.37234, + "grad_norm": 0.9130083609562124, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 37234 + }, + { + "epoch": 0.37235, + "grad_norm": 0.8911888090752191, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 37235 + }, + { + "epoch": 0.37236, + "grad_norm": 0.8605643713701565, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 37236 + }, + { + "epoch": 0.37237, + "grad_norm": 0.8173818346912681, + "learning_rate": 0.003, + "loss": 3.987, + "step": 37237 + }, + { + "epoch": 0.37238, + "grad_norm": 0.9537752094140795, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 37238 + }, + { + "epoch": 0.37239, + "grad_norm": 0.9592454388646733, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 37239 + }, + { + "epoch": 0.3724, + "grad_norm": 0.8422531911101884, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 37240 + }, + { + "epoch": 0.37241, + "grad_norm": 0.80055192534878, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 37241 + }, + { + "epoch": 0.37242, + "grad_norm": 0.7710020639514956, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 37242 + }, + { + "epoch": 0.37243, + "grad_norm": 0.700099091836724, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 37243 + }, + { + "epoch": 0.37244, + "grad_norm": 0.6671888504129259, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 37244 + }, + { + "epoch": 0.37245, + "grad_norm": 0.5858875072856548, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 37245 + }, + { + "epoch": 0.37246, + "grad_norm": 0.6095849795521492, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 37246 + }, + { + "epoch": 0.37247, + "grad_norm": 0.6340796943048824, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 37247 + }, + { + "epoch": 0.37248, + "grad_norm": 0.7542445202053955, + "learning_rate": 0.003, + "loss": 3.981, + "step": 37248 + }, + { + "epoch": 0.37249, + "grad_norm": 0.8768597108492602, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 37249 + }, + { + "epoch": 0.3725, + "grad_norm": 0.8663168561587347, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 37250 + }, + { + "epoch": 0.37251, + "grad_norm": 0.7634506725154843, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 37251 + }, + { + "epoch": 0.37252, + "grad_norm": 0.7518116032115035, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 37252 + }, + { + "epoch": 0.37253, + "grad_norm": 0.8577197408168549, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 37253 + }, + { + "epoch": 0.37254, + "grad_norm": 0.865701636763388, + "learning_rate": 0.003, + "loss": 3.9421, + "step": 37254 + }, + { + "epoch": 0.37255, + "grad_norm": 0.7320520429900182, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 37255 + }, + { + "epoch": 0.37256, + "grad_norm": 0.6348448185897848, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 37256 + }, + { + "epoch": 0.37257, + "grad_norm": 0.5877109437509861, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 37257 + }, + { + "epoch": 0.37258, + "grad_norm": 0.656127320642739, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 37258 + }, + { + "epoch": 0.37259, + "grad_norm": 0.712007197568936, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 37259 + }, + { + "epoch": 0.3726, + "grad_norm": 0.6912745780875651, + "learning_rate": 0.003, + "loss": 3.979, + "step": 37260 + }, + { + "epoch": 0.37261, + "grad_norm": 0.8535147093385371, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 37261 + }, + { + "epoch": 0.37262, + "grad_norm": 0.8763769454695224, + "learning_rate": 0.003, + "loss": 3.999, + "step": 37262 + }, + { + "epoch": 0.37263, + "grad_norm": 1.0619285969090917, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 37263 + }, + { + "epoch": 0.37264, + "grad_norm": 1.1202303240971399, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 37264 + }, + { + "epoch": 0.37265, + "grad_norm": 0.8536708872359994, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 37265 + }, + { + "epoch": 0.37266, + "grad_norm": 0.8252582105359072, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 37266 + }, + { + "epoch": 0.37267, + "grad_norm": 0.7626878988784744, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 37267 + }, + { + "epoch": 0.37268, + "grad_norm": 0.6525027767196135, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 37268 + }, + { + "epoch": 0.37269, + "grad_norm": 0.6519959007425183, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 37269 + }, + { + "epoch": 0.3727, + "grad_norm": 0.6748036955296202, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 37270 + }, + { + "epoch": 0.37271, + "grad_norm": 0.6350939416500438, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 37271 + }, + { + "epoch": 0.37272, + "grad_norm": 0.7067030199035378, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 37272 + }, + { + "epoch": 0.37273, + "grad_norm": 0.6506569489437788, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 37273 + }, + { + "epoch": 0.37274, + "grad_norm": 0.5975928559224511, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 37274 + }, + { + "epoch": 0.37275, + "grad_norm": 0.6285213511330894, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 37275 + }, + { + "epoch": 0.37276, + "grad_norm": 0.6103950918281116, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 37276 + }, + { + "epoch": 0.37277, + "grad_norm": 0.5890562119319009, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 37277 + }, + { + "epoch": 0.37278, + "grad_norm": 0.6150696183850882, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 37278 + }, + { + "epoch": 0.37279, + "grad_norm": 0.6131532117580231, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 37279 + }, + { + "epoch": 0.3728, + "grad_norm": 0.6455853561734526, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 37280 + }, + { + "epoch": 0.37281, + "grad_norm": 0.6131717490260385, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 37281 + }, + { + "epoch": 0.37282, + "grad_norm": 0.628654405062177, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 37282 + }, + { + "epoch": 0.37283, + "grad_norm": 0.6573635549400964, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 37283 + }, + { + "epoch": 0.37284, + "grad_norm": 0.7257836113024826, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 37284 + }, + { + "epoch": 0.37285, + "grad_norm": 0.8743113309641429, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 37285 + }, + { + "epoch": 0.37286, + "grad_norm": 1.1463378666294355, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 37286 + }, + { + "epoch": 0.37287, + "grad_norm": 0.8798814296096963, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 37287 + }, + { + "epoch": 0.37288, + "grad_norm": 0.7874416926378569, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 37288 + }, + { + "epoch": 0.37289, + "grad_norm": 0.7847762483764255, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 37289 + }, + { + "epoch": 0.3729, + "grad_norm": 0.8675054968344432, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 37290 + }, + { + "epoch": 0.37291, + "grad_norm": 0.9089692626583352, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 37291 + }, + { + "epoch": 0.37292, + "grad_norm": 0.9480222939182901, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 37292 + }, + { + "epoch": 0.37293, + "grad_norm": 0.877969393684879, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 37293 + }, + { + "epoch": 0.37294, + "grad_norm": 0.9632928162570167, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 37294 + }, + { + "epoch": 0.37295, + "grad_norm": 0.9271384957084665, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 37295 + }, + { + "epoch": 0.37296, + "grad_norm": 1.047827999740022, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37296 + }, + { + "epoch": 0.37297, + "grad_norm": 0.9944534450366007, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37297 + }, + { + "epoch": 0.37298, + "grad_norm": 0.9623335445618908, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 37298 + }, + { + "epoch": 0.37299, + "grad_norm": 0.938975332548934, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 37299 + }, + { + "epoch": 0.373, + "grad_norm": 1.000726780667716, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 37300 + }, + { + "epoch": 0.37301, + "grad_norm": 0.9520729951822574, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 37301 + }, + { + "epoch": 0.37302, + "grad_norm": 0.8394946946694267, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 37302 + }, + { + "epoch": 0.37303, + "grad_norm": 0.8656242357828023, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 37303 + }, + { + "epoch": 0.37304, + "grad_norm": 0.9798018926453167, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 37304 + }, + { + "epoch": 0.37305, + "grad_norm": 1.0347348666957321, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 37305 + }, + { + "epoch": 0.37306, + "grad_norm": 0.9654870261290052, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 37306 + }, + { + "epoch": 0.37307, + "grad_norm": 1.0154225497983462, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 37307 + }, + { + "epoch": 0.37308, + "grad_norm": 0.8763705729212907, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 37308 + }, + { + "epoch": 0.37309, + "grad_norm": 0.8395286708255739, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 37309 + }, + { + "epoch": 0.3731, + "grad_norm": 0.9774566086153257, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 37310 + }, + { + "epoch": 0.37311, + "grad_norm": 0.9530077123141971, + "learning_rate": 0.003, + "loss": 4.0434, + "step": 37311 + }, + { + "epoch": 0.37312, + "grad_norm": 0.7091304764300096, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 37312 + }, + { + "epoch": 0.37313, + "grad_norm": 0.6815470083048043, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 37313 + }, + { + "epoch": 0.37314, + "grad_norm": 0.6281634544186739, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 37314 + }, + { + "epoch": 0.37315, + "grad_norm": 0.7354536464241006, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 37315 + }, + { + "epoch": 0.37316, + "grad_norm": 0.75925589625108, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 37316 + }, + { + "epoch": 0.37317, + "grad_norm": 0.6524802647561788, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 37317 + }, + { + "epoch": 0.37318, + "grad_norm": 0.6348189242266901, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 37318 + }, + { + "epoch": 0.37319, + "grad_norm": 0.8274991206187768, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 37319 + }, + { + "epoch": 0.3732, + "grad_norm": 0.8563890259352667, + "learning_rate": 0.003, + "loss": 3.986, + "step": 37320 + }, + { + "epoch": 0.37321, + "grad_norm": 0.8659487101507977, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 37321 + }, + { + "epoch": 0.37322, + "grad_norm": 0.8893027179155709, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 37322 + }, + { + "epoch": 0.37323, + "grad_norm": 1.1080005806017625, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 37323 + }, + { + "epoch": 0.37324, + "grad_norm": 0.8585824559181597, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 37324 + }, + { + "epoch": 0.37325, + "grad_norm": 0.645444461621431, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 37325 + }, + { + "epoch": 0.37326, + "grad_norm": 0.5952501548713619, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 37326 + }, + { + "epoch": 0.37327, + "grad_norm": 0.75551946839552, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 37327 + }, + { + "epoch": 0.37328, + "grad_norm": 0.8174857551353808, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 37328 + }, + { + "epoch": 0.37329, + "grad_norm": 0.7486444920169942, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 37329 + }, + { + "epoch": 0.3733, + "grad_norm": 0.5618854233533205, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 37330 + }, + { + "epoch": 0.37331, + "grad_norm": 0.5018116352128354, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 37331 + }, + { + "epoch": 0.37332, + "grad_norm": 0.5485742675776232, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 37332 + }, + { + "epoch": 0.37333, + "grad_norm": 0.6101038985168526, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 37333 + }, + { + "epoch": 0.37334, + "grad_norm": 0.5967715562671757, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 37334 + }, + { + "epoch": 0.37335, + "grad_norm": 0.6500868564678257, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 37335 + }, + { + "epoch": 0.37336, + "grad_norm": 0.6359979089572391, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 37336 + }, + { + "epoch": 0.37337, + "grad_norm": 0.5821761386707297, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 37337 + }, + { + "epoch": 0.37338, + "grad_norm": 0.46876214038476194, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 37338 + }, + { + "epoch": 0.37339, + "grad_norm": 0.4846515847832008, + "learning_rate": 0.003, + "loss": 3.9423, + "step": 37339 + }, + { + "epoch": 0.3734, + "grad_norm": 0.5439433947517618, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 37340 + }, + { + "epoch": 0.37341, + "grad_norm": 0.5951467970173181, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 37341 + }, + { + "epoch": 0.37342, + "grad_norm": 0.6935672242237882, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 37342 + }, + { + "epoch": 0.37343, + "grad_norm": 0.7134542348279271, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 37343 + }, + { + "epoch": 0.37344, + "grad_norm": 0.8634412065086391, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 37344 + }, + { + "epoch": 0.37345, + "grad_norm": 1.049967896544687, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 37345 + }, + { + "epoch": 0.37346, + "grad_norm": 0.9880802894059691, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 37346 + }, + { + "epoch": 0.37347, + "grad_norm": 0.8450202574814536, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 37347 + }, + { + "epoch": 0.37348, + "grad_norm": 0.7757310620980441, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 37348 + }, + { + "epoch": 0.37349, + "grad_norm": 0.8641250782129447, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 37349 + }, + { + "epoch": 0.3735, + "grad_norm": 0.9877147473890417, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 37350 + }, + { + "epoch": 0.37351, + "grad_norm": 1.0100920178011623, + "learning_rate": 0.003, + "loss": 4.015, + "step": 37351 + }, + { + "epoch": 0.37352, + "grad_norm": 1.0700724707202567, + "learning_rate": 0.003, + "loss": 3.973, + "step": 37352 + }, + { + "epoch": 0.37353, + "grad_norm": 0.8684590270227756, + "learning_rate": 0.003, + "loss": 4.017, + "step": 37353 + }, + { + "epoch": 0.37354, + "grad_norm": 0.9170826093653327, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 37354 + }, + { + "epoch": 0.37355, + "grad_norm": 0.8909618572906112, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 37355 + }, + { + "epoch": 0.37356, + "grad_norm": 0.8306664691364247, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 37356 + }, + { + "epoch": 0.37357, + "grad_norm": 0.8264454376519775, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 37357 + }, + { + "epoch": 0.37358, + "grad_norm": 0.8653930982602971, + "learning_rate": 0.003, + "loss": 3.999, + "step": 37358 + }, + { + "epoch": 0.37359, + "grad_norm": 0.8452242259509569, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 37359 + }, + { + "epoch": 0.3736, + "grad_norm": 0.699114176648461, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 37360 + }, + { + "epoch": 0.37361, + "grad_norm": 0.6651754641783343, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 37361 + }, + { + "epoch": 0.37362, + "grad_norm": 0.6106875221696445, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 37362 + }, + { + "epoch": 0.37363, + "grad_norm": 0.6529164240253157, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 37363 + }, + { + "epoch": 0.37364, + "grad_norm": 0.6976705362809796, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 37364 + }, + { + "epoch": 0.37365, + "grad_norm": 0.7060587785501755, + "learning_rate": 0.003, + "loss": 3.998, + "step": 37365 + }, + { + "epoch": 0.37366, + "grad_norm": 0.7652217106878832, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 37366 + }, + { + "epoch": 0.37367, + "grad_norm": 0.7438192187480438, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 37367 + }, + { + "epoch": 0.37368, + "grad_norm": 0.6962597099835177, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 37368 + }, + { + "epoch": 0.37369, + "grad_norm": 0.8253705660601298, + "learning_rate": 0.003, + "loss": 4.003, + "step": 37369 + }, + { + "epoch": 0.3737, + "grad_norm": 0.8335859106132695, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 37370 + }, + { + "epoch": 0.37371, + "grad_norm": 0.8049552832425905, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 37371 + }, + { + "epoch": 0.37372, + "grad_norm": 0.8207161382343537, + "learning_rate": 0.003, + "loss": 3.985, + "step": 37372 + }, + { + "epoch": 0.37373, + "grad_norm": 0.7409444176575817, + "learning_rate": 0.003, + "loss": 3.991, + "step": 37373 + }, + { + "epoch": 0.37374, + "grad_norm": 0.753542796108346, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 37374 + }, + { + "epoch": 0.37375, + "grad_norm": 0.7563733703568998, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 37375 + }, + { + "epoch": 0.37376, + "grad_norm": 0.6591407542877353, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 37376 + }, + { + "epoch": 0.37377, + "grad_norm": 0.7057537380075949, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 37377 + }, + { + "epoch": 0.37378, + "grad_norm": 0.7780033950778171, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 37378 + }, + { + "epoch": 0.37379, + "grad_norm": 0.8542363490180581, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 37379 + }, + { + "epoch": 0.3738, + "grad_norm": 0.9173352128186889, + "learning_rate": 0.003, + "loss": 4.0249, + "step": 37380 + }, + { + "epoch": 0.37381, + "grad_norm": 0.9394540886968845, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 37381 + }, + { + "epoch": 0.37382, + "grad_norm": 0.8716373918413468, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 37382 + }, + { + "epoch": 0.37383, + "grad_norm": 0.9469313914296389, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 37383 + }, + { + "epoch": 0.37384, + "grad_norm": 0.9585890953895665, + "learning_rate": 0.003, + "loss": 3.987, + "step": 37384 + }, + { + "epoch": 0.37385, + "grad_norm": 0.9396190174418507, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 37385 + }, + { + "epoch": 0.37386, + "grad_norm": 0.8267538482890197, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 37386 + }, + { + "epoch": 0.37387, + "grad_norm": 0.7615099585352625, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 37387 + }, + { + "epoch": 0.37388, + "grad_norm": 0.710340946581221, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 37388 + }, + { + "epoch": 0.37389, + "grad_norm": 0.7623160952139183, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 37389 + }, + { + "epoch": 0.3739, + "grad_norm": 0.7536998796099466, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 37390 + }, + { + "epoch": 0.37391, + "grad_norm": 0.7745730505909763, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 37391 + }, + { + "epoch": 0.37392, + "grad_norm": 0.7857100348529295, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 37392 + }, + { + "epoch": 0.37393, + "grad_norm": 0.7877904205295481, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 37393 + }, + { + "epoch": 0.37394, + "grad_norm": 0.8582251755456481, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 37394 + }, + { + "epoch": 0.37395, + "grad_norm": 0.861397086765479, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 37395 + }, + { + "epoch": 0.37396, + "grad_norm": 0.8324161140515305, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 37396 + }, + { + "epoch": 0.37397, + "grad_norm": 0.8826864476066514, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 37397 + }, + { + "epoch": 0.37398, + "grad_norm": 0.8953443632127868, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 37398 + }, + { + "epoch": 0.37399, + "grad_norm": 0.7546157332498569, + "learning_rate": 0.003, + "loss": 3.997, + "step": 37399 + }, + { + "epoch": 0.374, + "grad_norm": 0.7427569552520432, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 37400 + }, + { + "epoch": 0.37401, + "grad_norm": 0.6643027580166215, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 37401 + }, + { + "epoch": 0.37402, + "grad_norm": 0.6282927870101597, + "learning_rate": 0.003, + "loss": 3.986, + "step": 37402 + }, + { + "epoch": 0.37403, + "grad_norm": 0.5440188843688981, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 37403 + }, + { + "epoch": 0.37404, + "grad_norm": 0.6181494752344099, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 37404 + }, + { + "epoch": 0.37405, + "grad_norm": 0.6712892267278168, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 37405 + }, + { + "epoch": 0.37406, + "grad_norm": 0.6229963005481294, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 37406 + }, + { + "epoch": 0.37407, + "grad_norm": 0.5720014942796242, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 37407 + }, + { + "epoch": 0.37408, + "grad_norm": 0.6313715866164077, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 37408 + }, + { + "epoch": 0.37409, + "grad_norm": 0.7449888167360625, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 37409 + }, + { + "epoch": 0.3741, + "grad_norm": 0.853699697336929, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 37410 + }, + { + "epoch": 0.37411, + "grad_norm": 0.9862036630518604, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 37411 + }, + { + "epoch": 0.37412, + "grad_norm": 1.0031403243147599, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 37412 + }, + { + "epoch": 0.37413, + "grad_norm": 1.1292386770194731, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 37413 + }, + { + "epoch": 0.37414, + "grad_norm": 1.1074741016484193, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 37414 + }, + { + "epoch": 0.37415, + "grad_norm": 1.0134931936455254, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 37415 + }, + { + "epoch": 0.37416, + "grad_norm": 0.8942680953907601, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 37416 + }, + { + "epoch": 0.37417, + "grad_norm": 0.7337453945200607, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 37417 + }, + { + "epoch": 0.37418, + "grad_norm": 0.7057514825811906, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 37418 + }, + { + "epoch": 0.37419, + "grad_norm": 0.7036983333747775, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 37419 + }, + { + "epoch": 0.3742, + "grad_norm": 0.7029506735353123, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 37420 + }, + { + "epoch": 0.37421, + "grad_norm": 0.6361064346521262, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 37421 + }, + { + "epoch": 0.37422, + "grad_norm": 0.6413093924005558, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 37422 + }, + { + "epoch": 0.37423, + "grad_norm": 0.6828794631824642, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 37423 + }, + { + "epoch": 0.37424, + "grad_norm": 0.7940284138952746, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 37424 + }, + { + "epoch": 0.37425, + "grad_norm": 0.8732852447895998, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 37425 + }, + { + "epoch": 0.37426, + "grad_norm": 0.8330053123263151, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 37426 + }, + { + "epoch": 0.37427, + "grad_norm": 0.6914047393559911, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 37427 + }, + { + "epoch": 0.37428, + "grad_norm": 0.5569067365288185, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 37428 + }, + { + "epoch": 0.37429, + "grad_norm": 0.6763883517104458, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 37429 + }, + { + "epoch": 0.3743, + "grad_norm": 0.6543130183908568, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 37430 + }, + { + "epoch": 0.37431, + "grad_norm": 0.6154449529856381, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 37431 + }, + { + "epoch": 0.37432, + "grad_norm": 0.6978250981928227, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 37432 + }, + { + "epoch": 0.37433, + "grad_norm": 0.7470234371852736, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 37433 + }, + { + "epoch": 0.37434, + "grad_norm": 0.7314159464132386, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 37434 + }, + { + "epoch": 0.37435, + "grad_norm": 0.7117055741482531, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 37435 + }, + { + "epoch": 0.37436, + "grad_norm": 0.6296178682553557, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 37436 + }, + { + "epoch": 0.37437, + "grad_norm": 0.6172378372099198, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 37437 + }, + { + "epoch": 0.37438, + "grad_norm": 0.6025908898751332, + "learning_rate": 0.003, + "loss": 3.992, + "step": 37438 + }, + { + "epoch": 0.37439, + "grad_norm": 0.6378600662235486, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 37439 + }, + { + "epoch": 0.3744, + "grad_norm": 0.8309265494796793, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 37440 + }, + { + "epoch": 0.37441, + "grad_norm": 1.0255576153021526, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 37441 + }, + { + "epoch": 0.37442, + "grad_norm": 1.1642512860018723, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 37442 + }, + { + "epoch": 0.37443, + "grad_norm": 0.6105963244245117, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 37443 + }, + { + "epoch": 0.37444, + "grad_norm": 0.6998554639056801, + "learning_rate": 0.003, + "loss": 3.9358, + "step": 37444 + }, + { + "epoch": 0.37445, + "grad_norm": 0.9545260997090552, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 37445 + }, + { + "epoch": 0.37446, + "grad_norm": 1.0160095109992273, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 37446 + }, + { + "epoch": 0.37447, + "grad_norm": 0.906347723354112, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 37447 + }, + { + "epoch": 0.37448, + "grad_norm": 1.0372689357547316, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 37448 + }, + { + "epoch": 0.37449, + "grad_norm": 0.9105461406458383, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 37449 + }, + { + "epoch": 0.3745, + "grad_norm": 0.9280334127937262, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 37450 + }, + { + "epoch": 0.37451, + "grad_norm": 0.9617207843618527, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 37451 + }, + { + "epoch": 0.37452, + "grad_norm": 1.0247104453551654, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 37452 + }, + { + "epoch": 0.37453, + "grad_norm": 0.8755556738719373, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 37453 + }, + { + "epoch": 0.37454, + "grad_norm": 0.8629144495306883, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 37454 + }, + { + "epoch": 0.37455, + "grad_norm": 0.9095156809688039, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 37455 + }, + { + "epoch": 0.37456, + "grad_norm": 0.8232414530948822, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 37456 + }, + { + "epoch": 0.37457, + "grad_norm": 0.7745231598925637, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 37457 + }, + { + "epoch": 0.37458, + "grad_norm": 0.7588660139212347, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 37458 + }, + { + "epoch": 0.37459, + "grad_norm": 0.7390212289180133, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 37459 + }, + { + "epoch": 0.3746, + "grad_norm": 0.6640130307187208, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 37460 + }, + { + "epoch": 0.37461, + "grad_norm": 0.6531615809068507, + "learning_rate": 0.003, + "loss": 3.966, + "step": 37461 + }, + { + "epoch": 0.37462, + "grad_norm": 0.5923000357493902, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 37462 + }, + { + "epoch": 0.37463, + "grad_norm": 0.648701626994966, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37463 + }, + { + "epoch": 0.37464, + "grad_norm": 0.7539844547617447, + "learning_rate": 0.003, + "loss": 3.971, + "step": 37464 + }, + { + "epoch": 0.37465, + "grad_norm": 0.7832980581359611, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 37465 + }, + { + "epoch": 0.37466, + "grad_norm": 0.6862918501895211, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 37466 + }, + { + "epoch": 0.37467, + "grad_norm": 0.6319418122254149, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 37467 + }, + { + "epoch": 0.37468, + "grad_norm": 0.6625156701893838, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 37468 + }, + { + "epoch": 0.37469, + "grad_norm": 0.6807598307574066, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 37469 + }, + { + "epoch": 0.3747, + "grad_norm": 0.7228152960049523, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 37470 + }, + { + "epoch": 0.37471, + "grad_norm": 0.8944039820630618, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 37471 + }, + { + "epoch": 0.37472, + "grad_norm": 0.9974841798139948, + "learning_rate": 0.003, + "loss": 3.979, + "step": 37472 + }, + { + "epoch": 0.37473, + "grad_norm": 1.267956121569078, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 37473 + }, + { + "epoch": 0.37474, + "grad_norm": 0.8046243768679462, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 37474 + }, + { + "epoch": 0.37475, + "grad_norm": 0.8177511026267983, + "learning_rate": 0.003, + "loss": 4.0318, + "step": 37475 + }, + { + "epoch": 0.37476, + "grad_norm": 0.8626941198381533, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 37476 + }, + { + "epoch": 0.37477, + "grad_norm": 1.0338168676543318, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 37477 + }, + { + "epoch": 0.37478, + "grad_norm": 1.0754975494915697, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 37478 + }, + { + "epoch": 0.37479, + "grad_norm": 0.9036919705137321, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 37479 + }, + { + "epoch": 0.3748, + "grad_norm": 0.9765898912848208, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 37480 + }, + { + "epoch": 0.37481, + "grad_norm": 1.0741835420645967, + "learning_rate": 0.003, + "loss": 4.008, + "step": 37481 + }, + { + "epoch": 0.37482, + "grad_norm": 0.9284306079543848, + "learning_rate": 0.003, + "loss": 4.019, + "step": 37482 + }, + { + "epoch": 0.37483, + "grad_norm": 0.6879398971528317, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 37483 + }, + { + "epoch": 0.37484, + "grad_norm": 0.7097320878053697, + "learning_rate": 0.003, + "loss": 4.003, + "step": 37484 + }, + { + "epoch": 0.37485, + "grad_norm": 0.8803653781037454, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 37485 + }, + { + "epoch": 0.37486, + "grad_norm": 0.885659891569434, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 37486 + }, + { + "epoch": 0.37487, + "grad_norm": 0.8933284173016445, + "learning_rate": 0.003, + "loss": 4.037, + "step": 37487 + }, + { + "epoch": 0.37488, + "grad_norm": 0.8487331020617223, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 37488 + }, + { + "epoch": 0.37489, + "grad_norm": 0.7568125363379143, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 37489 + }, + { + "epoch": 0.3749, + "grad_norm": 0.6947405413754655, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 37490 + }, + { + "epoch": 0.37491, + "grad_norm": 0.6747183013164517, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 37491 + }, + { + "epoch": 0.37492, + "grad_norm": 0.6180235095771918, + "learning_rate": 0.003, + "loss": 3.974, + "step": 37492 + }, + { + "epoch": 0.37493, + "grad_norm": 0.5469494819295646, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 37493 + }, + { + "epoch": 0.37494, + "grad_norm": 0.553433541760024, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 37494 + }, + { + "epoch": 0.37495, + "grad_norm": 0.5400550999649506, + "learning_rate": 0.003, + "loss": 3.988, + "step": 37495 + }, + { + "epoch": 0.37496, + "grad_norm": 0.5461802078355622, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 37496 + }, + { + "epoch": 0.37497, + "grad_norm": 0.5845754887308152, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 37497 + }, + { + "epoch": 0.37498, + "grad_norm": 0.5419473369998568, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 37498 + }, + { + "epoch": 0.37499, + "grad_norm": 0.502947236777261, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 37499 + }, + { + "epoch": 0.375, + "grad_norm": 0.47942490711058733, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 37500 + }, + { + "epoch": 0.37501, + "grad_norm": 0.5410017155922114, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 37501 + }, + { + "epoch": 0.37502, + "grad_norm": 0.6266887156319505, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 37502 + }, + { + "epoch": 0.37503, + "grad_norm": 0.6069969746346214, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 37503 + }, + { + "epoch": 0.37504, + "grad_norm": 0.6063893531129788, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 37504 + }, + { + "epoch": 0.37505, + "grad_norm": 0.6675516940798849, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 37505 + }, + { + "epoch": 0.37506, + "grad_norm": 0.6905393957308059, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 37506 + }, + { + "epoch": 0.37507, + "grad_norm": 0.6725178569637036, + "learning_rate": 0.003, + "loss": 3.959, + "step": 37507 + }, + { + "epoch": 0.37508, + "grad_norm": 0.7224173117381232, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 37508 + }, + { + "epoch": 0.37509, + "grad_norm": 0.7135924151678457, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 37509 + }, + { + "epoch": 0.3751, + "grad_norm": 0.7613208290071297, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 37510 + }, + { + "epoch": 0.37511, + "grad_norm": 0.7445628699271816, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 37511 + }, + { + "epoch": 0.37512, + "grad_norm": 0.7292328724643606, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 37512 + }, + { + "epoch": 0.37513, + "grad_norm": 0.7284569032357863, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 37513 + }, + { + "epoch": 0.37514, + "grad_norm": 0.7941545521087524, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 37514 + }, + { + "epoch": 0.37515, + "grad_norm": 0.9555626062552246, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 37515 + }, + { + "epoch": 0.37516, + "grad_norm": 1.2699080005360888, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 37516 + }, + { + "epoch": 0.37517, + "grad_norm": 0.7291901973562469, + "learning_rate": 0.003, + "loss": 3.993, + "step": 37517 + }, + { + "epoch": 0.37518, + "grad_norm": 0.7200506672244228, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 37518 + }, + { + "epoch": 0.37519, + "grad_norm": 0.8241954916198887, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37519 + }, + { + "epoch": 0.3752, + "grad_norm": 0.885687524925398, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 37520 + }, + { + "epoch": 0.37521, + "grad_norm": 0.9885963943532603, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 37521 + }, + { + "epoch": 0.37522, + "grad_norm": 0.94475438856232, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 37522 + }, + { + "epoch": 0.37523, + "grad_norm": 0.994570931531351, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 37523 + }, + { + "epoch": 0.37524, + "grad_norm": 1.044653271424228, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 37524 + }, + { + "epoch": 0.37525, + "grad_norm": 1.0005434324875775, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 37525 + }, + { + "epoch": 0.37526, + "grad_norm": 1.064727409154718, + "learning_rate": 0.003, + "loss": 4.0539, + "step": 37526 + }, + { + "epoch": 0.37527, + "grad_norm": 0.9326772635808867, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 37527 + }, + { + "epoch": 0.37528, + "grad_norm": 0.8814456689660898, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 37528 + }, + { + "epoch": 0.37529, + "grad_norm": 0.7833397231446023, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 37529 + }, + { + "epoch": 0.3753, + "grad_norm": 0.7911068199359043, + "learning_rate": 0.003, + "loss": 3.996, + "step": 37530 + }, + { + "epoch": 0.37531, + "grad_norm": 0.8652744759789625, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 37531 + }, + { + "epoch": 0.37532, + "grad_norm": 0.8708493869748539, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 37532 + }, + { + "epoch": 0.37533, + "grad_norm": 0.96506183124248, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 37533 + }, + { + "epoch": 0.37534, + "grad_norm": 1.125188282589192, + "learning_rate": 0.003, + "loss": 4.0362, + "step": 37534 + }, + { + "epoch": 0.37535, + "grad_norm": 0.8175922310842288, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 37535 + }, + { + "epoch": 0.37536, + "grad_norm": 0.7773892683139042, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 37536 + }, + { + "epoch": 0.37537, + "grad_norm": 0.7186482483948315, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 37537 + }, + { + "epoch": 0.37538, + "grad_norm": 0.7749651844327817, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 37538 + }, + { + "epoch": 0.37539, + "grad_norm": 0.7942442698784332, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 37539 + }, + { + "epoch": 0.3754, + "grad_norm": 0.7994814308952909, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 37540 + }, + { + "epoch": 0.37541, + "grad_norm": 0.7502563353676109, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 37541 + }, + { + "epoch": 0.37542, + "grad_norm": 0.6843744918062599, + "learning_rate": 0.003, + "loss": 3.995, + "step": 37542 + }, + { + "epoch": 0.37543, + "grad_norm": 0.5491588708090013, + "learning_rate": 0.003, + "loss": 3.988, + "step": 37543 + }, + { + "epoch": 0.37544, + "grad_norm": 0.5980898125538066, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 37544 + }, + { + "epoch": 0.37545, + "grad_norm": 0.6983631097203208, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 37545 + }, + { + "epoch": 0.37546, + "grad_norm": 0.7187855234410777, + "learning_rate": 0.003, + "loss": 3.967, + "step": 37546 + }, + { + "epoch": 0.37547, + "grad_norm": 0.6583500069048308, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 37547 + }, + { + "epoch": 0.37548, + "grad_norm": 0.6245118576350976, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 37548 + }, + { + "epoch": 0.37549, + "grad_norm": 0.5757568868589783, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 37549 + }, + { + "epoch": 0.3755, + "grad_norm": 0.5982906097228816, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 37550 + }, + { + "epoch": 0.37551, + "grad_norm": 0.5974965725249125, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 37551 + }, + { + "epoch": 0.37552, + "grad_norm": 0.7066535513173714, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 37552 + }, + { + "epoch": 0.37553, + "grad_norm": 0.8195009035054723, + "learning_rate": 0.003, + "loss": 3.995, + "step": 37553 + }, + { + "epoch": 0.37554, + "grad_norm": 0.9533353774922597, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 37554 + }, + { + "epoch": 0.37555, + "grad_norm": 0.9287702972202824, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 37555 + }, + { + "epoch": 0.37556, + "grad_norm": 0.706817809878817, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 37556 + }, + { + "epoch": 0.37557, + "grad_norm": 0.780915315016022, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 37557 + }, + { + "epoch": 0.37558, + "grad_norm": 0.682849899055085, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 37558 + }, + { + "epoch": 0.37559, + "grad_norm": 0.6835666422626024, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 37559 + }, + { + "epoch": 0.3756, + "grad_norm": 0.7769639739567, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 37560 + }, + { + "epoch": 0.37561, + "grad_norm": 0.8529514208528357, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 37561 + }, + { + "epoch": 0.37562, + "grad_norm": 0.7632144096022576, + "learning_rate": 0.003, + "loss": 3.979, + "step": 37562 + }, + { + "epoch": 0.37563, + "grad_norm": 0.8921451039865895, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 37563 + }, + { + "epoch": 0.37564, + "grad_norm": 1.0787828429998947, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 37564 + }, + { + "epoch": 0.37565, + "grad_norm": 0.8598191161988294, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 37565 + }, + { + "epoch": 0.37566, + "grad_norm": 0.6717790097569921, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 37566 + }, + { + "epoch": 0.37567, + "grad_norm": 0.6942548111493001, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 37567 + }, + { + "epoch": 0.37568, + "grad_norm": 0.7747983206324404, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 37568 + }, + { + "epoch": 0.37569, + "grad_norm": 0.8041704603116417, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 37569 + }, + { + "epoch": 0.3757, + "grad_norm": 0.6534295944585151, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 37570 + }, + { + "epoch": 0.37571, + "grad_norm": 0.5446255355432879, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 37571 + }, + { + "epoch": 0.37572, + "grad_norm": 0.5250098782336887, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 37572 + }, + { + "epoch": 0.37573, + "grad_norm": 0.5808294136926647, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 37573 + }, + { + "epoch": 0.37574, + "grad_norm": 0.6710774996071235, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 37574 + }, + { + "epoch": 0.37575, + "grad_norm": 0.7670511303168616, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37575 + }, + { + "epoch": 0.37576, + "grad_norm": 0.8218847962054424, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 37576 + }, + { + "epoch": 0.37577, + "grad_norm": 0.6970919136289484, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 37577 + }, + { + "epoch": 0.37578, + "grad_norm": 0.5764434449606441, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 37578 + }, + { + "epoch": 0.37579, + "grad_norm": 0.6541158743118796, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 37579 + }, + { + "epoch": 0.3758, + "grad_norm": 0.7759856351622544, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 37580 + }, + { + "epoch": 0.37581, + "grad_norm": 0.9022082314541743, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 37581 + }, + { + "epoch": 0.37582, + "grad_norm": 1.020249636450374, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 37582 + }, + { + "epoch": 0.37583, + "grad_norm": 1.012794278051464, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 37583 + }, + { + "epoch": 0.37584, + "grad_norm": 0.9952047240759406, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 37584 + }, + { + "epoch": 0.37585, + "grad_norm": 1.035232368513082, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 37585 + }, + { + "epoch": 0.37586, + "grad_norm": 1.2096251451004814, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 37586 + }, + { + "epoch": 0.37587, + "grad_norm": 0.9731865464300764, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 37587 + }, + { + "epoch": 0.37588, + "grad_norm": 1.122897895011807, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 37588 + }, + { + "epoch": 0.37589, + "grad_norm": 0.9957576168725669, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 37589 + }, + { + "epoch": 0.3759, + "grad_norm": 1.0562185474796113, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 37590 + }, + { + "epoch": 0.37591, + "grad_norm": 0.9353503876704962, + "learning_rate": 0.003, + "loss": 3.993, + "step": 37591 + }, + { + "epoch": 0.37592, + "grad_norm": 0.8194747003752035, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 37592 + }, + { + "epoch": 0.37593, + "grad_norm": 0.751349085750394, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 37593 + }, + { + "epoch": 0.37594, + "grad_norm": 0.6397078405750574, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 37594 + }, + { + "epoch": 0.37595, + "grad_norm": 0.762888503939805, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 37595 + }, + { + "epoch": 0.37596, + "grad_norm": 0.8254778221489716, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 37596 + }, + { + "epoch": 0.37597, + "grad_norm": 0.8197824710098478, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 37597 + }, + { + "epoch": 0.37598, + "grad_norm": 0.6652876717089997, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 37598 + }, + { + "epoch": 0.37599, + "grad_norm": 0.6103737678598473, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 37599 + }, + { + "epoch": 0.376, + "grad_norm": 0.6129141598461, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 37600 + }, + { + "epoch": 0.37601, + "grad_norm": 0.572957938767319, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 37601 + }, + { + "epoch": 0.37602, + "grad_norm": 0.5907780013099033, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 37602 + }, + { + "epoch": 0.37603, + "grad_norm": 0.5966550608942868, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 37603 + }, + { + "epoch": 0.37604, + "grad_norm": 0.5977961911418912, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 37604 + }, + { + "epoch": 0.37605, + "grad_norm": 0.6852699677184776, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 37605 + }, + { + "epoch": 0.37606, + "grad_norm": 0.6581179476339979, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 37606 + }, + { + "epoch": 0.37607, + "grad_norm": 0.5562055120285148, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 37607 + }, + { + "epoch": 0.37608, + "grad_norm": 0.5140666388771046, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 37608 + }, + { + "epoch": 0.37609, + "grad_norm": 0.557228695395016, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 37609 + }, + { + "epoch": 0.3761, + "grad_norm": 0.5554853358648362, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 37610 + }, + { + "epoch": 0.37611, + "grad_norm": 0.588955280144161, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 37611 + }, + { + "epoch": 0.37612, + "grad_norm": 0.6333973774331685, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 37612 + }, + { + "epoch": 0.37613, + "grad_norm": 0.8094870104712669, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 37613 + }, + { + "epoch": 0.37614, + "grad_norm": 1.1654052511935242, + "learning_rate": 0.003, + "loss": 4.0207, + "step": 37614 + }, + { + "epoch": 0.37615, + "grad_norm": 1.0912647273279117, + "learning_rate": 0.003, + "loss": 3.992, + "step": 37615 + }, + { + "epoch": 0.37616, + "grad_norm": 0.7960610984826438, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 37616 + }, + { + "epoch": 0.37617, + "grad_norm": 0.6892947228045568, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 37617 + }, + { + "epoch": 0.37618, + "grad_norm": 0.7369120890409627, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 37618 + }, + { + "epoch": 0.37619, + "grad_norm": 0.7980748023439, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 37619 + }, + { + "epoch": 0.3762, + "grad_norm": 0.9282966502589589, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 37620 + }, + { + "epoch": 0.37621, + "grad_norm": 0.9684905089791018, + "learning_rate": 0.003, + "loss": 3.981, + "step": 37621 + }, + { + "epoch": 0.37622, + "grad_norm": 0.9392250970633965, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 37622 + }, + { + "epoch": 0.37623, + "grad_norm": 0.8539235742057135, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 37623 + }, + { + "epoch": 0.37624, + "grad_norm": 0.6295197260251323, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 37624 + }, + { + "epoch": 0.37625, + "grad_norm": 0.636668128623914, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 37625 + }, + { + "epoch": 0.37626, + "grad_norm": 0.7611649013802346, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 37626 + }, + { + "epoch": 0.37627, + "grad_norm": 0.8944421618756662, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 37627 + }, + { + "epoch": 0.37628, + "grad_norm": 0.8877311435130922, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 37628 + }, + { + "epoch": 0.37629, + "grad_norm": 0.7814934635815074, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 37629 + }, + { + "epoch": 0.3763, + "grad_norm": 0.8186343737447455, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 37630 + }, + { + "epoch": 0.37631, + "grad_norm": 0.7682682563671901, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 37631 + }, + { + "epoch": 0.37632, + "grad_norm": 0.8119285681669176, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 37632 + }, + { + "epoch": 0.37633, + "grad_norm": 0.9897863175133467, + "learning_rate": 0.003, + "loss": 3.98, + "step": 37633 + }, + { + "epoch": 0.37634, + "grad_norm": 1.0215271819011316, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 37634 + }, + { + "epoch": 0.37635, + "grad_norm": 1.110912504256488, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 37635 + }, + { + "epoch": 0.37636, + "grad_norm": 0.9456697242111394, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 37636 + }, + { + "epoch": 0.37637, + "grad_norm": 0.9042153964224602, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 37637 + }, + { + "epoch": 0.37638, + "grad_norm": 0.7804948094428329, + "learning_rate": 0.003, + "loss": 3.998, + "step": 37638 + }, + { + "epoch": 0.37639, + "grad_norm": 0.8301971019293063, + "learning_rate": 0.003, + "loss": 3.971, + "step": 37639 + }, + { + "epoch": 0.3764, + "grad_norm": 0.6762554073093059, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 37640 + }, + { + "epoch": 0.37641, + "grad_norm": 0.6517210573266736, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 37641 + }, + { + "epoch": 0.37642, + "grad_norm": 0.669419207639262, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 37642 + }, + { + "epoch": 0.37643, + "grad_norm": 0.6619188715544693, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 37643 + }, + { + "epoch": 0.37644, + "grad_norm": 0.6307247069047152, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 37644 + }, + { + "epoch": 0.37645, + "grad_norm": 0.6986857343828898, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 37645 + }, + { + "epoch": 0.37646, + "grad_norm": 0.7579464940185475, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 37646 + }, + { + "epoch": 0.37647, + "grad_norm": 0.8936035870971202, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 37647 + }, + { + "epoch": 0.37648, + "grad_norm": 1.0710524818881337, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 37648 + }, + { + "epoch": 0.37649, + "grad_norm": 0.9722551375282845, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 37649 + }, + { + "epoch": 0.3765, + "grad_norm": 0.7573531251313054, + "learning_rate": 0.003, + "loss": 4.0, + "step": 37650 + }, + { + "epoch": 0.37651, + "grad_norm": 0.6552160610693785, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 37651 + }, + { + "epoch": 0.37652, + "grad_norm": 0.6561800818443376, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 37652 + }, + { + "epoch": 0.37653, + "grad_norm": 0.7012412367120938, + "learning_rate": 0.003, + "loss": 3.968, + "step": 37653 + }, + { + "epoch": 0.37654, + "grad_norm": 0.7297483881134879, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 37654 + }, + { + "epoch": 0.37655, + "grad_norm": 0.6679752326421335, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 37655 + }, + { + "epoch": 0.37656, + "grad_norm": 0.6144330725196042, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 37656 + }, + { + "epoch": 0.37657, + "grad_norm": 0.6707455486377908, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 37657 + }, + { + "epoch": 0.37658, + "grad_norm": 0.6394325301568832, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 37658 + }, + { + "epoch": 0.37659, + "grad_norm": 0.631848258953076, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 37659 + }, + { + "epoch": 0.3766, + "grad_norm": 0.7194158280055424, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 37660 + }, + { + "epoch": 0.37661, + "grad_norm": 0.8533927084894468, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 37661 + }, + { + "epoch": 0.37662, + "grad_norm": 0.9781013603672237, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 37662 + }, + { + "epoch": 0.37663, + "grad_norm": 0.9988519467905981, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 37663 + }, + { + "epoch": 0.37664, + "grad_norm": 0.9742207306117695, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 37664 + }, + { + "epoch": 0.37665, + "grad_norm": 0.8388203215094868, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 37665 + }, + { + "epoch": 0.37666, + "grad_norm": 0.8136384336530343, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 37666 + }, + { + "epoch": 0.37667, + "grad_norm": 0.6729312140454653, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 37667 + }, + { + "epoch": 0.37668, + "grad_norm": 0.5997924205103363, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 37668 + }, + { + "epoch": 0.37669, + "grad_norm": 0.5259025848744172, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 37669 + }, + { + "epoch": 0.3767, + "grad_norm": 0.5846026004606122, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 37670 + }, + { + "epoch": 0.37671, + "grad_norm": 0.5697912184704432, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 37671 + }, + { + "epoch": 0.37672, + "grad_norm": 0.6093409309486736, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 37672 + }, + { + "epoch": 0.37673, + "grad_norm": 0.6807838898742328, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 37673 + }, + { + "epoch": 0.37674, + "grad_norm": 0.8844249938460842, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 37674 + }, + { + "epoch": 0.37675, + "grad_norm": 1.1455295019563383, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 37675 + }, + { + "epoch": 0.37676, + "grad_norm": 0.7083025260710485, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 37676 + }, + { + "epoch": 0.37677, + "grad_norm": 0.5840343850400623, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 37677 + }, + { + "epoch": 0.37678, + "grad_norm": 0.637459540836412, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 37678 + }, + { + "epoch": 0.37679, + "grad_norm": 0.7083365240652156, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 37679 + }, + { + "epoch": 0.3768, + "grad_norm": 0.7408254558952887, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 37680 + }, + { + "epoch": 0.37681, + "grad_norm": 0.7710513136660371, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 37681 + }, + { + "epoch": 0.37682, + "grad_norm": 0.6279541443977413, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 37682 + }, + { + "epoch": 0.37683, + "grad_norm": 0.6093239146852881, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 37683 + }, + { + "epoch": 0.37684, + "grad_norm": 0.5768179366514272, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 37684 + }, + { + "epoch": 0.37685, + "grad_norm": 0.649993895849267, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 37685 + }, + { + "epoch": 0.37686, + "grad_norm": 0.683662641450512, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 37686 + }, + { + "epoch": 0.37687, + "grad_norm": 0.8267845999742509, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 37687 + }, + { + "epoch": 0.37688, + "grad_norm": 0.9175336585210522, + "learning_rate": 0.003, + "loss": 3.979, + "step": 37688 + }, + { + "epoch": 0.37689, + "grad_norm": 0.9146836779206693, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 37689 + }, + { + "epoch": 0.3769, + "grad_norm": 0.8425561367088785, + "learning_rate": 0.003, + "loss": 3.987, + "step": 37690 + }, + { + "epoch": 0.37691, + "grad_norm": 0.7743803056697719, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 37691 + }, + { + "epoch": 0.37692, + "grad_norm": 0.7404425953051121, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 37692 + }, + { + "epoch": 0.37693, + "grad_norm": 0.9784028046198953, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 37693 + }, + { + "epoch": 0.37694, + "grad_norm": 1.0682480183601362, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 37694 + }, + { + "epoch": 0.37695, + "grad_norm": 0.8866468552547515, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 37695 + }, + { + "epoch": 0.37696, + "grad_norm": 1.0136399129473164, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 37696 + }, + { + "epoch": 0.37697, + "grad_norm": 1.179660427940457, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 37697 + }, + { + "epoch": 0.37698, + "grad_norm": 0.9406282567659788, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 37698 + }, + { + "epoch": 0.37699, + "grad_norm": 0.9654978483243494, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 37699 + }, + { + "epoch": 0.377, + "grad_norm": 0.8978582730898722, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 37700 + }, + { + "epoch": 0.37701, + "grad_norm": 0.9287821092225419, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 37701 + }, + { + "epoch": 0.37702, + "grad_norm": 0.9630711404913033, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 37702 + }, + { + "epoch": 0.37703, + "grad_norm": 1.146845809091769, + "learning_rate": 0.003, + "loss": 3.996, + "step": 37703 + }, + { + "epoch": 0.37704, + "grad_norm": 0.9915738463205419, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 37704 + }, + { + "epoch": 0.37705, + "grad_norm": 0.9815049464794681, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 37705 + }, + { + "epoch": 0.37706, + "grad_norm": 0.91890926639466, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 37706 + }, + { + "epoch": 0.37707, + "grad_norm": 0.9135981592086554, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 37707 + }, + { + "epoch": 0.37708, + "grad_norm": 0.8710937006316204, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 37708 + }, + { + "epoch": 0.37709, + "grad_norm": 0.8459638639437944, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 37709 + }, + { + "epoch": 0.3771, + "grad_norm": 0.8001944000652103, + "learning_rate": 0.003, + "loss": 3.991, + "step": 37710 + }, + { + "epoch": 0.37711, + "grad_norm": 0.7243342108882026, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 37711 + }, + { + "epoch": 0.37712, + "grad_norm": 0.7620348628112503, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 37712 + }, + { + "epoch": 0.37713, + "grad_norm": 0.804111147368821, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 37713 + }, + { + "epoch": 0.37714, + "grad_norm": 0.7063789710236513, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 37714 + }, + { + "epoch": 0.37715, + "grad_norm": 0.60780335743029, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 37715 + }, + { + "epoch": 0.37716, + "grad_norm": 0.5622180527646446, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 37716 + }, + { + "epoch": 0.37717, + "grad_norm": 0.55285750215725, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 37717 + }, + { + "epoch": 0.37718, + "grad_norm": 0.4808100275206687, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 37718 + }, + { + "epoch": 0.37719, + "grad_norm": 0.48001380135846355, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 37719 + }, + { + "epoch": 0.3772, + "grad_norm": 0.47027519296599823, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 37720 + }, + { + "epoch": 0.37721, + "grad_norm": 0.4528848333127901, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 37721 + }, + { + "epoch": 0.37722, + "grad_norm": 0.42144729273179077, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 37722 + }, + { + "epoch": 0.37723, + "grad_norm": 0.4232470710314611, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 37723 + }, + { + "epoch": 0.37724, + "grad_norm": 0.43945855406204704, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 37724 + }, + { + "epoch": 0.37725, + "grad_norm": 0.43725567013844363, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 37725 + }, + { + "epoch": 0.37726, + "grad_norm": 0.47739737759773965, + "learning_rate": 0.003, + "loss": 3.9364, + "step": 37726 + }, + { + "epoch": 0.37727, + "grad_norm": 0.5778850061148421, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 37727 + }, + { + "epoch": 0.37728, + "grad_norm": 0.6859141278440767, + "learning_rate": 0.003, + "loss": 3.9463, + "step": 37728 + }, + { + "epoch": 0.37729, + "grad_norm": 0.8515749440183192, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 37729 + }, + { + "epoch": 0.3773, + "grad_norm": 1.1265483209793057, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 37730 + }, + { + "epoch": 0.37731, + "grad_norm": 0.932457497426721, + "learning_rate": 0.003, + "loss": 3.97, + "step": 37731 + }, + { + "epoch": 0.37732, + "grad_norm": 0.8472521620350162, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 37732 + }, + { + "epoch": 0.37733, + "grad_norm": 0.8492331549790643, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 37733 + }, + { + "epoch": 0.37734, + "grad_norm": 0.8449581797947622, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 37734 + }, + { + "epoch": 0.37735, + "grad_norm": 0.7757937042007832, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 37735 + }, + { + "epoch": 0.37736, + "grad_norm": 0.7557787333842608, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 37736 + }, + { + "epoch": 0.37737, + "grad_norm": 0.7550404082859846, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 37737 + }, + { + "epoch": 0.37738, + "grad_norm": 0.828005919574937, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 37738 + }, + { + "epoch": 0.37739, + "grad_norm": 0.919669152313108, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 37739 + }, + { + "epoch": 0.3774, + "grad_norm": 0.9138978034878885, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 37740 + }, + { + "epoch": 0.37741, + "grad_norm": 1.1670882712042376, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 37741 + }, + { + "epoch": 0.37742, + "grad_norm": 0.899212213276405, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 37742 + }, + { + "epoch": 0.37743, + "grad_norm": 1.0543032210680687, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 37743 + }, + { + "epoch": 0.37744, + "grad_norm": 1.0189338663526413, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 37744 + }, + { + "epoch": 0.37745, + "grad_norm": 1.0656932453446357, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 37745 + }, + { + "epoch": 0.37746, + "grad_norm": 1.0449127790960697, + "learning_rate": 0.003, + "loss": 4.038, + "step": 37746 + }, + { + "epoch": 0.37747, + "grad_norm": 0.9322228208802472, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 37747 + }, + { + "epoch": 0.37748, + "grad_norm": 0.7850715607544348, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 37748 + }, + { + "epoch": 0.37749, + "grad_norm": 0.758616489016226, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 37749 + }, + { + "epoch": 0.3775, + "grad_norm": 0.811351751736349, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 37750 + }, + { + "epoch": 0.37751, + "grad_norm": 0.8318998586171474, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 37751 + }, + { + "epoch": 0.37752, + "grad_norm": 0.8082939101877747, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 37752 + }, + { + "epoch": 0.37753, + "grad_norm": 0.7509677699510824, + "learning_rate": 0.003, + "loss": 3.985, + "step": 37753 + }, + { + "epoch": 0.37754, + "grad_norm": 0.6782329515425797, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 37754 + }, + { + "epoch": 0.37755, + "grad_norm": 0.7140135969006861, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 37755 + }, + { + "epoch": 0.37756, + "grad_norm": 0.7005357159433724, + "learning_rate": 0.003, + "loss": 3.978, + "step": 37756 + }, + { + "epoch": 0.37757, + "grad_norm": 0.6699749806189571, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 37757 + }, + { + "epoch": 0.37758, + "grad_norm": 0.7030013869530659, + "learning_rate": 0.003, + "loss": 3.986, + "step": 37758 + }, + { + "epoch": 0.37759, + "grad_norm": 0.76978256969331, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 37759 + }, + { + "epoch": 0.3776, + "grad_norm": 0.8274160093397616, + "learning_rate": 0.003, + "loss": 3.997, + "step": 37760 + }, + { + "epoch": 0.37761, + "grad_norm": 0.8352994213266255, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 37761 + }, + { + "epoch": 0.37762, + "grad_norm": 0.8606319943452992, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 37762 + }, + { + "epoch": 0.37763, + "grad_norm": 0.7275113349946178, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 37763 + }, + { + "epoch": 0.37764, + "grad_norm": 0.6391493298280895, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 37764 + }, + { + "epoch": 0.37765, + "grad_norm": 0.697980962978027, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 37765 + }, + { + "epoch": 0.37766, + "grad_norm": 0.6144116227848626, + "learning_rate": 0.003, + "loss": 4.005, + "step": 37766 + }, + { + "epoch": 0.37767, + "grad_norm": 0.5657918124838077, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 37767 + }, + { + "epoch": 0.37768, + "grad_norm": 0.6176245107512606, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 37768 + }, + { + "epoch": 0.37769, + "grad_norm": 0.5757352266949798, + "learning_rate": 0.003, + "loss": 3.979, + "step": 37769 + }, + { + "epoch": 0.3777, + "grad_norm": 0.4860441661946625, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 37770 + }, + { + "epoch": 0.37771, + "grad_norm": 0.43916512012564457, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 37771 + }, + { + "epoch": 0.37772, + "grad_norm": 0.4557565856193411, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 37772 + }, + { + "epoch": 0.37773, + "grad_norm": 0.5325333971259022, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 37773 + }, + { + "epoch": 0.37774, + "grad_norm": 0.5842510854870503, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 37774 + }, + { + "epoch": 0.37775, + "grad_norm": 0.6287075101891108, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 37775 + }, + { + "epoch": 0.37776, + "grad_norm": 0.6827862281282577, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 37776 + }, + { + "epoch": 0.37777, + "grad_norm": 0.7395103571327696, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 37777 + }, + { + "epoch": 0.37778, + "grad_norm": 0.8595103328929242, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 37778 + }, + { + "epoch": 0.37779, + "grad_norm": 0.91068768810323, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 37779 + }, + { + "epoch": 0.3778, + "grad_norm": 0.9130485865957447, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 37780 + }, + { + "epoch": 0.37781, + "grad_norm": 0.9257368304469504, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 37781 + }, + { + "epoch": 0.37782, + "grad_norm": 0.8195741799990969, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 37782 + }, + { + "epoch": 0.37783, + "grad_norm": 1.0758971962959885, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 37783 + }, + { + "epoch": 0.37784, + "grad_norm": 0.9701824919944285, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 37784 + }, + { + "epoch": 0.37785, + "grad_norm": 0.903022143092857, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 37785 + }, + { + "epoch": 0.37786, + "grad_norm": 0.9623300028757145, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 37786 + }, + { + "epoch": 0.37787, + "grad_norm": 0.9737887673494513, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 37787 + }, + { + "epoch": 0.37788, + "grad_norm": 0.9945231692567551, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 37788 + }, + { + "epoch": 0.37789, + "grad_norm": 0.9185156323064324, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 37789 + }, + { + "epoch": 0.3779, + "grad_norm": 0.9799865457381863, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 37790 + }, + { + "epoch": 0.37791, + "grad_norm": 1.1454960211535583, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 37791 + }, + { + "epoch": 0.37792, + "grad_norm": 0.9505174252273345, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 37792 + }, + { + "epoch": 0.37793, + "grad_norm": 0.999436099484006, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 37793 + }, + { + "epoch": 0.37794, + "grad_norm": 0.817256317002761, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 37794 + }, + { + "epoch": 0.37795, + "grad_norm": 0.7946627346509123, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 37795 + }, + { + "epoch": 0.37796, + "grad_norm": 0.7044778044403943, + "learning_rate": 0.003, + "loss": 4.0361, + "step": 37796 + }, + { + "epoch": 0.37797, + "grad_norm": 0.7468914441103848, + "learning_rate": 0.003, + "loss": 3.997, + "step": 37797 + }, + { + "epoch": 0.37798, + "grad_norm": 0.6653653865684904, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 37798 + }, + { + "epoch": 0.37799, + "grad_norm": 0.674685591004292, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 37799 + }, + { + "epoch": 0.378, + "grad_norm": 0.7758556234611457, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 37800 + }, + { + "epoch": 0.37801, + "grad_norm": 0.8228195894199913, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 37801 + }, + { + "epoch": 0.37802, + "grad_norm": 0.9235441596845545, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 37802 + }, + { + "epoch": 0.37803, + "grad_norm": 1.0435601544939344, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 37803 + }, + { + "epoch": 0.37804, + "grad_norm": 0.9412852687805469, + "learning_rate": 0.003, + "loss": 4.002, + "step": 37804 + }, + { + "epoch": 0.37805, + "grad_norm": 0.9347262889470677, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 37805 + }, + { + "epoch": 0.37806, + "grad_norm": 0.8077807736812975, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 37806 + }, + { + "epoch": 0.37807, + "grad_norm": 0.7064849869403987, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 37807 + }, + { + "epoch": 0.37808, + "grad_norm": 0.7031613323562076, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 37808 + }, + { + "epoch": 0.37809, + "grad_norm": 0.6896553348037956, + "learning_rate": 0.003, + "loss": 3.991, + "step": 37809 + }, + { + "epoch": 0.3781, + "grad_norm": 0.6468813963357081, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 37810 + }, + { + "epoch": 0.37811, + "grad_norm": 0.5919489355671707, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 37811 + }, + { + "epoch": 0.37812, + "grad_norm": 0.5515563171395352, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 37812 + }, + { + "epoch": 0.37813, + "grad_norm": 0.5928150566998326, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 37813 + }, + { + "epoch": 0.37814, + "grad_norm": 0.6667434389994062, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 37814 + }, + { + "epoch": 0.37815, + "grad_norm": 0.6648556782114546, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 37815 + }, + { + "epoch": 0.37816, + "grad_norm": 0.6073618210948812, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 37816 + }, + { + "epoch": 0.37817, + "grad_norm": 0.629646454480699, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 37817 + }, + { + "epoch": 0.37818, + "grad_norm": 0.7987048319284403, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 37818 + }, + { + "epoch": 0.37819, + "grad_norm": 0.9186425908495656, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 37819 + }, + { + "epoch": 0.3782, + "grad_norm": 1.0387604984919496, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 37820 + }, + { + "epoch": 0.37821, + "grad_norm": 0.9072094757890167, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 37821 + }, + { + "epoch": 0.37822, + "grad_norm": 0.9372780966677112, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 37822 + }, + { + "epoch": 0.37823, + "grad_norm": 0.8616390694615782, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 37823 + }, + { + "epoch": 0.37824, + "grad_norm": 0.7356884613909451, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 37824 + }, + { + "epoch": 0.37825, + "grad_norm": 0.7437002864478125, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 37825 + }, + { + "epoch": 0.37826, + "grad_norm": 0.8298801351634448, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 37826 + }, + { + "epoch": 0.37827, + "grad_norm": 0.8898339975946483, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 37827 + }, + { + "epoch": 0.37828, + "grad_norm": 0.948433543462478, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 37828 + }, + { + "epoch": 0.37829, + "grad_norm": 1.1777260786623964, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 37829 + }, + { + "epoch": 0.3783, + "grad_norm": 0.893474251722228, + "learning_rate": 0.003, + "loss": 4.003, + "step": 37830 + }, + { + "epoch": 0.37831, + "grad_norm": 0.7310430232785141, + "learning_rate": 0.003, + "loss": 4.003, + "step": 37831 + }, + { + "epoch": 0.37832, + "grad_norm": 0.6904026784599502, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 37832 + }, + { + "epoch": 0.37833, + "grad_norm": 0.7290148994610548, + "learning_rate": 0.003, + "loss": 4.005, + "step": 37833 + }, + { + "epoch": 0.37834, + "grad_norm": 0.775011623383399, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 37834 + }, + { + "epoch": 0.37835, + "grad_norm": 0.8181534963599199, + "learning_rate": 0.003, + "loss": 3.98, + "step": 37835 + }, + { + "epoch": 0.37836, + "grad_norm": 0.8451439093829015, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 37836 + }, + { + "epoch": 0.37837, + "grad_norm": 0.9056418697428571, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 37837 + }, + { + "epoch": 0.37838, + "grad_norm": 0.9716998941264041, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 37838 + }, + { + "epoch": 0.37839, + "grad_norm": 1.0054962261325455, + "learning_rate": 0.003, + "loss": 3.989, + "step": 37839 + }, + { + "epoch": 0.3784, + "grad_norm": 0.882632588464395, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 37840 + }, + { + "epoch": 0.37841, + "grad_norm": 0.7693692735610547, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 37841 + }, + { + "epoch": 0.37842, + "grad_norm": 0.7472826480143309, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 37842 + }, + { + "epoch": 0.37843, + "grad_norm": 0.7453380713679216, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 37843 + }, + { + "epoch": 0.37844, + "grad_norm": 0.7892084571973835, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 37844 + }, + { + "epoch": 0.37845, + "grad_norm": 0.8739331894102118, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 37845 + }, + { + "epoch": 0.37846, + "grad_norm": 0.8746957307625215, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 37846 + }, + { + "epoch": 0.37847, + "grad_norm": 0.888988882706778, + "learning_rate": 0.003, + "loss": 4.013, + "step": 37847 + }, + { + "epoch": 0.37848, + "grad_norm": 0.806313121922836, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 37848 + }, + { + "epoch": 0.37849, + "grad_norm": 0.7517268892177343, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 37849 + }, + { + "epoch": 0.3785, + "grad_norm": 0.760281903208848, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 37850 + }, + { + "epoch": 0.37851, + "grad_norm": 0.743396820405079, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 37851 + }, + { + "epoch": 0.37852, + "grad_norm": 0.7897308211703274, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 37852 + }, + { + "epoch": 0.37853, + "grad_norm": 0.8084948814937445, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 37853 + }, + { + "epoch": 0.37854, + "grad_norm": 0.6571506414457826, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 37854 + }, + { + "epoch": 0.37855, + "grad_norm": 0.5529861582063079, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 37855 + }, + { + "epoch": 0.37856, + "grad_norm": 0.4976848689278767, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 37856 + }, + { + "epoch": 0.37857, + "grad_norm": 0.5963880878903028, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 37857 + }, + { + "epoch": 0.37858, + "grad_norm": 0.6078839667973482, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 37858 + }, + { + "epoch": 0.37859, + "grad_norm": 0.6379997539138692, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 37859 + }, + { + "epoch": 0.3786, + "grad_norm": 0.7205644280573656, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 37860 + }, + { + "epoch": 0.37861, + "grad_norm": 0.8203350815815349, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 37861 + }, + { + "epoch": 0.37862, + "grad_norm": 0.9002420824796343, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 37862 + }, + { + "epoch": 0.37863, + "grad_norm": 0.9231471985730564, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 37863 + }, + { + "epoch": 0.37864, + "grad_norm": 0.9570480143359696, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 37864 + }, + { + "epoch": 0.37865, + "grad_norm": 0.8179359262638501, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 37865 + }, + { + "epoch": 0.37866, + "grad_norm": 0.70844894376593, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 37866 + }, + { + "epoch": 0.37867, + "grad_norm": 0.7538091974102565, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 37867 + }, + { + "epoch": 0.37868, + "grad_norm": 0.6726981735954163, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 37868 + }, + { + "epoch": 0.37869, + "grad_norm": 0.673616362959558, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 37869 + }, + { + "epoch": 0.3787, + "grad_norm": 0.8043661366271967, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 37870 + }, + { + "epoch": 0.37871, + "grad_norm": 0.6648407764619257, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 37871 + }, + { + "epoch": 0.37872, + "grad_norm": 0.6011557106549654, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 37872 + }, + { + "epoch": 0.37873, + "grad_norm": 0.6066471541672047, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 37873 + }, + { + "epoch": 0.37874, + "grad_norm": 0.5988169983490612, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 37874 + }, + { + "epoch": 0.37875, + "grad_norm": 0.6612992786865874, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 37875 + }, + { + "epoch": 0.37876, + "grad_norm": 0.7524044192808187, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 37876 + }, + { + "epoch": 0.37877, + "grad_norm": 0.7094213567008241, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 37877 + }, + { + "epoch": 0.37878, + "grad_norm": 0.809337284329775, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 37878 + }, + { + "epoch": 0.37879, + "grad_norm": 1.0328994358407388, + "learning_rate": 0.003, + "loss": 4.0, + "step": 37879 + }, + { + "epoch": 0.3788, + "grad_norm": 1.1095765982214179, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 37880 + }, + { + "epoch": 0.37881, + "grad_norm": 0.959492738328909, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 37881 + }, + { + "epoch": 0.37882, + "grad_norm": 0.8863466367315679, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 37882 + }, + { + "epoch": 0.37883, + "grad_norm": 0.793201807604062, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 37883 + }, + { + "epoch": 0.37884, + "grad_norm": 0.7769619708124925, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 37884 + }, + { + "epoch": 0.37885, + "grad_norm": 0.899125584102146, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 37885 + }, + { + "epoch": 0.37886, + "grad_norm": 1.043662261540568, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 37886 + }, + { + "epoch": 0.37887, + "grad_norm": 0.9528164368399423, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 37887 + }, + { + "epoch": 0.37888, + "grad_norm": 0.8170564927225548, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 37888 + }, + { + "epoch": 0.37889, + "grad_norm": 0.7745346920245966, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 37889 + }, + { + "epoch": 0.3789, + "grad_norm": 0.8778626357370708, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 37890 + }, + { + "epoch": 0.37891, + "grad_norm": 0.9299785889239963, + "learning_rate": 0.003, + "loss": 4.0366, + "step": 37891 + }, + { + "epoch": 0.37892, + "grad_norm": 1.0656201751897136, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 37892 + }, + { + "epoch": 0.37893, + "grad_norm": 0.9752926402659042, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 37893 + }, + { + "epoch": 0.37894, + "grad_norm": 0.9441796649365259, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 37894 + }, + { + "epoch": 0.37895, + "grad_norm": 0.8865183664517561, + "learning_rate": 0.003, + "loss": 4.001, + "step": 37895 + }, + { + "epoch": 0.37896, + "grad_norm": 0.7415532888881353, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 37896 + }, + { + "epoch": 0.37897, + "grad_norm": 0.7158053515864566, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 37897 + }, + { + "epoch": 0.37898, + "grad_norm": 0.7094094554124685, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 37898 + }, + { + "epoch": 0.37899, + "grad_norm": 0.7774476682506718, + "learning_rate": 0.003, + "loss": 3.989, + "step": 37899 + }, + { + "epoch": 0.379, + "grad_norm": 0.834525019675757, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 37900 + }, + { + "epoch": 0.37901, + "grad_norm": 0.7759224726855993, + "learning_rate": 0.003, + "loss": 3.976, + "step": 37901 + }, + { + "epoch": 0.37902, + "grad_norm": 0.8100996204897122, + "learning_rate": 0.003, + "loss": 3.997, + "step": 37902 + }, + { + "epoch": 0.37903, + "grad_norm": 0.6929512857940097, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 37903 + }, + { + "epoch": 0.37904, + "grad_norm": 0.5649434943212999, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 37904 + }, + { + "epoch": 0.37905, + "grad_norm": 0.62737111366098, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 37905 + }, + { + "epoch": 0.37906, + "grad_norm": 0.6497958890483849, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 37906 + }, + { + "epoch": 0.37907, + "grad_norm": 0.6917457064190997, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 37907 + }, + { + "epoch": 0.37908, + "grad_norm": 0.7000363079164226, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 37908 + }, + { + "epoch": 0.37909, + "grad_norm": 0.7782095908883413, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 37909 + }, + { + "epoch": 0.3791, + "grad_norm": 0.9807447087223495, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 37910 + }, + { + "epoch": 0.37911, + "grad_norm": 1.162016326076343, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 37911 + }, + { + "epoch": 0.37912, + "grad_norm": 0.6927083587278285, + "learning_rate": 0.003, + "loss": 3.985, + "step": 37912 + }, + { + "epoch": 0.37913, + "grad_norm": 0.5848933544581498, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 37913 + }, + { + "epoch": 0.37914, + "grad_norm": 0.6521062716543771, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 37914 + }, + { + "epoch": 0.37915, + "grad_norm": 0.6881418066053108, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 37915 + }, + { + "epoch": 0.37916, + "grad_norm": 0.6328211972486141, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 37916 + }, + { + "epoch": 0.37917, + "grad_norm": 0.5990493158307689, + "learning_rate": 0.003, + "loss": 3.982, + "step": 37917 + }, + { + "epoch": 0.37918, + "grad_norm": 0.636213357514235, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 37918 + }, + { + "epoch": 0.37919, + "grad_norm": 0.5839596588787119, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 37919 + }, + { + "epoch": 0.3792, + "grad_norm": 0.6394119892371762, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 37920 + }, + { + "epoch": 0.37921, + "grad_norm": 0.6205332334567891, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 37921 + }, + { + "epoch": 0.37922, + "grad_norm": 0.59048568612785, + "learning_rate": 0.003, + "loss": 3.983, + "step": 37922 + }, + { + "epoch": 0.37923, + "grad_norm": 0.6288523546488699, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 37923 + }, + { + "epoch": 0.37924, + "grad_norm": 0.7212908894561169, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 37924 + }, + { + "epoch": 0.37925, + "grad_norm": 0.9340633400215055, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 37925 + }, + { + "epoch": 0.37926, + "grad_norm": 1.1308749340436144, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 37926 + }, + { + "epoch": 0.37927, + "grad_norm": 0.6853531807139027, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 37927 + }, + { + "epoch": 0.37928, + "grad_norm": 0.8358229859520456, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 37928 + }, + { + "epoch": 0.37929, + "grad_norm": 0.9247681006270237, + "learning_rate": 0.003, + "loss": 3.943, + "step": 37929 + }, + { + "epoch": 0.3793, + "grad_norm": 0.940978682442885, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 37930 + }, + { + "epoch": 0.37931, + "grad_norm": 0.8300718830724751, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 37931 + }, + { + "epoch": 0.37932, + "grad_norm": 0.8538110259692988, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 37932 + }, + { + "epoch": 0.37933, + "grad_norm": 0.821396437355987, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 37933 + }, + { + "epoch": 0.37934, + "grad_norm": 0.9368115187229461, + "learning_rate": 0.003, + "loss": 3.978, + "step": 37934 + }, + { + "epoch": 0.37935, + "grad_norm": 0.9794429125168955, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 37935 + }, + { + "epoch": 0.37936, + "grad_norm": 1.0563428223534457, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 37936 + }, + { + "epoch": 0.37937, + "grad_norm": 0.8440014693229381, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 37937 + }, + { + "epoch": 0.37938, + "grad_norm": 0.7919589316422403, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 37938 + }, + { + "epoch": 0.37939, + "grad_norm": 0.7496153510829137, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 37939 + }, + { + "epoch": 0.3794, + "grad_norm": 0.7117378959040088, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 37940 + }, + { + "epoch": 0.37941, + "grad_norm": 0.7683327294280534, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 37941 + }, + { + "epoch": 0.37942, + "grad_norm": 0.7511689810514286, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 37942 + }, + { + "epoch": 0.37943, + "grad_norm": 0.8611873178108461, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 37943 + }, + { + "epoch": 0.37944, + "grad_norm": 0.8917929772138319, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 37944 + }, + { + "epoch": 0.37945, + "grad_norm": 1.0178421969642812, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 37945 + }, + { + "epoch": 0.37946, + "grad_norm": 1.136111667438351, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 37946 + }, + { + "epoch": 0.37947, + "grad_norm": 0.9175655137608808, + "learning_rate": 0.003, + "loss": 3.987, + "step": 37947 + }, + { + "epoch": 0.37948, + "grad_norm": 0.7724225994766251, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 37948 + }, + { + "epoch": 0.37949, + "grad_norm": 0.6130411041074276, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 37949 + }, + { + "epoch": 0.3795, + "grad_norm": 0.5630658560625095, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 37950 + }, + { + "epoch": 0.37951, + "grad_norm": 0.49893867298040934, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 37951 + }, + { + "epoch": 0.37952, + "grad_norm": 0.5129282661625038, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 37952 + }, + { + "epoch": 0.37953, + "grad_norm": 0.5090091531183553, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 37953 + }, + { + "epoch": 0.37954, + "grad_norm": 0.5198586631088649, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 37954 + }, + { + "epoch": 0.37955, + "grad_norm": 0.5893485434560322, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 37955 + }, + { + "epoch": 0.37956, + "grad_norm": 0.6731253033842587, + "learning_rate": 0.003, + "loss": 3.992, + "step": 37956 + }, + { + "epoch": 0.37957, + "grad_norm": 0.8102044113895517, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 37957 + }, + { + "epoch": 0.37958, + "grad_norm": 0.9207522726116862, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 37958 + }, + { + "epoch": 0.37959, + "grad_norm": 0.6882127731075938, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 37959 + }, + { + "epoch": 0.3796, + "grad_norm": 0.5099101947837342, + "learning_rate": 0.003, + "loss": 3.967, + "step": 37960 + }, + { + "epoch": 0.37961, + "grad_norm": 0.6089041836406105, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 37961 + }, + { + "epoch": 0.37962, + "grad_norm": 0.6926149755692004, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 37962 + }, + { + "epoch": 0.37963, + "grad_norm": 0.7218487390464181, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 37963 + }, + { + "epoch": 0.37964, + "grad_norm": 0.7648058273798769, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 37964 + }, + { + "epoch": 0.37965, + "grad_norm": 0.853439127470998, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 37965 + }, + { + "epoch": 0.37966, + "grad_norm": 0.8277019486332082, + "learning_rate": 0.003, + "loss": 3.99, + "step": 37966 + }, + { + "epoch": 0.37967, + "grad_norm": 0.8828734804849827, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 37967 + }, + { + "epoch": 0.37968, + "grad_norm": 0.8940703560858428, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 37968 + }, + { + "epoch": 0.37969, + "grad_norm": 0.9622956256957419, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 37969 + }, + { + "epoch": 0.3797, + "grad_norm": 0.8096371277286264, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 37970 + }, + { + "epoch": 0.37971, + "grad_norm": 0.7397013802636494, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 37971 + }, + { + "epoch": 0.37972, + "grad_norm": 0.7075677464005032, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 37972 + }, + { + "epoch": 0.37973, + "grad_norm": 0.7506706029531338, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 37973 + }, + { + "epoch": 0.37974, + "grad_norm": 0.8443667347160524, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 37974 + }, + { + "epoch": 0.37975, + "grad_norm": 0.8322630126949182, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 37975 + }, + { + "epoch": 0.37976, + "grad_norm": 0.7930938867182977, + "learning_rate": 0.003, + "loss": 3.995, + "step": 37976 + }, + { + "epoch": 0.37977, + "grad_norm": 0.8988180681795724, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 37977 + }, + { + "epoch": 0.37978, + "grad_norm": 0.9773957898398931, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 37978 + }, + { + "epoch": 0.37979, + "grad_norm": 1.0272960081273668, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 37979 + }, + { + "epoch": 0.3798, + "grad_norm": 1.0449670707774537, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 37980 + }, + { + "epoch": 0.37981, + "grad_norm": 1.040639601409558, + "learning_rate": 0.003, + "loss": 4.0254, + "step": 37981 + }, + { + "epoch": 0.37982, + "grad_norm": 1.0420334690361086, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 37982 + }, + { + "epoch": 0.37983, + "grad_norm": 0.8561151610723602, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 37983 + }, + { + "epoch": 0.37984, + "grad_norm": 0.904621389579488, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 37984 + }, + { + "epoch": 0.37985, + "grad_norm": 0.8642257987607783, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 37985 + }, + { + "epoch": 0.37986, + "grad_norm": 0.856401758077838, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 37986 + }, + { + "epoch": 0.37987, + "grad_norm": 0.7364898075615456, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 37987 + }, + { + "epoch": 0.37988, + "grad_norm": 0.7231412512212834, + "learning_rate": 0.003, + "loss": 4.016, + "step": 37988 + }, + { + "epoch": 0.37989, + "grad_norm": 0.7105727145242039, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 37989 + }, + { + "epoch": 0.3799, + "grad_norm": 0.7814897500356025, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 37990 + }, + { + "epoch": 0.37991, + "grad_norm": 0.8633296235298666, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 37991 + }, + { + "epoch": 0.37992, + "grad_norm": 0.7774261818152868, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 37992 + }, + { + "epoch": 0.37993, + "grad_norm": 0.7572589132961159, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 37993 + }, + { + "epoch": 0.37994, + "grad_norm": 0.7991111951287301, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 37994 + }, + { + "epoch": 0.37995, + "grad_norm": 0.7793766192913089, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 37995 + }, + { + "epoch": 0.37996, + "grad_norm": 0.780767495256318, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 37996 + }, + { + "epoch": 0.37997, + "grad_norm": 0.8090010052661532, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 37997 + }, + { + "epoch": 0.37998, + "grad_norm": 0.7696384166391984, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 37998 + }, + { + "epoch": 0.37999, + "grad_norm": 0.8079098886162137, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 37999 + }, + { + "epoch": 0.38, + "grad_norm": 0.680639662877687, + "learning_rate": 0.003, + "loss": 4.003, + "step": 38000 + }, + { + "epoch": 0.38001, + "grad_norm": 0.544293450902653, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 38001 + }, + { + "epoch": 0.38002, + "grad_norm": 0.6439864720927451, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 38002 + }, + { + "epoch": 0.38003, + "grad_norm": 0.7025527673575265, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 38003 + }, + { + "epoch": 0.38004, + "grad_norm": 0.638224447122092, + "learning_rate": 0.003, + "loss": 3.968, + "step": 38004 + }, + { + "epoch": 0.38005, + "grad_norm": 0.5747571495945384, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 38005 + }, + { + "epoch": 0.38006, + "grad_norm": 0.5570850121715085, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 38006 + }, + { + "epoch": 0.38007, + "grad_norm": 0.5137986886131461, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 38007 + }, + { + "epoch": 0.38008, + "grad_norm": 0.5235242507690236, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 38008 + }, + { + "epoch": 0.38009, + "grad_norm": 0.5174777285575577, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 38009 + }, + { + "epoch": 0.3801, + "grad_norm": 0.5852552889245491, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 38010 + }, + { + "epoch": 0.38011, + "grad_norm": 0.6788498977411043, + "learning_rate": 0.003, + "loss": 3.999, + "step": 38011 + }, + { + "epoch": 0.38012, + "grad_norm": 0.6574424537684819, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 38012 + }, + { + "epoch": 0.38013, + "grad_norm": 0.6252237324085448, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 38013 + }, + { + "epoch": 0.38014, + "grad_norm": 0.6575082096234379, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 38014 + }, + { + "epoch": 0.38015, + "grad_norm": 0.6399024351692227, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 38015 + }, + { + "epoch": 0.38016, + "grad_norm": 0.6350201978439339, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 38016 + }, + { + "epoch": 0.38017, + "grad_norm": 0.6875465827131708, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 38017 + }, + { + "epoch": 0.38018, + "grad_norm": 0.7452451714703568, + "learning_rate": 0.003, + "loss": 3.969, + "step": 38018 + }, + { + "epoch": 0.38019, + "grad_norm": 0.9203468653268189, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 38019 + }, + { + "epoch": 0.3802, + "grad_norm": 1.2493617267894286, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 38020 + }, + { + "epoch": 0.38021, + "grad_norm": 0.9024437508411147, + "learning_rate": 0.003, + "loss": 3.954, + "step": 38021 + }, + { + "epoch": 0.38022, + "grad_norm": 0.842622085845868, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 38022 + }, + { + "epoch": 0.38023, + "grad_norm": 0.8883610198275168, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 38023 + }, + { + "epoch": 0.38024, + "grad_norm": 1.072099014549423, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 38024 + }, + { + "epoch": 0.38025, + "grad_norm": 0.9663616222514926, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 38025 + }, + { + "epoch": 0.38026, + "grad_norm": 0.8446584504519249, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 38026 + }, + { + "epoch": 0.38027, + "grad_norm": 0.8783487896782413, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 38027 + }, + { + "epoch": 0.38028, + "grad_norm": 0.9380120755622207, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 38028 + }, + { + "epoch": 0.38029, + "grad_norm": 1.0073058470440566, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 38029 + }, + { + "epoch": 0.3803, + "grad_norm": 1.060487907733484, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 38030 + }, + { + "epoch": 0.38031, + "grad_norm": 0.9302961388030331, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 38031 + }, + { + "epoch": 0.38032, + "grad_norm": 0.9304855479005008, + "learning_rate": 0.003, + "loss": 4.007, + "step": 38032 + }, + { + "epoch": 0.38033, + "grad_norm": 0.9801337742930532, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 38033 + }, + { + "epoch": 0.38034, + "grad_norm": 0.9445880236263632, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 38034 + }, + { + "epoch": 0.38035, + "grad_norm": 0.868973095154773, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 38035 + }, + { + "epoch": 0.38036, + "grad_norm": 0.9573935118038015, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 38036 + }, + { + "epoch": 0.38037, + "grad_norm": 1.1102453105332883, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 38037 + }, + { + "epoch": 0.38038, + "grad_norm": 0.9492467213110993, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 38038 + }, + { + "epoch": 0.38039, + "grad_norm": 0.8604671667369681, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 38039 + }, + { + "epoch": 0.3804, + "grad_norm": 0.8237148759949765, + "learning_rate": 0.003, + "loss": 4.012, + "step": 38040 + }, + { + "epoch": 0.38041, + "grad_norm": 0.8391196455748905, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 38041 + }, + { + "epoch": 0.38042, + "grad_norm": 0.8242632134362492, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 38042 + }, + { + "epoch": 0.38043, + "grad_norm": 0.8763049821467538, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38043 + }, + { + "epoch": 0.38044, + "grad_norm": 0.8442312579789918, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 38044 + }, + { + "epoch": 0.38045, + "grad_norm": 0.8881451654468165, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 38045 + }, + { + "epoch": 0.38046, + "grad_norm": 0.7743572194693573, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 38046 + }, + { + "epoch": 0.38047, + "grad_norm": 0.7817953546789947, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 38047 + }, + { + "epoch": 0.38048, + "grad_norm": 0.7083247539670817, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 38048 + }, + { + "epoch": 0.38049, + "grad_norm": 0.6553940647645508, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 38049 + }, + { + "epoch": 0.3805, + "grad_norm": 0.6548225595112475, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 38050 + }, + { + "epoch": 0.38051, + "grad_norm": 0.6805997272063031, + "learning_rate": 0.003, + "loss": 3.987, + "step": 38051 + }, + { + "epoch": 0.38052, + "grad_norm": 0.573738662775434, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 38052 + }, + { + "epoch": 0.38053, + "grad_norm": 0.4665826995640333, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 38053 + }, + { + "epoch": 0.38054, + "grad_norm": 0.5220336968058233, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 38054 + }, + { + "epoch": 0.38055, + "grad_norm": 0.4928459825488656, + "learning_rate": 0.003, + "loss": 3.97, + "step": 38055 + }, + { + "epoch": 0.38056, + "grad_norm": 0.5525984275962194, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 38056 + }, + { + "epoch": 0.38057, + "grad_norm": 0.6002382315325332, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 38057 + }, + { + "epoch": 0.38058, + "grad_norm": 0.7424242668145115, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 38058 + }, + { + "epoch": 0.38059, + "grad_norm": 0.9144162789345561, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 38059 + }, + { + "epoch": 0.3806, + "grad_norm": 0.9925202071862205, + "learning_rate": 0.003, + "loss": 4.0, + "step": 38060 + }, + { + "epoch": 0.38061, + "grad_norm": 0.8913259902256402, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 38061 + }, + { + "epoch": 0.38062, + "grad_norm": 0.6838680916740438, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 38062 + }, + { + "epoch": 0.38063, + "grad_norm": 0.6671545177401047, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 38063 + }, + { + "epoch": 0.38064, + "grad_norm": 0.7080245686758815, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 38064 + }, + { + "epoch": 0.38065, + "grad_norm": 0.6471874515010556, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 38065 + }, + { + "epoch": 0.38066, + "grad_norm": 0.6246845672649199, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 38066 + }, + { + "epoch": 0.38067, + "grad_norm": 0.6055872906243166, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 38067 + }, + { + "epoch": 0.38068, + "grad_norm": 0.5998772712402958, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 38068 + }, + { + "epoch": 0.38069, + "grad_norm": 0.6277052219039043, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 38069 + }, + { + "epoch": 0.3807, + "grad_norm": 0.6148585030531132, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 38070 + }, + { + "epoch": 0.38071, + "grad_norm": 0.6210901522187436, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 38071 + }, + { + "epoch": 0.38072, + "grad_norm": 0.6457275748373554, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 38072 + }, + { + "epoch": 0.38073, + "grad_norm": 0.6206186590715901, + "learning_rate": 0.003, + "loss": 3.9414, + "step": 38073 + }, + { + "epoch": 0.38074, + "grad_norm": 0.5710300574903515, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 38074 + }, + { + "epoch": 0.38075, + "grad_norm": 0.5731171466411819, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 38075 + }, + { + "epoch": 0.38076, + "grad_norm": 0.5645644573630199, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 38076 + }, + { + "epoch": 0.38077, + "grad_norm": 0.5531201116282707, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 38077 + }, + { + "epoch": 0.38078, + "grad_norm": 0.5598197353032514, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 38078 + }, + { + "epoch": 0.38079, + "grad_norm": 0.6304659760981612, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 38079 + }, + { + "epoch": 0.3808, + "grad_norm": 0.7583028122726064, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 38080 + }, + { + "epoch": 0.38081, + "grad_norm": 0.975589210247603, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 38081 + }, + { + "epoch": 0.38082, + "grad_norm": 1.3083741368325292, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 38082 + }, + { + "epoch": 0.38083, + "grad_norm": 0.7885393272511328, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 38083 + }, + { + "epoch": 0.38084, + "grad_norm": 0.7186516049069336, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 38084 + }, + { + "epoch": 0.38085, + "grad_norm": 0.7636280739858945, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 38085 + }, + { + "epoch": 0.38086, + "grad_norm": 0.768688833641533, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 38086 + }, + { + "epoch": 0.38087, + "grad_norm": 0.7414071834030448, + "learning_rate": 0.003, + "loss": 4.002, + "step": 38087 + }, + { + "epoch": 0.38088, + "grad_norm": 0.7398937469469656, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 38088 + }, + { + "epoch": 0.38089, + "grad_norm": 0.7011864211171012, + "learning_rate": 0.003, + "loss": 3.967, + "step": 38089 + }, + { + "epoch": 0.3809, + "grad_norm": 0.8406806417371531, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 38090 + }, + { + "epoch": 0.38091, + "grad_norm": 0.9852881075361701, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 38091 + }, + { + "epoch": 0.38092, + "grad_norm": 1.1077085439042136, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 38092 + }, + { + "epoch": 0.38093, + "grad_norm": 0.9380384253323122, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 38093 + }, + { + "epoch": 0.38094, + "grad_norm": 0.9396964842622861, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 38094 + }, + { + "epoch": 0.38095, + "grad_norm": 0.818164225352015, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 38095 + }, + { + "epoch": 0.38096, + "grad_norm": 0.946233255262969, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 38096 + }, + { + "epoch": 0.38097, + "grad_norm": 0.8786905041500939, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 38097 + }, + { + "epoch": 0.38098, + "grad_norm": 0.9463267983497763, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 38098 + }, + { + "epoch": 0.38099, + "grad_norm": 1.0190501970302275, + "learning_rate": 0.003, + "loss": 3.979, + "step": 38099 + }, + { + "epoch": 0.381, + "grad_norm": 0.9672124854441374, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 38100 + }, + { + "epoch": 0.38101, + "grad_norm": 1.0655877318608786, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 38101 + }, + { + "epoch": 0.38102, + "grad_norm": 0.8539477560916741, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 38102 + }, + { + "epoch": 0.38103, + "grad_norm": 0.6749813116591513, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 38103 + }, + { + "epoch": 0.38104, + "grad_norm": 0.8086374296079829, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 38104 + }, + { + "epoch": 0.38105, + "grad_norm": 1.0668449753067275, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 38105 + }, + { + "epoch": 0.38106, + "grad_norm": 0.860756258605184, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 38106 + }, + { + "epoch": 0.38107, + "grad_norm": 0.8270114729504099, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 38107 + }, + { + "epoch": 0.38108, + "grad_norm": 0.6860707601594532, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 38108 + }, + { + "epoch": 0.38109, + "grad_norm": 0.7640810599768855, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 38109 + }, + { + "epoch": 0.3811, + "grad_norm": 0.8056550511952054, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 38110 + }, + { + "epoch": 0.38111, + "grad_norm": 0.7490012778027929, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 38111 + }, + { + "epoch": 0.38112, + "grad_norm": 0.7969165662112274, + "learning_rate": 0.003, + "loss": 3.981, + "step": 38112 + }, + { + "epoch": 0.38113, + "grad_norm": 0.8704330775783122, + "learning_rate": 0.003, + "loss": 3.996, + "step": 38113 + }, + { + "epoch": 0.38114, + "grad_norm": 0.9832965674347018, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 38114 + }, + { + "epoch": 0.38115, + "grad_norm": 1.0239899691890253, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 38115 + }, + { + "epoch": 0.38116, + "grad_norm": 0.9464416016675822, + "learning_rate": 0.003, + "loss": 3.992, + "step": 38116 + }, + { + "epoch": 0.38117, + "grad_norm": 0.9410250702399342, + "learning_rate": 0.003, + "loss": 4.0303, + "step": 38117 + }, + { + "epoch": 0.38118, + "grad_norm": 0.9031818925194234, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 38118 + }, + { + "epoch": 0.38119, + "grad_norm": 0.7248331815358083, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 38119 + }, + { + "epoch": 0.3812, + "grad_norm": 0.6563286943871488, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 38120 + }, + { + "epoch": 0.38121, + "grad_norm": 0.7347492775365311, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 38121 + }, + { + "epoch": 0.38122, + "grad_norm": 0.7482301559144211, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 38122 + }, + { + "epoch": 0.38123, + "grad_norm": 0.7125007392988711, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 38123 + }, + { + "epoch": 0.38124, + "grad_norm": 0.7254218422010297, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 38124 + }, + { + "epoch": 0.38125, + "grad_norm": 0.7220537897131535, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 38125 + }, + { + "epoch": 0.38126, + "grad_norm": 0.7429612527110674, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 38126 + }, + { + "epoch": 0.38127, + "grad_norm": 0.6477510799314072, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 38127 + }, + { + "epoch": 0.38128, + "grad_norm": 0.6651121878930271, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 38128 + }, + { + "epoch": 0.38129, + "grad_norm": 0.63043831466129, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 38129 + }, + { + "epoch": 0.3813, + "grad_norm": 0.6301863470864633, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 38130 + }, + { + "epoch": 0.38131, + "grad_norm": 0.6487205648063521, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 38131 + }, + { + "epoch": 0.38132, + "grad_norm": 0.8361999162453612, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 38132 + }, + { + "epoch": 0.38133, + "grad_norm": 1.0297960324472, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 38133 + }, + { + "epoch": 0.38134, + "grad_norm": 1.0426023601147465, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 38134 + }, + { + "epoch": 0.38135, + "grad_norm": 0.7318433251672815, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 38135 + }, + { + "epoch": 0.38136, + "grad_norm": 0.7230962069738985, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 38136 + }, + { + "epoch": 0.38137, + "grad_norm": 0.761062919009629, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 38137 + }, + { + "epoch": 0.38138, + "grad_norm": 0.800214766646808, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 38138 + }, + { + "epoch": 0.38139, + "grad_norm": 0.7457375178773074, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 38139 + }, + { + "epoch": 0.3814, + "grad_norm": 0.891005573245429, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 38140 + }, + { + "epoch": 0.38141, + "grad_norm": 0.8888819194463203, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 38141 + }, + { + "epoch": 0.38142, + "grad_norm": 0.9784360358736225, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 38142 + }, + { + "epoch": 0.38143, + "grad_norm": 1.081627788231294, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 38143 + }, + { + "epoch": 0.38144, + "grad_norm": 0.8665733871639267, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 38144 + }, + { + "epoch": 0.38145, + "grad_norm": 0.7800713380445049, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 38145 + }, + { + "epoch": 0.38146, + "grad_norm": 0.7530799250579291, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 38146 + }, + { + "epoch": 0.38147, + "grad_norm": 0.7454978194484276, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 38147 + }, + { + "epoch": 0.38148, + "grad_norm": 0.6637291582759427, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 38148 + }, + { + "epoch": 0.38149, + "grad_norm": 0.7150923911501271, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 38149 + }, + { + "epoch": 0.3815, + "grad_norm": 0.766982199133606, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 38150 + }, + { + "epoch": 0.38151, + "grad_norm": 0.8600681672219717, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 38151 + }, + { + "epoch": 0.38152, + "grad_norm": 0.9268209247405135, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 38152 + }, + { + "epoch": 0.38153, + "grad_norm": 0.8277194978885103, + "learning_rate": 0.003, + "loss": 3.944, + "step": 38153 + }, + { + "epoch": 0.38154, + "grad_norm": 0.6722841061983209, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 38154 + }, + { + "epoch": 0.38155, + "grad_norm": 0.7066330138877528, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 38155 + }, + { + "epoch": 0.38156, + "grad_norm": 0.6537911666743924, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 38156 + }, + { + "epoch": 0.38157, + "grad_norm": 0.6203097275355597, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 38157 + }, + { + "epoch": 0.38158, + "grad_norm": 0.6183042926140996, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 38158 + }, + { + "epoch": 0.38159, + "grad_norm": 0.7327160548372114, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 38159 + }, + { + "epoch": 0.3816, + "grad_norm": 0.7304434539811504, + "learning_rate": 0.003, + "loss": 3.975, + "step": 38160 + }, + { + "epoch": 0.38161, + "grad_norm": 0.7057628030918416, + "learning_rate": 0.003, + "loss": 3.984, + "step": 38161 + }, + { + "epoch": 0.38162, + "grad_norm": 0.7309795252353662, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 38162 + }, + { + "epoch": 0.38163, + "grad_norm": 0.706934800061663, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 38163 + }, + { + "epoch": 0.38164, + "grad_norm": 0.6981355216672323, + "learning_rate": 0.003, + "loss": 3.957, + "step": 38164 + }, + { + "epoch": 0.38165, + "grad_norm": 0.7056586196882499, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 38165 + }, + { + "epoch": 0.38166, + "grad_norm": 0.7185409339311551, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 38166 + }, + { + "epoch": 0.38167, + "grad_norm": 0.7864262154904355, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 38167 + }, + { + "epoch": 0.38168, + "grad_norm": 0.7790287559198511, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 38168 + }, + { + "epoch": 0.38169, + "grad_norm": 0.7975405667270618, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 38169 + }, + { + "epoch": 0.3817, + "grad_norm": 0.8780076532056739, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 38170 + }, + { + "epoch": 0.38171, + "grad_norm": 0.9102851093611765, + "learning_rate": 0.003, + "loss": 3.995, + "step": 38171 + }, + { + "epoch": 0.38172, + "grad_norm": 0.8155162845797034, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 38172 + }, + { + "epoch": 0.38173, + "grad_norm": 0.7880682019387922, + "learning_rate": 0.003, + "loss": 3.972, + "step": 38173 + }, + { + "epoch": 0.38174, + "grad_norm": 0.7553100537382662, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 38174 + }, + { + "epoch": 0.38175, + "grad_norm": 0.6618902266044735, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 38175 + }, + { + "epoch": 0.38176, + "grad_norm": 0.6456555551081613, + "learning_rate": 0.003, + "loss": 4.004, + "step": 38176 + }, + { + "epoch": 0.38177, + "grad_norm": 0.6416201832139371, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 38177 + }, + { + "epoch": 0.38178, + "grad_norm": 0.7169204052981143, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 38178 + }, + { + "epoch": 0.38179, + "grad_norm": 0.6891533686987389, + "learning_rate": 0.003, + "loss": 3.962, + "step": 38179 + }, + { + "epoch": 0.3818, + "grad_norm": 0.682422065724245, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 38180 + }, + { + "epoch": 0.38181, + "grad_norm": 0.8116810443833542, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 38181 + }, + { + "epoch": 0.38182, + "grad_norm": 0.838572964748612, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 38182 + }, + { + "epoch": 0.38183, + "grad_norm": 0.9117773347477458, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 38183 + }, + { + "epoch": 0.38184, + "grad_norm": 0.9794470285281275, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 38184 + }, + { + "epoch": 0.38185, + "grad_norm": 0.9959141967165328, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 38185 + }, + { + "epoch": 0.38186, + "grad_norm": 0.9228978366711355, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 38186 + }, + { + "epoch": 0.38187, + "grad_norm": 0.9404837376941096, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 38187 + }, + { + "epoch": 0.38188, + "grad_norm": 0.8876403555323696, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 38188 + }, + { + "epoch": 0.38189, + "grad_norm": 0.8770940368612438, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 38189 + }, + { + "epoch": 0.3819, + "grad_norm": 0.946330672784003, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 38190 + }, + { + "epoch": 0.38191, + "grad_norm": 1.0048927288791378, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 38191 + }, + { + "epoch": 0.38192, + "grad_norm": 1.0713970292113135, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 38192 + }, + { + "epoch": 0.38193, + "grad_norm": 0.9781782351628128, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 38193 + }, + { + "epoch": 0.38194, + "grad_norm": 0.9655532761676165, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 38194 + }, + { + "epoch": 0.38195, + "grad_norm": 0.9110312323331953, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 38195 + }, + { + "epoch": 0.38196, + "grad_norm": 0.7848804325359915, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 38196 + }, + { + "epoch": 0.38197, + "grad_norm": 0.7070214231821674, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 38197 + }, + { + "epoch": 0.38198, + "grad_norm": 0.6846954340635475, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 38198 + }, + { + "epoch": 0.38199, + "grad_norm": 0.7212465832211515, + "learning_rate": 0.003, + "loss": 4.004, + "step": 38199 + }, + { + "epoch": 0.382, + "grad_norm": 0.7219160199425769, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 38200 + }, + { + "epoch": 0.38201, + "grad_norm": 0.6942293297426388, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 38201 + }, + { + "epoch": 0.38202, + "grad_norm": 0.6452155424546235, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 38202 + }, + { + "epoch": 0.38203, + "grad_norm": 0.6623812253926563, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 38203 + }, + { + "epoch": 0.38204, + "grad_norm": 0.7346444963414712, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 38204 + }, + { + "epoch": 0.38205, + "grad_norm": 0.6726626696493861, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 38205 + }, + { + "epoch": 0.38206, + "grad_norm": 0.7000008834249554, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 38206 + }, + { + "epoch": 0.38207, + "grad_norm": 0.7980281298873146, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 38207 + }, + { + "epoch": 0.38208, + "grad_norm": 0.8987670748348751, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 38208 + }, + { + "epoch": 0.38209, + "grad_norm": 1.0117868139975628, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 38209 + }, + { + "epoch": 0.3821, + "grad_norm": 0.9176242770570625, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 38210 + }, + { + "epoch": 0.38211, + "grad_norm": 0.973306623921461, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 38211 + }, + { + "epoch": 0.38212, + "grad_norm": 0.8553427662893254, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38212 + }, + { + "epoch": 0.38213, + "grad_norm": 0.8589345432450991, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 38213 + }, + { + "epoch": 0.38214, + "grad_norm": 0.9084363740528418, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 38214 + }, + { + "epoch": 0.38215, + "grad_norm": 0.9574522579299085, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 38215 + }, + { + "epoch": 0.38216, + "grad_norm": 0.9863056889697038, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 38216 + }, + { + "epoch": 0.38217, + "grad_norm": 0.9423939234827402, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 38217 + }, + { + "epoch": 0.38218, + "grad_norm": 0.9417693239113901, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 38218 + }, + { + "epoch": 0.38219, + "grad_norm": 1.1341806785733899, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 38219 + }, + { + "epoch": 0.3822, + "grad_norm": 0.985200468482593, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 38220 + }, + { + "epoch": 0.38221, + "grad_norm": 0.816089647792984, + "learning_rate": 0.003, + "loss": 4.003, + "step": 38221 + }, + { + "epoch": 0.38222, + "grad_norm": 0.9633603362162801, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 38222 + }, + { + "epoch": 0.38223, + "grad_norm": 1.163466763875891, + "learning_rate": 0.003, + "loss": 4.0334, + "step": 38223 + }, + { + "epoch": 0.38224, + "grad_norm": 1.0479122629938535, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 38224 + }, + { + "epoch": 0.38225, + "grad_norm": 1.0597283368775592, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 38225 + }, + { + "epoch": 0.38226, + "grad_norm": 0.8750035964436241, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 38226 + }, + { + "epoch": 0.38227, + "grad_norm": 0.8275923784397596, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 38227 + }, + { + "epoch": 0.38228, + "grad_norm": 0.8480558227834896, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 38228 + }, + { + "epoch": 0.38229, + "grad_norm": 0.9757537582177784, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 38229 + }, + { + "epoch": 0.3823, + "grad_norm": 1.0793321907012805, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 38230 + }, + { + "epoch": 0.38231, + "grad_norm": 1.053202274195791, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 38231 + }, + { + "epoch": 0.38232, + "grad_norm": 0.83127465074928, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 38232 + }, + { + "epoch": 0.38233, + "grad_norm": 0.8727308350674455, + "learning_rate": 0.003, + "loss": 4.009, + "step": 38233 + }, + { + "epoch": 0.38234, + "grad_norm": 1.0613300036804467, + "learning_rate": 0.003, + "loss": 4.0344, + "step": 38234 + }, + { + "epoch": 0.38235, + "grad_norm": 1.153762321671268, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 38235 + }, + { + "epoch": 0.38236, + "grad_norm": 0.8980832703589573, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 38236 + }, + { + "epoch": 0.38237, + "grad_norm": 0.7462262936658602, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 38237 + }, + { + "epoch": 0.38238, + "grad_norm": 0.6104626875890696, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 38238 + }, + { + "epoch": 0.38239, + "grad_norm": 0.6150022661609421, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 38239 + }, + { + "epoch": 0.3824, + "grad_norm": 0.6635051148302341, + "learning_rate": 0.003, + "loss": 3.9419, + "step": 38240 + }, + { + "epoch": 0.38241, + "grad_norm": 0.7415922949407115, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 38241 + }, + { + "epoch": 0.38242, + "grad_norm": 0.865012901896612, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 38242 + }, + { + "epoch": 0.38243, + "grad_norm": 0.8989408835522011, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 38243 + }, + { + "epoch": 0.38244, + "grad_norm": 0.7315942247454367, + "learning_rate": 0.003, + "loss": 3.996, + "step": 38244 + }, + { + "epoch": 0.38245, + "grad_norm": 0.6442636251374294, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 38245 + }, + { + "epoch": 0.38246, + "grad_norm": 0.6809262584944545, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 38246 + }, + { + "epoch": 0.38247, + "grad_norm": 0.7302607637698377, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 38247 + }, + { + "epoch": 0.38248, + "grad_norm": 0.7276883520104748, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 38248 + }, + { + "epoch": 0.38249, + "grad_norm": 0.6809477838014649, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 38249 + }, + { + "epoch": 0.3825, + "grad_norm": 0.693225705502149, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 38250 + }, + { + "epoch": 0.38251, + "grad_norm": 0.6849183312023261, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 38251 + }, + { + "epoch": 0.38252, + "grad_norm": 0.7668656052558896, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 38252 + }, + { + "epoch": 0.38253, + "grad_norm": 0.7116606241632168, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 38253 + }, + { + "epoch": 0.38254, + "grad_norm": 0.6198794480412949, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 38254 + }, + { + "epoch": 0.38255, + "grad_norm": 0.5203944060502018, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 38255 + }, + { + "epoch": 0.38256, + "grad_norm": 0.5762448668159991, + "learning_rate": 0.003, + "loss": 3.984, + "step": 38256 + }, + { + "epoch": 0.38257, + "grad_norm": 0.5929946571182206, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 38257 + }, + { + "epoch": 0.38258, + "grad_norm": 0.6813500693204154, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 38258 + }, + { + "epoch": 0.38259, + "grad_norm": 0.6422795656215591, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 38259 + }, + { + "epoch": 0.3826, + "grad_norm": 0.5738794343683625, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 38260 + }, + { + "epoch": 0.38261, + "grad_norm": 0.529742723452253, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 38261 + }, + { + "epoch": 0.38262, + "grad_norm": 0.5439960953504007, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 38262 + }, + { + "epoch": 0.38263, + "grad_norm": 0.6049118964898869, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 38263 + }, + { + "epoch": 0.38264, + "grad_norm": 0.6736703946122576, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 38264 + }, + { + "epoch": 0.38265, + "grad_norm": 0.8893277341025122, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 38265 + }, + { + "epoch": 0.38266, + "grad_norm": 1.0190668821274822, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 38266 + }, + { + "epoch": 0.38267, + "grad_norm": 0.9582491286065264, + "learning_rate": 0.003, + "loss": 4.015, + "step": 38267 + }, + { + "epoch": 0.38268, + "grad_norm": 0.8905306784061151, + "learning_rate": 0.003, + "loss": 3.994, + "step": 38268 + }, + { + "epoch": 0.38269, + "grad_norm": 0.9278845507310071, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 38269 + }, + { + "epoch": 0.3827, + "grad_norm": 0.908209358696024, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 38270 + }, + { + "epoch": 0.38271, + "grad_norm": 0.9195876440253716, + "learning_rate": 0.003, + "loss": 4.015, + "step": 38271 + }, + { + "epoch": 0.38272, + "grad_norm": 0.9300921195333317, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 38272 + }, + { + "epoch": 0.38273, + "grad_norm": 1.1032595143096016, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 38273 + }, + { + "epoch": 0.38274, + "grad_norm": 1.0186033726857107, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 38274 + }, + { + "epoch": 0.38275, + "grad_norm": 0.8121748189464645, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 38275 + }, + { + "epoch": 0.38276, + "grad_norm": 0.6389895786091431, + "learning_rate": 0.003, + "loss": 3.9473, + "step": 38276 + }, + { + "epoch": 0.38277, + "grad_norm": 0.7478145092794876, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 38277 + }, + { + "epoch": 0.38278, + "grad_norm": 0.8132454200304586, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 38278 + }, + { + "epoch": 0.38279, + "grad_norm": 0.8027350332595876, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 38279 + }, + { + "epoch": 0.3828, + "grad_norm": 0.7571154621317943, + "learning_rate": 0.003, + "loss": 3.958, + "step": 38280 + }, + { + "epoch": 0.38281, + "grad_norm": 0.6810691809800355, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 38281 + }, + { + "epoch": 0.38282, + "grad_norm": 0.6899431303590243, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 38282 + }, + { + "epoch": 0.38283, + "grad_norm": 0.697523089937598, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 38283 + }, + { + "epoch": 0.38284, + "grad_norm": 0.6595799726458652, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 38284 + }, + { + "epoch": 0.38285, + "grad_norm": 0.6917283762065306, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 38285 + }, + { + "epoch": 0.38286, + "grad_norm": 0.6826225849775285, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 38286 + }, + { + "epoch": 0.38287, + "grad_norm": 0.5712106473303833, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 38287 + }, + { + "epoch": 0.38288, + "grad_norm": 0.5845302840433165, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 38288 + }, + { + "epoch": 0.38289, + "grad_norm": 0.7117553017986284, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 38289 + }, + { + "epoch": 0.3829, + "grad_norm": 0.8372196449090548, + "learning_rate": 0.003, + "loss": 3.989, + "step": 38290 + }, + { + "epoch": 0.38291, + "grad_norm": 0.874822658010707, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 38291 + }, + { + "epoch": 0.38292, + "grad_norm": 0.9998752887073707, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 38292 + }, + { + "epoch": 0.38293, + "grad_norm": 1.0487853809929073, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 38293 + }, + { + "epoch": 0.38294, + "grad_norm": 0.8562455957030045, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 38294 + }, + { + "epoch": 0.38295, + "grad_norm": 0.7040774634547238, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 38295 + }, + { + "epoch": 0.38296, + "grad_norm": 0.7135913616922035, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 38296 + }, + { + "epoch": 0.38297, + "grad_norm": 0.6402915828169875, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 38297 + }, + { + "epoch": 0.38298, + "grad_norm": 0.6666742823804896, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 38298 + }, + { + "epoch": 0.38299, + "grad_norm": 0.6157429404453688, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 38299 + }, + { + "epoch": 0.383, + "grad_norm": 0.6731305803780947, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 38300 + }, + { + "epoch": 0.38301, + "grad_norm": 0.7421474598525224, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 38301 + }, + { + "epoch": 0.38302, + "grad_norm": 0.8329957324743444, + "learning_rate": 0.003, + "loss": 3.964, + "step": 38302 + }, + { + "epoch": 0.38303, + "grad_norm": 0.8056064958049536, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 38303 + }, + { + "epoch": 0.38304, + "grad_norm": 0.8837699568709921, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 38304 + }, + { + "epoch": 0.38305, + "grad_norm": 0.9041291307297372, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 38305 + }, + { + "epoch": 0.38306, + "grad_norm": 0.9163713369834909, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 38306 + }, + { + "epoch": 0.38307, + "grad_norm": 0.9023977687536848, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 38307 + }, + { + "epoch": 0.38308, + "grad_norm": 0.6994124756609702, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 38308 + }, + { + "epoch": 0.38309, + "grad_norm": 0.6930974846457012, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 38309 + }, + { + "epoch": 0.3831, + "grad_norm": 0.617213607070421, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 38310 + }, + { + "epoch": 0.38311, + "grad_norm": 0.6576770719579843, + "learning_rate": 0.003, + "loss": 3.977, + "step": 38311 + }, + { + "epoch": 0.38312, + "grad_norm": 0.6446195531207207, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 38312 + }, + { + "epoch": 0.38313, + "grad_norm": 0.6502513279161872, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 38313 + }, + { + "epoch": 0.38314, + "grad_norm": 0.6238568480544372, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 38314 + }, + { + "epoch": 0.38315, + "grad_norm": 0.6053036372874858, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 38315 + }, + { + "epoch": 0.38316, + "grad_norm": 0.6611656795757074, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 38316 + }, + { + "epoch": 0.38317, + "grad_norm": 0.7707582094775066, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 38317 + }, + { + "epoch": 0.38318, + "grad_norm": 0.9551185922732249, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 38318 + }, + { + "epoch": 0.38319, + "grad_norm": 1.0298360315816952, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 38319 + }, + { + "epoch": 0.3832, + "grad_norm": 1.0593388805875148, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 38320 + }, + { + "epoch": 0.38321, + "grad_norm": 1.021033462168398, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 38321 + }, + { + "epoch": 0.38322, + "grad_norm": 0.9078466011358126, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 38322 + }, + { + "epoch": 0.38323, + "grad_norm": 0.8834977562218115, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 38323 + }, + { + "epoch": 0.38324, + "grad_norm": 0.9452848558263865, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 38324 + }, + { + "epoch": 0.38325, + "grad_norm": 1.1023073338115486, + "learning_rate": 0.003, + "loss": 4.003, + "step": 38325 + }, + { + "epoch": 0.38326, + "grad_norm": 0.9588058559423898, + "learning_rate": 0.003, + "loss": 4.018, + "step": 38326 + }, + { + "epoch": 0.38327, + "grad_norm": 0.9427890022007319, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 38327 + }, + { + "epoch": 0.38328, + "grad_norm": 0.8485240358434655, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 38328 + }, + { + "epoch": 0.38329, + "grad_norm": 0.7998698143147713, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 38329 + }, + { + "epoch": 0.3833, + "grad_norm": 0.809758654495169, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 38330 + }, + { + "epoch": 0.38331, + "grad_norm": 0.840016131929622, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 38331 + }, + { + "epoch": 0.38332, + "grad_norm": 0.8062836751597142, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 38332 + }, + { + "epoch": 0.38333, + "grad_norm": 0.6611775584571887, + "learning_rate": 0.003, + "loss": 3.978, + "step": 38333 + }, + { + "epoch": 0.38334, + "grad_norm": 0.6666188603334234, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 38334 + }, + { + "epoch": 0.38335, + "grad_norm": 0.6422365365982683, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 38335 + }, + { + "epoch": 0.38336, + "grad_norm": 0.6177105815257007, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 38336 + }, + { + "epoch": 0.38337, + "grad_norm": 0.5830744453423877, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 38337 + }, + { + "epoch": 0.38338, + "grad_norm": 0.5424742696324515, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 38338 + }, + { + "epoch": 0.38339, + "grad_norm": 0.5214297431779016, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 38339 + }, + { + "epoch": 0.3834, + "grad_norm": 0.5226470474419074, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 38340 + }, + { + "epoch": 0.38341, + "grad_norm": 0.5879066274524195, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 38341 + }, + { + "epoch": 0.38342, + "grad_norm": 0.7003681945917781, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 38342 + }, + { + "epoch": 0.38343, + "grad_norm": 0.7495888289494775, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 38343 + }, + { + "epoch": 0.38344, + "grad_norm": 0.9444509317528769, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 38344 + }, + { + "epoch": 0.38345, + "grad_norm": 1.0016121672513194, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 38345 + }, + { + "epoch": 0.38346, + "grad_norm": 0.9119378214921137, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 38346 + }, + { + "epoch": 0.38347, + "grad_norm": 0.7510325892457559, + "learning_rate": 0.003, + "loss": 4.009, + "step": 38347 + }, + { + "epoch": 0.38348, + "grad_norm": 0.7120234427021368, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 38348 + }, + { + "epoch": 0.38349, + "grad_norm": 0.6921806932479764, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 38349 + }, + { + "epoch": 0.3835, + "grad_norm": 0.7459681012231708, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 38350 + }, + { + "epoch": 0.38351, + "grad_norm": 0.866022988477525, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 38351 + }, + { + "epoch": 0.38352, + "grad_norm": 0.9594917598230828, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 38352 + }, + { + "epoch": 0.38353, + "grad_norm": 1.0300208003971447, + "learning_rate": 0.003, + "loss": 3.98, + "step": 38353 + }, + { + "epoch": 0.38354, + "grad_norm": 0.843155369559974, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 38354 + }, + { + "epoch": 0.38355, + "grad_norm": 0.6934602461038744, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 38355 + }, + { + "epoch": 0.38356, + "grad_norm": 0.6947327968737659, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 38356 + }, + { + "epoch": 0.38357, + "grad_norm": 0.761255789970044, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 38357 + }, + { + "epoch": 0.38358, + "grad_norm": 0.7385562760860226, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 38358 + }, + { + "epoch": 0.38359, + "grad_norm": 0.6678468151727546, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 38359 + }, + { + "epoch": 0.3836, + "grad_norm": 0.6576812923558943, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 38360 + }, + { + "epoch": 0.38361, + "grad_norm": 0.6195837124678791, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 38361 + }, + { + "epoch": 0.38362, + "grad_norm": 0.5698746718466097, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 38362 + }, + { + "epoch": 0.38363, + "grad_norm": 0.5103238663262337, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 38363 + }, + { + "epoch": 0.38364, + "grad_norm": 0.654778605694799, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 38364 + }, + { + "epoch": 0.38365, + "grad_norm": 0.7810442720368103, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 38365 + }, + { + "epoch": 0.38366, + "grad_norm": 0.9423331254590209, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 38366 + }, + { + "epoch": 0.38367, + "grad_norm": 0.9343160992223027, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 38367 + }, + { + "epoch": 0.38368, + "grad_norm": 0.7470160186650143, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 38368 + }, + { + "epoch": 0.38369, + "grad_norm": 0.777210889810732, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 38369 + }, + { + "epoch": 0.3837, + "grad_norm": 0.8651500347223002, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 38370 + }, + { + "epoch": 0.38371, + "grad_norm": 0.9982452691566503, + "learning_rate": 0.003, + "loss": 4.0392, + "step": 38371 + }, + { + "epoch": 0.38372, + "grad_norm": 1.0723887879057772, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 38372 + }, + { + "epoch": 0.38373, + "grad_norm": 0.9170482094912734, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 38373 + }, + { + "epoch": 0.38374, + "grad_norm": 0.785832808009697, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 38374 + }, + { + "epoch": 0.38375, + "grad_norm": 0.7644852192746205, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 38375 + }, + { + "epoch": 0.38376, + "grad_norm": 0.8610833331893006, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 38376 + }, + { + "epoch": 0.38377, + "grad_norm": 0.893461351589776, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 38377 + }, + { + "epoch": 0.38378, + "grad_norm": 0.9488730596608022, + "learning_rate": 0.003, + "loss": 4.0367, + "step": 38378 + }, + { + "epoch": 0.38379, + "grad_norm": 0.952607781800174, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 38379 + }, + { + "epoch": 0.3838, + "grad_norm": 0.9368273407320094, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 38380 + }, + { + "epoch": 0.38381, + "grad_norm": 0.8945071504379223, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 38381 + }, + { + "epoch": 0.38382, + "grad_norm": 0.8364811498200848, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 38382 + }, + { + "epoch": 0.38383, + "grad_norm": 0.9556525117838559, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 38383 + }, + { + "epoch": 0.38384, + "grad_norm": 0.9895573535449904, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 38384 + }, + { + "epoch": 0.38385, + "grad_norm": 0.9380084660152628, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 38385 + }, + { + "epoch": 0.38386, + "grad_norm": 0.879592614341129, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 38386 + }, + { + "epoch": 0.38387, + "grad_norm": 0.8687275102387908, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 38387 + }, + { + "epoch": 0.38388, + "grad_norm": 1.0301719771613957, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 38388 + }, + { + "epoch": 0.38389, + "grad_norm": 0.9850194547869084, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 38389 + }, + { + "epoch": 0.3839, + "grad_norm": 0.9806890612038851, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 38390 + }, + { + "epoch": 0.38391, + "grad_norm": 0.7882519370880798, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 38391 + }, + { + "epoch": 0.38392, + "grad_norm": 0.7396739564058907, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 38392 + }, + { + "epoch": 0.38393, + "grad_norm": 0.6967947214540483, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 38393 + }, + { + "epoch": 0.38394, + "grad_norm": 0.6925278341899757, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 38394 + }, + { + "epoch": 0.38395, + "grad_norm": 0.6322363042676342, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 38395 + }, + { + "epoch": 0.38396, + "grad_norm": 0.5616736226397192, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 38396 + }, + { + "epoch": 0.38397, + "grad_norm": 0.5360231051069562, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 38397 + }, + { + "epoch": 0.38398, + "grad_norm": 0.5582959955491481, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 38398 + }, + { + "epoch": 0.38399, + "grad_norm": 0.5047585312351758, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 38399 + }, + { + "epoch": 0.384, + "grad_norm": 0.5402075903470634, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 38400 + }, + { + "epoch": 0.38401, + "grad_norm": 0.5383200068033238, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 38401 + }, + { + "epoch": 0.38402, + "grad_norm": 0.5702671608666043, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 38402 + }, + { + "epoch": 0.38403, + "grad_norm": 0.5895605434697262, + "learning_rate": 0.003, + "loss": 3.961, + "step": 38403 + }, + { + "epoch": 0.38404, + "grad_norm": 0.6471785557682271, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 38404 + }, + { + "epoch": 0.38405, + "grad_norm": 0.7535031765470356, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 38405 + }, + { + "epoch": 0.38406, + "grad_norm": 0.8179710070099083, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 38406 + }, + { + "epoch": 0.38407, + "grad_norm": 0.8230870007268118, + "learning_rate": 0.003, + "loss": 3.936, + "step": 38407 + }, + { + "epoch": 0.38408, + "grad_norm": 0.8503766843767346, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 38408 + }, + { + "epoch": 0.38409, + "grad_norm": 0.8988930734532637, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 38409 + }, + { + "epoch": 0.3841, + "grad_norm": 0.7660577696521761, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 38410 + }, + { + "epoch": 0.38411, + "grad_norm": 0.7063534473776156, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 38411 + }, + { + "epoch": 0.38412, + "grad_norm": 0.6993650372719925, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 38412 + }, + { + "epoch": 0.38413, + "grad_norm": 0.7494973125496492, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 38413 + }, + { + "epoch": 0.38414, + "grad_norm": 0.7238835048105695, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 38414 + }, + { + "epoch": 0.38415, + "grad_norm": 0.7730956900577478, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 38415 + }, + { + "epoch": 0.38416, + "grad_norm": 0.9391032740753352, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 38416 + }, + { + "epoch": 0.38417, + "grad_norm": 1.0352191428871347, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 38417 + }, + { + "epoch": 0.38418, + "grad_norm": 0.9627629149751036, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 38418 + }, + { + "epoch": 0.38419, + "grad_norm": 0.8999519632828411, + "learning_rate": 0.003, + "loss": 3.993, + "step": 38419 + }, + { + "epoch": 0.3842, + "grad_norm": 0.8659580169368171, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 38420 + }, + { + "epoch": 0.38421, + "grad_norm": 0.8327729708160061, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 38421 + }, + { + "epoch": 0.38422, + "grad_norm": 0.8025952003302248, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 38422 + }, + { + "epoch": 0.38423, + "grad_norm": 0.8364939237458956, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 38423 + }, + { + "epoch": 0.38424, + "grad_norm": 0.8686249953465359, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 38424 + }, + { + "epoch": 0.38425, + "grad_norm": 1.0374444687323838, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 38425 + }, + { + "epoch": 0.38426, + "grad_norm": 0.9782532793909557, + "learning_rate": 0.003, + "loss": 4.0282, + "step": 38426 + }, + { + "epoch": 0.38427, + "grad_norm": 0.8193566966289, + "learning_rate": 0.003, + "loss": 3.974, + "step": 38427 + }, + { + "epoch": 0.38428, + "grad_norm": 0.9078703492160549, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 38428 + }, + { + "epoch": 0.38429, + "grad_norm": 0.9067417044057916, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 38429 + }, + { + "epoch": 0.3843, + "grad_norm": 0.7612627448604877, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 38430 + }, + { + "epoch": 0.38431, + "grad_norm": 0.7966334880315818, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 38431 + }, + { + "epoch": 0.38432, + "grad_norm": 0.7646079399713888, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 38432 + }, + { + "epoch": 0.38433, + "grad_norm": 0.6847468702747747, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 38433 + }, + { + "epoch": 0.38434, + "grad_norm": 0.74826027238128, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 38434 + }, + { + "epoch": 0.38435, + "grad_norm": 0.8592844335032169, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 38435 + }, + { + "epoch": 0.38436, + "grad_norm": 0.987283117472227, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 38436 + }, + { + "epoch": 0.38437, + "grad_norm": 0.9210434290218215, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 38437 + }, + { + "epoch": 0.38438, + "grad_norm": 0.8615449696543145, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 38438 + }, + { + "epoch": 0.38439, + "grad_norm": 0.7256503133307821, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 38439 + }, + { + "epoch": 0.3844, + "grad_norm": 0.7004296584897484, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 38440 + }, + { + "epoch": 0.38441, + "grad_norm": 0.6503265892186398, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 38441 + }, + { + "epoch": 0.38442, + "grad_norm": 0.5448293356066316, + "learning_rate": 0.003, + "loss": 4.008, + "step": 38442 + }, + { + "epoch": 0.38443, + "grad_norm": 0.5088120208086294, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 38443 + }, + { + "epoch": 0.38444, + "grad_norm": 0.537203846804458, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 38444 + }, + { + "epoch": 0.38445, + "grad_norm": 0.5567619894884762, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 38445 + }, + { + "epoch": 0.38446, + "grad_norm": 0.5647442796662462, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 38446 + }, + { + "epoch": 0.38447, + "grad_norm": 0.5685578523603217, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 38447 + }, + { + "epoch": 0.38448, + "grad_norm": 0.6055925912836008, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 38448 + }, + { + "epoch": 0.38449, + "grad_norm": 0.7088736265848713, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 38449 + }, + { + "epoch": 0.3845, + "grad_norm": 0.7634900846928737, + "learning_rate": 0.003, + "loss": 4.016, + "step": 38450 + }, + { + "epoch": 0.38451, + "grad_norm": 0.7559771733795305, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 38451 + }, + { + "epoch": 0.38452, + "grad_norm": 0.696963268225481, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 38452 + }, + { + "epoch": 0.38453, + "grad_norm": 0.6930794547921288, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 38453 + }, + { + "epoch": 0.38454, + "grad_norm": 0.6539675313075488, + "learning_rate": 0.003, + "loss": 3.99, + "step": 38454 + }, + { + "epoch": 0.38455, + "grad_norm": 0.6211182292896609, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 38455 + }, + { + "epoch": 0.38456, + "grad_norm": 0.6775440365170178, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 38456 + }, + { + "epoch": 0.38457, + "grad_norm": 0.7666636506131562, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 38457 + }, + { + "epoch": 0.38458, + "grad_norm": 0.8806152930501583, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 38458 + }, + { + "epoch": 0.38459, + "grad_norm": 0.9828372188366236, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 38459 + }, + { + "epoch": 0.3846, + "grad_norm": 0.9703863746626122, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 38460 + }, + { + "epoch": 0.38461, + "grad_norm": 0.9173475823749344, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 38461 + }, + { + "epoch": 0.38462, + "grad_norm": 0.8551452490802769, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 38462 + }, + { + "epoch": 0.38463, + "grad_norm": 0.786717925379788, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 38463 + }, + { + "epoch": 0.38464, + "grad_norm": 0.7572983623346782, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 38464 + }, + { + "epoch": 0.38465, + "grad_norm": 0.8246480246180884, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 38465 + }, + { + "epoch": 0.38466, + "grad_norm": 0.982983020891683, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 38466 + }, + { + "epoch": 0.38467, + "grad_norm": 1.18408403934421, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 38467 + }, + { + "epoch": 0.38468, + "grad_norm": 0.8009593843641809, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 38468 + }, + { + "epoch": 0.38469, + "grad_norm": 0.7290975668954555, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 38469 + }, + { + "epoch": 0.3847, + "grad_norm": 0.6738830084472994, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 38470 + }, + { + "epoch": 0.38471, + "grad_norm": 0.7345093557594733, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 38471 + }, + { + "epoch": 0.38472, + "grad_norm": 0.8115057094187953, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 38472 + }, + { + "epoch": 0.38473, + "grad_norm": 0.7826813695352557, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 38473 + }, + { + "epoch": 0.38474, + "grad_norm": 0.8376744962598831, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 38474 + }, + { + "epoch": 0.38475, + "grad_norm": 0.8533814257877842, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 38475 + }, + { + "epoch": 0.38476, + "grad_norm": 0.804993355738369, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 38476 + }, + { + "epoch": 0.38477, + "grad_norm": 0.8122204868546978, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 38477 + }, + { + "epoch": 0.38478, + "grad_norm": 0.9503965075670343, + "learning_rate": 0.003, + "loss": 3.984, + "step": 38478 + }, + { + "epoch": 0.38479, + "grad_norm": 0.9432207213976123, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 38479 + }, + { + "epoch": 0.3848, + "grad_norm": 0.9933851826504916, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 38480 + }, + { + "epoch": 0.38481, + "grad_norm": 1.134853601193645, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 38481 + }, + { + "epoch": 0.38482, + "grad_norm": 0.8251492370417043, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 38482 + }, + { + "epoch": 0.38483, + "grad_norm": 0.7578158735989492, + "learning_rate": 0.003, + "loss": 4.024, + "step": 38483 + }, + { + "epoch": 0.38484, + "grad_norm": 0.7395377209470984, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 38484 + }, + { + "epoch": 0.38485, + "grad_norm": 0.7235367868108944, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 38485 + }, + { + "epoch": 0.38486, + "grad_norm": 0.6670592524351279, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 38486 + }, + { + "epoch": 0.38487, + "grad_norm": 0.6895128879257001, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 38487 + }, + { + "epoch": 0.38488, + "grad_norm": 0.6101824841908998, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 38488 + }, + { + "epoch": 0.38489, + "grad_norm": 0.6778744878005233, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 38489 + }, + { + "epoch": 0.3849, + "grad_norm": 0.6854384310629152, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 38490 + }, + { + "epoch": 0.38491, + "grad_norm": 0.7758961826058383, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 38491 + }, + { + "epoch": 0.38492, + "grad_norm": 0.7822284895429592, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 38492 + }, + { + "epoch": 0.38493, + "grad_norm": 0.9091224490951963, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 38493 + }, + { + "epoch": 0.38494, + "grad_norm": 0.9421902209110492, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 38494 + }, + { + "epoch": 0.38495, + "grad_norm": 1.0053095229450313, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 38495 + }, + { + "epoch": 0.38496, + "grad_norm": 0.8201571540397335, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 38496 + }, + { + "epoch": 0.38497, + "grad_norm": 0.7159874096823838, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 38497 + }, + { + "epoch": 0.38498, + "grad_norm": 0.6508283809682662, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 38498 + }, + { + "epoch": 0.38499, + "grad_norm": 0.6667370152604674, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 38499 + }, + { + "epoch": 0.385, + "grad_norm": 0.7667653443656782, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 38500 + }, + { + "epoch": 0.38501, + "grad_norm": 0.8162074798615182, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 38501 + }, + { + "epoch": 0.38502, + "grad_norm": 0.7701610661841405, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 38502 + }, + { + "epoch": 0.38503, + "grad_norm": 0.7107398930819439, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 38503 + }, + { + "epoch": 0.38504, + "grad_norm": 0.7324080944113073, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 38504 + }, + { + "epoch": 0.38505, + "grad_norm": 0.8487332346907824, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 38505 + }, + { + "epoch": 0.38506, + "grad_norm": 0.9373989284200993, + "learning_rate": 0.003, + "loss": 3.9341, + "step": 38506 + }, + { + "epoch": 0.38507, + "grad_norm": 0.9533058429159336, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 38507 + }, + { + "epoch": 0.38508, + "grad_norm": 0.9955669664010913, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 38508 + }, + { + "epoch": 0.38509, + "grad_norm": 0.9520871486242144, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 38509 + }, + { + "epoch": 0.3851, + "grad_norm": 0.8669920576298896, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 38510 + }, + { + "epoch": 0.38511, + "grad_norm": 0.7108727062955049, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 38511 + }, + { + "epoch": 0.38512, + "grad_norm": 0.6125121372380273, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 38512 + }, + { + "epoch": 0.38513, + "grad_norm": 0.6293656020993494, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 38513 + }, + { + "epoch": 0.38514, + "grad_norm": 0.6641968694504502, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 38514 + }, + { + "epoch": 0.38515, + "grad_norm": 0.7530237743774769, + "learning_rate": 0.003, + "loss": 3.995, + "step": 38515 + }, + { + "epoch": 0.38516, + "grad_norm": 0.6859990701806934, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 38516 + }, + { + "epoch": 0.38517, + "grad_norm": 0.7779674045771376, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 38517 + }, + { + "epoch": 0.38518, + "grad_norm": 0.883354604198942, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 38518 + }, + { + "epoch": 0.38519, + "grad_norm": 0.9283765725363083, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 38519 + }, + { + "epoch": 0.3852, + "grad_norm": 0.8991205166399159, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 38520 + }, + { + "epoch": 0.38521, + "grad_norm": 0.8964318562987852, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 38521 + }, + { + "epoch": 0.38522, + "grad_norm": 0.9656880002427702, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 38522 + }, + { + "epoch": 0.38523, + "grad_norm": 1.0057204911157658, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 38523 + }, + { + "epoch": 0.38524, + "grad_norm": 0.7973583481835916, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 38524 + }, + { + "epoch": 0.38525, + "grad_norm": 0.6687872996907689, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 38525 + }, + { + "epoch": 0.38526, + "grad_norm": 0.5948963756460779, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 38526 + }, + { + "epoch": 0.38527, + "grad_norm": 0.5916091970357626, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 38527 + }, + { + "epoch": 0.38528, + "grad_norm": 0.573971307808133, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 38528 + }, + { + "epoch": 0.38529, + "grad_norm": 0.5453152287202735, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 38529 + }, + { + "epoch": 0.3853, + "grad_norm": 0.5801464289264779, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 38530 + }, + { + "epoch": 0.38531, + "grad_norm": 0.5276660745139905, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 38531 + }, + { + "epoch": 0.38532, + "grad_norm": 0.5376846467274194, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 38532 + }, + { + "epoch": 0.38533, + "grad_norm": 0.45234637458712007, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 38533 + }, + { + "epoch": 0.38534, + "grad_norm": 0.4527439061703508, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 38534 + }, + { + "epoch": 0.38535, + "grad_norm": 0.4621565326707428, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 38535 + }, + { + "epoch": 0.38536, + "grad_norm": 0.49755562450142493, + "learning_rate": 0.003, + "loss": 3.974, + "step": 38536 + }, + { + "epoch": 0.38537, + "grad_norm": 0.6119057504011868, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 38537 + }, + { + "epoch": 0.38538, + "grad_norm": 0.800887556523863, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 38538 + }, + { + "epoch": 0.38539, + "grad_norm": 0.9337611536466184, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 38539 + }, + { + "epoch": 0.3854, + "grad_norm": 0.9092612442035458, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 38540 + }, + { + "epoch": 0.38541, + "grad_norm": 0.9720595915666885, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 38541 + }, + { + "epoch": 0.38542, + "grad_norm": 1.090833531425513, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 38542 + }, + { + "epoch": 0.38543, + "grad_norm": 0.7509339944378054, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 38543 + }, + { + "epoch": 0.38544, + "grad_norm": 0.7513483755429178, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 38544 + }, + { + "epoch": 0.38545, + "grad_norm": 0.7310649174749737, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 38545 + }, + { + "epoch": 0.38546, + "grad_norm": 0.7710787252636054, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 38546 + }, + { + "epoch": 0.38547, + "grad_norm": 0.979452436895872, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 38547 + }, + { + "epoch": 0.38548, + "grad_norm": 1.268468617192534, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 38548 + }, + { + "epoch": 0.38549, + "grad_norm": 0.8376191290097692, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 38549 + }, + { + "epoch": 0.3855, + "grad_norm": 0.9247155548609963, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 38550 + }, + { + "epoch": 0.38551, + "grad_norm": 1.0584898990395026, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 38551 + }, + { + "epoch": 0.38552, + "grad_norm": 1.003455998806238, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 38552 + }, + { + "epoch": 0.38553, + "grad_norm": 1.0750794145130422, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 38553 + }, + { + "epoch": 0.38554, + "grad_norm": 1.1699758750886762, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 38554 + }, + { + "epoch": 0.38555, + "grad_norm": 1.025597350045339, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 38555 + }, + { + "epoch": 0.38556, + "grad_norm": 1.2137852414863044, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 38556 + }, + { + "epoch": 0.38557, + "grad_norm": 0.9720911645870691, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 38557 + }, + { + "epoch": 0.38558, + "grad_norm": 1.093621471730534, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 38558 + }, + { + "epoch": 0.38559, + "grad_norm": 1.0968155493870932, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 38559 + }, + { + "epoch": 0.3856, + "grad_norm": 1.1643207108977216, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 38560 + }, + { + "epoch": 0.38561, + "grad_norm": 1.1641697388763252, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 38561 + }, + { + "epoch": 0.38562, + "grad_norm": 0.9075642978747752, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 38562 + }, + { + "epoch": 0.38563, + "grad_norm": 1.107449956072842, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 38563 + }, + { + "epoch": 0.38564, + "grad_norm": 0.9692316151159924, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 38564 + }, + { + "epoch": 0.38565, + "grad_norm": 0.8614455297010484, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 38565 + }, + { + "epoch": 0.38566, + "grad_norm": 1.023208152792714, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 38566 + }, + { + "epoch": 0.38567, + "grad_norm": 1.0243187927990287, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 38567 + }, + { + "epoch": 0.38568, + "grad_norm": 1.016767974611277, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 38568 + }, + { + "epoch": 0.38569, + "grad_norm": 1.0035544160591137, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 38569 + }, + { + "epoch": 0.3857, + "grad_norm": 1.0351944699515871, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 38570 + }, + { + "epoch": 0.38571, + "grad_norm": 0.8222473463968172, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 38571 + }, + { + "epoch": 0.38572, + "grad_norm": 0.6632980031805603, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 38572 + }, + { + "epoch": 0.38573, + "grad_norm": 0.6003656768510222, + "learning_rate": 0.003, + "loss": 3.998, + "step": 38573 + }, + { + "epoch": 0.38574, + "grad_norm": 0.6250077724910719, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 38574 + }, + { + "epoch": 0.38575, + "grad_norm": 0.6173666999047964, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 38575 + }, + { + "epoch": 0.38576, + "grad_norm": 0.6872793381515736, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 38576 + }, + { + "epoch": 0.38577, + "grad_norm": 0.7646790699984197, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 38577 + }, + { + "epoch": 0.38578, + "grad_norm": 0.8902263338198475, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 38578 + }, + { + "epoch": 0.38579, + "grad_norm": 0.9311128983016178, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 38579 + }, + { + "epoch": 0.3858, + "grad_norm": 0.9552691427966994, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 38580 + }, + { + "epoch": 0.38581, + "grad_norm": 0.8544918583647035, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 38581 + }, + { + "epoch": 0.38582, + "grad_norm": 0.8279459393246449, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 38582 + }, + { + "epoch": 0.38583, + "grad_norm": 0.9307249040240164, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 38583 + }, + { + "epoch": 0.38584, + "grad_norm": 0.939867685354018, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 38584 + }, + { + "epoch": 0.38585, + "grad_norm": 0.8981291801729596, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 38585 + }, + { + "epoch": 0.38586, + "grad_norm": 0.8424772668361619, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 38586 + }, + { + "epoch": 0.38587, + "grad_norm": 0.7214063479628856, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 38587 + }, + { + "epoch": 0.38588, + "grad_norm": 0.6998447267774941, + "learning_rate": 0.003, + "loss": 4.025, + "step": 38588 + }, + { + "epoch": 0.38589, + "grad_norm": 0.6464384835807269, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 38589 + }, + { + "epoch": 0.3859, + "grad_norm": 0.5923785997438828, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 38590 + }, + { + "epoch": 0.38591, + "grad_norm": 0.566495198083911, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 38591 + }, + { + "epoch": 0.38592, + "grad_norm": 0.569616327198443, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 38592 + }, + { + "epoch": 0.38593, + "grad_norm": 0.5891175209810032, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 38593 + }, + { + "epoch": 0.38594, + "grad_norm": 0.6093064049744262, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 38594 + }, + { + "epoch": 0.38595, + "grad_norm": 0.6044748215909483, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 38595 + }, + { + "epoch": 0.38596, + "grad_norm": 0.6960487765781451, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 38596 + }, + { + "epoch": 0.38597, + "grad_norm": 0.7245824407578914, + "learning_rate": 0.003, + "loss": 3.966, + "step": 38597 + }, + { + "epoch": 0.38598, + "grad_norm": 0.7008411642936784, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 38598 + }, + { + "epoch": 0.38599, + "grad_norm": 0.7096182281227067, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 38599 + }, + { + "epoch": 0.386, + "grad_norm": 0.7449755578934614, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 38600 + }, + { + "epoch": 0.38601, + "grad_norm": 0.7194433810998644, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 38601 + }, + { + "epoch": 0.38602, + "grad_norm": 0.6213192194338528, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 38602 + }, + { + "epoch": 0.38603, + "grad_norm": 0.5609425297264409, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 38603 + }, + { + "epoch": 0.38604, + "grad_norm": 0.6593967489246929, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 38604 + }, + { + "epoch": 0.38605, + "grad_norm": 0.7878680280907576, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 38605 + }, + { + "epoch": 0.38606, + "grad_norm": 0.9950607273211727, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 38606 + }, + { + "epoch": 0.38607, + "grad_norm": 1.2078080363740609, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 38607 + }, + { + "epoch": 0.38608, + "grad_norm": 0.7483935962978124, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 38608 + }, + { + "epoch": 0.38609, + "grad_norm": 0.7368097309979548, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 38609 + }, + { + "epoch": 0.3861, + "grad_norm": 0.8876728422442111, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 38610 + }, + { + "epoch": 0.38611, + "grad_norm": 0.9555520239557256, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 38611 + }, + { + "epoch": 0.38612, + "grad_norm": 0.9473447642340246, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 38612 + }, + { + "epoch": 0.38613, + "grad_norm": 0.8036240622679759, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 38613 + }, + { + "epoch": 0.38614, + "grad_norm": 0.6724342612405049, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 38614 + }, + { + "epoch": 0.38615, + "grad_norm": 0.6313099879345109, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 38615 + }, + { + "epoch": 0.38616, + "grad_norm": 0.6288426582460797, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 38616 + }, + { + "epoch": 0.38617, + "grad_norm": 0.6681803480122629, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 38617 + }, + { + "epoch": 0.38618, + "grad_norm": 0.7302458356439694, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 38618 + }, + { + "epoch": 0.38619, + "grad_norm": 0.7452278167683383, + "learning_rate": 0.003, + "loss": 3.9431, + "step": 38619 + }, + { + "epoch": 0.3862, + "grad_norm": 0.7424501954707129, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 38620 + }, + { + "epoch": 0.38621, + "grad_norm": 0.8671109570464639, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 38621 + }, + { + "epoch": 0.38622, + "grad_norm": 0.8942841854399672, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 38622 + }, + { + "epoch": 0.38623, + "grad_norm": 0.8737391628618366, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 38623 + }, + { + "epoch": 0.38624, + "grad_norm": 0.9324022396964928, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 38624 + }, + { + "epoch": 0.38625, + "grad_norm": 0.8873006017722292, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 38625 + }, + { + "epoch": 0.38626, + "grad_norm": 0.8179611292100413, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 38626 + }, + { + "epoch": 0.38627, + "grad_norm": 0.8763716111532526, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 38627 + }, + { + "epoch": 0.38628, + "grad_norm": 0.8977287077049488, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 38628 + }, + { + "epoch": 0.38629, + "grad_norm": 0.8136362598762861, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 38629 + }, + { + "epoch": 0.3863, + "grad_norm": 0.7968583201562408, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 38630 + }, + { + "epoch": 0.38631, + "grad_norm": 0.6118260951685601, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 38631 + }, + { + "epoch": 0.38632, + "grad_norm": 0.6663190903190405, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 38632 + }, + { + "epoch": 0.38633, + "grad_norm": 0.6970945695363499, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 38633 + }, + { + "epoch": 0.38634, + "grad_norm": 0.7477834312441525, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 38634 + }, + { + "epoch": 0.38635, + "grad_norm": 0.6931481435504935, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 38635 + }, + { + "epoch": 0.38636, + "grad_norm": 0.6429404670556897, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 38636 + }, + { + "epoch": 0.38637, + "grad_norm": 0.7562348138663546, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 38637 + }, + { + "epoch": 0.38638, + "grad_norm": 0.8149193912414361, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 38638 + }, + { + "epoch": 0.38639, + "grad_norm": 0.9803137456939586, + "learning_rate": 0.003, + "loss": 4.0, + "step": 38639 + }, + { + "epoch": 0.3864, + "grad_norm": 1.1251711260281059, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 38640 + }, + { + "epoch": 0.38641, + "grad_norm": 0.7631816603372144, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 38641 + }, + { + "epoch": 0.38642, + "grad_norm": 0.5932300666614952, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 38642 + }, + { + "epoch": 0.38643, + "grad_norm": 0.6328662296247675, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 38643 + }, + { + "epoch": 0.38644, + "grad_norm": 0.6226111097283205, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 38644 + }, + { + "epoch": 0.38645, + "grad_norm": 0.6026436338813632, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 38645 + }, + { + "epoch": 0.38646, + "grad_norm": 0.5513698320243958, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 38646 + }, + { + "epoch": 0.38647, + "grad_norm": 0.5854393978019898, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 38647 + }, + { + "epoch": 0.38648, + "grad_norm": 0.5778466833630757, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 38648 + }, + { + "epoch": 0.38649, + "grad_norm": 0.6633018324711688, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 38649 + }, + { + "epoch": 0.3865, + "grad_norm": 0.603927398396661, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 38650 + }, + { + "epoch": 0.38651, + "grad_norm": 0.5431923494287594, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 38651 + }, + { + "epoch": 0.38652, + "grad_norm": 0.5403052033168594, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 38652 + }, + { + "epoch": 0.38653, + "grad_norm": 0.5324097574935155, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 38653 + }, + { + "epoch": 0.38654, + "grad_norm": 0.5541904910298733, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 38654 + }, + { + "epoch": 0.38655, + "grad_norm": 0.6453248328590059, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 38655 + }, + { + "epoch": 0.38656, + "grad_norm": 0.8265365516254961, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 38656 + }, + { + "epoch": 0.38657, + "grad_norm": 0.962524116850464, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 38657 + }, + { + "epoch": 0.38658, + "grad_norm": 0.9541939110697664, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 38658 + }, + { + "epoch": 0.38659, + "grad_norm": 0.9915523725631599, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 38659 + }, + { + "epoch": 0.3866, + "grad_norm": 1.099282433285853, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 38660 + }, + { + "epoch": 0.38661, + "grad_norm": 0.8828940243686407, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 38661 + }, + { + "epoch": 0.38662, + "grad_norm": 0.9312815104837294, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 38662 + }, + { + "epoch": 0.38663, + "grad_norm": 0.9808152984838736, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 38663 + }, + { + "epoch": 0.38664, + "grad_norm": 0.9941779168354317, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 38664 + }, + { + "epoch": 0.38665, + "grad_norm": 0.898330612078955, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 38665 + }, + { + "epoch": 0.38666, + "grad_norm": 0.9788981124240889, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 38666 + }, + { + "epoch": 0.38667, + "grad_norm": 0.8568967570986609, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 38667 + }, + { + "epoch": 0.38668, + "grad_norm": 0.8373481201815066, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 38668 + }, + { + "epoch": 0.38669, + "grad_norm": 0.885868752373239, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 38669 + }, + { + "epoch": 0.3867, + "grad_norm": 0.9199686844468681, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 38670 + }, + { + "epoch": 0.38671, + "grad_norm": 0.8762716668862998, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 38671 + }, + { + "epoch": 0.38672, + "grad_norm": 0.8365471295649889, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 38672 + }, + { + "epoch": 0.38673, + "grad_norm": 0.7054912032445169, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 38673 + }, + { + "epoch": 0.38674, + "grad_norm": 0.735157593922258, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 38674 + }, + { + "epoch": 0.38675, + "grad_norm": 0.8202502267207679, + "learning_rate": 0.003, + "loss": 4.013, + "step": 38675 + }, + { + "epoch": 0.38676, + "grad_norm": 0.8961736263286375, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 38676 + }, + { + "epoch": 0.38677, + "grad_norm": 0.8995172680818202, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 38677 + }, + { + "epoch": 0.38678, + "grad_norm": 0.7865180908033907, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 38678 + }, + { + "epoch": 0.38679, + "grad_norm": 0.6903044630705325, + "learning_rate": 0.003, + "loss": 3.994, + "step": 38679 + }, + { + "epoch": 0.3868, + "grad_norm": 0.6300947184264726, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 38680 + }, + { + "epoch": 0.38681, + "grad_norm": 0.7128644520302552, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 38681 + }, + { + "epoch": 0.38682, + "grad_norm": 0.6382383799493063, + "learning_rate": 0.003, + "loss": 3.994, + "step": 38682 + }, + { + "epoch": 0.38683, + "grad_norm": 0.6175658718745116, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 38683 + }, + { + "epoch": 0.38684, + "grad_norm": 0.6224598224437282, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 38684 + }, + { + "epoch": 0.38685, + "grad_norm": 0.6869596738577789, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 38685 + }, + { + "epoch": 0.38686, + "grad_norm": 0.7536750240827425, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 38686 + }, + { + "epoch": 0.38687, + "grad_norm": 0.761837996887082, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 38687 + }, + { + "epoch": 0.38688, + "grad_norm": 0.7290574321601587, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 38688 + }, + { + "epoch": 0.38689, + "grad_norm": 0.7427274499511074, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 38689 + }, + { + "epoch": 0.3869, + "grad_norm": 0.7461283867646604, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 38690 + }, + { + "epoch": 0.38691, + "grad_norm": 0.6783243415496115, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 38691 + }, + { + "epoch": 0.38692, + "grad_norm": 0.6654520976633801, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 38692 + }, + { + "epoch": 0.38693, + "grad_norm": 0.6863949641084518, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 38693 + }, + { + "epoch": 0.38694, + "grad_norm": 0.846857375817505, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 38694 + }, + { + "epoch": 0.38695, + "grad_norm": 1.0016862910384956, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 38695 + }, + { + "epoch": 0.38696, + "grad_norm": 1.0307926782886476, + "learning_rate": 0.003, + "loss": 4.003, + "step": 38696 + }, + { + "epoch": 0.38697, + "grad_norm": 0.7786207476017057, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 38697 + }, + { + "epoch": 0.38698, + "grad_norm": 0.73789097431824, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 38698 + }, + { + "epoch": 0.38699, + "grad_norm": 0.72428498779766, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 38699 + }, + { + "epoch": 0.387, + "grad_norm": 0.7348187942572505, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 38700 + }, + { + "epoch": 0.38701, + "grad_norm": 0.9160676772353422, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 38701 + }, + { + "epoch": 0.38702, + "grad_norm": 0.9317544921300378, + "learning_rate": 0.003, + "loss": 3.9424, + "step": 38702 + }, + { + "epoch": 0.38703, + "grad_norm": 0.8668917532922444, + "learning_rate": 0.003, + "loss": 3.9409, + "step": 38703 + }, + { + "epoch": 0.38704, + "grad_norm": 0.8255075814984282, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 38704 + }, + { + "epoch": 0.38705, + "grad_norm": 0.785910112310139, + "learning_rate": 0.003, + "loss": 3.973, + "step": 38705 + }, + { + "epoch": 0.38706, + "grad_norm": 0.7457114691920691, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 38706 + }, + { + "epoch": 0.38707, + "grad_norm": 0.7206477772053185, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 38707 + }, + { + "epoch": 0.38708, + "grad_norm": 0.6949607468127798, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 38708 + }, + { + "epoch": 0.38709, + "grad_norm": 0.6488901986355733, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 38709 + }, + { + "epoch": 0.3871, + "grad_norm": 0.6358382279945657, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 38710 + }, + { + "epoch": 0.38711, + "grad_norm": 0.7123329614558971, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 38711 + }, + { + "epoch": 0.38712, + "grad_norm": 0.735274207149715, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 38712 + }, + { + "epoch": 0.38713, + "grad_norm": 0.736387158556655, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 38713 + }, + { + "epoch": 0.38714, + "grad_norm": 0.9271359022994539, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 38714 + }, + { + "epoch": 0.38715, + "grad_norm": 0.989511586427934, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 38715 + }, + { + "epoch": 0.38716, + "grad_norm": 0.8875173297529868, + "learning_rate": 0.003, + "loss": 3.985, + "step": 38716 + }, + { + "epoch": 0.38717, + "grad_norm": 0.7941425478750817, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 38717 + }, + { + "epoch": 0.38718, + "grad_norm": 0.9453945719442144, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 38718 + }, + { + "epoch": 0.38719, + "grad_norm": 0.9516631566662369, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 38719 + }, + { + "epoch": 0.3872, + "grad_norm": 0.9173315766296285, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 38720 + }, + { + "epoch": 0.38721, + "grad_norm": 1.0100770287299279, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 38721 + }, + { + "epoch": 0.38722, + "grad_norm": 1.0751094975442554, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 38722 + }, + { + "epoch": 0.38723, + "grad_norm": 0.7713754779787534, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 38723 + }, + { + "epoch": 0.38724, + "grad_norm": 0.7164681477055772, + "learning_rate": 0.003, + "loss": 4.005, + "step": 38724 + }, + { + "epoch": 0.38725, + "grad_norm": 0.8542690655601684, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 38725 + }, + { + "epoch": 0.38726, + "grad_norm": 0.9456395418992027, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 38726 + }, + { + "epoch": 0.38727, + "grad_norm": 0.9592955751360123, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 38727 + }, + { + "epoch": 0.38728, + "grad_norm": 0.9240826851122333, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 38728 + }, + { + "epoch": 0.38729, + "grad_norm": 0.9479204044041811, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 38729 + }, + { + "epoch": 0.3873, + "grad_norm": 0.9759589984533497, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 38730 + }, + { + "epoch": 0.38731, + "grad_norm": 0.813974840038849, + "learning_rate": 0.003, + "loss": 4.0319, + "step": 38731 + }, + { + "epoch": 0.38732, + "grad_norm": 0.7718508372164498, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 38732 + }, + { + "epoch": 0.38733, + "grad_norm": 0.8377087777118436, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 38733 + }, + { + "epoch": 0.38734, + "grad_norm": 0.8465701745071381, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 38734 + }, + { + "epoch": 0.38735, + "grad_norm": 0.753128505304011, + "learning_rate": 0.003, + "loss": 4.022, + "step": 38735 + }, + { + "epoch": 0.38736, + "grad_norm": 0.8322745954935454, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 38736 + }, + { + "epoch": 0.38737, + "grad_norm": 0.8106667213529353, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 38737 + }, + { + "epoch": 0.38738, + "grad_norm": 0.9036698151221835, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 38738 + }, + { + "epoch": 0.38739, + "grad_norm": 1.0150317958002482, + "learning_rate": 0.003, + "loss": 4.001, + "step": 38739 + }, + { + "epoch": 0.3874, + "grad_norm": 0.9904143298429844, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 38740 + }, + { + "epoch": 0.38741, + "grad_norm": 0.8938328216034751, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 38741 + }, + { + "epoch": 0.38742, + "grad_norm": 0.8962927988444439, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 38742 + }, + { + "epoch": 0.38743, + "grad_norm": 0.8169641451537204, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 38743 + }, + { + "epoch": 0.38744, + "grad_norm": 0.750795750486783, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 38744 + }, + { + "epoch": 0.38745, + "grad_norm": 0.6128308483206149, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 38745 + }, + { + "epoch": 0.38746, + "grad_norm": 0.4946344946145023, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 38746 + }, + { + "epoch": 0.38747, + "grad_norm": 0.44384464059883993, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 38747 + }, + { + "epoch": 0.38748, + "grad_norm": 0.4598568819477488, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 38748 + }, + { + "epoch": 0.38749, + "grad_norm": 0.4152856291027401, + "learning_rate": 0.003, + "loss": 3.9444, + "step": 38749 + }, + { + "epoch": 0.3875, + "grad_norm": 0.4154589255097264, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 38750 + }, + { + "epoch": 0.38751, + "grad_norm": 0.414123552026003, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 38751 + }, + { + "epoch": 0.38752, + "grad_norm": 0.38328908847033377, + "learning_rate": 0.003, + "loss": 4.004, + "step": 38752 + }, + { + "epoch": 0.38753, + "grad_norm": 0.3305960908395654, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 38753 + }, + { + "epoch": 0.38754, + "grad_norm": 0.33559768614609053, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 38754 + }, + { + "epoch": 0.38755, + "grad_norm": 0.3722749991619875, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 38755 + }, + { + "epoch": 0.38756, + "grad_norm": 0.44774486069762387, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 38756 + }, + { + "epoch": 0.38757, + "grad_norm": 0.601900211463878, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 38757 + }, + { + "epoch": 0.38758, + "grad_norm": 0.8462872525277473, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 38758 + }, + { + "epoch": 0.38759, + "grad_norm": 1.0218605882240304, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 38759 + }, + { + "epoch": 0.3876, + "grad_norm": 1.0127409334582522, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 38760 + }, + { + "epoch": 0.38761, + "grad_norm": 0.7577658229005374, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 38761 + }, + { + "epoch": 0.38762, + "grad_norm": 0.5846230035166444, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 38762 + }, + { + "epoch": 0.38763, + "grad_norm": 0.6572262069394805, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 38763 + }, + { + "epoch": 0.38764, + "grad_norm": 0.7396544829113192, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 38764 + }, + { + "epoch": 0.38765, + "grad_norm": 0.7560354206811364, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 38765 + }, + { + "epoch": 0.38766, + "grad_norm": 0.7711335790216832, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 38766 + }, + { + "epoch": 0.38767, + "grad_norm": 0.7729311843903763, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 38767 + }, + { + "epoch": 0.38768, + "grad_norm": 0.7751895974841231, + "learning_rate": 0.003, + "loss": 3.985, + "step": 38768 + }, + { + "epoch": 0.38769, + "grad_norm": 0.8894457782230388, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 38769 + }, + { + "epoch": 0.3877, + "grad_norm": 0.9917067532899909, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 38770 + }, + { + "epoch": 0.38771, + "grad_norm": 1.0830328374208438, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 38771 + }, + { + "epoch": 0.38772, + "grad_norm": 0.8775857414906459, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 38772 + }, + { + "epoch": 0.38773, + "grad_norm": 0.8388820628864398, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 38773 + }, + { + "epoch": 0.38774, + "grad_norm": 0.9230233315836717, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 38774 + }, + { + "epoch": 0.38775, + "grad_norm": 0.840336358531563, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 38775 + }, + { + "epoch": 0.38776, + "grad_norm": 0.9379780425780302, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 38776 + }, + { + "epoch": 0.38777, + "grad_norm": 1.0596336885977087, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 38777 + }, + { + "epoch": 0.38778, + "grad_norm": 0.9699614693283618, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 38778 + }, + { + "epoch": 0.38779, + "grad_norm": 1.0269572911334612, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 38779 + }, + { + "epoch": 0.3878, + "grad_norm": 1.0175349502288642, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 38780 + }, + { + "epoch": 0.38781, + "grad_norm": 0.9421518007696974, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 38781 + }, + { + "epoch": 0.38782, + "grad_norm": 1.021000284278893, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 38782 + }, + { + "epoch": 0.38783, + "grad_norm": 0.8700674370524426, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 38783 + }, + { + "epoch": 0.38784, + "grad_norm": 0.9484207488070465, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 38784 + }, + { + "epoch": 0.38785, + "grad_norm": 1.0656249190207239, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 38785 + }, + { + "epoch": 0.38786, + "grad_norm": 1.0273081430892423, + "learning_rate": 0.003, + "loss": 4.03, + "step": 38786 + }, + { + "epoch": 0.38787, + "grad_norm": 1.0063032865925092, + "learning_rate": 0.003, + "loss": 4.019, + "step": 38787 + }, + { + "epoch": 0.38788, + "grad_norm": 0.983548031115232, + "learning_rate": 0.003, + "loss": 4.019, + "step": 38788 + }, + { + "epoch": 0.38789, + "grad_norm": 0.7914808550630678, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 38789 + }, + { + "epoch": 0.3879, + "grad_norm": 0.6809750791113232, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 38790 + }, + { + "epoch": 0.38791, + "grad_norm": 0.6886443764783478, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 38791 + }, + { + "epoch": 0.38792, + "grad_norm": 0.6856989613459902, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 38792 + }, + { + "epoch": 0.38793, + "grad_norm": 0.770199832781716, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 38793 + }, + { + "epoch": 0.38794, + "grad_norm": 0.8590788387637011, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 38794 + }, + { + "epoch": 0.38795, + "grad_norm": 0.8488503232682285, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 38795 + }, + { + "epoch": 0.38796, + "grad_norm": 0.7953499959000947, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 38796 + }, + { + "epoch": 0.38797, + "grad_norm": 0.6339725283229531, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 38797 + }, + { + "epoch": 0.38798, + "grad_norm": 0.6121149661196073, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 38798 + }, + { + "epoch": 0.38799, + "grad_norm": 0.6251231411333579, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 38799 + }, + { + "epoch": 0.388, + "grad_norm": 0.5842625029360741, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 38800 + }, + { + "epoch": 0.38801, + "grad_norm": 0.6332601413858834, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 38801 + }, + { + "epoch": 0.38802, + "grad_norm": 0.6725538967622752, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 38802 + }, + { + "epoch": 0.38803, + "grad_norm": 0.6706278075980855, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 38803 + }, + { + "epoch": 0.38804, + "grad_norm": 0.7915753263780262, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 38804 + }, + { + "epoch": 0.38805, + "grad_norm": 0.7613815206272312, + "learning_rate": 0.003, + "loss": 3.9419, + "step": 38805 + }, + { + "epoch": 0.38806, + "grad_norm": 0.7102855831279494, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 38806 + }, + { + "epoch": 0.38807, + "grad_norm": 0.6881145459825135, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 38807 + }, + { + "epoch": 0.38808, + "grad_norm": 0.734837975206811, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 38808 + }, + { + "epoch": 0.38809, + "grad_norm": 0.6805097294588597, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 38809 + }, + { + "epoch": 0.3881, + "grad_norm": 0.7214344305376849, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 38810 + }, + { + "epoch": 0.38811, + "grad_norm": 0.7083531472045641, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 38811 + }, + { + "epoch": 0.38812, + "grad_norm": 0.7674992009820534, + "learning_rate": 0.003, + "loss": 3.9468, + "step": 38812 + }, + { + "epoch": 0.38813, + "grad_norm": 0.8304361353851268, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 38813 + }, + { + "epoch": 0.38814, + "grad_norm": 0.7871864367800777, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 38814 + }, + { + "epoch": 0.38815, + "grad_norm": 0.68904722569733, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 38815 + }, + { + "epoch": 0.38816, + "grad_norm": 0.6644507979893197, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 38816 + }, + { + "epoch": 0.38817, + "grad_norm": 0.6305247979111276, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 38817 + }, + { + "epoch": 0.38818, + "grad_norm": 0.5657061066685622, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 38818 + }, + { + "epoch": 0.38819, + "grad_norm": 0.6088274786139243, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 38819 + }, + { + "epoch": 0.3882, + "grad_norm": 0.5882113033866285, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 38820 + }, + { + "epoch": 0.38821, + "grad_norm": 0.6587524360127528, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 38821 + }, + { + "epoch": 0.38822, + "grad_norm": 0.7715184679137368, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 38822 + }, + { + "epoch": 0.38823, + "grad_norm": 0.8781931249787345, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 38823 + }, + { + "epoch": 0.38824, + "grad_norm": 0.9979566805193771, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 38824 + }, + { + "epoch": 0.38825, + "grad_norm": 1.1019073319072785, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 38825 + }, + { + "epoch": 0.38826, + "grad_norm": 0.9052883813809218, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 38826 + }, + { + "epoch": 0.38827, + "grad_norm": 0.8012417138180391, + "learning_rate": 0.003, + "loss": 3.958, + "step": 38827 + }, + { + "epoch": 0.38828, + "grad_norm": 0.8222922598349867, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 38828 + }, + { + "epoch": 0.38829, + "grad_norm": 0.7954548998557852, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 38829 + }, + { + "epoch": 0.3883, + "grad_norm": 0.8137380208219783, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 38830 + }, + { + "epoch": 0.38831, + "grad_norm": 0.8966997728360785, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 38831 + }, + { + "epoch": 0.38832, + "grad_norm": 0.8946457802175848, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 38832 + }, + { + "epoch": 0.38833, + "grad_norm": 0.9472109302917885, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 38833 + }, + { + "epoch": 0.38834, + "grad_norm": 0.9415225099915121, + "learning_rate": 0.003, + "loss": 3.976, + "step": 38834 + }, + { + "epoch": 0.38835, + "grad_norm": 0.8789179867960124, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 38835 + }, + { + "epoch": 0.38836, + "grad_norm": 1.02461599248857, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 38836 + }, + { + "epoch": 0.38837, + "grad_norm": 0.9028378044364689, + "learning_rate": 0.003, + "loss": 3.977, + "step": 38837 + }, + { + "epoch": 0.38838, + "grad_norm": 0.9273717659181246, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 38838 + }, + { + "epoch": 0.38839, + "grad_norm": 1.1233204735833564, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 38839 + }, + { + "epoch": 0.3884, + "grad_norm": 1.0082972179902312, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 38840 + }, + { + "epoch": 0.38841, + "grad_norm": 0.8590795350257147, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 38841 + }, + { + "epoch": 0.38842, + "grad_norm": 0.7405152178390805, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 38842 + }, + { + "epoch": 0.38843, + "grad_norm": 0.7524789198404057, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 38843 + }, + { + "epoch": 0.38844, + "grad_norm": 0.7615116623620359, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 38844 + }, + { + "epoch": 0.38845, + "grad_norm": 0.8153826631336797, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 38845 + }, + { + "epoch": 0.38846, + "grad_norm": 0.9957645181155608, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 38846 + }, + { + "epoch": 0.38847, + "grad_norm": 1.0091882668128471, + "learning_rate": 0.003, + "loss": 4.002, + "step": 38847 + }, + { + "epoch": 0.38848, + "grad_norm": 0.7363740147121801, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 38848 + }, + { + "epoch": 0.38849, + "grad_norm": 0.6148387173975091, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 38849 + }, + { + "epoch": 0.3885, + "grad_norm": 0.5874614865554514, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 38850 + }, + { + "epoch": 0.38851, + "grad_norm": 0.605406686891821, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 38851 + }, + { + "epoch": 0.38852, + "grad_norm": 0.7668680924146729, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 38852 + }, + { + "epoch": 0.38853, + "grad_norm": 0.8167436559113834, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 38853 + }, + { + "epoch": 0.38854, + "grad_norm": 0.6874964586608129, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 38854 + }, + { + "epoch": 0.38855, + "grad_norm": 0.570250718374726, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 38855 + }, + { + "epoch": 0.38856, + "grad_norm": 0.4776637996032867, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38856 + }, + { + "epoch": 0.38857, + "grad_norm": 0.5415804990833227, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 38857 + }, + { + "epoch": 0.38858, + "grad_norm": 0.552017509039536, + "learning_rate": 0.003, + "loss": 3.98, + "step": 38858 + }, + { + "epoch": 0.38859, + "grad_norm": 0.5155698417675236, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 38859 + }, + { + "epoch": 0.3886, + "grad_norm": 0.5500894657808449, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 38860 + }, + { + "epoch": 0.38861, + "grad_norm": 0.5665925334730404, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 38861 + }, + { + "epoch": 0.38862, + "grad_norm": 0.6171641640533062, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 38862 + }, + { + "epoch": 0.38863, + "grad_norm": 0.644058293702115, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 38863 + }, + { + "epoch": 0.38864, + "grad_norm": 0.7681050998944102, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 38864 + }, + { + "epoch": 0.38865, + "grad_norm": 0.9062091799930242, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 38865 + }, + { + "epoch": 0.38866, + "grad_norm": 0.8179913444690574, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 38866 + }, + { + "epoch": 0.38867, + "grad_norm": 0.6373155476464869, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 38867 + }, + { + "epoch": 0.38868, + "grad_norm": 0.5748145723438935, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 38868 + }, + { + "epoch": 0.38869, + "grad_norm": 0.576626178393959, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 38869 + }, + { + "epoch": 0.3887, + "grad_norm": 0.7790866441398379, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 38870 + }, + { + "epoch": 0.38871, + "grad_norm": 0.9268618007239168, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 38871 + }, + { + "epoch": 0.38872, + "grad_norm": 1.0431771152123113, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 38872 + }, + { + "epoch": 0.38873, + "grad_norm": 0.8145447092135218, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 38873 + }, + { + "epoch": 0.38874, + "grad_norm": 0.7681977137994277, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 38874 + }, + { + "epoch": 0.38875, + "grad_norm": 0.9317969291672361, + "learning_rate": 0.003, + "loss": 3.99, + "step": 38875 + }, + { + "epoch": 0.38876, + "grad_norm": 0.8459433695264081, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 38876 + }, + { + "epoch": 0.38877, + "grad_norm": 0.7672818855252862, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 38877 + }, + { + "epoch": 0.38878, + "grad_norm": 0.76321207019284, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 38878 + }, + { + "epoch": 0.38879, + "grad_norm": 0.7272403680119423, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 38879 + }, + { + "epoch": 0.3888, + "grad_norm": 0.7706739192447665, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 38880 + }, + { + "epoch": 0.38881, + "grad_norm": 0.7795256123807676, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 38881 + }, + { + "epoch": 0.38882, + "grad_norm": 0.7690245876284633, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 38882 + }, + { + "epoch": 0.38883, + "grad_norm": 0.7984945453776389, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 38883 + }, + { + "epoch": 0.38884, + "grad_norm": 0.7660822872538331, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 38884 + }, + { + "epoch": 0.38885, + "grad_norm": 0.7842042870894237, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 38885 + }, + { + "epoch": 0.38886, + "grad_norm": 0.8077435084025317, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 38886 + }, + { + "epoch": 0.38887, + "grad_norm": 0.7381810254556048, + "learning_rate": 0.003, + "loss": 4.011, + "step": 38887 + }, + { + "epoch": 0.38888, + "grad_norm": 0.8126540670462277, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 38888 + }, + { + "epoch": 0.38889, + "grad_norm": 0.8388941463479644, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 38889 + }, + { + "epoch": 0.3889, + "grad_norm": 0.8263851401654541, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 38890 + }, + { + "epoch": 0.38891, + "grad_norm": 0.6663275974222985, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 38891 + }, + { + "epoch": 0.38892, + "grad_norm": 0.6645308872067235, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 38892 + }, + { + "epoch": 0.38893, + "grad_norm": 0.5563108972853081, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 38893 + }, + { + "epoch": 0.38894, + "grad_norm": 0.6015572045661434, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 38894 + }, + { + "epoch": 0.38895, + "grad_norm": 0.6716558102486119, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 38895 + }, + { + "epoch": 0.38896, + "grad_norm": 0.803709831107598, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 38896 + }, + { + "epoch": 0.38897, + "grad_norm": 0.8139385509942345, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 38897 + }, + { + "epoch": 0.38898, + "grad_norm": 0.7949823379054382, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 38898 + }, + { + "epoch": 0.38899, + "grad_norm": 0.9243899093385087, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 38899 + }, + { + "epoch": 0.389, + "grad_norm": 1.0505356577379215, + "learning_rate": 0.003, + "loss": 4.022, + "step": 38900 + }, + { + "epoch": 0.38901, + "grad_norm": 0.9632625055313276, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 38901 + }, + { + "epoch": 0.38902, + "grad_norm": 0.8714702772895644, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 38902 + }, + { + "epoch": 0.38903, + "grad_norm": 0.7367533673416515, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 38903 + }, + { + "epoch": 0.38904, + "grad_norm": 0.7792841037823507, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 38904 + }, + { + "epoch": 0.38905, + "grad_norm": 0.6864383774105394, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 38905 + }, + { + "epoch": 0.38906, + "grad_norm": 0.6918710247187948, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 38906 + }, + { + "epoch": 0.38907, + "grad_norm": 0.7559391632286406, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 38907 + }, + { + "epoch": 0.38908, + "grad_norm": 0.8436168796934648, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 38908 + }, + { + "epoch": 0.38909, + "grad_norm": 0.8679869783696712, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 38909 + }, + { + "epoch": 0.3891, + "grad_norm": 0.7663305910536167, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 38910 + }, + { + "epoch": 0.38911, + "grad_norm": 0.6965954215029775, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 38911 + }, + { + "epoch": 0.38912, + "grad_norm": 0.691119286042028, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 38912 + }, + { + "epoch": 0.38913, + "grad_norm": 0.7199298650859408, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 38913 + }, + { + "epoch": 0.38914, + "grad_norm": 0.7513401706643191, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 38914 + }, + { + "epoch": 0.38915, + "grad_norm": 0.7658932898306556, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 38915 + }, + { + "epoch": 0.38916, + "grad_norm": 0.8304202136630514, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 38916 + }, + { + "epoch": 0.38917, + "grad_norm": 0.8461033733525124, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 38917 + }, + { + "epoch": 0.38918, + "grad_norm": 0.8734350410851281, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 38918 + }, + { + "epoch": 0.38919, + "grad_norm": 1.0363427752129197, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 38919 + }, + { + "epoch": 0.3892, + "grad_norm": 0.9026291188051893, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 38920 + }, + { + "epoch": 0.38921, + "grad_norm": 0.9096375028904868, + "learning_rate": 0.003, + "loss": 3.968, + "step": 38921 + }, + { + "epoch": 0.38922, + "grad_norm": 0.8326445063089987, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 38922 + }, + { + "epoch": 0.38923, + "grad_norm": 0.776380014154207, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 38923 + }, + { + "epoch": 0.38924, + "grad_norm": 0.6956140205524919, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 38924 + }, + { + "epoch": 0.38925, + "grad_norm": 0.6491400255567439, + "learning_rate": 0.003, + "loss": 3.978, + "step": 38925 + }, + { + "epoch": 0.38926, + "grad_norm": 0.6359384954655883, + "learning_rate": 0.003, + "loss": 4.01, + "step": 38926 + }, + { + "epoch": 0.38927, + "grad_norm": 0.7087950573759099, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 38927 + }, + { + "epoch": 0.38928, + "grad_norm": 0.7502911799239914, + "learning_rate": 0.003, + "loss": 4.0, + "step": 38928 + }, + { + "epoch": 0.38929, + "grad_norm": 0.7093050549766606, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 38929 + }, + { + "epoch": 0.3893, + "grad_norm": 0.7105708256871383, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 38930 + }, + { + "epoch": 0.38931, + "grad_norm": 0.6600100004590332, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 38931 + }, + { + "epoch": 0.38932, + "grad_norm": 0.6165474466486666, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 38932 + }, + { + "epoch": 0.38933, + "grad_norm": 0.6839892898257746, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 38933 + }, + { + "epoch": 0.38934, + "grad_norm": 0.8694615978608256, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 38934 + }, + { + "epoch": 0.38935, + "grad_norm": 0.9828198369966139, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 38935 + }, + { + "epoch": 0.38936, + "grad_norm": 0.8883148438729517, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 38936 + }, + { + "epoch": 0.38937, + "grad_norm": 0.7527927027314262, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 38937 + }, + { + "epoch": 0.38938, + "grad_norm": 0.7137767692391537, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 38938 + }, + { + "epoch": 0.38939, + "grad_norm": 0.7278975179757211, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 38939 + }, + { + "epoch": 0.3894, + "grad_norm": 0.7427859352984651, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 38940 + }, + { + "epoch": 0.38941, + "grad_norm": 0.8146257770187462, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 38941 + }, + { + "epoch": 0.38942, + "grad_norm": 1.0241606862035117, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 38942 + }, + { + "epoch": 0.38943, + "grad_norm": 1.1053562352751314, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 38943 + }, + { + "epoch": 0.38944, + "grad_norm": 0.7443845126176396, + "learning_rate": 0.003, + "loss": 3.981, + "step": 38944 + }, + { + "epoch": 0.38945, + "grad_norm": 0.6802062182034447, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 38945 + }, + { + "epoch": 0.38946, + "grad_norm": 0.7103821935339816, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 38946 + }, + { + "epoch": 0.38947, + "grad_norm": 0.7068526779631239, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38947 + }, + { + "epoch": 0.38948, + "grad_norm": 0.7394363424339179, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 38948 + }, + { + "epoch": 0.38949, + "grad_norm": 0.8183246256871932, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 38949 + }, + { + "epoch": 0.3895, + "grad_norm": 0.8512258966326337, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 38950 + }, + { + "epoch": 0.38951, + "grad_norm": 0.7557225243854859, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 38951 + }, + { + "epoch": 0.38952, + "grad_norm": 0.7973908428127476, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 38952 + }, + { + "epoch": 0.38953, + "grad_norm": 0.9871858296380348, + "learning_rate": 0.003, + "loss": 3.955, + "step": 38953 + }, + { + "epoch": 0.38954, + "grad_norm": 1.0465533543259957, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 38954 + }, + { + "epoch": 0.38955, + "grad_norm": 1.0922975631190315, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 38955 + }, + { + "epoch": 0.38956, + "grad_norm": 0.9410148124969266, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 38956 + }, + { + "epoch": 0.38957, + "grad_norm": 0.8454861803362907, + "learning_rate": 0.003, + "loss": 4.012, + "step": 38957 + }, + { + "epoch": 0.38958, + "grad_norm": 0.8608805941487094, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 38958 + }, + { + "epoch": 0.38959, + "grad_norm": 0.8294074756150371, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 38959 + }, + { + "epoch": 0.3896, + "grad_norm": 0.7821786062947643, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 38960 + }, + { + "epoch": 0.38961, + "grad_norm": 0.7288333922558015, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 38961 + }, + { + "epoch": 0.38962, + "grad_norm": 0.6886215372307468, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 38962 + }, + { + "epoch": 0.38963, + "grad_norm": 0.710917596104318, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 38963 + }, + { + "epoch": 0.38964, + "grad_norm": 0.7478034086019928, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 38964 + }, + { + "epoch": 0.38965, + "grad_norm": 0.7809677973997871, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 38965 + }, + { + "epoch": 0.38966, + "grad_norm": 0.7699001369885181, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 38966 + }, + { + "epoch": 0.38967, + "grad_norm": 0.8561316120231904, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 38967 + }, + { + "epoch": 0.38968, + "grad_norm": 0.856888171479896, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 38968 + }, + { + "epoch": 0.38969, + "grad_norm": 0.937450057003872, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 38969 + }, + { + "epoch": 0.3897, + "grad_norm": 0.9694327183485799, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 38970 + }, + { + "epoch": 0.38971, + "grad_norm": 0.9586279724993617, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 38971 + }, + { + "epoch": 0.38972, + "grad_norm": 0.879363879247839, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 38972 + }, + { + "epoch": 0.38973, + "grad_norm": 0.7699658323422887, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 38973 + }, + { + "epoch": 0.38974, + "grad_norm": 0.8108432720158002, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 38974 + }, + { + "epoch": 0.38975, + "grad_norm": 0.7113298433128816, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 38975 + }, + { + "epoch": 0.38976, + "grad_norm": 0.6898674021243089, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 38976 + }, + { + "epoch": 0.38977, + "grad_norm": 0.6454489265558884, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 38977 + }, + { + "epoch": 0.38978, + "grad_norm": 0.6614210303360643, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 38978 + }, + { + "epoch": 0.38979, + "grad_norm": 0.6445320780350634, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 38979 + }, + { + "epoch": 0.3898, + "grad_norm": 0.7359777703893613, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 38980 + }, + { + "epoch": 0.38981, + "grad_norm": 0.6556590918817149, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 38981 + }, + { + "epoch": 0.38982, + "grad_norm": 0.5957874505069737, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 38982 + }, + { + "epoch": 0.38983, + "grad_norm": 0.5348274400670503, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 38983 + }, + { + "epoch": 0.38984, + "grad_norm": 0.46390644998733904, + "learning_rate": 0.003, + "loss": 3.965, + "step": 38984 + }, + { + "epoch": 0.38985, + "grad_norm": 0.4600662678543835, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 38985 + }, + { + "epoch": 0.38986, + "grad_norm": 0.48894717777334634, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 38986 + }, + { + "epoch": 0.38987, + "grad_norm": 0.50880893557077, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 38987 + }, + { + "epoch": 0.38988, + "grad_norm": 0.5759896505716041, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 38988 + }, + { + "epoch": 0.38989, + "grad_norm": 0.6637538965589673, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 38989 + }, + { + "epoch": 0.3899, + "grad_norm": 0.8373231647962104, + "learning_rate": 0.003, + "loss": 3.9324, + "step": 38990 + }, + { + "epoch": 0.38991, + "grad_norm": 0.9645347169550935, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 38991 + }, + { + "epoch": 0.38992, + "grad_norm": 0.8799773490382418, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 38992 + }, + { + "epoch": 0.38993, + "grad_norm": 0.8529900777795488, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 38993 + }, + { + "epoch": 0.38994, + "grad_norm": 0.8769640966713778, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 38994 + }, + { + "epoch": 0.38995, + "grad_norm": 0.8879229257077269, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 38995 + }, + { + "epoch": 0.38996, + "grad_norm": 0.8357669929907824, + "learning_rate": 0.003, + "loss": 3.984, + "step": 38996 + }, + { + "epoch": 0.38997, + "grad_norm": 0.7798132982537312, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 38997 + }, + { + "epoch": 0.38998, + "grad_norm": 0.8213992185774748, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 38998 + }, + { + "epoch": 0.38999, + "grad_norm": 0.8922094757283864, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 38999 + }, + { + "epoch": 0.39, + "grad_norm": 0.8658265478447392, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 39000 + }, + { + "epoch": 0.39001, + "grad_norm": 0.8870746216489465, + "learning_rate": 0.003, + "loss": 4.011, + "step": 39001 + }, + { + "epoch": 0.39002, + "grad_norm": 0.9551161296781069, + "learning_rate": 0.003, + "loss": 3.9483, + "step": 39002 + }, + { + "epoch": 0.39003, + "grad_norm": 0.9566739875189366, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 39003 + }, + { + "epoch": 0.39004, + "grad_norm": 0.7688764785734126, + "learning_rate": 0.003, + "loss": 3.979, + "step": 39004 + }, + { + "epoch": 0.39005, + "grad_norm": 0.803898233282558, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 39005 + }, + { + "epoch": 0.39006, + "grad_norm": 0.8426853290561998, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 39006 + }, + { + "epoch": 0.39007, + "grad_norm": 0.9748908669307271, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 39007 + }, + { + "epoch": 0.39008, + "grad_norm": 1.076176287654905, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 39008 + }, + { + "epoch": 0.39009, + "grad_norm": 0.871524591806645, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 39009 + }, + { + "epoch": 0.3901, + "grad_norm": 0.7227618770926751, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 39010 + }, + { + "epoch": 0.39011, + "grad_norm": 0.6989052622677726, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 39011 + }, + { + "epoch": 0.39012, + "grad_norm": 0.8043948270527324, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 39012 + }, + { + "epoch": 0.39013, + "grad_norm": 0.7697474154086367, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 39013 + }, + { + "epoch": 0.39014, + "grad_norm": 0.7793285009411929, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 39014 + }, + { + "epoch": 0.39015, + "grad_norm": 0.7432965374494532, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 39015 + }, + { + "epoch": 0.39016, + "grad_norm": 0.8399139455022584, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 39016 + }, + { + "epoch": 0.39017, + "grad_norm": 0.9192404123247064, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 39017 + }, + { + "epoch": 0.39018, + "grad_norm": 0.8541501397605128, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 39018 + }, + { + "epoch": 0.39019, + "grad_norm": 0.7840577044263324, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 39019 + }, + { + "epoch": 0.3902, + "grad_norm": 0.6709165039073952, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 39020 + }, + { + "epoch": 0.39021, + "grad_norm": 0.6558239455760769, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 39021 + }, + { + "epoch": 0.39022, + "grad_norm": 0.6907156833143245, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 39022 + }, + { + "epoch": 0.39023, + "grad_norm": 0.7241780593275207, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 39023 + }, + { + "epoch": 0.39024, + "grad_norm": 0.8178743101072966, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 39024 + }, + { + "epoch": 0.39025, + "grad_norm": 0.9813853816577992, + "learning_rate": 0.003, + "loss": 3.997, + "step": 39025 + }, + { + "epoch": 0.39026, + "grad_norm": 1.0098902322972465, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 39026 + }, + { + "epoch": 0.39027, + "grad_norm": 0.8592704179371988, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 39027 + }, + { + "epoch": 0.39028, + "grad_norm": 0.8193596907842369, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 39028 + }, + { + "epoch": 0.39029, + "grad_norm": 0.6678459794874941, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 39029 + }, + { + "epoch": 0.3903, + "grad_norm": 0.7302756953789404, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 39030 + }, + { + "epoch": 0.39031, + "grad_norm": 0.6191415114794578, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 39031 + }, + { + "epoch": 0.39032, + "grad_norm": 0.7025984406655019, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 39032 + }, + { + "epoch": 0.39033, + "grad_norm": 0.9211818779865231, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 39033 + }, + { + "epoch": 0.39034, + "grad_norm": 1.0622494568154348, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 39034 + }, + { + "epoch": 0.39035, + "grad_norm": 0.8293805489395525, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 39035 + }, + { + "epoch": 0.39036, + "grad_norm": 0.7661023284073702, + "learning_rate": 0.003, + "loss": 3.989, + "step": 39036 + }, + { + "epoch": 0.39037, + "grad_norm": 0.7210714235441215, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 39037 + }, + { + "epoch": 0.39038, + "grad_norm": 0.755025677031958, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 39038 + }, + { + "epoch": 0.39039, + "grad_norm": 0.7897954690943892, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 39039 + }, + { + "epoch": 0.3904, + "grad_norm": 0.7791184161476681, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 39040 + }, + { + "epoch": 0.39041, + "grad_norm": 0.7375340340105367, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 39041 + }, + { + "epoch": 0.39042, + "grad_norm": 0.8583533921334595, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 39042 + }, + { + "epoch": 0.39043, + "grad_norm": 1.0690362997321128, + "learning_rate": 0.003, + "loss": 3.999, + "step": 39043 + }, + { + "epoch": 0.39044, + "grad_norm": 0.9795421846727029, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 39044 + }, + { + "epoch": 0.39045, + "grad_norm": 0.8621564878012538, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 39045 + }, + { + "epoch": 0.39046, + "grad_norm": 0.8487984031260682, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 39046 + }, + { + "epoch": 0.39047, + "grad_norm": 0.7352175677566362, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 39047 + }, + { + "epoch": 0.39048, + "grad_norm": 0.6465035802113501, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 39048 + }, + { + "epoch": 0.39049, + "grad_norm": 0.5954224465413875, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 39049 + }, + { + "epoch": 0.3905, + "grad_norm": 0.6231035860027723, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 39050 + }, + { + "epoch": 0.39051, + "grad_norm": 0.6610621680239317, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 39051 + }, + { + "epoch": 0.39052, + "grad_norm": 0.7996493021856819, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 39052 + }, + { + "epoch": 0.39053, + "grad_norm": 0.8423100647565794, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 39053 + }, + { + "epoch": 0.39054, + "grad_norm": 0.8577861772236287, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 39054 + }, + { + "epoch": 0.39055, + "grad_norm": 0.7448824602855045, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 39055 + }, + { + "epoch": 0.39056, + "grad_norm": 0.5019144323971751, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 39056 + }, + { + "epoch": 0.39057, + "grad_norm": 0.5228729989057496, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 39057 + }, + { + "epoch": 0.39058, + "grad_norm": 0.5918357092431565, + "learning_rate": 0.003, + "loss": 3.973, + "step": 39058 + }, + { + "epoch": 0.39059, + "grad_norm": 0.6379107714900727, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 39059 + }, + { + "epoch": 0.3906, + "grad_norm": 0.6682691657511521, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 39060 + }, + { + "epoch": 0.39061, + "grad_norm": 0.7157042912772753, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 39061 + }, + { + "epoch": 0.39062, + "grad_norm": 0.7834221316149629, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 39062 + }, + { + "epoch": 0.39063, + "grad_norm": 0.8019084532992358, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 39063 + }, + { + "epoch": 0.39064, + "grad_norm": 0.8268968505051802, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 39064 + }, + { + "epoch": 0.39065, + "grad_norm": 0.6965766604939877, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 39065 + }, + { + "epoch": 0.39066, + "grad_norm": 0.6447558488456822, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 39066 + }, + { + "epoch": 0.39067, + "grad_norm": 0.702638997969116, + "learning_rate": 0.003, + "loss": 3.926, + "step": 39067 + }, + { + "epoch": 0.39068, + "grad_norm": 0.6925928877319604, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 39068 + }, + { + "epoch": 0.39069, + "grad_norm": 0.681267406566776, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 39069 + }, + { + "epoch": 0.3907, + "grad_norm": 0.6939740898394271, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 39070 + }, + { + "epoch": 0.39071, + "grad_norm": 0.8201268888657514, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 39071 + }, + { + "epoch": 0.39072, + "grad_norm": 0.8133829516109613, + "learning_rate": 0.003, + "loss": 3.9417, + "step": 39072 + }, + { + "epoch": 0.39073, + "grad_norm": 0.7063854325595821, + "learning_rate": 0.003, + "loss": 3.9379, + "step": 39073 + }, + { + "epoch": 0.39074, + "grad_norm": 0.8314502261945391, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 39074 + }, + { + "epoch": 0.39075, + "grad_norm": 0.8673311155371918, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 39075 + }, + { + "epoch": 0.39076, + "grad_norm": 0.809458581654585, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 39076 + }, + { + "epoch": 0.39077, + "grad_norm": 0.798244273729946, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 39077 + }, + { + "epoch": 0.39078, + "grad_norm": 0.8190039667474184, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 39078 + }, + { + "epoch": 0.39079, + "grad_norm": 0.7784784787026008, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 39079 + }, + { + "epoch": 0.3908, + "grad_norm": 0.8569116858519544, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 39080 + }, + { + "epoch": 0.39081, + "grad_norm": 0.9485468527546488, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 39081 + }, + { + "epoch": 0.39082, + "grad_norm": 1.2431703909102976, + "learning_rate": 0.003, + "loss": 4.0284, + "step": 39082 + }, + { + "epoch": 0.39083, + "grad_norm": 0.8286274431358079, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 39083 + }, + { + "epoch": 0.39084, + "grad_norm": 0.7222030825418273, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 39084 + }, + { + "epoch": 0.39085, + "grad_norm": 0.7330112901475097, + "learning_rate": 0.003, + "loss": 3.963, + "step": 39085 + }, + { + "epoch": 0.39086, + "grad_norm": 0.7041205638554033, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 39086 + }, + { + "epoch": 0.39087, + "grad_norm": 0.7947428944122105, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 39087 + }, + { + "epoch": 0.39088, + "grad_norm": 0.9335695515322476, + "learning_rate": 0.003, + "loss": 3.998, + "step": 39088 + }, + { + "epoch": 0.39089, + "grad_norm": 0.9813308205406347, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 39089 + }, + { + "epoch": 0.3909, + "grad_norm": 0.9159281305708327, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 39090 + }, + { + "epoch": 0.39091, + "grad_norm": 0.949109649184094, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 39091 + }, + { + "epoch": 0.39092, + "grad_norm": 0.881545104463364, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 39092 + }, + { + "epoch": 0.39093, + "grad_norm": 0.8787582318311751, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 39093 + }, + { + "epoch": 0.39094, + "grad_norm": 0.8359862940691538, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 39094 + }, + { + "epoch": 0.39095, + "grad_norm": 0.8168964399755546, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 39095 + }, + { + "epoch": 0.39096, + "grad_norm": 0.7517558994586813, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 39096 + }, + { + "epoch": 0.39097, + "grad_norm": 0.8434582099587458, + "learning_rate": 0.003, + "loss": 3.986, + "step": 39097 + }, + { + "epoch": 0.39098, + "grad_norm": 0.915572339219069, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 39098 + }, + { + "epoch": 0.39099, + "grad_norm": 0.9333581683860549, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 39099 + }, + { + "epoch": 0.391, + "grad_norm": 0.8567516691177677, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 39100 + }, + { + "epoch": 0.39101, + "grad_norm": 0.6805622544197193, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 39101 + }, + { + "epoch": 0.39102, + "grad_norm": 0.7104008426697064, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 39102 + }, + { + "epoch": 0.39103, + "grad_norm": 0.7695907161194934, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 39103 + }, + { + "epoch": 0.39104, + "grad_norm": 0.662538528924418, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 39104 + }, + { + "epoch": 0.39105, + "grad_norm": 0.6168535364962525, + "learning_rate": 0.003, + "loss": 3.998, + "step": 39105 + }, + { + "epoch": 0.39106, + "grad_norm": 0.621678019348185, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 39106 + }, + { + "epoch": 0.39107, + "grad_norm": 0.6424808833840391, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 39107 + }, + { + "epoch": 0.39108, + "grad_norm": 0.5886375778998183, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 39108 + }, + { + "epoch": 0.39109, + "grad_norm": 0.6219980492218924, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 39109 + }, + { + "epoch": 0.3911, + "grad_norm": 0.584851710694595, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 39110 + }, + { + "epoch": 0.39111, + "grad_norm": 0.5424857907951587, + "learning_rate": 0.003, + "loss": 3.993, + "step": 39111 + }, + { + "epoch": 0.39112, + "grad_norm": 0.5734868821494784, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 39112 + }, + { + "epoch": 0.39113, + "grad_norm": 0.5993944014464961, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 39113 + }, + { + "epoch": 0.39114, + "grad_norm": 0.6221722276581553, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 39114 + }, + { + "epoch": 0.39115, + "grad_norm": 0.7571897371937836, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 39115 + }, + { + "epoch": 0.39116, + "grad_norm": 0.9654800243551432, + "learning_rate": 0.003, + "loss": 3.968, + "step": 39116 + }, + { + "epoch": 0.39117, + "grad_norm": 1.2711368328260175, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 39117 + }, + { + "epoch": 0.39118, + "grad_norm": 0.7304649679052327, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 39118 + }, + { + "epoch": 0.39119, + "grad_norm": 0.7104608080037872, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 39119 + }, + { + "epoch": 0.3912, + "grad_norm": 0.8363868520055177, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 39120 + }, + { + "epoch": 0.39121, + "grad_norm": 1.169766870213273, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 39121 + }, + { + "epoch": 0.39122, + "grad_norm": 1.0118732542623443, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 39122 + }, + { + "epoch": 0.39123, + "grad_norm": 0.829082320683579, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 39123 + }, + { + "epoch": 0.39124, + "grad_norm": 0.689913396084513, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 39124 + }, + { + "epoch": 0.39125, + "grad_norm": 0.8894388410155515, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 39125 + }, + { + "epoch": 0.39126, + "grad_norm": 0.9872536893804993, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 39126 + }, + { + "epoch": 0.39127, + "grad_norm": 0.9373769894376722, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 39127 + }, + { + "epoch": 0.39128, + "grad_norm": 0.9510875311385742, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 39128 + }, + { + "epoch": 0.39129, + "grad_norm": 0.9412626496593252, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 39129 + }, + { + "epoch": 0.3913, + "grad_norm": 0.947718875948672, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 39130 + }, + { + "epoch": 0.39131, + "grad_norm": 0.962333801654647, + "learning_rate": 0.003, + "loss": 4.026, + "step": 39131 + }, + { + "epoch": 0.39132, + "grad_norm": 0.8556741077660458, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 39132 + }, + { + "epoch": 0.39133, + "grad_norm": 0.777235325378262, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 39133 + }, + { + "epoch": 0.39134, + "grad_norm": 0.7060865667879703, + "learning_rate": 0.003, + "loss": 3.995, + "step": 39134 + }, + { + "epoch": 0.39135, + "grad_norm": 0.6931383605426656, + "learning_rate": 0.003, + "loss": 3.965, + "step": 39135 + }, + { + "epoch": 0.39136, + "grad_norm": 0.6902455897937122, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 39136 + }, + { + "epoch": 0.39137, + "grad_norm": 0.8058633768336421, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 39137 + }, + { + "epoch": 0.39138, + "grad_norm": 0.8643122239940175, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 39138 + }, + { + "epoch": 0.39139, + "grad_norm": 0.9560498948374605, + "learning_rate": 0.003, + "loss": 4.003, + "step": 39139 + }, + { + "epoch": 0.3914, + "grad_norm": 1.0081287092398181, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 39140 + }, + { + "epoch": 0.39141, + "grad_norm": 1.009298415468302, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 39141 + }, + { + "epoch": 0.39142, + "grad_norm": 0.9364365602264737, + "learning_rate": 0.003, + "loss": 4.021, + "step": 39142 + }, + { + "epoch": 0.39143, + "grad_norm": 0.8478396551007293, + "learning_rate": 0.003, + "loss": 3.992, + "step": 39143 + }, + { + "epoch": 0.39144, + "grad_norm": 0.8067913782634621, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 39144 + }, + { + "epoch": 0.39145, + "grad_norm": 0.6889561596195167, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 39145 + }, + { + "epoch": 0.39146, + "grad_norm": 0.6898146087805891, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 39146 + }, + { + "epoch": 0.39147, + "grad_norm": 0.6249679404742001, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 39147 + }, + { + "epoch": 0.39148, + "grad_norm": 0.6259710027974399, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 39148 + }, + { + "epoch": 0.39149, + "grad_norm": 0.6293431884143242, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 39149 + }, + { + "epoch": 0.3915, + "grad_norm": 0.6759862536027652, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 39150 + }, + { + "epoch": 0.39151, + "grad_norm": 0.6930184402577146, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 39151 + }, + { + "epoch": 0.39152, + "grad_norm": 0.8103869970643075, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 39152 + }, + { + "epoch": 0.39153, + "grad_norm": 0.8935884891567846, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 39153 + }, + { + "epoch": 0.39154, + "grad_norm": 0.9983055857180514, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 39154 + }, + { + "epoch": 0.39155, + "grad_norm": 0.9489294995408127, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 39155 + }, + { + "epoch": 0.39156, + "grad_norm": 0.7546963131062612, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 39156 + }, + { + "epoch": 0.39157, + "grad_norm": 0.6915279719077833, + "learning_rate": 0.003, + "loss": 3.984, + "step": 39157 + }, + { + "epoch": 0.39158, + "grad_norm": 0.596978117771879, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 39158 + }, + { + "epoch": 0.39159, + "grad_norm": 0.6106665617542333, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 39159 + }, + { + "epoch": 0.3916, + "grad_norm": 0.5567813304475554, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 39160 + }, + { + "epoch": 0.39161, + "grad_norm": 0.6679911864243342, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 39161 + }, + { + "epoch": 0.39162, + "grad_norm": 0.648262411908176, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 39162 + }, + { + "epoch": 0.39163, + "grad_norm": 0.6618994230956817, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 39163 + }, + { + "epoch": 0.39164, + "grad_norm": 0.6224430068588402, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 39164 + }, + { + "epoch": 0.39165, + "grad_norm": 0.7056821088520688, + "learning_rate": 0.003, + "loss": 3.976, + "step": 39165 + }, + { + "epoch": 0.39166, + "grad_norm": 0.7580421117402342, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 39166 + }, + { + "epoch": 0.39167, + "grad_norm": 0.8324636415922649, + "learning_rate": 0.003, + "loss": 4.004, + "step": 39167 + }, + { + "epoch": 0.39168, + "grad_norm": 0.8293958038531865, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 39168 + }, + { + "epoch": 0.39169, + "grad_norm": 0.7937177330832668, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 39169 + }, + { + "epoch": 0.3917, + "grad_norm": 0.7840622563738237, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 39170 + }, + { + "epoch": 0.39171, + "grad_norm": 0.771320832292232, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 39171 + }, + { + "epoch": 0.39172, + "grad_norm": 0.7534421072078468, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 39172 + }, + { + "epoch": 0.39173, + "grad_norm": 0.8275189038525906, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 39173 + }, + { + "epoch": 0.39174, + "grad_norm": 0.883673149420201, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 39174 + }, + { + "epoch": 0.39175, + "grad_norm": 0.9792954566185439, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 39175 + }, + { + "epoch": 0.39176, + "grad_norm": 1.134443008282323, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 39176 + }, + { + "epoch": 0.39177, + "grad_norm": 0.805291105619837, + "learning_rate": 0.003, + "loss": 3.978, + "step": 39177 + }, + { + "epoch": 0.39178, + "grad_norm": 0.8077557181664825, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 39178 + }, + { + "epoch": 0.39179, + "grad_norm": 0.6290124045248553, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 39179 + }, + { + "epoch": 0.3918, + "grad_norm": 0.6481844929443464, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 39180 + }, + { + "epoch": 0.39181, + "grad_norm": 0.6209764840781353, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 39181 + }, + { + "epoch": 0.39182, + "grad_norm": 0.6858863217682474, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 39182 + }, + { + "epoch": 0.39183, + "grad_norm": 0.5892617611496179, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 39183 + }, + { + "epoch": 0.39184, + "grad_norm": 0.5421215300659908, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 39184 + }, + { + "epoch": 0.39185, + "grad_norm": 0.631691103967474, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 39185 + }, + { + "epoch": 0.39186, + "grad_norm": 0.7052938865269501, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 39186 + }, + { + "epoch": 0.39187, + "grad_norm": 0.8723329980075487, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 39187 + }, + { + "epoch": 0.39188, + "grad_norm": 1.1135278632571732, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 39188 + }, + { + "epoch": 0.39189, + "grad_norm": 0.9645283018397451, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 39189 + }, + { + "epoch": 0.3919, + "grad_norm": 1.0043871092041623, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 39190 + }, + { + "epoch": 0.39191, + "grad_norm": 0.9012480763462661, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 39191 + }, + { + "epoch": 0.39192, + "grad_norm": 0.8794043477711514, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 39192 + }, + { + "epoch": 0.39193, + "grad_norm": 0.8950946258033937, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 39193 + }, + { + "epoch": 0.39194, + "grad_norm": 0.8438713315352648, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 39194 + }, + { + "epoch": 0.39195, + "grad_norm": 0.7689094469879424, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 39195 + }, + { + "epoch": 0.39196, + "grad_norm": 0.7240307421450913, + "learning_rate": 0.003, + "loss": 3.978, + "step": 39196 + }, + { + "epoch": 0.39197, + "grad_norm": 0.7600978178146329, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 39197 + }, + { + "epoch": 0.39198, + "grad_norm": 0.7591376376352817, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 39198 + }, + { + "epoch": 0.39199, + "grad_norm": 0.7840709073607364, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 39199 + }, + { + "epoch": 0.392, + "grad_norm": 0.9793529562112981, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 39200 + }, + { + "epoch": 0.39201, + "grad_norm": 1.0785027485460845, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 39201 + }, + { + "epoch": 0.39202, + "grad_norm": 0.7606023060768752, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 39202 + }, + { + "epoch": 0.39203, + "grad_norm": 0.7735916094928301, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 39203 + }, + { + "epoch": 0.39204, + "grad_norm": 0.7776117681815669, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 39204 + }, + { + "epoch": 0.39205, + "grad_norm": 0.7962984654573197, + "learning_rate": 0.003, + "loss": 4.0, + "step": 39205 + }, + { + "epoch": 0.39206, + "grad_norm": 0.8794050766610625, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 39206 + }, + { + "epoch": 0.39207, + "grad_norm": 0.8813928290945543, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 39207 + }, + { + "epoch": 0.39208, + "grad_norm": 0.8855296456209529, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 39208 + }, + { + "epoch": 0.39209, + "grad_norm": 0.943856420172909, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 39209 + }, + { + "epoch": 0.3921, + "grad_norm": 0.9327230176759909, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 39210 + }, + { + "epoch": 0.39211, + "grad_norm": 0.804696838382851, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 39211 + }, + { + "epoch": 0.39212, + "grad_norm": 0.6790496793443822, + "learning_rate": 0.003, + "loss": 3.9445, + "step": 39212 + }, + { + "epoch": 0.39213, + "grad_norm": 0.6679448172228053, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 39213 + }, + { + "epoch": 0.39214, + "grad_norm": 0.6580102322015433, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 39214 + }, + { + "epoch": 0.39215, + "grad_norm": 0.6828926861227175, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 39215 + }, + { + "epoch": 0.39216, + "grad_norm": 0.738203671790916, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 39216 + }, + { + "epoch": 0.39217, + "grad_norm": 0.6467786951894816, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 39217 + }, + { + "epoch": 0.39218, + "grad_norm": 0.6697474321776857, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 39218 + }, + { + "epoch": 0.39219, + "grad_norm": 0.6641843582328233, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 39219 + }, + { + "epoch": 0.3922, + "grad_norm": 0.7850740697743089, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 39220 + }, + { + "epoch": 0.39221, + "grad_norm": 0.9284622185123602, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 39221 + }, + { + "epoch": 0.39222, + "grad_norm": 1.0623309966871095, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 39222 + }, + { + "epoch": 0.39223, + "grad_norm": 0.922468947152912, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 39223 + }, + { + "epoch": 0.39224, + "grad_norm": 0.7796063331070627, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 39224 + }, + { + "epoch": 0.39225, + "grad_norm": 0.7773942512074319, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 39225 + }, + { + "epoch": 0.39226, + "grad_norm": 0.8010298431628535, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 39226 + }, + { + "epoch": 0.39227, + "grad_norm": 0.891535187000427, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 39227 + }, + { + "epoch": 0.39228, + "grad_norm": 0.8869685720759525, + "learning_rate": 0.003, + "loss": 3.971, + "step": 39228 + }, + { + "epoch": 0.39229, + "grad_norm": 0.8416464691214476, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 39229 + }, + { + "epoch": 0.3923, + "grad_norm": 0.847914143489617, + "learning_rate": 0.003, + "loss": 3.9391, + "step": 39230 + }, + { + "epoch": 0.39231, + "grad_norm": 0.9002971289697099, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 39231 + }, + { + "epoch": 0.39232, + "grad_norm": 0.8561242819607001, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 39232 + }, + { + "epoch": 0.39233, + "grad_norm": 0.7512083370128855, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 39233 + }, + { + "epoch": 0.39234, + "grad_norm": 0.6867832112181383, + "learning_rate": 0.003, + "loss": 3.9346, + "step": 39234 + }, + { + "epoch": 0.39235, + "grad_norm": 0.6840407326420207, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 39235 + }, + { + "epoch": 0.39236, + "grad_norm": 0.6144909083692798, + "learning_rate": 0.003, + "loss": 3.977, + "step": 39236 + }, + { + "epoch": 0.39237, + "grad_norm": 0.7463927070336841, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 39237 + }, + { + "epoch": 0.39238, + "grad_norm": 0.8035112894119967, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 39238 + }, + { + "epoch": 0.39239, + "grad_norm": 0.8709326001169083, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 39239 + }, + { + "epoch": 0.3924, + "grad_norm": 0.7464140281450297, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 39240 + }, + { + "epoch": 0.39241, + "grad_norm": 0.6042589433713436, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 39241 + }, + { + "epoch": 0.39242, + "grad_norm": 0.5440922908069884, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 39242 + }, + { + "epoch": 0.39243, + "grad_norm": 0.6137780439131494, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 39243 + }, + { + "epoch": 0.39244, + "grad_norm": 0.6714163334086062, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 39244 + }, + { + "epoch": 0.39245, + "grad_norm": 0.6772279949278148, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 39245 + }, + { + "epoch": 0.39246, + "grad_norm": 0.7047399869178195, + "learning_rate": 0.003, + "loss": 3.9408, + "step": 39246 + }, + { + "epoch": 0.39247, + "grad_norm": 0.7169940861173251, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 39247 + }, + { + "epoch": 0.39248, + "grad_norm": 0.7383838560190931, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 39248 + }, + { + "epoch": 0.39249, + "grad_norm": 0.7932779729099412, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 39249 + }, + { + "epoch": 0.3925, + "grad_norm": 0.9065407564084702, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 39250 + }, + { + "epoch": 0.39251, + "grad_norm": 1.0479247551906063, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 39251 + }, + { + "epoch": 0.39252, + "grad_norm": 1.0933003898252378, + "learning_rate": 0.003, + "loss": 3.971, + "step": 39252 + }, + { + "epoch": 0.39253, + "grad_norm": 1.0549726459002904, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 39253 + }, + { + "epoch": 0.39254, + "grad_norm": 0.8890517204793957, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 39254 + }, + { + "epoch": 0.39255, + "grad_norm": 0.9263771804254171, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 39255 + }, + { + "epoch": 0.39256, + "grad_norm": 0.9071297239435373, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 39256 + }, + { + "epoch": 0.39257, + "grad_norm": 0.948915612891233, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 39257 + }, + { + "epoch": 0.39258, + "grad_norm": 1.0655631408898576, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 39258 + }, + { + "epoch": 0.39259, + "grad_norm": 1.1406767595434175, + "learning_rate": 0.003, + "loss": 4.0309, + "step": 39259 + }, + { + "epoch": 0.3926, + "grad_norm": 0.7881659627566664, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 39260 + }, + { + "epoch": 0.39261, + "grad_norm": 0.6486235903229407, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 39261 + }, + { + "epoch": 0.39262, + "grad_norm": 0.6597947920272167, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 39262 + }, + { + "epoch": 0.39263, + "grad_norm": 0.5983246935577645, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 39263 + }, + { + "epoch": 0.39264, + "grad_norm": 0.6591777710823548, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 39264 + }, + { + "epoch": 0.39265, + "grad_norm": 0.7156801340106215, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 39265 + }, + { + "epoch": 0.39266, + "grad_norm": 0.7097377808190172, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 39266 + }, + { + "epoch": 0.39267, + "grad_norm": 0.6263395433464036, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 39267 + }, + { + "epoch": 0.39268, + "grad_norm": 0.6231501494047903, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 39268 + }, + { + "epoch": 0.39269, + "grad_norm": 0.6029680950673826, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 39269 + }, + { + "epoch": 0.3927, + "grad_norm": 0.6088125591361052, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 39270 + }, + { + "epoch": 0.39271, + "grad_norm": 0.6013772847230914, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 39271 + }, + { + "epoch": 0.39272, + "grad_norm": 0.6118547376990295, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 39272 + }, + { + "epoch": 0.39273, + "grad_norm": 0.5928887270625377, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 39273 + }, + { + "epoch": 0.39274, + "grad_norm": 0.564299444236864, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 39274 + }, + { + "epoch": 0.39275, + "grad_norm": 0.6336509489644842, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 39275 + }, + { + "epoch": 0.39276, + "grad_norm": 0.6265878341398852, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 39276 + }, + { + "epoch": 0.39277, + "grad_norm": 0.5364433307867613, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 39277 + }, + { + "epoch": 0.39278, + "grad_norm": 0.5409694042384479, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 39278 + }, + { + "epoch": 0.39279, + "grad_norm": 0.6174624282745492, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 39279 + }, + { + "epoch": 0.3928, + "grad_norm": 0.6740014927020924, + "learning_rate": 0.003, + "loss": 3.9416, + "step": 39280 + }, + { + "epoch": 0.39281, + "grad_norm": 0.6940410236860167, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 39281 + }, + { + "epoch": 0.39282, + "grad_norm": 0.742431854777002, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 39282 + }, + { + "epoch": 0.39283, + "grad_norm": 0.8509558865801771, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 39283 + }, + { + "epoch": 0.39284, + "grad_norm": 1.048890509421592, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 39284 + }, + { + "epoch": 0.39285, + "grad_norm": 1.002513269967796, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 39285 + }, + { + "epoch": 0.39286, + "grad_norm": 0.8975001882486616, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 39286 + }, + { + "epoch": 0.39287, + "grad_norm": 0.9060645072528445, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 39287 + }, + { + "epoch": 0.39288, + "grad_norm": 0.9508141756020326, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 39288 + }, + { + "epoch": 0.39289, + "grad_norm": 0.9964687072067765, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 39289 + }, + { + "epoch": 0.3929, + "grad_norm": 1.0820373578894695, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 39290 + }, + { + "epoch": 0.39291, + "grad_norm": 0.7824102710742279, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 39291 + }, + { + "epoch": 0.39292, + "grad_norm": 0.7393892391047351, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 39292 + }, + { + "epoch": 0.39293, + "grad_norm": 0.7460809969569709, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 39293 + }, + { + "epoch": 0.39294, + "grad_norm": 0.6784817384179362, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 39294 + }, + { + "epoch": 0.39295, + "grad_norm": 0.6738269756592506, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 39295 + }, + { + "epoch": 0.39296, + "grad_norm": 0.7534878276389347, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 39296 + }, + { + "epoch": 0.39297, + "grad_norm": 0.7062417941957252, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 39297 + }, + { + "epoch": 0.39298, + "grad_norm": 0.7044109511649445, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 39298 + }, + { + "epoch": 0.39299, + "grad_norm": 0.8067460735497186, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 39299 + }, + { + "epoch": 0.393, + "grad_norm": 0.7843026799464127, + "learning_rate": 0.003, + "loss": 3.969, + "step": 39300 + }, + { + "epoch": 0.39301, + "grad_norm": 0.6955808067112438, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 39301 + }, + { + "epoch": 0.39302, + "grad_norm": 0.7791736715792437, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 39302 + }, + { + "epoch": 0.39303, + "grad_norm": 0.8568116585011496, + "learning_rate": 0.003, + "loss": 4.013, + "step": 39303 + }, + { + "epoch": 0.39304, + "grad_norm": 0.9451218959501914, + "learning_rate": 0.003, + "loss": 3.993, + "step": 39304 + }, + { + "epoch": 0.39305, + "grad_norm": 1.0800515241558475, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 39305 + }, + { + "epoch": 0.39306, + "grad_norm": 0.9420462004586565, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 39306 + }, + { + "epoch": 0.39307, + "grad_norm": 0.9408875900088634, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 39307 + }, + { + "epoch": 0.39308, + "grad_norm": 0.8893899696972548, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 39308 + }, + { + "epoch": 0.39309, + "grad_norm": 0.7879171183046556, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 39309 + }, + { + "epoch": 0.3931, + "grad_norm": 0.7608931152780315, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 39310 + }, + { + "epoch": 0.39311, + "grad_norm": 0.7846719900621121, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 39311 + }, + { + "epoch": 0.39312, + "grad_norm": 0.8680316871084157, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 39312 + }, + { + "epoch": 0.39313, + "grad_norm": 0.8337800361334301, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 39313 + }, + { + "epoch": 0.39314, + "grad_norm": 0.7736535387361381, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 39314 + }, + { + "epoch": 0.39315, + "grad_norm": 0.7802420966633032, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 39315 + }, + { + "epoch": 0.39316, + "grad_norm": 0.7992417393940159, + "learning_rate": 0.003, + "loss": 3.985, + "step": 39316 + }, + { + "epoch": 0.39317, + "grad_norm": 0.9293970091094853, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 39317 + }, + { + "epoch": 0.39318, + "grad_norm": 0.9913622957585273, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 39318 + }, + { + "epoch": 0.39319, + "grad_norm": 0.9596041594529513, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 39319 + }, + { + "epoch": 0.3932, + "grad_norm": 0.7810030045879666, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 39320 + }, + { + "epoch": 0.39321, + "grad_norm": 0.8627957178255666, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 39321 + }, + { + "epoch": 0.39322, + "grad_norm": 0.7683381298492634, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 39322 + }, + { + "epoch": 0.39323, + "grad_norm": 0.808716258578542, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 39323 + }, + { + "epoch": 0.39324, + "grad_norm": 0.7905350128490027, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 39324 + }, + { + "epoch": 0.39325, + "grad_norm": 0.8030841169767494, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 39325 + }, + { + "epoch": 0.39326, + "grad_norm": 0.8264410488573631, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 39326 + }, + { + "epoch": 0.39327, + "grad_norm": 0.9477056000778545, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 39327 + }, + { + "epoch": 0.39328, + "grad_norm": 1.0196681079566816, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 39328 + }, + { + "epoch": 0.39329, + "grad_norm": 1.0728976812482633, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 39329 + }, + { + "epoch": 0.3933, + "grad_norm": 0.9744994809721518, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 39330 + }, + { + "epoch": 0.39331, + "grad_norm": 0.9321004020980749, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 39331 + }, + { + "epoch": 0.39332, + "grad_norm": 0.8597395326520532, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 39332 + }, + { + "epoch": 0.39333, + "grad_norm": 0.8310092029219119, + "learning_rate": 0.003, + "loss": 3.982, + "step": 39333 + }, + { + "epoch": 0.39334, + "grad_norm": 0.842692358839694, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 39334 + }, + { + "epoch": 0.39335, + "grad_norm": 0.7255795217285639, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 39335 + }, + { + "epoch": 0.39336, + "grad_norm": 0.7609850946708865, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 39336 + }, + { + "epoch": 0.39337, + "grad_norm": 0.7542662772961133, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 39337 + }, + { + "epoch": 0.39338, + "grad_norm": 0.7351987979537785, + "learning_rate": 0.003, + "loss": 3.978, + "step": 39338 + }, + { + "epoch": 0.39339, + "grad_norm": 0.6825681598923844, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 39339 + }, + { + "epoch": 0.3934, + "grad_norm": 0.6423046486053836, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 39340 + }, + { + "epoch": 0.39341, + "grad_norm": 0.6076723778767404, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 39341 + }, + { + "epoch": 0.39342, + "grad_norm": 0.5633377282924855, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 39342 + }, + { + "epoch": 0.39343, + "grad_norm": 0.5737862134687207, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 39343 + }, + { + "epoch": 0.39344, + "grad_norm": 0.5884237795552406, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 39344 + }, + { + "epoch": 0.39345, + "grad_norm": 0.5973107622961268, + "learning_rate": 0.003, + "loss": 3.994, + "step": 39345 + }, + { + "epoch": 0.39346, + "grad_norm": 0.5883820252622808, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 39346 + }, + { + "epoch": 0.39347, + "grad_norm": 0.6731948674857925, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 39347 + }, + { + "epoch": 0.39348, + "grad_norm": 0.7353580641455384, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 39348 + }, + { + "epoch": 0.39349, + "grad_norm": 0.7342363930206881, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 39349 + }, + { + "epoch": 0.3935, + "grad_norm": 0.6850106072312787, + "learning_rate": 0.003, + "loss": 3.978, + "step": 39350 + }, + { + "epoch": 0.39351, + "grad_norm": 0.8254468293188646, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 39351 + }, + { + "epoch": 0.39352, + "grad_norm": 0.9436812495731799, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 39352 + }, + { + "epoch": 0.39353, + "grad_norm": 1.174392478693455, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 39353 + }, + { + "epoch": 0.39354, + "grad_norm": 0.9726668590376157, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 39354 + }, + { + "epoch": 0.39355, + "grad_norm": 0.9772253018736792, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 39355 + }, + { + "epoch": 0.39356, + "grad_norm": 1.179808489852096, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 39356 + }, + { + "epoch": 0.39357, + "grad_norm": 0.9111053147360924, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 39357 + }, + { + "epoch": 0.39358, + "grad_norm": 0.8297834119109093, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 39358 + }, + { + "epoch": 0.39359, + "grad_norm": 0.785975999022587, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 39359 + }, + { + "epoch": 0.3936, + "grad_norm": 0.7105700136768867, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 39360 + }, + { + "epoch": 0.39361, + "grad_norm": 0.6123167669251738, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 39361 + }, + { + "epoch": 0.39362, + "grad_norm": 0.5857275635389092, + "learning_rate": 0.003, + "loss": 3.981, + "step": 39362 + }, + { + "epoch": 0.39363, + "grad_norm": 0.6258875977715957, + "learning_rate": 0.003, + "loss": 3.967, + "step": 39363 + }, + { + "epoch": 0.39364, + "grad_norm": 0.5988255068793353, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 39364 + }, + { + "epoch": 0.39365, + "grad_norm": 0.645804720617894, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 39365 + }, + { + "epoch": 0.39366, + "grad_norm": 0.5980083557151162, + "learning_rate": 0.003, + "loss": 3.989, + "step": 39366 + }, + { + "epoch": 0.39367, + "grad_norm": 0.6668284770095849, + "learning_rate": 0.003, + "loss": 3.973, + "step": 39367 + }, + { + "epoch": 0.39368, + "grad_norm": 0.7755000935612606, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 39368 + }, + { + "epoch": 0.39369, + "grad_norm": 0.7133241156887473, + "learning_rate": 0.003, + "loss": 3.973, + "step": 39369 + }, + { + "epoch": 0.3937, + "grad_norm": 0.6500714960008103, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 39370 + }, + { + "epoch": 0.39371, + "grad_norm": 0.7022677814930418, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 39371 + }, + { + "epoch": 0.39372, + "grad_norm": 0.8030546394155174, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 39372 + }, + { + "epoch": 0.39373, + "grad_norm": 0.8637863657844953, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 39373 + }, + { + "epoch": 0.39374, + "grad_norm": 0.881657680411266, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 39374 + }, + { + "epoch": 0.39375, + "grad_norm": 0.7873481464858261, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 39375 + }, + { + "epoch": 0.39376, + "grad_norm": 0.6446876533459455, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 39376 + }, + { + "epoch": 0.39377, + "grad_norm": 0.701391181800368, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 39377 + }, + { + "epoch": 0.39378, + "grad_norm": 0.7547456172246942, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 39378 + }, + { + "epoch": 0.39379, + "grad_norm": 0.7834527499655657, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 39379 + }, + { + "epoch": 0.3938, + "grad_norm": 0.827882924302223, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 39380 + }, + { + "epoch": 0.39381, + "grad_norm": 0.8609957941280557, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 39381 + }, + { + "epoch": 0.39382, + "grad_norm": 0.777450589303046, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 39382 + }, + { + "epoch": 0.39383, + "grad_norm": 0.7227002904826125, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 39383 + }, + { + "epoch": 0.39384, + "grad_norm": 0.9312109504499226, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 39384 + }, + { + "epoch": 0.39385, + "grad_norm": 1.3028292683442078, + "learning_rate": 0.003, + "loss": 3.987, + "step": 39385 + }, + { + "epoch": 0.39386, + "grad_norm": 0.966376586630482, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 39386 + }, + { + "epoch": 0.39387, + "grad_norm": 0.8505594530205914, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 39387 + }, + { + "epoch": 0.39388, + "grad_norm": 0.921986602234735, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 39388 + }, + { + "epoch": 0.39389, + "grad_norm": 0.9646549370261779, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 39389 + }, + { + "epoch": 0.3939, + "grad_norm": 1.0194664398844593, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 39390 + }, + { + "epoch": 0.39391, + "grad_norm": 0.9948118358891712, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 39391 + }, + { + "epoch": 0.39392, + "grad_norm": 0.8964183534840556, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 39392 + }, + { + "epoch": 0.39393, + "grad_norm": 0.8957738170926395, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 39393 + }, + { + "epoch": 0.39394, + "grad_norm": 0.9109749579327798, + "learning_rate": 0.003, + "loss": 3.987, + "step": 39394 + }, + { + "epoch": 0.39395, + "grad_norm": 0.7570093528818572, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 39395 + }, + { + "epoch": 0.39396, + "grad_norm": 0.708193863741981, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 39396 + }, + { + "epoch": 0.39397, + "grad_norm": 0.7462900195422356, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 39397 + }, + { + "epoch": 0.39398, + "grad_norm": 0.661503630646736, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 39398 + }, + { + "epoch": 0.39399, + "grad_norm": 0.6604655778368552, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 39399 + }, + { + "epoch": 0.394, + "grad_norm": 0.7173675279881095, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 39400 + }, + { + "epoch": 0.39401, + "grad_norm": 0.817177367944073, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 39401 + }, + { + "epoch": 0.39402, + "grad_norm": 0.9587932895571465, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 39402 + }, + { + "epoch": 0.39403, + "grad_norm": 0.9887268186137612, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 39403 + }, + { + "epoch": 0.39404, + "grad_norm": 1.0018951393777984, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 39404 + }, + { + "epoch": 0.39405, + "grad_norm": 0.8541276644043119, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 39405 + }, + { + "epoch": 0.39406, + "grad_norm": 0.6922789873637166, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 39406 + }, + { + "epoch": 0.39407, + "grad_norm": 0.7668670788586334, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 39407 + }, + { + "epoch": 0.39408, + "grad_norm": 0.8733158719640729, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 39408 + }, + { + "epoch": 0.39409, + "grad_norm": 0.9599811379391038, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 39409 + }, + { + "epoch": 0.3941, + "grad_norm": 0.7086132682136488, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 39410 + }, + { + "epoch": 0.39411, + "grad_norm": 0.6203003411963648, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 39411 + }, + { + "epoch": 0.39412, + "grad_norm": 0.5595036202447506, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 39412 + }, + { + "epoch": 0.39413, + "grad_norm": 0.5239086832322637, + "learning_rate": 0.003, + "loss": 3.945, + "step": 39413 + }, + { + "epoch": 0.39414, + "grad_norm": 0.4402714622367152, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 39414 + }, + { + "epoch": 0.39415, + "grad_norm": 0.49076748799338704, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 39415 + }, + { + "epoch": 0.39416, + "grad_norm": 0.5235694687628076, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 39416 + }, + { + "epoch": 0.39417, + "grad_norm": 0.5132881284741135, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 39417 + }, + { + "epoch": 0.39418, + "grad_norm": 0.5107799252828404, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 39418 + }, + { + "epoch": 0.39419, + "grad_norm": 0.5791925136547587, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 39419 + }, + { + "epoch": 0.3942, + "grad_norm": 0.7157073433777125, + "learning_rate": 0.003, + "loss": 3.972, + "step": 39420 + }, + { + "epoch": 0.39421, + "grad_norm": 0.7958094720351333, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 39421 + }, + { + "epoch": 0.39422, + "grad_norm": 0.9839435315077569, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 39422 + }, + { + "epoch": 0.39423, + "grad_norm": 1.0711439545643242, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 39423 + }, + { + "epoch": 0.39424, + "grad_norm": 0.7935072525216705, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 39424 + }, + { + "epoch": 0.39425, + "grad_norm": 0.6948917230353294, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 39425 + }, + { + "epoch": 0.39426, + "grad_norm": 0.701914277778327, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 39426 + }, + { + "epoch": 0.39427, + "grad_norm": 0.7714230807167858, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 39427 + }, + { + "epoch": 0.39428, + "grad_norm": 0.907174666034603, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 39428 + }, + { + "epoch": 0.39429, + "grad_norm": 0.9978033568976652, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 39429 + }, + { + "epoch": 0.3943, + "grad_norm": 1.0367800882232996, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 39430 + }, + { + "epoch": 0.39431, + "grad_norm": 1.1207652433494075, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 39431 + }, + { + "epoch": 0.39432, + "grad_norm": 0.9892992264755062, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 39432 + }, + { + "epoch": 0.39433, + "grad_norm": 0.917626187544469, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 39433 + }, + { + "epoch": 0.39434, + "grad_norm": 0.7008165037555155, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 39434 + }, + { + "epoch": 0.39435, + "grad_norm": 0.6230099433834483, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 39435 + }, + { + "epoch": 0.39436, + "grad_norm": 0.6745586402137151, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 39436 + }, + { + "epoch": 0.39437, + "grad_norm": 0.6345961913820062, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 39437 + }, + { + "epoch": 0.39438, + "grad_norm": 0.5207378845093672, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 39438 + }, + { + "epoch": 0.39439, + "grad_norm": 0.5583189465133712, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 39439 + }, + { + "epoch": 0.3944, + "grad_norm": 0.5742164916788268, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 39440 + }, + { + "epoch": 0.39441, + "grad_norm": 0.5785664678325063, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 39441 + }, + { + "epoch": 0.39442, + "grad_norm": 0.6527977745767359, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 39442 + }, + { + "epoch": 0.39443, + "grad_norm": 0.7278957992848901, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 39443 + }, + { + "epoch": 0.39444, + "grad_norm": 0.880344544233569, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 39444 + }, + { + "epoch": 0.39445, + "grad_norm": 1.0009368315757468, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 39445 + }, + { + "epoch": 0.39446, + "grad_norm": 1.0087895180917392, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 39446 + }, + { + "epoch": 0.39447, + "grad_norm": 0.9134653212630681, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 39447 + }, + { + "epoch": 0.39448, + "grad_norm": 0.7874177960489602, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 39448 + }, + { + "epoch": 0.39449, + "grad_norm": 0.7556555681692293, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 39449 + }, + { + "epoch": 0.3945, + "grad_norm": 0.6990315720913007, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 39450 + }, + { + "epoch": 0.39451, + "grad_norm": 0.7216855762004037, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 39451 + }, + { + "epoch": 0.39452, + "grad_norm": 0.690558353739745, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 39452 + }, + { + "epoch": 0.39453, + "grad_norm": 0.6584560630756278, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 39453 + }, + { + "epoch": 0.39454, + "grad_norm": 0.5849788021287344, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 39454 + }, + { + "epoch": 0.39455, + "grad_norm": 0.6072689339403552, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 39455 + }, + { + "epoch": 0.39456, + "grad_norm": 0.6457386343659395, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 39456 + }, + { + "epoch": 0.39457, + "grad_norm": 0.5911899457388764, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 39457 + }, + { + "epoch": 0.39458, + "grad_norm": 0.6864007095831071, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 39458 + }, + { + "epoch": 0.39459, + "grad_norm": 0.8724875463524688, + "learning_rate": 0.003, + "loss": 3.961, + "step": 39459 + }, + { + "epoch": 0.3946, + "grad_norm": 1.0928221367692474, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 39460 + }, + { + "epoch": 0.39461, + "grad_norm": 0.9849038025110116, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 39461 + }, + { + "epoch": 0.39462, + "grad_norm": 0.989415632669842, + "learning_rate": 0.003, + "loss": 3.995, + "step": 39462 + }, + { + "epoch": 0.39463, + "grad_norm": 0.8097127942985104, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 39463 + }, + { + "epoch": 0.39464, + "grad_norm": 0.7299770925194492, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 39464 + }, + { + "epoch": 0.39465, + "grad_norm": 0.7422828986141008, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 39465 + }, + { + "epoch": 0.39466, + "grad_norm": 0.6990134522772092, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 39466 + }, + { + "epoch": 0.39467, + "grad_norm": 0.6845952464738084, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 39467 + }, + { + "epoch": 0.39468, + "grad_norm": 0.743845171541107, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 39468 + }, + { + "epoch": 0.39469, + "grad_norm": 0.8097474921534514, + "learning_rate": 0.003, + "loss": 3.954, + "step": 39469 + }, + { + "epoch": 0.3947, + "grad_norm": 0.7127090297853852, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 39470 + }, + { + "epoch": 0.39471, + "grad_norm": 0.6889256319328744, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 39471 + }, + { + "epoch": 0.39472, + "grad_norm": 0.7237060486703122, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 39472 + }, + { + "epoch": 0.39473, + "grad_norm": 0.759677737834576, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 39473 + }, + { + "epoch": 0.39474, + "grad_norm": 0.721952833922785, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 39474 + }, + { + "epoch": 0.39475, + "grad_norm": 0.7519518545443924, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 39475 + }, + { + "epoch": 0.39476, + "grad_norm": 0.824894959613605, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 39476 + }, + { + "epoch": 0.39477, + "grad_norm": 0.8500740068627425, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 39477 + }, + { + "epoch": 0.39478, + "grad_norm": 0.8764049575610233, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 39478 + }, + { + "epoch": 0.39479, + "grad_norm": 0.8875527881115001, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 39479 + }, + { + "epoch": 0.3948, + "grad_norm": 0.7986446048221564, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 39480 + }, + { + "epoch": 0.39481, + "grad_norm": 0.8190058962058898, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 39481 + }, + { + "epoch": 0.39482, + "grad_norm": 0.9076710717735788, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 39482 + }, + { + "epoch": 0.39483, + "grad_norm": 1.104616791269931, + "learning_rate": 0.003, + "loss": 3.972, + "step": 39483 + }, + { + "epoch": 0.39484, + "grad_norm": 1.0520133993907472, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 39484 + }, + { + "epoch": 0.39485, + "grad_norm": 0.8982506214260291, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 39485 + }, + { + "epoch": 0.39486, + "grad_norm": 0.741540784920957, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 39486 + }, + { + "epoch": 0.39487, + "grad_norm": 0.7617462736814523, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 39487 + }, + { + "epoch": 0.39488, + "grad_norm": 0.8583696954707549, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 39488 + }, + { + "epoch": 0.39489, + "grad_norm": 0.8237169391895462, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 39489 + }, + { + "epoch": 0.3949, + "grad_norm": 0.7455096644253006, + "learning_rate": 0.003, + "loss": 3.988, + "step": 39490 + }, + { + "epoch": 0.39491, + "grad_norm": 0.6906269568321391, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 39491 + }, + { + "epoch": 0.39492, + "grad_norm": 0.677722484969479, + "learning_rate": 0.003, + "loss": 3.998, + "step": 39492 + }, + { + "epoch": 0.39493, + "grad_norm": 0.6269455347420425, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 39493 + }, + { + "epoch": 0.39494, + "grad_norm": 0.6195520091936978, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 39494 + }, + { + "epoch": 0.39495, + "grad_norm": 0.6451828521231512, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 39495 + }, + { + "epoch": 0.39496, + "grad_norm": 0.771441507172754, + "learning_rate": 0.003, + "loss": 3.989, + "step": 39496 + }, + { + "epoch": 0.39497, + "grad_norm": 0.9197569229841257, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 39497 + }, + { + "epoch": 0.39498, + "grad_norm": 1.1223432222963232, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 39498 + }, + { + "epoch": 0.39499, + "grad_norm": 0.9256994001365226, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 39499 + }, + { + "epoch": 0.395, + "grad_norm": 0.8645848459455359, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 39500 + }, + { + "epoch": 0.39501, + "grad_norm": 0.7822009050988895, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 39501 + }, + { + "epoch": 0.39502, + "grad_norm": 0.8467566428648259, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 39502 + }, + { + "epoch": 0.39503, + "grad_norm": 0.8059008822995206, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 39503 + }, + { + "epoch": 0.39504, + "grad_norm": 0.8724600697061268, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 39504 + }, + { + "epoch": 0.39505, + "grad_norm": 0.8332983474300975, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 39505 + }, + { + "epoch": 0.39506, + "grad_norm": 0.828375802500982, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 39506 + }, + { + "epoch": 0.39507, + "grad_norm": 0.8830475585843365, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 39507 + }, + { + "epoch": 0.39508, + "grad_norm": 0.8969966487342493, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 39508 + }, + { + "epoch": 0.39509, + "grad_norm": 0.8019923513107797, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 39509 + }, + { + "epoch": 0.3951, + "grad_norm": 0.8136307674574476, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 39510 + }, + { + "epoch": 0.39511, + "grad_norm": 0.8978170547759478, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 39511 + }, + { + "epoch": 0.39512, + "grad_norm": 0.9008602039516698, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 39512 + }, + { + "epoch": 0.39513, + "grad_norm": 0.8046007312079339, + "learning_rate": 0.003, + "loss": 3.997, + "step": 39513 + }, + { + "epoch": 0.39514, + "grad_norm": 0.8717080059130577, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 39514 + }, + { + "epoch": 0.39515, + "grad_norm": 0.8791121946124126, + "learning_rate": 0.003, + "loss": 4.008, + "step": 39515 + }, + { + "epoch": 0.39516, + "grad_norm": 0.7682161536500648, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 39516 + }, + { + "epoch": 0.39517, + "grad_norm": 0.8114067641693136, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 39517 + }, + { + "epoch": 0.39518, + "grad_norm": 0.875829813648051, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 39518 + }, + { + "epoch": 0.39519, + "grad_norm": 0.9276377848270014, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 39519 + }, + { + "epoch": 0.3952, + "grad_norm": 0.8786616422431952, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 39520 + }, + { + "epoch": 0.39521, + "grad_norm": 0.8090497170652347, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 39521 + }, + { + "epoch": 0.39522, + "grad_norm": 0.7918117481258703, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 39522 + }, + { + "epoch": 0.39523, + "grad_norm": 0.9011723902755099, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 39523 + }, + { + "epoch": 0.39524, + "grad_norm": 0.9577430418647895, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 39524 + }, + { + "epoch": 0.39525, + "grad_norm": 0.8334425490234015, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 39525 + }, + { + "epoch": 0.39526, + "grad_norm": 0.8184678091435967, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 39526 + }, + { + "epoch": 0.39527, + "grad_norm": 0.9035689232779389, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 39527 + }, + { + "epoch": 0.39528, + "grad_norm": 0.8551499116112012, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 39528 + }, + { + "epoch": 0.39529, + "grad_norm": 0.7284373161905512, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 39529 + }, + { + "epoch": 0.3953, + "grad_norm": 0.67721742034311, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 39530 + }, + { + "epoch": 0.39531, + "grad_norm": 0.6580097113511707, + "learning_rate": 0.003, + "loss": 3.987, + "step": 39531 + }, + { + "epoch": 0.39532, + "grad_norm": 0.7298960743663471, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 39532 + }, + { + "epoch": 0.39533, + "grad_norm": 0.8450530141412633, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 39533 + }, + { + "epoch": 0.39534, + "grad_norm": 0.9114361862304913, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 39534 + }, + { + "epoch": 0.39535, + "grad_norm": 0.9222340253489839, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 39535 + }, + { + "epoch": 0.39536, + "grad_norm": 0.7382785956861501, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 39536 + }, + { + "epoch": 0.39537, + "grad_norm": 0.5881600359605963, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 39537 + }, + { + "epoch": 0.39538, + "grad_norm": 0.565641458571018, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 39538 + }, + { + "epoch": 0.39539, + "grad_norm": 0.5765095451103858, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 39539 + }, + { + "epoch": 0.3954, + "grad_norm": 0.6258166942924283, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 39540 + }, + { + "epoch": 0.39541, + "grad_norm": 0.6552015308672909, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 39541 + }, + { + "epoch": 0.39542, + "grad_norm": 0.6692099583363191, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 39542 + }, + { + "epoch": 0.39543, + "grad_norm": 0.6759645121242068, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 39543 + }, + { + "epoch": 0.39544, + "grad_norm": 0.6215566304912582, + "learning_rate": 0.003, + "loss": 3.983, + "step": 39544 + }, + { + "epoch": 0.39545, + "grad_norm": 0.58864824010763, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 39545 + }, + { + "epoch": 0.39546, + "grad_norm": 0.538399003073878, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 39546 + }, + { + "epoch": 0.39547, + "grad_norm": 0.49188697561399053, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 39547 + }, + { + "epoch": 0.39548, + "grad_norm": 0.5292109911986655, + "learning_rate": 0.003, + "loss": 3.9371, + "step": 39548 + }, + { + "epoch": 0.39549, + "grad_norm": 0.5484682469510094, + "learning_rate": 0.003, + "loss": 3.9406, + "step": 39549 + }, + { + "epoch": 0.3955, + "grad_norm": 0.5846655872203801, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 39550 + }, + { + "epoch": 0.39551, + "grad_norm": 0.7712939465139929, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 39551 + }, + { + "epoch": 0.39552, + "grad_norm": 0.888141349230115, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 39552 + }, + { + "epoch": 0.39553, + "grad_norm": 1.1654488266566794, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 39553 + }, + { + "epoch": 0.39554, + "grad_norm": 0.7407508387000746, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 39554 + }, + { + "epoch": 0.39555, + "grad_norm": 0.6924725922598464, + "learning_rate": 0.003, + "loss": 3.9488, + "step": 39555 + }, + { + "epoch": 0.39556, + "grad_norm": 0.7498175255716854, + "learning_rate": 0.003, + "loss": 4.003, + "step": 39556 + }, + { + "epoch": 0.39557, + "grad_norm": 0.7274064165265018, + "learning_rate": 0.003, + "loss": 3.976, + "step": 39557 + }, + { + "epoch": 0.39558, + "grad_norm": 0.682660540827337, + "learning_rate": 0.003, + "loss": 3.965, + "step": 39558 + }, + { + "epoch": 0.39559, + "grad_norm": 0.6360580907803589, + "learning_rate": 0.003, + "loss": 3.9454, + "step": 39559 + }, + { + "epoch": 0.3956, + "grad_norm": 0.7113834472969335, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 39560 + }, + { + "epoch": 0.39561, + "grad_norm": 0.8932441896526341, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 39561 + }, + { + "epoch": 0.39562, + "grad_norm": 1.0182920551663233, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 39562 + }, + { + "epoch": 0.39563, + "grad_norm": 1.1305192702300466, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 39563 + }, + { + "epoch": 0.39564, + "grad_norm": 0.8293672622447607, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 39564 + }, + { + "epoch": 0.39565, + "grad_norm": 0.7357618902477636, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 39565 + }, + { + "epoch": 0.39566, + "grad_norm": 0.7307392683468891, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 39566 + }, + { + "epoch": 0.39567, + "grad_norm": 0.7361654998164043, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 39567 + }, + { + "epoch": 0.39568, + "grad_norm": 0.8316243264608212, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 39568 + }, + { + "epoch": 0.39569, + "grad_norm": 0.9669549597754639, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 39569 + }, + { + "epoch": 0.3957, + "grad_norm": 1.1059134495893312, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 39570 + }, + { + "epoch": 0.39571, + "grad_norm": 1.0061451785574023, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 39571 + }, + { + "epoch": 0.39572, + "grad_norm": 1.0663280620540343, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 39572 + }, + { + "epoch": 0.39573, + "grad_norm": 0.8703671952222566, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 39573 + }, + { + "epoch": 0.39574, + "grad_norm": 0.9614484048401445, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 39574 + }, + { + "epoch": 0.39575, + "grad_norm": 0.9325487603099263, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 39575 + }, + { + "epoch": 0.39576, + "grad_norm": 0.922882682994143, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 39576 + }, + { + "epoch": 0.39577, + "grad_norm": 0.9206395538643257, + "learning_rate": 0.003, + "loss": 3.969, + "step": 39577 + }, + { + "epoch": 0.39578, + "grad_norm": 0.9785224664174802, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 39578 + }, + { + "epoch": 0.39579, + "grad_norm": 0.9062018516568625, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 39579 + }, + { + "epoch": 0.3958, + "grad_norm": 0.8360201247862841, + "learning_rate": 0.003, + "loss": 4.0452, + "step": 39580 + }, + { + "epoch": 0.39581, + "grad_norm": 0.781182055363143, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 39581 + }, + { + "epoch": 0.39582, + "grad_norm": 0.7614946084518658, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 39582 + }, + { + "epoch": 0.39583, + "grad_norm": 0.7559991554992785, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 39583 + }, + { + "epoch": 0.39584, + "grad_norm": 0.7681959006975142, + "learning_rate": 0.003, + "loss": 3.967, + "step": 39584 + }, + { + "epoch": 0.39585, + "grad_norm": 0.781515575308298, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 39585 + }, + { + "epoch": 0.39586, + "grad_norm": 0.7315583913017869, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 39586 + }, + { + "epoch": 0.39587, + "grad_norm": 0.7245637016339459, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 39587 + }, + { + "epoch": 0.39588, + "grad_norm": 0.8064592257554076, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 39588 + }, + { + "epoch": 0.39589, + "grad_norm": 0.7834260917824224, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 39589 + }, + { + "epoch": 0.3959, + "grad_norm": 0.8174886899826886, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 39590 + }, + { + "epoch": 0.39591, + "grad_norm": 0.8489828809682625, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 39591 + }, + { + "epoch": 0.39592, + "grad_norm": 0.760465974097475, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 39592 + }, + { + "epoch": 0.39593, + "grad_norm": 0.7082466927041494, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 39593 + }, + { + "epoch": 0.39594, + "grad_norm": 0.8095322411706645, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 39594 + }, + { + "epoch": 0.39595, + "grad_norm": 0.8745069810395336, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 39595 + }, + { + "epoch": 0.39596, + "grad_norm": 0.8681686672198416, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 39596 + }, + { + "epoch": 0.39597, + "grad_norm": 0.7877499166662667, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 39597 + }, + { + "epoch": 0.39598, + "grad_norm": 0.6574970711566591, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 39598 + }, + { + "epoch": 0.39599, + "grad_norm": 0.7056641112319518, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 39599 + }, + { + "epoch": 0.396, + "grad_norm": 0.7184849449009334, + "learning_rate": 0.003, + "loss": 3.963, + "step": 39600 + }, + { + "epoch": 0.39601, + "grad_norm": 0.775622242511099, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 39601 + }, + { + "epoch": 0.39602, + "grad_norm": 0.7038571607901138, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 39602 + }, + { + "epoch": 0.39603, + "grad_norm": 0.6922190948635454, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 39603 + }, + { + "epoch": 0.39604, + "grad_norm": 0.6890537825045864, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 39604 + }, + { + "epoch": 0.39605, + "grad_norm": 0.6805103191850745, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 39605 + }, + { + "epoch": 0.39606, + "grad_norm": 0.6524378222384405, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 39606 + }, + { + "epoch": 0.39607, + "grad_norm": 0.6342932923947844, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 39607 + }, + { + "epoch": 0.39608, + "grad_norm": 0.6389219123627077, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 39608 + }, + { + "epoch": 0.39609, + "grad_norm": 0.5882245325479187, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 39609 + }, + { + "epoch": 0.3961, + "grad_norm": 0.554969315385092, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 39610 + }, + { + "epoch": 0.39611, + "grad_norm": 0.6087972096981987, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 39611 + }, + { + "epoch": 0.39612, + "grad_norm": 0.6090358153938016, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 39612 + }, + { + "epoch": 0.39613, + "grad_norm": 0.5753184527901043, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 39613 + }, + { + "epoch": 0.39614, + "grad_norm": 0.589417413452244, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 39614 + }, + { + "epoch": 0.39615, + "grad_norm": 0.59517137858903, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 39615 + }, + { + "epoch": 0.39616, + "grad_norm": 0.7665716022841212, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 39616 + }, + { + "epoch": 0.39617, + "grad_norm": 0.9192710438356341, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 39617 + }, + { + "epoch": 0.39618, + "grad_norm": 1.080894792558098, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 39618 + }, + { + "epoch": 0.39619, + "grad_norm": 1.1146589370497493, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 39619 + }, + { + "epoch": 0.3962, + "grad_norm": 0.768516152457283, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 39620 + }, + { + "epoch": 0.39621, + "grad_norm": 0.7065991541100429, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 39621 + }, + { + "epoch": 0.39622, + "grad_norm": 0.7412628455295078, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 39622 + }, + { + "epoch": 0.39623, + "grad_norm": 0.736074870632814, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 39623 + }, + { + "epoch": 0.39624, + "grad_norm": 0.6591021739942442, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 39624 + }, + { + "epoch": 0.39625, + "grad_norm": 0.6749308054630881, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 39625 + }, + { + "epoch": 0.39626, + "grad_norm": 0.7440069294647035, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 39626 + }, + { + "epoch": 0.39627, + "grad_norm": 0.7232887413443803, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 39627 + }, + { + "epoch": 0.39628, + "grad_norm": 0.6741576490406286, + "learning_rate": 0.003, + "loss": 4.0, + "step": 39628 + }, + { + "epoch": 0.39629, + "grad_norm": 0.6734449688978683, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 39629 + }, + { + "epoch": 0.3963, + "grad_norm": 0.743032070021682, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 39630 + }, + { + "epoch": 0.39631, + "grad_norm": 0.9004116413661514, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 39631 + }, + { + "epoch": 0.39632, + "grad_norm": 1.0585294602649236, + "learning_rate": 0.003, + "loss": 3.966, + "step": 39632 + }, + { + "epoch": 0.39633, + "grad_norm": 0.9875064802896999, + "learning_rate": 0.003, + "loss": 3.975, + "step": 39633 + }, + { + "epoch": 0.39634, + "grad_norm": 1.0224575874182078, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 39634 + }, + { + "epoch": 0.39635, + "grad_norm": 1.0441346996197374, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 39635 + }, + { + "epoch": 0.39636, + "grad_norm": 0.9690925571274809, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 39636 + }, + { + "epoch": 0.39637, + "grad_norm": 1.024686092098744, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 39637 + }, + { + "epoch": 0.39638, + "grad_norm": 1.0967971630163134, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 39638 + }, + { + "epoch": 0.39639, + "grad_norm": 1.0305999315385337, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 39639 + }, + { + "epoch": 0.3964, + "grad_norm": 0.9944815934964593, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 39640 + }, + { + "epoch": 0.39641, + "grad_norm": 0.9998705860725583, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 39641 + }, + { + "epoch": 0.39642, + "grad_norm": 1.0776053950868325, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 39642 + }, + { + "epoch": 0.39643, + "grad_norm": 0.9211294952418053, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 39643 + }, + { + "epoch": 0.39644, + "grad_norm": 0.8725563101777135, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 39644 + }, + { + "epoch": 0.39645, + "grad_norm": 0.8218914236770642, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 39645 + }, + { + "epoch": 0.39646, + "grad_norm": 0.840994945055507, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 39646 + }, + { + "epoch": 0.39647, + "grad_norm": 0.8993161395414363, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 39647 + }, + { + "epoch": 0.39648, + "grad_norm": 0.8454503955837577, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 39648 + }, + { + "epoch": 0.39649, + "grad_norm": 0.6923037217219368, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 39649 + }, + { + "epoch": 0.3965, + "grad_norm": 0.6267628075107327, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 39650 + }, + { + "epoch": 0.39651, + "grad_norm": 0.6100054334184135, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 39651 + }, + { + "epoch": 0.39652, + "grad_norm": 0.6574198976672591, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 39652 + }, + { + "epoch": 0.39653, + "grad_norm": 0.7039886370689546, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 39653 + }, + { + "epoch": 0.39654, + "grad_norm": 0.6945367376506041, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 39654 + }, + { + "epoch": 0.39655, + "grad_norm": 0.6455735289009799, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 39655 + }, + { + "epoch": 0.39656, + "grad_norm": 0.6352180810027367, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 39656 + }, + { + "epoch": 0.39657, + "grad_norm": 0.5914284747277307, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 39657 + }, + { + "epoch": 0.39658, + "grad_norm": 0.6699586648075225, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 39658 + }, + { + "epoch": 0.39659, + "grad_norm": 0.6955929782407007, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 39659 + }, + { + "epoch": 0.3966, + "grad_norm": 0.6689958141074472, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 39660 + }, + { + "epoch": 0.39661, + "grad_norm": 0.5962720401208219, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 39661 + }, + { + "epoch": 0.39662, + "grad_norm": 0.6144895570824092, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 39662 + }, + { + "epoch": 0.39663, + "grad_norm": 0.7360552940282692, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 39663 + }, + { + "epoch": 0.39664, + "grad_norm": 0.7803575070293307, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 39664 + }, + { + "epoch": 0.39665, + "grad_norm": 0.8857736173430455, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 39665 + }, + { + "epoch": 0.39666, + "grad_norm": 1.0206291402827412, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 39666 + }, + { + "epoch": 0.39667, + "grad_norm": 1.0715440682366206, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 39667 + }, + { + "epoch": 0.39668, + "grad_norm": 0.7134695694354908, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 39668 + }, + { + "epoch": 0.39669, + "grad_norm": 0.6855241507728559, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 39669 + }, + { + "epoch": 0.3967, + "grad_norm": 0.6642779416759467, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 39670 + }, + { + "epoch": 0.39671, + "grad_norm": 0.7077504163971439, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 39671 + }, + { + "epoch": 0.39672, + "grad_norm": 0.6825540223882696, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 39672 + }, + { + "epoch": 0.39673, + "grad_norm": 0.783561478030962, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 39673 + }, + { + "epoch": 0.39674, + "grad_norm": 0.8194219693411915, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 39674 + }, + { + "epoch": 0.39675, + "grad_norm": 0.9153994398391228, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 39675 + }, + { + "epoch": 0.39676, + "grad_norm": 1.0439242737256267, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 39676 + }, + { + "epoch": 0.39677, + "grad_norm": 1.035830681583259, + "learning_rate": 0.003, + "loss": 3.981, + "step": 39677 + }, + { + "epoch": 0.39678, + "grad_norm": 0.8345283581957904, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 39678 + }, + { + "epoch": 0.39679, + "grad_norm": 0.8634576265425898, + "learning_rate": 0.003, + "loss": 4.0, + "step": 39679 + }, + { + "epoch": 0.3968, + "grad_norm": 0.6900604709246955, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 39680 + }, + { + "epoch": 0.39681, + "grad_norm": 0.7341021379132009, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 39681 + }, + { + "epoch": 0.39682, + "grad_norm": 0.8124941572930604, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 39682 + }, + { + "epoch": 0.39683, + "grad_norm": 0.9146471402518351, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 39683 + }, + { + "epoch": 0.39684, + "grad_norm": 0.8574084120350758, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 39684 + }, + { + "epoch": 0.39685, + "grad_norm": 0.7229983282187624, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 39685 + }, + { + "epoch": 0.39686, + "grad_norm": 0.6876663267338038, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 39686 + }, + { + "epoch": 0.39687, + "grad_norm": 0.6581952509294167, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 39687 + }, + { + "epoch": 0.39688, + "grad_norm": 0.6698264261329576, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 39688 + }, + { + "epoch": 0.39689, + "grad_norm": 0.7843994825497469, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 39689 + }, + { + "epoch": 0.3969, + "grad_norm": 0.8849158329782435, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 39690 + }, + { + "epoch": 0.39691, + "grad_norm": 0.9479474044492602, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 39691 + }, + { + "epoch": 0.39692, + "grad_norm": 0.8432702187028907, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 39692 + }, + { + "epoch": 0.39693, + "grad_norm": 0.764233045723463, + "learning_rate": 0.003, + "loss": 3.9265, + "step": 39693 + }, + { + "epoch": 0.39694, + "grad_norm": 0.7062442727452667, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 39694 + }, + { + "epoch": 0.39695, + "grad_norm": 0.7281749403019216, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 39695 + }, + { + "epoch": 0.39696, + "grad_norm": 0.7894340892218594, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 39696 + }, + { + "epoch": 0.39697, + "grad_norm": 0.8122875087452947, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 39697 + }, + { + "epoch": 0.39698, + "grad_norm": 0.7451680009657224, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 39698 + }, + { + "epoch": 0.39699, + "grad_norm": 0.744991678595856, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 39699 + }, + { + "epoch": 0.397, + "grad_norm": 0.7733498009525441, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 39700 + }, + { + "epoch": 0.39701, + "grad_norm": 0.7161440205135546, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 39701 + }, + { + "epoch": 0.39702, + "grad_norm": 0.5620151611399379, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 39702 + }, + { + "epoch": 0.39703, + "grad_norm": 0.6522821792217526, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 39703 + }, + { + "epoch": 0.39704, + "grad_norm": 0.6851965360738722, + "learning_rate": 0.003, + "loss": 4.0, + "step": 39704 + }, + { + "epoch": 0.39705, + "grad_norm": 0.6241012488054754, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 39705 + }, + { + "epoch": 0.39706, + "grad_norm": 0.6955173249905567, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 39706 + }, + { + "epoch": 0.39707, + "grad_norm": 0.7104161621864015, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 39707 + }, + { + "epoch": 0.39708, + "grad_norm": 0.6769283161608818, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 39708 + }, + { + "epoch": 0.39709, + "grad_norm": 0.6561603157173987, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 39709 + }, + { + "epoch": 0.3971, + "grad_norm": 0.8165412463369809, + "learning_rate": 0.003, + "loss": 3.962, + "step": 39710 + }, + { + "epoch": 0.39711, + "grad_norm": 1.0770017065212794, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 39711 + }, + { + "epoch": 0.39712, + "grad_norm": 1.1874717771120833, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 39712 + }, + { + "epoch": 0.39713, + "grad_norm": 0.7732060711023268, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 39713 + }, + { + "epoch": 0.39714, + "grad_norm": 0.6741826420625007, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 39714 + }, + { + "epoch": 0.39715, + "grad_norm": 0.7788322576392941, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 39715 + }, + { + "epoch": 0.39716, + "grad_norm": 0.9700569453265081, + "learning_rate": 0.003, + "loss": 3.975, + "step": 39716 + }, + { + "epoch": 0.39717, + "grad_norm": 1.1675989967632976, + "learning_rate": 0.003, + "loss": 4.013, + "step": 39717 + }, + { + "epoch": 0.39718, + "grad_norm": 0.8342447894918883, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 39718 + }, + { + "epoch": 0.39719, + "grad_norm": 0.8270692482547519, + "learning_rate": 0.003, + "loss": 3.976, + "step": 39719 + }, + { + "epoch": 0.3972, + "grad_norm": 1.0200365869875254, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 39720 + }, + { + "epoch": 0.39721, + "grad_norm": 1.050493080660704, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 39721 + }, + { + "epoch": 0.39722, + "grad_norm": 0.8884082277084492, + "learning_rate": 0.003, + "loss": 4.001, + "step": 39722 + }, + { + "epoch": 0.39723, + "grad_norm": 0.832475893862065, + "learning_rate": 0.003, + "loss": 4.0, + "step": 39723 + }, + { + "epoch": 0.39724, + "grad_norm": 0.8380965331858319, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 39724 + }, + { + "epoch": 0.39725, + "grad_norm": 0.8912939259557716, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 39725 + }, + { + "epoch": 0.39726, + "grad_norm": 0.8028919021196573, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 39726 + }, + { + "epoch": 0.39727, + "grad_norm": 0.797020484118947, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 39727 + }, + { + "epoch": 0.39728, + "grad_norm": 0.8175570973475803, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 39728 + }, + { + "epoch": 0.39729, + "grad_norm": 0.7939347076134174, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 39729 + }, + { + "epoch": 0.3973, + "grad_norm": 0.7947742955638248, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 39730 + }, + { + "epoch": 0.39731, + "grad_norm": 0.809019495161874, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 39731 + }, + { + "epoch": 0.39732, + "grad_norm": 0.8586612959894423, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 39732 + }, + { + "epoch": 0.39733, + "grad_norm": 0.9986854810683098, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 39733 + }, + { + "epoch": 0.39734, + "grad_norm": 0.9667504348599734, + "learning_rate": 0.003, + "loss": 4.016, + "step": 39734 + }, + { + "epoch": 0.39735, + "grad_norm": 0.7476454969469702, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 39735 + }, + { + "epoch": 0.39736, + "grad_norm": 0.6955072588436282, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 39736 + }, + { + "epoch": 0.39737, + "grad_norm": 0.6927837226309491, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 39737 + }, + { + "epoch": 0.39738, + "grad_norm": 0.6717689897144629, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 39738 + }, + { + "epoch": 0.39739, + "grad_norm": 0.6954768596202456, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 39739 + }, + { + "epoch": 0.3974, + "grad_norm": 0.7914084331026315, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 39740 + }, + { + "epoch": 0.39741, + "grad_norm": 0.7807765278891323, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 39741 + }, + { + "epoch": 0.39742, + "grad_norm": 0.8450644395120438, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 39742 + }, + { + "epoch": 0.39743, + "grad_norm": 0.8498823811447678, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 39743 + }, + { + "epoch": 0.39744, + "grad_norm": 0.7283932470349808, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 39744 + }, + { + "epoch": 0.39745, + "grad_norm": 0.6614331432102813, + "learning_rate": 0.003, + "loss": 3.9453, + "step": 39745 + }, + { + "epoch": 0.39746, + "grad_norm": 0.6890825928064334, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 39746 + }, + { + "epoch": 0.39747, + "grad_norm": 0.6556846809616569, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 39747 + }, + { + "epoch": 0.39748, + "grad_norm": 0.6092477745701709, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 39748 + }, + { + "epoch": 0.39749, + "grad_norm": 0.6137763104873021, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 39749 + }, + { + "epoch": 0.3975, + "grad_norm": 0.6348684593407508, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 39750 + }, + { + "epoch": 0.39751, + "grad_norm": 0.6458597481257199, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 39751 + }, + { + "epoch": 0.39752, + "grad_norm": 0.6633908711365661, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 39752 + }, + { + "epoch": 0.39753, + "grad_norm": 0.6640955082921254, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 39753 + }, + { + "epoch": 0.39754, + "grad_norm": 0.6187534008429658, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 39754 + }, + { + "epoch": 0.39755, + "grad_norm": 0.6935412838220811, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 39755 + }, + { + "epoch": 0.39756, + "grad_norm": 0.788866003571788, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 39756 + }, + { + "epoch": 0.39757, + "grad_norm": 0.9329215382291682, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 39757 + }, + { + "epoch": 0.39758, + "grad_norm": 0.9792956920301693, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 39758 + }, + { + "epoch": 0.39759, + "grad_norm": 0.8626716481765239, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 39759 + }, + { + "epoch": 0.3976, + "grad_norm": 0.7329681915330466, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 39760 + }, + { + "epoch": 0.39761, + "grad_norm": 0.863614457297805, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 39761 + }, + { + "epoch": 0.39762, + "grad_norm": 0.9677950413297348, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 39762 + }, + { + "epoch": 0.39763, + "grad_norm": 1.1088964215527763, + "learning_rate": 0.003, + "loss": 4.007, + "step": 39763 + }, + { + "epoch": 0.39764, + "grad_norm": 0.74457440754463, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 39764 + }, + { + "epoch": 0.39765, + "grad_norm": 0.682977771657616, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 39765 + }, + { + "epoch": 0.39766, + "grad_norm": 0.7130570755798586, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 39766 + }, + { + "epoch": 0.39767, + "grad_norm": 0.8084937109674117, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 39767 + }, + { + "epoch": 0.39768, + "grad_norm": 0.8324190863882506, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 39768 + }, + { + "epoch": 0.39769, + "grad_norm": 0.8078092928700852, + "learning_rate": 0.003, + "loss": 3.984, + "step": 39769 + }, + { + "epoch": 0.3977, + "grad_norm": 0.7860786742483872, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 39770 + }, + { + "epoch": 0.39771, + "grad_norm": 0.6881112654590279, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 39771 + }, + { + "epoch": 0.39772, + "grad_norm": 0.5615188405068564, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 39772 + }, + { + "epoch": 0.39773, + "grad_norm": 0.5993297451097617, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 39773 + }, + { + "epoch": 0.39774, + "grad_norm": 0.6464601004330521, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 39774 + }, + { + "epoch": 0.39775, + "grad_norm": 0.656118189660195, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 39775 + }, + { + "epoch": 0.39776, + "grad_norm": 0.6580540378383489, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 39776 + }, + { + "epoch": 0.39777, + "grad_norm": 0.5962729517514107, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 39777 + }, + { + "epoch": 0.39778, + "grad_norm": 0.6137697815686265, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 39778 + }, + { + "epoch": 0.39779, + "grad_norm": 0.5613045668728094, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 39779 + }, + { + "epoch": 0.3978, + "grad_norm": 0.5164465280327138, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 39780 + }, + { + "epoch": 0.39781, + "grad_norm": 0.5752920479185821, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 39781 + }, + { + "epoch": 0.39782, + "grad_norm": 0.7766470151733323, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 39782 + }, + { + "epoch": 0.39783, + "grad_norm": 0.9230426801044574, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 39783 + }, + { + "epoch": 0.39784, + "grad_norm": 1.109832489323812, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 39784 + }, + { + "epoch": 0.39785, + "grad_norm": 1.0942736406697366, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 39785 + }, + { + "epoch": 0.39786, + "grad_norm": 0.8155515089462227, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 39786 + }, + { + "epoch": 0.39787, + "grad_norm": 0.7009685497596815, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 39787 + }, + { + "epoch": 0.39788, + "grad_norm": 0.7711041279940567, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 39788 + }, + { + "epoch": 0.39789, + "grad_norm": 0.8793116266027122, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 39789 + }, + { + "epoch": 0.3979, + "grad_norm": 0.9127216678405482, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 39790 + }, + { + "epoch": 0.39791, + "grad_norm": 0.8416364897999481, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 39791 + }, + { + "epoch": 0.39792, + "grad_norm": 0.7990153334895898, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 39792 + }, + { + "epoch": 0.39793, + "grad_norm": 0.7725351115816156, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 39793 + }, + { + "epoch": 0.39794, + "grad_norm": 0.7577968462580165, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 39794 + }, + { + "epoch": 0.39795, + "grad_norm": 0.7370502322108881, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 39795 + }, + { + "epoch": 0.39796, + "grad_norm": 1.0067803351639324, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 39796 + }, + { + "epoch": 0.39797, + "grad_norm": 1.2557908629140313, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 39797 + }, + { + "epoch": 0.39798, + "grad_norm": 0.736996185440963, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 39798 + }, + { + "epoch": 0.39799, + "grad_norm": 0.8797963383475411, + "learning_rate": 0.003, + "loss": 3.979, + "step": 39799 + }, + { + "epoch": 0.398, + "grad_norm": 1.0684921423309774, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 39800 + }, + { + "epoch": 0.39801, + "grad_norm": 1.0160298705369903, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 39801 + }, + { + "epoch": 0.39802, + "grad_norm": 0.9561053316438844, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 39802 + }, + { + "epoch": 0.39803, + "grad_norm": 0.9060061787898728, + "learning_rate": 0.003, + "loss": 3.987, + "step": 39803 + }, + { + "epoch": 0.39804, + "grad_norm": 0.8729118650734118, + "learning_rate": 0.003, + "loss": 4.0287, + "step": 39804 + }, + { + "epoch": 0.39805, + "grad_norm": 0.9015665331252899, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 39805 + }, + { + "epoch": 0.39806, + "grad_norm": 0.8274441333206904, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 39806 + }, + { + "epoch": 0.39807, + "grad_norm": 0.7697201503091751, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 39807 + }, + { + "epoch": 0.39808, + "grad_norm": 0.7768568354475375, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 39808 + }, + { + "epoch": 0.39809, + "grad_norm": 0.8108831335135005, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 39809 + }, + { + "epoch": 0.3981, + "grad_norm": 0.8098948233551951, + "learning_rate": 0.003, + "loss": 3.979, + "step": 39810 + }, + { + "epoch": 0.39811, + "grad_norm": 0.76021729979516, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 39811 + }, + { + "epoch": 0.39812, + "grad_norm": 0.62365687916884, + "learning_rate": 0.003, + "loss": 3.9386, + "step": 39812 + }, + { + "epoch": 0.39813, + "grad_norm": 0.5607323771259873, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 39813 + }, + { + "epoch": 0.39814, + "grad_norm": 0.5706746475301334, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 39814 + }, + { + "epoch": 0.39815, + "grad_norm": 0.6017533086085989, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 39815 + }, + { + "epoch": 0.39816, + "grad_norm": 0.673228635299865, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 39816 + }, + { + "epoch": 0.39817, + "grad_norm": 0.7059215692569634, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 39817 + }, + { + "epoch": 0.39818, + "grad_norm": 0.7154657389877431, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 39818 + }, + { + "epoch": 0.39819, + "grad_norm": 0.6905328052766977, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 39819 + }, + { + "epoch": 0.3982, + "grad_norm": 0.6282560383829576, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 39820 + }, + { + "epoch": 0.39821, + "grad_norm": 0.6893490741962801, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 39821 + }, + { + "epoch": 0.39822, + "grad_norm": 1.0492071023620095, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 39822 + }, + { + "epoch": 0.39823, + "grad_norm": 1.3328676057524325, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 39823 + }, + { + "epoch": 0.39824, + "grad_norm": 0.6426419675154005, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 39824 + }, + { + "epoch": 0.39825, + "grad_norm": 0.8686000182038071, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 39825 + }, + { + "epoch": 0.39826, + "grad_norm": 0.8922867579828925, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 39826 + }, + { + "epoch": 0.39827, + "grad_norm": 0.8056828012316337, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 39827 + }, + { + "epoch": 0.39828, + "grad_norm": 0.7058541693445759, + "learning_rate": 0.003, + "loss": 3.9411, + "step": 39828 + }, + { + "epoch": 0.39829, + "grad_norm": 0.6937497501468876, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 39829 + }, + { + "epoch": 0.3983, + "grad_norm": 0.7952504071890669, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 39830 + }, + { + "epoch": 0.39831, + "grad_norm": 0.8013497483649713, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 39831 + }, + { + "epoch": 0.39832, + "grad_norm": 0.8468229650768051, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 39832 + }, + { + "epoch": 0.39833, + "grad_norm": 0.917642991114594, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 39833 + }, + { + "epoch": 0.39834, + "grad_norm": 0.9050175845054508, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 39834 + }, + { + "epoch": 0.39835, + "grad_norm": 0.7839621173939656, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 39835 + }, + { + "epoch": 0.39836, + "grad_norm": 0.7116276113907638, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 39836 + }, + { + "epoch": 0.39837, + "grad_norm": 0.7403643317095043, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 39837 + }, + { + "epoch": 0.39838, + "grad_norm": 0.9146118701091658, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 39838 + }, + { + "epoch": 0.39839, + "grad_norm": 0.9935270445720542, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 39839 + }, + { + "epoch": 0.3984, + "grad_norm": 1.0178049524310169, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 39840 + }, + { + "epoch": 0.39841, + "grad_norm": 0.8726394395457112, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 39841 + }, + { + "epoch": 0.39842, + "grad_norm": 0.8528839255129915, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 39842 + }, + { + "epoch": 0.39843, + "grad_norm": 0.7833311193538277, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 39843 + }, + { + "epoch": 0.39844, + "grad_norm": 0.7104439716485897, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 39844 + }, + { + "epoch": 0.39845, + "grad_norm": 0.6804744496031414, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 39845 + }, + { + "epoch": 0.39846, + "grad_norm": 0.6638626244539717, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 39846 + }, + { + "epoch": 0.39847, + "grad_norm": 0.6678084385450478, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 39847 + }, + { + "epoch": 0.39848, + "grad_norm": 0.68559972315439, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 39848 + }, + { + "epoch": 0.39849, + "grad_norm": 0.7645442130300675, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 39849 + }, + { + "epoch": 0.3985, + "grad_norm": 0.8921305657429804, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 39850 + }, + { + "epoch": 0.39851, + "grad_norm": 1.0125111186192324, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 39851 + }, + { + "epoch": 0.39852, + "grad_norm": 0.9959741126950838, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 39852 + }, + { + "epoch": 0.39853, + "grad_norm": 1.0040483133154048, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 39853 + }, + { + "epoch": 0.39854, + "grad_norm": 0.8280048557180737, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 39854 + }, + { + "epoch": 0.39855, + "grad_norm": 0.7800374418257531, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 39855 + }, + { + "epoch": 0.39856, + "grad_norm": 0.7137274593569821, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 39856 + }, + { + "epoch": 0.39857, + "grad_norm": 0.808913573596959, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 39857 + }, + { + "epoch": 0.39858, + "grad_norm": 0.7987279934131544, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 39858 + }, + { + "epoch": 0.39859, + "grad_norm": 0.6381490698383248, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 39859 + }, + { + "epoch": 0.3986, + "grad_norm": 0.6210847231005885, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 39860 + }, + { + "epoch": 0.39861, + "grad_norm": 0.6432224840078333, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 39861 + }, + { + "epoch": 0.39862, + "grad_norm": 0.5667288356381252, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 39862 + }, + { + "epoch": 0.39863, + "grad_norm": 0.5682964299383447, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 39863 + }, + { + "epoch": 0.39864, + "grad_norm": 0.6355635384660113, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 39864 + }, + { + "epoch": 0.39865, + "grad_norm": 0.6262618662977015, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 39865 + }, + { + "epoch": 0.39866, + "grad_norm": 0.6857012380491585, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 39866 + }, + { + "epoch": 0.39867, + "grad_norm": 0.7472187284931684, + "learning_rate": 0.003, + "loss": 3.983, + "step": 39867 + }, + { + "epoch": 0.39868, + "grad_norm": 0.8801197376352438, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 39868 + }, + { + "epoch": 0.39869, + "grad_norm": 0.8495018387469551, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 39869 + }, + { + "epoch": 0.3987, + "grad_norm": 0.6972891590607908, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 39870 + }, + { + "epoch": 0.39871, + "grad_norm": 0.7726157348497869, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 39871 + }, + { + "epoch": 0.39872, + "grad_norm": 0.7390051977171274, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 39872 + }, + { + "epoch": 0.39873, + "grad_norm": 0.7884896223931855, + "learning_rate": 0.003, + "loss": 3.9369, + "step": 39873 + }, + { + "epoch": 0.39874, + "grad_norm": 0.8149478863379139, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 39874 + }, + { + "epoch": 0.39875, + "grad_norm": 0.8139346717624326, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 39875 + }, + { + "epoch": 0.39876, + "grad_norm": 0.7652663020757836, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 39876 + }, + { + "epoch": 0.39877, + "grad_norm": 0.8200161607683685, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 39877 + }, + { + "epoch": 0.39878, + "grad_norm": 0.7568431131098537, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 39878 + }, + { + "epoch": 0.39879, + "grad_norm": 0.7909328250192299, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 39879 + }, + { + "epoch": 0.3988, + "grad_norm": 0.9321208665781456, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 39880 + }, + { + "epoch": 0.39881, + "grad_norm": 0.9697641178170464, + "learning_rate": 0.003, + "loss": 3.981, + "step": 39881 + }, + { + "epoch": 0.39882, + "grad_norm": 0.9988490765686858, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 39882 + }, + { + "epoch": 0.39883, + "grad_norm": 0.9412021417901243, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 39883 + }, + { + "epoch": 0.39884, + "grad_norm": 0.9123377160813095, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 39884 + }, + { + "epoch": 0.39885, + "grad_norm": 0.852078845223012, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 39885 + }, + { + "epoch": 0.39886, + "grad_norm": 0.7642537409712934, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 39886 + }, + { + "epoch": 0.39887, + "grad_norm": 0.8639186281703339, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 39887 + }, + { + "epoch": 0.39888, + "grad_norm": 0.8167748830937721, + "learning_rate": 0.003, + "loss": 3.989, + "step": 39888 + }, + { + "epoch": 0.39889, + "grad_norm": 0.7792016382858492, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 39889 + }, + { + "epoch": 0.3989, + "grad_norm": 0.7336745825151928, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 39890 + }, + { + "epoch": 0.39891, + "grad_norm": 0.6477749484680313, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 39891 + }, + { + "epoch": 0.39892, + "grad_norm": 0.5917004481482596, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 39892 + }, + { + "epoch": 0.39893, + "grad_norm": 0.6241912498532283, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 39893 + }, + { + "epoch": 0.39894, + "grad_norm": 0.644902718324806, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 39894 + }, + { + "epoch": 0.39895, + "grad_norm": 0.6528764022189656, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 39895 + }, + { + "epoch": 0.39896, + "grad_norm": 0.6794188418599101, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 39896 + }, + { + "epoch": 0.39897, + "grad_norm": 0.6826045693499668, + "learning_rate": 0.003, + "loss": 3.969, + "step": 39897 + }, + { + "epoch": 0.39898, + "grad_norm": 0.6459009162076589, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 39898 + }, + { + "epoch": 0.39899, + "grad_norm": 0.70326951807901, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 39899 + }, + { + "epoch": 0.399, + "grad_norm": 0.7957744267627912, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 39900 + }, + { + "epoch": 0.39901, + "grad_norm": 0.8702347165688599, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 39901 + }, + { + "epoch": 0.39902, + "grad_norm": 0.9209701283211635, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 39902 + }, + { + "epoch": 0.39903, + "grad_norm": 0.8920276045618085, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 39903 + }, + { + "epoch": 0.39904, + "grad_norm": 0.817900416253833, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 39904 + }, + { + "epoch": 0.39905, + "grad_norm": 0.7479176117750655, + "learning_rate": 0.003, + "loss": 3.989, + "step": 39905 + }, + { + "epoch": 0.39906, + "grad_norm": 0.731341434413093, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 39906 + }, + { + "epoch": 0.39907, + "grad_norm": 0.7203716416376931, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 39907 + }, + { + "epoch": 0.39908, + "grad_norm": 0.7292859007250295, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 39908 + }, + { + "epoch": 0.39909, + "grad_norm": 0.8380657483947522, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 39909 + }, + { + "epoch": 0.3991, + "grad_norm": 0.9578861005394548, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 39910 + }, + { + "epoch": 0.39911, + "grad_norm": 0.9415382377206762, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 39911 + }, + { + "epoch": 0.39912, + "grad_norm": 0.762069396745696, + "learning_rate": 0.003, + "loss": 3.9468, + "step": 39912 + }, + { + "epoch": 0.39913, + "grad_norm": 0.6830645188309974, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 39913 + }, + { + "epoch": 0.39914, + "grad_norm": 0.7206657509454438, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 39914 + }, + { + "epoch": 0.39915, + "grad_norm": 0.790403087788367, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 39915 + }, + { + "epoch": 0.39916, + "grad_norm": 0.908620680305042, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 39916 + }, + { + "epoch": 0.39917, + "grad_norm": 1.0082433852259853, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 39917 + }, + { + "epoch": 0.39918, + "grad_norm": 0.9250498944153883, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 39918 + }, + { + "epoch": 0.39919, + "grad_norm": 0.8447503188034885, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 39919 + }, + { + "epoch": 0.3992, + "grad_norm": 0.754399340944829, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 39920 + }, + { + "epoch": 0.39921, + "grad_norm": 0.6902536998490334, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 39921 + }, + { + "epoch": 0.39922, + "grad_norm": 0.7697280646911292, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 39922 + }, + { + "epoch": 0.39923, + "grad_norm": 0.9295153509546545, + "learning_rate": 0.003, + "loss": 3.984, + "step": 39923 + }, + { + "epoch": 0.39924, + "grad_norm": 0.9605164244037172, + "learning_rate": 0.003, + "loss": 3.985, + "step": 39924 + }, + { + "epoch": 0.39925, + "grad_norm": 0.9265436149320744, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 39925 + }, + { + "epoch": 0.39926, + "grad_norm": 1.0058736740036127, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 39926 + }, + { + "epoch": 0.39927, + "grad_norm": 1.1892230317284291, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 39927 + }, + { + "epoch": 0.39928, + "grad_norm": 0.696833937957159, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 39928 + }, + { + "epoch": 0.39929, + "grad_norm": 0.6941764922831738, + "learning_rate": 0.003, + "loss": 3.9358, + "step": 39929 + }, + { + "epoch": 0.3993, + "grad_norm": 0.5952754664426638, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 39930 + }, + { + "epoch": 0.39931, + "grad_norm": 0.7115569393115266, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 39931 + }, + { + "epoch": 0.39932, + "grad_norm": 0.666006672195958, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 39932 + }, + { + "epoch": 0.39933, + "grad_norm": 0.5865016764969183, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 39933 + }, + { + "epoch": 0.39934, + "grad_norm": 0.5742388306732886, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 39934 + }, + { + "epoch": 0.39935, + "grad_norm": 0.6614770209287697, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 39935 + }, + { + "epoch": 0.39936, + "grad_norm": 0.7747034419175227, + "learning_rate": 0.003, + "loss": 3.967, + "step": 39936 + }, + { + "epoch": 0.39937, + "grad_norm": 0.787287022753523, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 39937 + }, + { + "epoch": 0.39938, + "grad_norm": 0.660976503460268, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 39938 + }, + { + "epoch": 0.39939, + "grad_norm": 0.627363382107839, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 39939 + }, + { + "epoch": 0.3994, + "grad_norm": 0.6529051241800513, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 39940 + }, + { + "epoch": 0.39941, + "grad_norm": 0.7035606773351605, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 39941 + }, + { + "epoch": 0.39942, + "grad_norm": 0.7955822069716286, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 39942 + }, + { + "epoch": 0.39943, + "grad_norm": 0.78741992351973, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 39943 + }, + { + "epoch": 0.39944, + "grad_norm": 0.9040403684108643, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 39944 + }, + { + "epoch": 0.39945, + "grad_norm": 1.0479345583428525, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 39945 + }, + { + "epoch": 0.39946, + "grad_norm": 1.059176852797213, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 39946 + }, + { + "epoch": 0.39947, + "grad_norm": 0.9123901853004124, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 39947 + }, + { + "epoch": 0.39948, + "grad_norm": 0.8167169304387238, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 39948 + }, + { + "epoch": 0.39949, + "grad_norm": 0.610858429122688, + "learning_rate": 0.003, + "loss": 3.998, + "step": 39949 + }, + { + "epoch": 0.3995, + "grad_norm": 0.6206624233471805, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 39950 + }, + { + "epoch": 0.39951, + "grad_norm": 0.5943530449388176, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 39951 + }, + { + "epoch": 0.39952, + "grad_norm": 0.5786351558097769, + "learning_rate": 0.003, + "loss": 3.925, + "step": 39952 + }, + { + "epoch": 0.39953, + "grad_norm": 0.5655617186465436, + "learning_rate": 0.003, + "loss": 3.971, + "step": 39953 + }, + { + "epoch": 0.39954, + "grad_norm": 0.6213773134100091, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 39954 + }, + { + "epoch": 0.39955, + "grad_norm": 0.6584896336608791, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 39955 + }, + { + "epoch": 0.39956, + "grad_norm": 0.6261249933315667, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 39956 + }, + { + "epoch": 0.39957, + "grad_norm": 0.6325334017463715, + "learning_rate": 0.003, + "loss": 3.987, + "step": 39957 + }, + { + "epoch": 0.39958, + "grad_norm": 0.5685685524694293, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 39958 + }, + { + "epoch": 0.39959, + "grad_norm": 0.6033432247587298, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 39959 + }, + { + "epoch": 0.3996, + "grad_norm": 0.6548132771659173, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 39960 + }, + { + "epoch": 0.39961, + "grad_norm": 0.7038703877944005, + "learning_rate": 0.003, + "loss": 3.9383, + "step": 39961 + }, + { + "epoch": 0.39962, + "grad_norm": 0.84641175776199, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 39962 + }, + { + "epoch": 0.39963, + "grad_norm": 0.8771301567834924, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 39963 + }, + { + "epoch": 0.39964, + "grad_norm": 0.8690081717488966, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 39964 + }, + { + "epoch": 0.39965, + "grad_norm": 0.7742945870047118, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 39965 + }, + { + "epoch": 0.39966, + "grad_norm": 0.8604227986720205, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 39966 + }, + { + "epoch": 0.39967, + "grad_norm": 0.8756339755160968, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 39967 + }, + { + "epoch": 0.39968, + "grad_norm": 0.8594685591475995, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 39968 + }, + { + "epoch": 0.39969, + "grad_norm": 0.8365392212947976, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 39969 + }, + { + "epoch": 0.3997, + "grad_norm": 0.9076391899334719, + "learning_rate": 0.003, + "loss": 3.986, + "step": 39970 + }, + { + "epoch": 0.39971, + "grad_norm": 1.045830897881126, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 39971 + }, + { + "epoch": 0.39972, + "grad_norm": 0.9750107547858448, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 39972 + }, + { + "epoch": 0.39973, + "grad_norm": 1.1396970513875202, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 39973 + }, + { + "epoch": 0.39974, + "grad_norm": 1.0015456748871105, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 39974 + }, + { + "epoch": 0.39975, + "grad_norm": 1.0575769940785977, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 39975 + }, + { + "epoch": 0.39976, + "grad_norm": 1.0869790665280221, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 39976 + }, + { + "epoch": 0.39977, + "grad_norm": 0.8957228842502997, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 39977 + }, + { + "epoch": 0.39978, + "grad_norm": 0.8475732876108886, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 39978 + }, + { + "epoch": 0.39979, + "grad_norm": 0.8118770417376443, + "learning_rate": 0.003, + "loss": 4.007, + "step": 39979 + }, + { + "epoch": 0.3998, + "grad_norm": 0.8145842087946781, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 39980 + }, + { + "epoch": 0.39981, + "grad_norm": 0.7899194770049085, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 39981 + }, + { + "epoch": 0.39982, + "grad_norm": 0.8294967719632813, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 39982 + }, + { + "epoch": 0.39983, + "grad_norm": 0.819246045299118, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 39983 + }, + { + "epoch": 0.39984, + "grad_norm": 0.9644644862107495, + "learning_rate": 0.003, + "loss": 4.0337, + "step": 39984 + }, + { + "epoch": 0.39985, + "grad_norm": 1.0245780989088609, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 39985 + }, + { + "epoch": 0.39986, + "grad_norm": 0.933568599825073, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 39986 + }, + { + "epoch": 0.39987, + "grad_norm": 0.9284774074656399, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 39987 + }, + { + "epoch": 0.39988, + "grad_norm": 0.9571127844033369, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 39988 + }, + { + "epoch": 0.39989, + "grad_norm": 0.9314167959378884, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 39989 + }, + { + "epoch": 0.3999, + "grad_norm": 1.0231165433505445, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 39990 + }, + { + "epoch": 0.39991, + "grad_norm": 0.9615156903443398, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 39991 + }, + { + "epoch": 0.39992, + "grad_norm": 0.8524809111461358, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 39992 + }, + { + "epoch": 0.39993, + "grad_norm": 0.6480485076223741, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 39993 + }, + { + "epoch": 0.39994, + "grad_norm": 0.7241140611570515, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 39994 + }, + { + "epoch": 0.39995, + "grad_norm": 0.644428236045371, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 39995 + }, + { + "epoch": 0.39996, + "grad_norm": 0.5899367477223468, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 39996 + }, + { + "epoch": 0.39997, + "grad_norm": 0.5584880585490967, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 39997 + }, + { + "epoch": 0.39998, + "grad_norm": 0.5877549931739657, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 39998 + }, + { + "epoch": 0.39999, + "grad_norm": 0.5270085547574621, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 39999 + }, + { + "epoch": 0.4, + "grad_norm": 0.5668809827792345, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 40000 + }, + { + "epoch": 0.40001, + "grad_norm": 0.6168874744183082, + "learning_rate": 0.003, + "loss": 3.986, + "step": 40001 + }, + { + "epoch": 0.40002, + "grad_norm": 0.5803160019163789, + "learning_rate": 0.003, + "loss": 3.979, + "step": 40002 + }, + { + "epoch": 0.40003, + "grad_norm": 0.6268354354005145, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 40003 + }, + { + "epoch": 0.40004, + "grad_norm": 0.7017911453208414, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 40004 + }, + { + "epoch": 0.40005, + "grad_norm": 0.7607767166301633, + "learning_rate": 0.003, + "loss": 4.0157, + "step": 40005 + }, + { + "epoch": 0.40006, + "grad_norm": 0.8543999088736312, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 40006 + }, + { + "epoch": 0.40007, + "grad_norm": 0.870481224654144, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 40007 + }, + { + "epoch": 0.40008, + "grad_norm": 0.6801811786192546, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 40008 + }, + { + "epoch": 0.40009, + "grad_norm": 0.686757736289389, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 40009 + }, + { + "epoch": 0.4001, + "grad_norm": 0.7193624136087858, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 40010 + }, + { + "epoch": 0.40011, + "grad_norm": 0.7189823997810285, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 40011 + }, + { + "epoch": 0.40012, + "grad_norm": 0.7443842357963023, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 40012 + }, + { + "epoch": 0.40013, + "grad_norm": 0.756413595207133, + "learning_rate": 0.003, + "loss": 3.9405, + "step": 40013 + }, + { + "epoch": 0.40014, + "grad_norm": 0.7735335941153443, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 40014 + }, + { + "epoch": 0.40015, + "grad_norm": 0.7284824104601961, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 40015 + }, + { + "epoch": 0.40016, + "grad_norm": 0.7283083185396169, + "learning_rate": 0.003, + "loss": 3.97, + "step": 40016 + }, + { + "epoch": 0.40017, + "grad_norm": 0.7987079486564186, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 40017 + }, + { + "epoch": 0.40018, + "grad_norm": 0.8605258771420375, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 40018 + }, + { + "epoch": 0.40019, + "grad_norm": 0.8778920263554479, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 40019 + }, + { + "epoch": 0.4002, + "grad_norm": 0.858799026853036, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 40020 + }, + { + "epoch": 0.40021, + "grad_norm": 0.8775812038080127, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 40021 + }, + { + "epoch": 0.40022, + "grad_norm": 0.7729106299371433, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 40022 + }, + { + "epoch": 0.40023, + "grad_norm": 0.7481360018061793, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 40023 + }, + { + "epoch": 0.40024, + "grad_norm": 0.7667433888593638, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 40024 + }, + { + "epoch": 0.40025, + "grad_norm": 0.7879451380656106, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 40025 + }, + { + "epoch": 0.40026, + "grad_norm": 0.7970992207085399, + "learning_rate": 0.003, + "loss": 3.982, + "step": 40026 + }, + { + "epoch": 0.40027, + "grad_norm": 0.7323401148721757, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 40027 + }, + { + "epoch": 0.40028, + "grad_norm": 0.6968454318572992, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 40028 + }, + { + "epoch": 0.40029, + "grad_norm": 0.7974805106657868, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 40029 + }, + { + "epoch": 0.4003, + "grad_norm": 0.8462340941299941, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 40030 + }, + { + "epoch": 0.40031, + "grad_norm": 0.899819095006699, + "learning_rate": 0.003, + "loss": 3.9421, + "step": 40031 + }, + { + "epoch": 0.40032, + "grad_norm": 0.7790892055900361, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 40032 + }, + { + "epoch": 0.40033, + "grad_norm": 0.8358303846392021, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 40033 + }, + { + "epoch": 0.40034, + "grad_norm": 0.9330440513327071, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 40034 + }, + { + "epoch": 0.40035, + "grad_norm": 1.0378578580685458, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 40035 + }, + { + "epoch": 0.40036, + "grad_norm": 1.0412895725098508, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 40036 + }, + { + "epoch": 0.40037, + "grad_norm": 0.8515370881753608, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 40037 + }, + { + "epoch": 0.40038, + "grad_norm": 0.8705335718657584, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 40038 + }, + { + "epoch": 0.40039, + "grad_norm": 0.9273326911556427, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 40039 + }, + { + "epoch": 0.4004, + "grad_norm": 0.9754606728036852, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 40040 + }, + { + "epoch": 0.40041, + "grad_norm": 0.9303551085115791, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 40041 + }, + { + "epoch": 0.40042, + "grad_norm": 0.9695524070810725, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 40042 + }, + { + "epoch": 0.40043, + "grad_norm": 0.9434740628250173, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 40043 + }, + { + "epoch": 0.40044, + "grad_norm": 0.8366518858416662, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 40044 + }, + { + "epoch": 0.40045, + "grad_norm": 0.9355848031104982, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 40045 + }, + { + "epoch": 0.40046, + "grad_norm": 0.9528137878701919, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 40046 + }, + { + "epoch": 0.40047, + "grad_norm": 0.9798014253356258, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 40047 + }, + { + "epoch": 0.40048, + "grad_norm": 0.9797450271095333, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 40048 + }, + { + "epoch": 0.40049, + "grad_norm": 0.8827883199399754, + "learning_rate": 0.003, + "loss": 3.991, + "step": 40049 + }, + { + "epoch": 0.4005, + "grad_norm": 0.7943996672125624, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 40050 + }, + { + "epoch": 0.40051, + "grad_norm": 0.7774727255962558, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 40051 + }, + { + "epoch": 0.40052, + "grad_norm": 0.7709804884483102, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 40052 + }, + { + "epoch": 0.40053, + "grad_norm": 0.8745411240463632, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 40053 + }, + { + "epoch": 0.40054, + "grad_norm": 0.9610188500363755, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 40054 + }, + { + "epoch": 0.40055, + "grad_norm": 0.8801779188029307, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 40055 + }, + { + "epoch": 0.40056, + "grad_norm": 0.7298626372339998, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 40056 + }, + { + "epoch": 0.40057, + "grad_norm": 0.6402846004445393, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 40057 + }, + { + "epoch": 0.40058, + "grad_norm": 0.561134504516546, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 40058 + }, + { + "epoch": 0.40059, + "grad_norm": 0.5517079669776254, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 40059 + }, + { + "epoch": 0.4006, + "grad_norm": 0.539638860257441, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 40060 + }, + { + "epoch": 0.40061, + "grad_norm": 0.5202248535224958, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 40061 + }, + { + "epoch": 0.40062, + "grad_norm": 0.579616278806306, + "learning_rate": 0.003, + "loss": 3.971, + "step": 40062 + }, + { + "epoch": 0.40063, + "grad_norm": 0.6243612752720821, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 40063 + }, + { + "epoch": 0.40064, + "grad_norm": 0.6016766824019842, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 40064 + }, + { + "epoch": 0.40065, + "grad_norm": 0.5728761392258546, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 40065 + }, + { + "epoch": 0.40066, + "grad_norm": 0.5199678229934449, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 40066 + }, + { + "epoch": 0.40067, + "grad_norm": 0.57322060104203, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 40067 + }, + { + "epoch": 0.40068, + "grad_norm": 0.6735362117802801, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 40068 + }, + { + "epoch": 0.40069, + "grad_norm": 0.8626431094510956, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 40069 + }, + { + "epoch": 0.4007, + "grad_norm": 1.06630015557156, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 40070 + }, + { + "epoch": 0.40071, + "grad_norm": 0.8934334851747471, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 40071 + }, + { + "epoch": 0.40072, + "grad_norm": 0.7681937601459786, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 40072 + }, + { + "epoch": 0.40073, + "grad_norm": 0.7370319604027105, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 40073 + }, + { + "epoch": 0.40074, + "grad_norm": 0.8119275104235075, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 40074 + }, + { + "epoch": 0.40075, + "grad_norm": 0.7959356156312696, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 40075 + }, + { + "epoch": 0.40076, + "grad_norm": 0.7542384399412217, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 40076 + }, + { + "epoch": 0.40077, + "grad_norm": 0.8029105918939411, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 40077 + }, + { + "epoch": 0.40078, + "grad_norm": 0.8125417329697238, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 40078 + }, + { + "epoch": 0.40079, + "grad_norm": 0.7000920276451176, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 40079 + }, + { + "epoch": 0.4008, + "grad_norm": 0.6551722271973323, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 40080 + }, + { + "epoch": 0.40081, + "grad_norm": 0.6772371193578214, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 40081 + }, + { + "epoch": 0.40082, + "grad_norm": 0.6892012428091355, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 40082 + }, + { + "epoch": 0.40083, + "grad_norm": 0.7212583338694478, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 40083 + }, + { + "epoch": 0.40084, + "grad_norm": 0.7684907913947876, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 40084 + }, + { + "epoch": 0.40085, + "grad_norm": 0.945822588262514, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 40085 + }, + { + "epoch": 0.40086, + "grad_norm": 1.1035514817873238, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 40086 + }, + { + "epoch": 0.40087, + "grad_norm": 0.6466933030812009, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 40087 + }, + { + "epoch": 0.40088, + "grad_norm": 0.7575920734505796, + "learning_rate": 0.003, + "loss": 3.9456, + "step": 40088 + }, + { + "epoch": 0.40089, + "grad_norm": 0.9553034810382366, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 40089 + }, + { + "epoch": 0.4009, + "grad_norm": 1.0249138248674843, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 40090 + }, + { + "epoch": 0.40091, + "grad_norm": 0.9262766090969894, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 40091 + }, + { + "epoch": 0.40092, + "grad_norm": 0.7215424705140334, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 40092 + }, + { + "epoch": 0.40093, + "grad_norm": 0.7552582226078144, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 40093 + }, + { + "epoch": 0.40094, + "grad_norm": 0.7620069716053157, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 40094 + }, + { + "epoch": 0.40095, + "grad_norm": 0.7597731989824007, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 40095 + }, + { + "epoch": 0.40096, + "grad_norm": 0.7776714025302783, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 40096 + }, + { + "epoch": 0.40097, + "grad_norm": 0.7873131560097103, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 40097 + }, + { + "epoch": 0.40098, + "grad_norm": 0.7373861459863148, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 40098 + }, + { + "epoch": 0.40099, + "grad_norm": 0.6816641158431762, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 40099 + }, + { + "epoch": 0.401, + "grad_norm": 0.7261540494579722, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 40100 + }, + { + "epoch": 0.40101, + "grad_norm": 0.8202673189370839, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 40101 + }, + { + "epoch": 0.40102, + "grad_norm": 0.9128508267094181, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 40102 + }, + { + "epoch": 0.40103, + "grad_norm": 0.9326385780912125, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 40103 + }, + { + "epoch": 0.40104, + "grad_norm": 0.8150129881620913, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 40104 + }, + { + "epoch": 0.40105, + "grad_norm": 0.7899787908943903, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 40105 + }, + { + "epoch": 0.40106, + "grad_norm": 0.8562978357815082, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 40106 + }, + { + "epoch": 0.40107, + "grad_norm": 0.8921530512340143, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 40107 + }, + { + "epoch": 0.40108, + "grad_norm": 0.8751523393839743, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 40108 + }, + { + "epoch": 0.40109, + "grad_norm": 0.9725818142251539, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 40109 + }, + { + "epoch": 0.4011, + "grad_norm": 0.9717724050692911, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 40110 + }, + { + "epoch": 0.40111, + "grad_norm": 0.9301451693922203, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 40111 + }, + { + "epoch": 0.40112, + "grad_norm": 0.8929962439281369, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 40112 + }, + { + "epoch": 0.40113, + "grad_norm": 0.9322877720200282, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 40113 + }, + { + "epoch": 0.40114, + "grad_norm": 0.8864234272376378, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 40114 + }, + { + "epoch": 0.40115, + "grad_norm": 0.9588805767514534, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 40115 + }, + { + "epoch": 0.40116, + "grad_norm": 0.8279271496814528, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 40116 + }, + { + "epoch": 0.40117, + "grad_norm": 0.7325233701085503, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 40117 + }, + { + "epoch": 0.40118, + "grad_norm": 0.7701134613839955, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 40118 + }, + { + "epoch": 0.40119, + "grad_norm": 0.8436093161675454, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 40119 + }, + { + "epoch": 0.4012, + "grad_norm": 0.8294268457196434, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 40120 + }, + { + "epoch": 0.40121, + "grad_norm": 0.8292583347871711, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 40121 + }, + { + "epoch": 0.40122, + "grad_norm": 0.724101610127991, + "learning_rate": 0.003, + "loss": 4.007, + "step": 40122 + }, + { + "epoch": 0.40123, + "grad_norm": 0.587005141128912, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 40123 + }, + { + "epoch": 0.40124, + "grad_norm": 0.6143588532121008, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 40124 + }, + { + "epoch": 0.40125, + "grad_norm": 0.726209755340411, + "learning_rate": 0.003, + "loss": 4.002, + "step": 40125 + }, + { + "epoch": 0.40126, + "grad_norm": 0.8036265269833425, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 40126 + }, + { + "epoch": 0.40127, + "grad_norm": 0.7437610321797189, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 40127 + }, + { + "epoch": 0.40128, + "grad_norm": 0.7264446397982101, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 40128 + }, + { + "epoch": 0.40129, + "grad_norm": 0.8028801117228375, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 40129 + }, + { + "epoch": 0.4013, + "grad_norm": 0.7269014289138088, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 40130 + }, + { + "epoch": 0.40131, + "grad_norm": 0.767324935921906, + "learning_rate": 0.003, + "loss": 3.997, + "step": 40131 + }, + { + "epoch": 0.40132, + "grad_norm": 0.8527509586137025, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 40132 + }, + { + "epoch": 0.40133, + "grad_norm": 0.8204541017906588, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 40133 + }, + { + "epoch": 0.40134, + "grad_norm": 0.8551400402875878, + "learning_rate": 0.003, + "loss": 3.999, + "step": 40134 + }, + { + "epoch": 0.40135, + "grad_norm": 0.9397815254727392, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 40135 + }, + { + "epoch": 0.40136, + "grad_norm": 1.1327819750744186, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 40136 + }, + { + "epoch": 0.40137, + "grad_norm": 0.7669435237103261, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 40137 + }, + { + "epoch": 0.40138, + "grad_norm": 0.6796367502027709, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 40138 + }, + { + "epoch": 0.40139, + "grad_norm": 0.7571464973044435, + "learning_rate": 0.003, + "loss": 3.959, + "step": 40139 + }, + { + "epoch": 0.4014, + "grad_norm": 0.7959320470100144, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 40140 + }, + { + "epoch": 0.40141, + "grad_norm": 0.9179839572756319, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 40141 + }, + { + "epoch": 0.40142, + "grad_norm": 1.0342653095525802, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 40142 + }, + { + "epoch": 0.40143, + "grad_norm": 0.8524004107244684, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 40143 + }, + { + "epoch": 0.40144, + "grad_norm": 0.7467247096242862, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 40144 + }, + { + "epoch": 0.40145, + "grad_norm": 0.7001547394394514, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 40145 + }, + { + "epoch": 0.40146, + "grad_norm": 0.7075628269048377, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 40146 + }, + { + "epoch": 0.40147, + "grad_norm": 0.7408152180706515, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 40147 + }, + { + "epoch": 0.40148, + "grad_norm": 0.8208590303995338, + "learning_rate": 0.003, + "loss": 3.984, + "step": 40148 + }, + { + "epoch": 0.40149, + "grad_norm": 0.7525629410003928, + "learning_rate": 0.003, + "loss": 3.97, + "step": 40149 + }, + { + "epoch": 0.4015, + "grad_norm": 0.6043782346404727, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 40150 + }, + { + "epoch": 0.40151, + "grad_norm": 0.6656707562835407, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 40151 + }, + { + "epoch": 0.40152, + "grad_norm": 0.7158274490898878, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 40152 + }, + { + "epoch": 0.40153, + "grad_norm": 0.6833153218214446, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 40153 + }, + { + "epoch": 0.40154, + "grad_norm": 0.7109720788347045, + "learning_rate": 0.003, + "loss": 3.967, + "step": 40154 + }, + { + "epoch": 0.40155, + "grad_norm": 0.6528647334768164, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 40155 + }, + { + "epoch": 0.40156, + "grad_norm": 0.6549325581070562, + "learning_rate": 0.003, + "loss": 3.9374, + "step": 40156 + }, + { + "epoch": 0.40157, + "grad_norm": 0.6723704832521099, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 40157 + }, + { + "epoch": 0.40158, + "grad_norm": 0.7143785362811802, + "learning_rate": 0.003, + "loss": 3.9457, + "step": 40158 + }, + { + "epoch": 0.40159, + "grad_norm": 0.6975471377916472, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 40159 + }, + { + "epoch": 0.4016, + "grad_norm": 0.6557050533566703, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 40160 + }, + { + "epoch": 0.40161, + "grad_norm": 0.5938896176915025, + "learning_rate": 0.003, + "loss": 3.9267, + "step": 40161 + }, + { + "epoch": 0.40162, + "grad_norm": 0.5830171589591959, + "learning_rate": 0.003, + "loss": 3.993, + "step": 40162 + }, + { + "epoch": 0.40163, + "grad_norm": 0.5748287824676046, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 40163 + }, + { + "epoch": 0.40164, + "grad_norm": 0.6157810013695463, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 40164 + }, + { + "epoch": 0.40165, + "grad_norm": 0.6117460898161552, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 40165 + }, + { + "epoch": 0.40166, + "grad_norm": 0.5858509945464321, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 40166 + }, + { + "epoch": 0.40167, + "grad_norm": 0.619602504537107, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 40167 + }, + { + "epoch": 0.40168, + "grad_norm": 0.626881432935991, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 40168 + }, + { + "epoch": 0.40169, + "grad_norm": 0.611573780154729, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 40169 + }, + { + "epoch": 0.4017, + "grad_norm": 0.7289161078077414, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 40170 + }, + { + "epoch": 0.40171, + "grad_norm": 0.8856093636568391, + "learning_rate": 0.003, + "loss": 3.98, + "step": 40171 + }, + { + "epoch": 0.40172, + "grad_norm": 0.9681813812003318, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 40172 + }, + { + "epoch": 0.40173, + "grad_norm": 0.9697294924177859, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 40173 + }, + { + "epoch": 0.40174, + "grad_norm": 0.9957927871338155, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 40174 + }, + { + "epoch": 0.40175, + "grad_norm": 1.225824691156812, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 40175 + }, + { + "epoch": 0.40176, + "grad_norm": 0.9313863005631812, + "learning_rate": 0.003, + "loss": 4.0424, + "step": 40176 + }, + { + "epoch": 0.40177, + "grad_norm": 0.9107212990251002, + "learning_rate": 0.003, + "loss": 4.02, + "step": 40177 + }, + { + "epoch": 0.40178, + "grad_norm": 0.9135711141085494, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 40178 + }, + { + "epoch": 0.40179, + "grad_norm": 0.8253698737385572, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 40179 + }, + { + "epoch": 0.4018, + "grad_norm": 0.8350384351701783, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 40180 + }, + { + "epoch": 0.40181, + "grad_norm": 0.7023255015485546, + "learning_rate": 0.003, + "loss": 3.9462, + "step": 40181 + }, + { + "epoch": 0.40182, + "grad_norm": 0.7273746424932483, + "learning_rate": 0.003, + "loss": 4.0335, + "step": 40182 + }, + { + "epoch": 0.40183, + "grad_norm": 0.6701094470831026, + "learning_rate": 0.003, + "loss": 4.0256, + "step": 40183 + }, + { + "epoch": 0.40184, + "grad_norm": 0.7088223584144443, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 40184 + }, + { + "epoch": 0.40185, + "grad_norm": 0.9509453108343793, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 40185 + }, + { + "epoch": 0.40186, + "grad_norm": 1.117500945428165, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 40186 + }, + { + "epoch": 0.40187, + "grad_norm": 0.9132909543779215, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 40187 + }, + { + "epoch": 0.40188, + "grad_norm": 1.004865859125726, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 40188 + }, + { + "epoch": 0.40189, + "grad_norm": 0.9621307496731343, + "learning_rate": 0.003, + "loss": 4.018, + "step": 40189 + }, + { + "epoch": 0.4019, + "grad_norm": 1.0728047189991619, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 40190 + }, + { + "epoch": 0.40191, + "grad_norm": 0.9324716795640233, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 40191 + }, + { + "epoch": 0.40192, + "grad_norm": 0.9076059843145348, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 40192 + }, + { + "epoch": 0.40193, + "grad_norm": 0.9016800332238367, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 40193 + }, + { + "epoch": 0.40194, + "grad_norm": 0.936947650827683, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 40194 + }, + { + "epoch": 0.40195, + "grad_norm": 0.8192852998548815, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 40195 + }, + { + "epoch": 0.40196, + "grad_norm": 0.8123997881358958, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 40196 + }, + { + "epoch": 0.40197, + "grad_norm": 0.7519906376103803, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 40197 + }, + { + "epoch": 0.40198, + "grad_norm": 0.6131357696932023, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 40198 + }, + { + "epoch": 0.40199, + "grad_norm": 0.5605835568650107, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 40199 + }, + { + "epoch": 0.402, + "grad_norm": 0.6849784249277981, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 40200 + }, + { + "epoch": 0.40201, + "grad_norm": 0.8753091502309182, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 40201 + }, + { + "epoch": 0.40202, + "grad_norm": 0.9457968316646556, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 40202 + }, + { + "epoch": 0.40203, + "grad_norm": 0.9167383008993041, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 40203 + }, + { + "epoch": 0.40204, + "grad_norm": 0.6969673562943839, + "learning_rate": 0.003, + "loss": 3.955, + "step": 40204 + }, + { + "epoch": 0.40205, + "grad_norm": 0.6700562792006659, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 40205 + }, + { + "epoch": 0.40206, + "grad_norm": 0.5945764438393836, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 40206 + }, + { + "epoch": 0.40207, + "grad_norm": 0.5712791868397351, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 40207 + }, + { + "epoch": 0.40208, + "grad_norm": 0.6181374674380377, + "learning_rate": 0.003, + "loss": 3.978, + "step": 40208 + }, + { + "epoch": 0.40209, + "grad_norm": 0.6383844956893562, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 40209 + }, + { + "epoch": 0.4021, + "grad_norm": 0.6611697972011799, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 40210 + }, + { + "epoch": 0.40211, + "grad_norm": 0.5684797110982297, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 40211 + }, + { + "epoch": 0.40212, + "grad_norm": 0.5801104731602257, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 40212 + }, + { + "epoch": 0.40213, + "grad_norm": 0.613549249258355, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 40213 + }, + { + "epoch": 0.40214, + "grad_norm": 0.6653856085512112, + "learning_rate": 0.003, + "loss": 3.94, + "step": 40214 + }, + { + "epoch": 0.40215, + "grad_norm": 0.770243245090276, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 40215 + }, + { + "epoch": 0.40216, + "grad_norm": 0.9947107310831187, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 40216 + }, + { + "epoch": 0.40217, + "grad_norm": 1.036096087213187, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 40217 + }, + { + "epoch": 0.40218, + "grad_norm": 1.0557420415007162, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 40218 + }, + { + "epoch": 0.40219, + "grad_norm": 0.928582585294056, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 40219 + }, + { + "epoch": 0.4022, + "grad_norm": 0.9040509649561969, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 40220 + }, + { + "epoch": 0.40221, + "grad_norm": 0.9575562435907448, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 40221 + }, + { + "epoch": 0.40222, + "grad_norm": 0.9256431570688143, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 40222 + }, + { + "epoch": 0.40223, + "grad_norm": 0.8904611936398852, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 40223 + }, + { + "epoch": 0.40224, + "grad_norm": 0.9013146003062118, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 40224 + }, + { + "epoch": 0.40225, + "grad_norm": 0.8399952759848518, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 40225 + }, + { + "epoch": 0.40226, + "grad_norm": 0.7707531893027534, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 40226 + }, + { + "epoch": 0.40227, + "grad_norm": 0.8734091056036699, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 40227 + }, + { + "epoch": 0.40228, + "grad_norm": 0.8948959377661145, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 40228 + }, + { + "epoch": 0.40229, + "grad_norm": 1.031488322646432, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 40229 + }, + { + "epoch": 0.4023, + "grad_norm": 0.8895995147450962, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 40230 + }, + { + "epoch": 0.40231, + "grad_norm": 0.7677800814690393, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 40231 + }, + { + "epoch": 0.40232, + "grad_norm": 0.6842869382201058, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 40232 + }, + { + "epoch": 0.40233, + "grad_norm": 0.7222626965571683, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 40233 + }, + { + "epoch": 0.40234, + "grad_norm": 0.8850637730759946, + "learning_rate": 0.003, + "loss": 4.001, + "step": 40234 + }, + { + "epoch": 0.40235, + "grad_norm": 0.9916422135708204, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 40235 + }, + { + "epoch": 0.40236, + "grad_norm": 0.852212799082059, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 40236 + }, + { + "epoch": 0.40237, + "grad_norm": 0.7804115686633631, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 40237 + }, + { + "epoch": 0.40238, + "grad_norm": 0.7928317344571626, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 40238 + }, + { + "epoch": 0.40239, + "grad_norm": 0.7152947907339733, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 40239 + }, + { + "epoch": 0.4024, + "grad_norm": 0.6471060576651131, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 40240 + }, + { + "epoch": 0.40241, + "grad_norm": 0.6553694617666942, + "learning_rate": 0.003, + "loss": 3.989, + "step": 40241 + }, + { + "epoch": 0.40242, + "grad_norm": 0.6140216697344536, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 40242 + }, + { + "epoch": 0.40243, + "grad_norm": 0.543593070701518, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 40243 + }, + { + "epoch": 0.40244, + "grad_norm": 0.5477525678303982, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 40244 + }, + { + "epoch": 0.40245, + "grad_norm": 0.5154161098842132, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 40245 + }, + { + "epoch": 0.40246, + "grad_norm": 0.514228639624844, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 40246 + }, + { + "epoch": 0.40247, + "grad_norm": 0.644110056718759, + "learning_rate": 0.003, + "loss": 3.965, + "step": 40247 + }, + { + "epoch": 0.40248, + "grad_norm": 0.7178410154270419, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 40248 + }, + { + "epoch": 0.40249, + "grad_norm": 0.8349615100550811, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 40249 + }, + { + "epoch": 0.4025, + "grad_norm": 0.8565462321225998, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 40250 + }, + { + "epoch": 0.40251, + "grad_norm": 0.7650209344038027, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 40251 + }, + { + "epoch": 0.40252, + "grad_norm": 0.71349098069144, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 40252 + }, + { + "epoch": 0.40253, + "grad_norm": 0.6737351035174199, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 40253 + }, + { + "epoch": 0.40254, + "grad_norm": 0.6924164361404989, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 40254 + }, + { + "epoch": 0.40255, + "grad_norm": 0.6882233283800404, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 40255 + }, + { + "epoch": 0.40256, + "grad_norm": 0.6732544828645098, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 40256 + }, + { + "epoch": 0.40257, + "grad_norm": 0.7310579418661384, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 40257 + }, + { + "epoch": 0.40258, + "grad_norm": 0.7451929510673737, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 40258 + }, + { + "epoch": 0.40259, + "grad_norm": 0.7462851588475409, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 40259 + }, + { + "epoch": 0.4026, + "grad_norm": 0.7003977812529885, + "learning_rate": 0.003, + "loss": 3.976, + "step": 40260 + }, + { + "epoch": 0.40261, + "grad_norm": 0.6551847956069784, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 40261 + }, + { + "epoch": 0.40262, + "grad_norm": 0.7586188213455101, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 40262 + }, + { + "epoch": 0.40263, + "grad_norm": 0.7959678130877985, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 40263 + }, + { + "epoch": 0.40264, + "grad_norm": 0.8945529619222847, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 40264 + }, + { + "epoch": 0.40265, + "grad_norm": 0.9559341976073519, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 40265 + }, + { + "epoch": 0.40266, + "grad_norm": 0.8982667924998562, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 40266 + }, + { + "epoch": 0.40267, + "grad_norm": 0.9211991799987646, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 40267 + }, + { + "epoch": 0.40268, + "grad_norm": 0.8253407032401062, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 40268 + }, + { + "epoch": 0.40269, + "grad_norm": 0.8544505368937546, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 40269 + }, + { + "epoch": 0.4027, + "grad_norm": 0.9937712005336868, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 40270 + }, + { + "epoch": 0.40271, + "grad_norm": 1.0658326383456505, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 40271 + }, + { + "epoch": 0.40272, + "grad_norm": 0.7938441168444171, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 40272 + }, + { + "epoch": 0.40273, + "grad_norm": 0.8488185585767496, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 40273 + }, + { + "epoch": 0.40274, + "grad_norm": 0.9391399691831547, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 40274 + }, + { + "epoch": 0.40275, + "grad_norm": 1.1755445166978222, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 40275 + }, + { + "epoch": 0.40276, + "grad_norm": 0.9480775958501855, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 40276 + }, + { + "epoch": 0.40277, + "grad_norm": 0.9677485939266, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 40277 + }, + { + "epoch": 0.40278, + "grad_norm": 0.9587132486574172, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 40278 + }, + { + "epoch": 0.40279, + "grad_norm": 0.9311946540123212, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 40279 + }, + { + "epoch": 0.4028, + "grad_norm": 0.9830498379798787, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 40280 + }, + { + "epoch": 0.40281, + "grad_norm": 0.9031164964745688, + "learning_rate": 0.003, + "loss": 4.004, + "step": 40281 + }, + { + "epoch": 0.40282, + "grad_norm": 0.9182513483626984, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 40282 + }, + { + "epoch": 0.40283, + "grad_norm": 0.9511547285987122, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 40283 + }, + { + "epoch": 0.40284, + "grad_norm": 0.9971232857814621, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 40284 + }, + { + "epoch": 0.40285, + "grad_norm": 0.9361155148005854, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 40285 + }, + { + "epoch": 0.40286, + "grad_norm": 0.8219428641053583, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 40286 + }, + { + "epoch": 0.40287, + "grad_norm": 0.7625535359575782, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 40287 + }, + { + "epoch": 0.40288, + "grad_norm": 0.714221427978873, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 40288 + }, + { + "epoch": 0.40289, + "grad_norm": 0.5764389967914999, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 40289 + }, + { + "epoch": 0.4029, + "grad_norm": 0.6295465493560918, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 40290 + }, + { + "epoch": 0.40291, + "grad_norm": 0.6243862703748986, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 40291 + }, + { + "epoch": 0.40292, + "grad_norm": 0.5975534794672404, + "learning_rate": 0.003, + "loss": 3.967, + "step": 40292 + }, + { + "epoch": 0.40293, + "grad_norm": 0.6088605177019601, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 40293 + }, + { + "epoch": 0.40294, + "grad_norm": 0.5581497101601097, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 40294 + }, + { + "epoch": 0.40295, + "grad_norm": 0.5415116065764185, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 40295 + }, + { + "epoch": 0.40296, + "grad_norm": 0.44788351106268537, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 40296 + }, + { + "epoch": 0.40297, + "grad_norm": 0.4912337734455677, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 40297 + }, + { + "epoch": 0.40298, + "grad_norm": 0.51265324681855, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 40298 + }, + { + "epoch": 0.40299, + "grad_norm": 0.6208739448185029, + "learning_rate": 0.003, + "loss": 3.997, + "step": 40299 + }, + { + "epoch": 0.403, + "grad_norm": 0.7137109551438303, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 40300 + }, + { + "epoch": 0.40301, + "grad_norm": 0.745598945456229, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 40301 + }, + { + "epoch": 0.40302, + "grad_norm": 0.787660053454992, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 40302 + }, + { + "epoch": 0.40303, + "grad_norm": 0.7478088499558307, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 40303 + }, + { + "epoch": 0.40304, + "grad_norm": 0.6709955582167634, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 40304 + }, + { + "epoch": 0.40305, + "grad_norm": 0.6577373979236927, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 40305 + }, + { + "epoch": 0.40306, + "grad_norm": 0.6353573644505569, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 40306 + }, + { + "epoch": 0.40307, + "grad_norm": 0.6363507176137982, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 40307 + }, + { + "epoch": 0.40308, + "grad_norm": 0.7119951552502989, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 40308 + }, + { + "epoch": 0.40309, + "grad_norm": 0.820091303386869, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 40309 + }, + { + "epoch": 0.4031, + "grad_norm": 0.9118063702816287, + "learning_rate": 0.003, + "loss": 3.976, + "step": 40310 + }, + { + "epoch": 0.40311, + "grad_norm": 0.9935127378236419, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 40311 + }, + { + "epoch": 0.40312, + "grad_norm": 1.117103694634018, + "learning_rate": 0.003, + "loss": 3.985, + "step": 40312 + }, + { + "epoch": 0.40313, + "grad_norm": 0.7111816990635891, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 40313 + }, + { + "epoch": 0.40314, + "grad_norm": 0.7497804153982609, + "learning_rate": 0.003, + "loss": 3.9377, + "step": 40314 + }, + { + "epoch": 0.40315, + "grad_norm": 0.9840029125492069, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 40315 + }, + { + "epoch": 0.40316, + "grad_norm": 1.0709561449192713, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 40316 + }, + { + "epoch": 0.40317, + "grad_norm": 0.8844282343159413, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 40317 + }, + { + "epoch": 0.40318, + "grad_norm": 0.824145373343518, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 40318 + }, + { + "epoch": 0.40319, + "grad_norm": 0.8437612534870452, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 40319 + }, + { + "epoch": 0.4032, + "grad_norm": 0.8652334861113496, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 40320 + }, + { + "epoch": 0.40321, + "grad_norm": 0.8570010454800868, + "learning_rate": 0.003, + "loss": 3.985, + "step": 40321 + }, + { + "epoch": 0.40322, + "grad_norm": 0.8514099097066022, + "learning_rate": 0.003, + "loss": 3.963, + "step": 40322 + }, + { + "epoch": 0.40323, + "grad_norm": 0.9088418748740795, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 40323 + }, + { + "epoch": 0.40324, + "grad_norm": 1.0452165005153407, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 40324 + }, + { + "epoch": 0.40325, + "grad_norm": 0.8535535896271489, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 40325 + }, + { + "epoch": 0.40326, + "grad_norm": 0.7192122918250018, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 40326 + }, + { + "epoch": 0.40327, + "grad_norm": 0.6927528764804015, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 40327 + }, + { + "epoch": 0.40328, + "grad_norm": 0.711989112847376, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 40328 + }, + { + "epoch": 0.40329, + "grad_norm": 0.7173641278449528, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 40329 + }, + { + "epoch": 0.4033, + "grad_norm": 0.7093779632972637, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 40330 + }, + { + "epoch": 0.40331, + "grad_norm": 0.6417491027709037, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 40331 + }, + { + "epoch": 0.40332, + "grad_norm": 0.6784769624254934, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 40332 + }, + { + "epoch": 0.40333, + "grad_norm": 0.6682431141339918, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 40333 + }, + { + "epoch": 0.40334, + "grad_norm": 0.8012585501404806, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 40334 + }, + { + "epoch": 0.40335, + "grad_norm": 0.8264687088249792, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 40335 + }, + { + "epoch": 0.40336, + "grad_norm": 0.948998552553575, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 40336 + }, + { + "epoch": 0.40337, + "grad_norm": 0.9187586856151857, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 40337 + }, + { + "epoch": 0.40338, + "grad_norm": 0.7676410126765411, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 40338 + }, + { + "epoch": 0.40339, + "grad_norm": 0.6275875579610952, + "learning_rate": 0.003, + "loss": 3.964, + "step": 40339 + }, + { + "epoch": 0.4034, + "grad_norm": 0.6177583976103461, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 40340 + }, + { + "epoch": 0.40341, + "grad_norm": 0.5795857992230063, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 40341 + }, + { + "epoch": 0.40342, + "grad_norm": 0.5729242372339142, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 40342 + }, + { + "epoch": 0.40343, + "grad_norm": 0.5534479447998512, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 40343 + }, + { + "epoch": 0.40344, + "grad_norm": 0.5903324216548779, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 40344 + }, + { + "epoch": 0.40345, + "grad_norm": 0.5914783041184403, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 40345 + }, + { + "epoch": 0.40346, + "grad_norm": 0.5246103366308923, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 40346 + }, + { + "epoch": 0.40347, + "grad_norm": 0.5373215907488252, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 40347 + }, + { + "epoch": 0.40348, + "grad_norm": 0.5889937954558283, + "learning_rate": 0.003, + "loss": 3.985, + "step": 40348 + }, + { + "epoch": 0.40349, + "grad_norm": 0.7221897930851594, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 40349 + }, + { + "epoch": 0.4035, + "grad_norm": 0.8818287025405819, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 40350 + }, + { + "epoch": 0.40351, + "grad_norm": 0.9465898070397875, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 40351 + }, + { + "epoch": 0.40352, + "grad_norm": 0.9569636026267961, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 40352 + }, + { + "epoch": 0.40353, + "grad_norm": 0.9763311430971944, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 40353 + }, + { + "epoch": 0.40354, + "grad_norm": 0.8981954514450654, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 40354 + }, + { + "epoch": 0.40355, + "grad_norm": 0.7741655990576661, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 40355 + }, + { + "epoch": 0.40356, + "grad_norm": 0.6900295080869485, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 40356 + }, + { + "epoch": 0.40357, + "grad_norm": 0.6534816417501824, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 40357 + }, + { + "epoch": 0.40358, + "grad_norm": 0.6819005640948225, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 40358 + }, + { + "epoch": 0.40359, + "grad_norm": 0.7290828095372816, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 40359 + }, + { + "epoch": 0.4036, + "grad_norm": 0.8137171679594737, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 40360 + }, + { + "epoch": 0.40361, + "grad_norm": 0.9835153480775435, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 40361 + }, + { + "epoch": 0.40362, + "grad_norm": 0.9274383080582367, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 40362 + }, + { + "epoch": 0.40363, + "grad_norm": 0.910675477224574, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 40363 + }, + { + "epoch": 0.40364, + "grad_norm": 1.0441346465765362, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 40364 + }, + { + "epoch": 0.40365, + "grad_norm": 1.0389265993615318, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 40365 + }, + { + "epoch": 0.40366, + "grad_norm": 0.9710140125982996, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 40366 + }, + { + "epoch": 0.40367, + "grad_norm": 1.0296738649912676, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 40367 + }, + { + "epoch": 0.40368, + "grad_norm": 0.9957985477274265, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 40368 + }, + { + "epoch": 0.40369, + "grad_norm": 0.9372245414273876, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 40369 + }, + { + "epoch": 0.4037, + "grad_norm": 0.9763072413030868, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 40370 + }, + { + "epoch": 0.40371, + "grad_norm": 0.9204441629307001, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 40371 + }, + { + "epoch": 0.40372, + "grad_norm": 0.9267233783897001, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 40372 + }, + { + "epoch": 0.40373, + "grad_norm": 1.0377894334144657, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 40373 + }, + { + "epoch": 0.40374, + "grad_norm": 1.1777886277167855, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 40374 + }, + { + "epoch": 0.40375, + "grad_norm": 0.7724742276897818, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 40375 + }, + { + "epoch": 0.40376, + "grad_norm": 0.6367130957831272, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 40376 + }, + { + "epoch": 0.40377, + "grad_norm": 0.6396513702656481, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 40377 + }, + { + "epoch": 0.40378, + "grad_norm": 0.6340543402458663, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 40378 + }, + { + "epoch": 0.40379, + "grad_norm": 0.6766021621874126, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 40379 + }, + { + "epoch": 0.4038, + "grad_norm": 0.6836965897773656, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 40380 + }, + { + "epoch": 0.40381, + "grad_norm": 0.7087633818583667, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 40381 + }, + { + "epoch": 0.40382, + "grad_norm": 0.7339651406198288, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 40382 + }, + { + "epoch": 0.40383, + "grad_norm": 0.6894623468169869, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 40383 + }, + { + "epoch": 0.40384, + "grad_norm": 0.779440871931475, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 40384 + }, + { + "epoch": 0.40385, + "grad_norm": 0.7348116000481157, + "learning_rate": 0.003, + "loss": 3.968, + "step": 40385 + }, + { + "epoch": 0.40386, + "grad_norm": 0.7077704162522523, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 40386 + }, + { + "epoch": 0.40387, + "grad_norm": 0.7067521266105816, + "learning_rate": 0.003, + "loss": 3.9462, + "step": 40387 + }, + { + "epoch": 0.40388, + "grad_norm": 0.5996860497461776, + "learning_rate": 0.003, + "loss": 3.972, + "step": 40388 + }, + { + "epoch": 0.40389, + "grad_norm": 0.532103867690196, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 40389 + }, + { + "epoch": 0.4039, + "grad_norm": 0.49671186217659036, + "learning_rate": 0.003, + "loss": 3.9436, + "step": 40390 + }, + { + "epoch": 0.40391, + "grad_norm": 0.5297884819169304, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 40391 + }, + { + "epoch": 0.40392, + "grad_norm": 0.5514434859704948, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 40392 + }, + { + "epoch": 0.40393, + "grad_norm": 0.5559795748334087, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 40393 + }, + { + "epoch": 0.40394, + "grad_norm": 0.5755904698235922, + "learning_rate": 0.003, + "loss": 3.9449, + "step": 40394 + }, + { + "epoch": 0.40395, + "grad_norm": 0.6237433778255732, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 40395 + }, + { + "epoch": 0.40396, + "grad_norm": 0.6067963525930761, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 40396 + }, + { + "epoch": 0.40397, + "grad_norm": 0.6986716917813053, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 40397 + }, + { + "epoch": 0.40398, + "grad_norm": 0.8059711408566294, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 40398 + }, + { + "epoch": 0.40399, + "grad_norm": 0.9588495333706016, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 40399 + }, + { + "epoch": 0.404, + "grad_norm": 0.9781996665767264, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 40400 + }, + { + "epoch": 0.40401, + "grad_norm": 0.8916403855721154, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 40401 + }, + { + "epoch": 0.40402, + "grad_norm": 0.7680867464572473, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 40402 + }, + { + "epoch": 0.40403, + "grad_norm": 0.8048078157830691, + "learning_rate": 0.003, + "loss": 3.95, + "step": 40403 + }, + { + "epoch": 0.40404, + "grad_norm": 0.7708792509772463, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 40404 + }, + { + "epoch": 0.40405, + "grad_norm": 0.7470115436818904, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 40405 + }, + { + "epoch": 0.40406, + "grad_norm": 0.6460924063322947, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 40406 + }, + { + "epoch": 0.40407, + "grad_norm": 0.6764234416906189, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 40407 + }, + { + "epoch": 0.40408, + "grad_norm": 0.5669815797621849, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 40408 + }, + { + "epoch": 0.40409, + "grad_norm": 0.6210867502680071, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 40409 + }, + { + "epoch": 0.4041, + "grad_norm": 0.7559750106640428, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 40410 + }, + { + "epoch": 0.40411, + "grad_norm": 0.7391902224701526, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 40411 + }, + { + "epoch": 0.40412, + "grad_norm": 0.6968578925906161, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 40412 + }, + { + "epoch": 0.40413, + "grad_norm": 0.755939570902906, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 40413 + }, + { + "epoch": 0.40414, + "grad_norm": 0.9606466227311546, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 40414 + }, + { + "epoch": 0.40415, + "grad_norm": 1.1235736881410168, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 40415 + }, + { + "epoch": 0.40416, + "grad_norm": 1.0067648459832415, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 40416 + }, + { + "epoch": 0.40417, + "grad_norm": 1.0561490878603097, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 40417 + }, + { + "epoch": 0.40418, + "grad_norm": 0.9266101452452462, + "learning_rate": 0.003, + "loss": 3.938, + "step": 40418 + }, + { + "epoch": 0.40419, + "grad_norm": 0.8301635690865871, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 40419 + }, + { + "epoch": 0.4042, + "grad_norm": 0.9619645665466254, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 40420 + }, + { + "epoch": 0.40421, + "grad_norm": 1.0218395498975474, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 40421 + }, + { + "epoch": 0.40422, + "grad_norm": 1.0724079664319843, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 40422 + }, + { + "epoch": 0.40423, + "grad_norm": 1.0597195503497905, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 40423 + }, + { + "epoch": 0.40424, + "grad_norm": 1.1340959997765323, + "learning_rate": 0.003, + "loss": 4.003, + "step": 40424 + }, + { + "epoch": 0.40425, + "grad_norm": 1.1042243673581116, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 40425 + }, + { + "epoch": 0.40426, + "grad_norm": 1.2008409155742008, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 40426 + }, + { + "epoch": 0.40427, + "grad_norm": 0.8791252092803368, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 40427 + }, + { + "epoch": 0.40428, + "grad_norm": 0.9283423992058517, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 40428 + }, + { + "epoch": 0.40429, + "grad_norm": 1.079564241984051, + "learning_rate": 0.003, + "loss": 3.998, + "step": 40429 + }, + { + "epoch": 0.4043, + "grad_norm": 0.9071063871184212, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 40430 + }, + { + "epoch": 0.40431, + "grad_norm": 1.0405970967408458, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 40431 + }, + { + "epoch": 0.40432, + "grad_norm": 0.9867037589130941, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 40432 + }, + { + "epoch": 0.40433, + "grad_norm": 1.0991199803407492, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 40433 + }, + { + "epoch": 0.40434, + "grad_norm": 0.983477761475096, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 40434 + }, + { + "epoch": 0.40435, + "grad_norm": 1.0336296618830156, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 40435 + }, + { + "epoch": 0.40436, + "grad_norm": 1.098785990985468, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 40436 + }, + { + "epoch": 0.40437, + "grad_norm": 0.9889419995313752, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 40437 + }, + { + "epoch": 0.40438, + "grad_norm": 0.9258239308577805, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 40438 + }, + { + "epoch": 0.40439, + "grad_norm": 0.8785983800903052, + "learning_rate": 0.003, + "loss": 3.974, + "step": 40439 + }, + { + "epoch": 0.4044, + "grad_norm": 0.9319249263750581, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 40440 + }, + { + "epoch": 0.40441, + "grad_norm": 0.9703463307952046, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 40441 + }, + { + "epoch": 0.40442, + "grad_norm": 1.0398193982531985, + "learning_rate": 0.003, + "loss": 3.997, + "step": 40442 + }, + { + "epoch": 0.40443, + "grad_norm": 0.9546356763805995, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 40443 + }, + { + "epoch": 0.40444, + "grad_norm": 1.069227345264588, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 40444 + }, + { + "epoch": 0.40445, + "grad_norm": 0.9493222473023823, + "learning_rate": 0.003, + "loss": 4.004, + "step": 40445 + }, + { + "epoch": 0.40446, + "grad_norm": 1.0557587600982123, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 40446 + }, + { + "epoch": 0.40447, + "grad_norm": 0.8204003939181759, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 40447 + }, + { + "epoch": 0.40448, + "grad_norm": 0.753034524663077, + "learning_rate": 0.003, + "loss": 4.015, + "step": 40448 + }, + { + "epoch": 0.40449, + "grad_norm": 0.6842259080943871, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 40449 + }, + { + "epoch": 0.4045, + "grad_norm": 0.6007245467769812, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 40450 + }, + { + "epoch": 0.40451, + "grad_norm": 0.6672369618428192, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 40451 + }, + { + "epoch": 0.40452, + "grad_norm": 0.7148997517145048, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 40452 + }, + { + "epoch": 0.40453, + "grad_norm": 0.677786982692826, + "learning_rate": 0.003, + "loss": 3.994, + "step": 40453 + }, + { + "epoch": 0.40454, + "grad_norm": 0.543438341721971, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 40454 + }, + { + "epoch": 0.40455, + "grad_norm": 0.5235961623627446, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 40455 + }, + { + "epoch": 0.40456, + "grad_norm": 0.561161988968209, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 40456 + }, + { + "epoch": 0.40457, + "grad_norm": 0.5878708395729805, + "learning_rate": 0.003, + "loss": 4.009, + "step": 40457 + }, + { + "epoch": 0.40458, + "grad_norm": 0.5452816921076217, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 40458 + }, + { + "epoch": 0.40459, + "grad_norm": 0.5349794643308354, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 40459 + }, + { + "epoch": 0.4046, + "grad_norm": 0.6117115690182655, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 40460 + }, + { + "epoch": 0.40461, + "grad_norm": 0.7050922329964324, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 40461 + }, + { + "epoch": 0.40462, + "grad_norm": 0.7087067647819227, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 40462 + }, + { + "epoch": 0.40463, + "grad_norm": 0.6755221833861778, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 40463 + }, + { + "epoch": 0.40464, + "grad_norm": 0.7516518603853357, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 40464 + }, + { + "epoch": 0.40465, + "grad_norm": 0.7875931556398561, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 40465 + }, + { + "epoch": 0.40466, + "grad_norm": 0.7761033575144628, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 40466 + }, + { + "epoch": 0.40467, + "grad_norm": 0.7432132057936958, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 40467 + }, + { + "epoch": 0.40468, + "grad_norm": 0.6551456633038442, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 40468 + }, + { + "epoch": 0.40469, + "grad_norm": 0.6316029253243972, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 40469 + }, + { + "epoch": 0.4047, + "grad_norm": 0.6481041312625335, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 40470 + }, + { + "epoch": 0.40471, + "grad_norm": 0.7087090913888856, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 40471 + }, + { + "epoch": 0.40472, + "grad_norm": 0.7953020939176924, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 40472 + }, + { + "epoch": 0.40473, + "grad_norm": 0.7945955763850252, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 40473 + }, + { + "epoch": 0.40474, + "grad_norm": 0.6732757131440771, + "learning_rate": 0.003, + "loss": 3.961, + "step": 40474 + }, + { + "epoch": 0.40475, + "grad_norm": 0.6540315285413172, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 40475 + }, + { + "epoch": 0.40476, + "grad_norm": 0.6521208677028845, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 40476 + }, + { + "epoch": 0.40477, + "grad_norm": 0.6618293313078143, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 40477 + }, + { + "epoch": 0.40478, + "grad_norm": 0.7216938927062692, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 40478 + }, + { + "epoch": 0.40479, + "grad_norm": 0.891268802999869, + "learning_rate": 0.003, + "loss": 3.964, + "step": 40479 + }, + { + "epoch": 0.4048, + "grad_norm": 0.9029740276378662, + "learning_rate": 0.003, + "loss": 3.99, + "step": 40480 + }, + { + "epoch": 0.40481, + "grad_norm": 0.9230699337918207, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 40481 + }, + { + "epoch": 0.40482, + "grad_norm": 0.8769719503032213, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 40482 + }, + { + "epoch": 0.40483, + "grad_norm": 0.7354042428354406, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 40483 + }, + { + "epoch": 0.40484, + "grad_norm": 0.6974465077999671, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 40484 + }, + { + "epoch": 0.40485, + "grad_norm": 0.6506904853083874, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 40485 + }, + { + "epoch": 0.40486, + "grad_norm": 0.6601231649470761, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 40486 + }, + { + "epoch": 0.40487, + "grad_norm": 0.6492838168222098, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 40487 + }, + { + "epoch": 0.40488, + "grad_norm": 0.6275117486331417, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 40488 + }, + { + "epoch": 0.40489, + "grad_norm": 0.7399107469964554, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 40489 + }, + { + "epoch": 0.4049, + "grad_norm": 0.8362894690511959, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 40490 + }, + { + "epoch": 0.40491, + "grad_norm": 0.8615441906790482, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 40491 + }, + { + "epoch": 0.40492, + "grad_norm": 0.7814774483921085, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 40492 + }, + { + "epoch": 0.40493, + "grad_norm": 0.7022290781696007, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 40493 + }, + { + "epoch": 0.40494, + "grad_norm": 0.683629481975243, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 40494 + }, + { + "epoch": 0.40495, + "grad_norm": 0.7201727600574717, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 40495 + }, + { + "epoch": 0.40496, + "grad_norm": 0.7661351448515337, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 40496 + }, + { + "epoch": 0.40497, + "grad_norm": 0.8607519747227019, + "learning_rate": 0.003, + "loss": 3.988, + "step": 40497 + }, + { + "epoch": 0.40498, + "grad_norm": 0.992013684219201, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 40498 + }, + { + "epoch": 0.40499, + "grad_norm": 1.0062685466769974, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 40499 + }, + { + "epoch": 0.405, + "grad_norm": 0.9873280690971661, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 40500 + }, + { + "epoch": 0.40501, + "grad_norm": 0.9070178347537489, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 40501 + }, + { + "epoch": 0.40502, + "grad_norm": 0.8600258778017537, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 40502 + }, + { + "epoch": 0.40503, + "grad_norm": 0.873464412145782, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 40503 + }, + { + "epoch": 0.40504, + "grad_norm": 0.8729987076473711, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 40504 + }, + { + "epoch": 0.40505, + "grad_norm": 0.9362703273404904, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 40505 + }, + { + "epoch": 0.40506, + "grad_norm": 0.7800956955192364, + "learning_rate": 0.003, + "loss": 4.006, + "step": 40506 + }, + { + "epoch": 0.40507, + "grad_norm": 0.9076943851031423, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 40507 + }, + { + "epoch": 0.40508, + "grad_norm": 1.0221669381570528, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 40508 + }, + { + "epoch": 0.40509, + "grad_norm": 1.0382364323168842, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 40509 + }, + { + "epoch": 0.4051, + "grad_norm": 1.0369600104372099, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 40510 + }, + { + "epoch": 0.40511, + "grad_norm": 1.143526486118189, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 40511 + }, + { + "epoch": 0.40512, + "grad_norm": 0.9136414704518185, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 40512 + }, + { + "epoch": 0.40513, + "grad_norm": 0.9522526251727315, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 40513 + }, + { + "epoch": 0.40514, + "grad_norm": 0.9591478229544385, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 40514 + }, + { + "epoch": 0.40515, + "grad_norm": 0.9624899630664128, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 40515 + }, + { + "epoch": 0.40516, + "grad_norm": 0.8746131448353892, + "learning_rate": 0.003, + "loss": 4.0371, + "step": 40516 + }, + { + "epoch": 0.40517, + "grad_norm": 0.7536786363640957, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 40517 + }, + { + "epoch": 0.40518, + "grad_norm": 0.7935018260117391, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 40518 + }, + { + "epoch": 0.40519, + "grad_norm": 0.7753499229904435, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 40519 + }, + { + "epoch": 0.4052, + "grad_norm": 0.7191433410026483, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 40520 + }, + { + "epoch": 0.40521, + "grad_norm": 0.6751497262842877, + "learning_rate": 0.003, + "loss": 3.976, + "step": 40521 + }, + { + "epoch": 0.40522, + "grad_norm": 0.7538970099170126, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 40522 + }, + { + "epoch": 0.40523, + "grad_norm": 0.8824366924272138, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 40523 + }, + { + "epoch": 0.40524, + "grad_norm": 0.93014126136086, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 40524 + }, + { + "epoch": 0.40525, + "grad_norm": 0.7876991906600105, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 40525 + }, + { + "epoch": 0.40526, + "grad_norm": 0.7210368001111638, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 40526 + }, + { + "epoch": 0.40527, + "grad_norm": 0.7258909728111493, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 40527 + }, + { + "epoch": 0.40528, + "grad_norm": 0.6458998369783201, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 40528 + }, + { + "epoch": 0.40529, + "grad_norm": 0.5886701529898211, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 40529 + }, + { + "epoch": 0.4053, + "grad_norm": 0.6645956185218845, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 40530 + }, + { + "epoch": 0.40531, + "grad_norm": 0.6466783773248577, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 40531 + }, + { + "epoch": 0.40532, + "grad_norm": 0.7118347146667857, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 40532 + }, + { + "epoch": 0.40533, + "grad_norm": 0.6628031415051199, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 40533 + }, + { + "epoch": 0.40534, + "grad_norm": 0.5494399213506107, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 40534 + }, + { + "epoch": 0.40535, + "grad_norm": 0.4791620888771338, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 40535 + }, + { + "epoch": 0.40536, + "grad_norm": 0.5790503363245053, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 40536 + }, + { + "epoch": 0.40537, + "grad_norm": 0.6573254233241242, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 40537 + }, + { + "epoch": 0.40538, + "grad_norm": 0.6049392246429791, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 40538 + }, + { + "epoch": 0.40539, + "grad_norm": 0.5691466987951546, + "learning_rate": 0.003, + "loss": 3.986, + "step": 40539 + }, + { + "epoch": 0.4054, + "grad_norm": 0.5864203710130835, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 40540 + }, + { + "epoch": 0.40541, + "grad_norm": 0.6436134666062258, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 40541 + }, + { + "epoch": 0.40542, + "grad_norm": 0.7345876000668249, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 40542 + }, + { + "epoch": 0.40543, + "grad_norm": 0.9374639804001444, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 40543 + }, + { + "epoch": 0.40544, + "grad_norm": 0.9457538211612159, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 40544 + }, + { + "epoch": 0.40545, + "grad_norm": 0.8126399047844379, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 40545 + }, + { + "epoch": 0.40546, + "grad_norm": 0.7656363643715347, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 40546 + }, + { + "epoch": 0.40547, + "grad_norm": 0.7775505348072896, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 40547 + }, + { + "epoch": 0.40548, + "grad_norm": 0.8668135175927207, + "learning_rate": 0.003, + "loss": 3.956, + "step": 40548 + }, + { + "epoch": 0.40549, + "grad_norm": 0.822589370061532, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 40549 + }, + { + "epoch": 0.4055, + "grad_norm": 0.855790526050487, + "learning_rate": 0.003, + "loss": 4.009, + "step": 40550 + }, + { + "epoch": 0.40551, + "grad_norm": 0.8040113447552724, + "learning_rate": 0.003, + "loss": 3.995, + "step": 40551 + }, + { + "epoch": 0.40552, + "grad_norm": 0.8404546240424331, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 40552 + }, + { + "epoch": 0.40553, + "grad_norm": 0.870458207287782, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 40553 + }, + { + "epoch": 0.40554, + "grad_norm": 0.8176533910794418, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 40554 + }, + { + "epoch": 0.40555, + "grad_norm": 0.8519513035812953, + "learning_rate": 0.003, + "loss": 4.006, + "step": 40555 + }, + { + "epoch": 0.40556, + "grad_norm": 0.8053073614999522, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 40556 + }, + { + "epoch": 0.40557, + "grad_norm": 0.8769084183307839, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 40557 + }, + { + "epoch": 0.40558, + "grad_norm": 1.0997546504874696, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 40558 + }, + { + "epoch": 0.40559, + "grad_norm": 1.0107823816195167, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 40559 + }, + { + "epoch": 0.4056, + "grad_norm": 0.884936071120066, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 40560 + }, + { + "epoch": 0.40561, + "grad_norm": 0.76447151197155, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 40561 + }, + { + "epoch": 0.40562, + "grad_norm": 0.7359116993289554, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 40562 + }, + { + "epoch": 0.40563, + "grad_norm": 0.8341579757720886, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 40563 + }, + { + "epoch": 0.40564, + "grad_norm": 1.0148690940174614, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 40564 + }, + { + "epoch": 0.40565, + "grad_norm": 0.9704532528907605, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 40565 + }, + { + "epoch": 0.40566, + "grad_norm": 0.9823983440375039, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 40566 + }, + { + "epoch": 0.40567, + "grad_norm": 0.9916231384219367, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 40567 + }, + { + "epoch": 0.40568, + "grad_norm": 0.8449730187446145, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 40568 + }, + { + "epoch": 0.40569, + "grad_norm": 0.7329197390919272, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 40569 + }, + { + "epoch": 0.4057, + "grad_norm": 0.637569191367666, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 40570 + }, + { + "epoch": 0.40571, + "grad_norm": 0.6812056730879765, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 40571 + }, + { + "epoch": 0.40572, + "grad_norm": 0.6282392342564954, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 40572 + }, + { + "epoch": 0.40573, + "grad_norm": 0.6522305186860852, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 40573 + }, + { + "epoch": 0.40574, + "grad_norm": 0.663586414011887, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 40574 + }, + { + "epoch": 0.40575, + "grad_norm": 0.6637559580104514, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 40575 + }, + { + "epoch": 0.40576, + "grad_norm": 0.5985274277417256, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 40576 + }, + { + "epoch": 0.40577, + "grad_norm": 0.6538280781644981, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 40577 + }, + { + "epoch": 0.40578, + "grad_norm": 0.6603644461555624, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 40578 + }, + { + "epoch": 0.40579, + "grad_norm": 0.6926464503912694, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 40579 + }, + { + "epoch": 0.4058, + "grad_norm": 0.6964248115424698, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 40580 + }, + { + "epoch": 0.40581, + "grad_norm": 0.7320353521849828, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 40581 + }, + { + "epoch": 0.40582, + "grad_norm": 0.6166136557482053, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 40582 + }, + { + "epoch": 0.40583, + "grad_norm": 0.5419454569747095, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 40583 + }, + { + "epoch": 0.40584, + "grad_norm": 0.5775564290769282, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 40584 + }, + { + "epoch": 0.40585, + "grad_norm": 0.559379342430454, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 40585 + }, + { + "epoch": 0.40586, + "grad_norm": 0.5360263381616966, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 40586 + }, + { + "epoch": 0.40587, + "grad_norm": 0.6057172068798834, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 40587 + }, + { + "epoch": 0.40588, + "grad_norm": 0.6228899593318735, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 40588 + }, + { + "epoch": 0.40589, + "grad_norm": 0.6260343518886117, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 40589 + }, + { + "epoch": 0.4059, + "grad_norm": 0.6951356889814206, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 40590 + }, + { + "epoch": 0.40591, + "grad_norm": 0.8834620304115398, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 40591 + }, + { + "epoch": 0.40592, + "grad_norm": 1.0998371327812888, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 40592 + }, + { + "epoch": 0.40593, + "grad_norm": 1.0123984650890052, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 40593 + }, + { + "epoch": 0.40594, + "grad_norm": 0.9110480909234523, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 40594 + }, + { + "epoch": 0.40595, + "grad_norm": 0.8366881874882587, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 40595 + }, + { + "epoch": 0.40596, + "grad_norm": 0.9055896589442115, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 40596 + }, + { + "epoch": 0.40597, + "grad_norm": 0.8008437778407469, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 40597 + }, + { + "epoch": 0.40598, + "grad_norm": 0.772256312016979, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 40598 + }, + { + "epoch": 0.40599, + "grad_norm": 0.6263501004969361, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 40599 + }, + { + "epoch": 0.406, + "grad_norm": 0.772962362268878, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 40600 + }, + { + "epoch": 0.40601, + "grad_norm": 0.8998685366295291, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 40601 + }, + { + "epoch": 0.40602, + "grad_norm": 1.0841657902873545, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 40602 + }, + { + "epoch": 0.40603, + "grad_norm": 0.98473079538761, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 40603 + }, + { + "epoch": 0.40604, + "grad_norm": 0.9513422840040693, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 40604 + }, + { + "epoch": 0.40605, + "grad_norm": 0.9546051639098201, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 40605 + }, + { + "epoch": 0.40606, + "grad_norm": 0.9506302998938144, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 40606 + }, + { + "epoch": 0.40607, + "grad_norm": 0.9535603414060966, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 40607 + }, + { + "epoch": 0.40608, + "grad_norm": 0.9941014431529163, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 40608 + }, + { + "epoch": 0.40609, + "grad_norm": 1.0057919831388913, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 40609 + }, + { + "epoch": 0.4061, + "grad_norm": 0.9393179284956495, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 40610 + }, + { + "epoch": 0.40611, + "grad_norm": 0.8974781562057342, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 40611 + }, + { + "epoch": 0.40612, + "grad_norm": 0.9078031831547598, + "learning_rate": 0.003, + "loss": 4.004, + "step": 40612 + }, + { + "epoch": 0.40613, + "grad_norm": 0.7441381027559676, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 40613 + }, + { + "epoch": 0.40614, + "grad_norm": 0.7665824103640757, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 40614 + }, + { + "epoch": 0.40615, + "grad_norm": 0.8655785717692598, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 40615 + }, + { + "epoch": 0.40616, + "grad_norm": 0.9099444013768291, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 40616 + }, + { + "epoch": 0.40617, + "grad_norm": 0.9227958952320909, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 40617 + }, + { + "epoch": 0.40618, + "grad_norm": 0.8215352781508688, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 40618 + }, + { + "epoch": 0.40619, + "grad_norm": 0.7638119110345121, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 40619 + }, + { + "epoch": 0.4062, + "grad_norm": 0.7553517343324477, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 40620 + }, + { + "epoch": 0.40621, + "grad_norm": 0.7741143121022676, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 40621 + }, + { + "epoch": 0.40622, + "grad_norm": 0.7071879879402521, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 40622 + }, + { + "epoch": 0.40623, + "grad_norm": 0.777714962165953, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 40623 + }, + { + "epoch": 0.40624, + "grad_norm": 0.7570417661171984, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 40624 + }, + { + "epoch": 0.40625, + "grad_norm": 0.7368557380206265, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 40625 + }, + { + "epoch": 0.40626, + "grad_norm": 0.6443483160413345, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 40626 + }, + { + "epoch": 0.40627, + "grad_norm": 0.6685155847728261, + "learning_rate": 0.003, + "loss": 4.003, + "step": 40627 + }, + { + "epoch": 0.40628, + "grad_norm": 0.7496097177030452, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 40628 + }, + { + "epoch": 0.40629, + "grad_norm": 0.7700177244341925, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 40629 + }, + { + "epoch": 0.4063, + "grad_norm": 0.8953039752598654, + "learning_rate": 0.003, + "loss": 3.9372, + "step": 40630 + }, + { + "epoch": 0.40631, + "grad_norm": 1.1151156551396126, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 40631 + }, + { + "epoch": 0.40632, + "grad_norm": 0.8570202827525834, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 40632 + }, + { + "epoch": 0.40633, + "grad_norm": 0.6285521717928543, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 40633 + }, + { + "epoch": 0.40634, + "grad_norm": 0.7081383508425971, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 40634 + }, + { + "epoch": 0.40635, + "grad_norm": 0.8145552316717058, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 40635 + }, + { + "epoch": 0.40636, + "grad_norm": 0.9264783953391464, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 40636 + }, + { + "epoch": 0.40637, + "grad_norm": 0.7815457826069185, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 40637 + }, + { + "epoch": 0.40638, + "grad_norm": 0.6354173291211258, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 40638 + }, + { + "epoch": 0.40639, + "grad_norm": 0.6350891432946487, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 40639 + }, + { + "epoch": 0.4064, + "grad_norm": 0.6513906901341276, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 40640 + }, + { + "epoch": 0.40641, + "grad_norm": 0.5900294470404799, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 40641 + }, + { + "epoch": 0.40642, + "grad_norm": 0.5927379207753227, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 40642 + }, + { + "epoch": 0.40643, + "grad_norm": 0.5763256647398405, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 40643 + }, + { + "epoch": 0.40644, + "grad_norm": 0.589158447377721, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 40644 + }, + { + "epoch": 0.40645, + "grad_norm": 0.5852691566409507, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 40645 + }, + { + "epoch": 0.40646, + "grad_norm": 0.6951318518049282, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 40646 + }, + { + "epoch": 0.40647, + "grad_norm": 0.686921939963535, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 40647 + }, + { + "epoch": 0.40648, + "grad_norm": 0.8219858203339919, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 40648 + }, + { + "epoch": 0.40649, + "grad_norm": 0.9020444113827694, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 40649 + }, + { + "epoch": 0.4065, + "grad_norm": 0.7701073764898707, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 40650 + }, + { + "epoch": 0.40651, + "grad_norm": 0.7629979740665177, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 40651 + }, + { + "epoch": 0.40652, + "grad_norm": 0.8105592524807376, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 40652 + }, + { + "epoch": 0.40653, + "grad_norm": 0.7838822678237396, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 40653 + }, + { + "epoch": 0.40654, + "grad_norm": 0.7705495502266089, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 40654 + }, + { + "epoch": 0.40655, + "grad_norm": 0.8649276404499948, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 40655 + }, + { + "epoch": 0.40656, + "grad_norm": 0.9030315579057899, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 40656 + }, + { + "epoch": 0.40657, + "grad_norm": 0.88646085220966, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 40657 + }, + { + "epoch": 0.40658, + "grad_norm": 0.8344617002359639, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 40658 + }, + { + "epoch": 0.40659, + "grad_norm": 0.7457130908060228, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 40659 + }, + { + "epoch": 0.4066, + "grad_norm": 0.7213205840095235, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 40660 + }, + { + "epoch": 0.40661, + "grad_norm": 0.7399910926759592, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 40661 + }, + { + "epoch": 0.40662, + "grad_norm": 0.7919890200227363, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 40662 + }, + { + "epoch": 0.40663, + "grad_norm": 0.8745079869104208, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 40663 + }, + { + "epoch": 0.40664, + "grad_norm": 0.7996736719375068, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 40664 + }, + { + "epoch": 0.40665, + "grad_norm": 0.7548895442317185, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 40665 + }, + { + "epoch": 0.40666, + "grad_norm": 0.6321107488510231, + "learning_rate": 0.003, + "loss": 3.957, + "step": 40666 + }, + { + "epoch": 0.40667, + "grad_norm": 0.6723111023604422, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 40667 + }, + { + "epoch": 0.40668, + "grad_norm": 0.7341646798050293, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 40668 + }, + { + "epoch": 0.40669, + "grad_norm": 0.7452692231382395, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 40669 + }, + { + "epoch": 0.4067, + "grad_norm": 0.7642834308037059, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 40670 + }, + { + "epoch": 0.40671, + "grad_norm": 0.9132746728579801, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 40671 + }, + { + "epoch": 0.40672, + "grad_norm": 1.1254406107441053, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 40672 + }, + { + "epoch": 0.40673, + "grad_norm": 0.8689304076089819, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 40673 + }, + { + "epoch": 0.40674, + "grad_norm": 0.7539769987268558, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 40674 + }, + { + "epoch": 0.40675, + "grad_norm": 0.8493687717242238, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 40675 + }, + { + "epoch": 0.40676, + "grad_norm": 0.9713533978450273, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 40676 + }, + { + "epoch": 0.40677, + "grad_norm": 1.0042391553173258, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 40677 + }, + { + "epoch": 0.40678, + "grad_norm": 0.9683400046671992, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 40678 + }, + { + "epoch": 0.40679, + "grad_norm": 0.8759931346887455, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 40679 + }, + { + "epoch": 0.4068, + "grad_norm": 0.7597139682256089, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 40680 + }, + { + "epoch": 0.40681, + "grad_norm": 0.7513258407824724, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 40681 + }, + { + "epoch": 0.40682, + "grad_norm": 0.7785244094128233, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 40682 + }, + { + "epoch": 0.40683, + "grad_norm": 0.7201177757851103, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 40683 + }, + { + "epoch": 0.40684, + "grad_norm": 0.7600676312542839, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 40684 + }, + { + "epoch": 0.40685, + "grad_norm": 0.7994803497914967, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 40685 + }, + { + "epoch": 0.40686, + "grad_norm": 0.7459888009908044, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 40686 + }, + { + "epoch": 0.40687, + "grad_norm": 0.7214536818919091, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 40687 + }, + { + "epoch": 0.40688, + "grad_norm": 0.7053360310248237, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 40688 + }, + { + "epoch": 0.40689, + "grad_norm": 0.7004393324319129, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 40689 + }, + { + "epoch": 0.4069, + "grad_norm": 0.7703925093753288, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 40690 + }, + { + "epoch": 0.40691, + "grad_norm": 0.863354912108571, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 40691 + }, + { + "epoch": 0.40692, + "grad_norm": 0.7752342282649712, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 40692 + }, + { + "epoch": 0.40693, + "grad_norm": 0.7180424926190423, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 40693 + }, + { + "epoch": 0.40694, + "grad_norm": 0.6694637164649234, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 40694 + }, + { + "epoch": 0.40695, + "grad_norm": 0.6735416410811009, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 40695 + }, + { + "epoch": 0.40696, + "grad_norm": 0.60861714691968, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 40696 + }, + { + "epoch": 0.40697, + "grad_norm": 0.5566388136237711, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 40697 + }, + { + "epoch": 0.40698, + "grad_norm": 0.6352783821567137, + "learning_rate": 0.003, + "loss": 3.973, + "step": 40698 + }, + { + "epoch": 0.40699, + "grad_norm": 0.7513980645167083, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 40699 + }, + { + "epoch": 0.407, + "grad_norm": 0.8764890014795438, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 40700 + }, + { + "epoch": 0.40701, + "grad_norm": 0.9544353991655886, + "learning_rate": 0.003, + "loss": 3.983, + "step": 40701 + }, + { + "epoch": 0.40702, + "grad_norm": 1.030827436485926, + "learning_rate": 0.003, + "loss": 3.979, + "step": 40702 + }, + { + "epoch": 0.40703, + "grad_norm": 1.0622848524903343, + "learning_rate": 0.003, + "loss": 3.996, + "step": 40703 + }, + { + "epoch": 0.40704, + "grad_norm": 1.0263594041920876, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 40704 + }, + { + "epoch": 0.40705, + "grad_norm": 0.9252137854389866, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 40705 + }, + { + "epoch": 0.40706, + "grad_norm": 0.8953350589019401, + "learning_rate": 0.003, + "loss": 3.999, + "step": 40706 + }, + { + "epoch": 0.40707, + "grad_norm": 0.8539346186157674, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 40707 + }, + { + "epoch": 0.40708, + "grad_norm": 0.8078363348947661, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 40708 + }, + { + "epoch": 0.40709, + "grad_norm": 0.7977097195460467, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 40709 + }, + { + "epoch": 0.4071, + "grad_norm": 0.6592238617157781, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 40710 + }, + { + "epoch": 0.40711, + "grad_norm": 0.5895461017733188, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 40711 + }, + { + "epoch": 0.40712, + "grad_norm": 0.5374637007778402, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 40712 + }, + { + "epoch": 0.40713, + "grad_norm": 0.5765231096078716, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 40713 + }, + { + "epoch": 0.40714, + "grad_norm": 0.5136762866952403, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 40714 + }, + { + "epoch": 0.40715, + "grad_norm": 0.5694461256769507, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 40715 + }, + { + "epoch": 0.40716, + "grad_norm": 0.6224300520819563, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 40716 + }, + { + "epoch": 0.40717, + "grad_norm": 0.6300283382645432, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 40717 + }, + { + "epoch": 0.40718, + "grad_norm": 0.6401921877371982, + "learning_rate": 0.003, + "loss": 3.981, + "step": 40718 + }, + { + "epoch": 0.40719, + "grad_norm": 0.7027598364002506, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 40719 + }, + { + "epoch": 0.4072, + "grad_norm": 0.7339990376988365, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 40720 + }, + { + "epoch": 0.40721, + "grad_norm": 0.757759233969656, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 40721 + }, + { + "epoch": 0.40722, + "grad_norm": 0.8858833398342276, + "learning_rate": 0.003, + "loss": 3.991, + "step": 40722 + }, + { + "epoch": 0.40723, + "grad_norm": 0.8834221253802722, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 40723 + }, + { + "epoch": 0.40724, + "grad_norm": 0.8599015057554948, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 40724 + }, + { + "epoch": 0.40725, + "grad_norm": 0.872763573237972, + "learning_rate": 0.003, + "loss": 4.0296, + "step": 40725 + }, + { + "epoch": 0.40726, + "grad_norm": 0.894970379530788, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 40726 + }, + { + "epoch": 0.40727, + "grad_norm": 0.8161080435058508, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 40727 + }, + { + "epoch": 0.40728, + "grad_norm": 0.873610458099427, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 40728 + }, + { + "epoch": 0.40729, + "grad_norm": 0.9482418117503963, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 40729 + }, + { + "epoch": 0.4073, + "grad_norm": 1.070459738150314, + "learning_rate": 0.003, + "loss": 3.999, + "step": 40730 + }, + { + "epoch": 0.40731, + "grad_norm": 1.0950964916385164, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 40731 + }, + { + "epoch": 0.40732, + "grad_norm": 0.8190420709476925, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 40732 + }, + { + "epoch": 0.40733, + "grad_norm": 0.8284603075574453, + "learning_rate": 0.003, + "loss": 3.993, + "step": 40733 + }, + { + "epoch": 0.40734, + "grad_norm": 0.895201457664333, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 40734 + }, + { + "epoch": 0.40735, + "grad_norm": 0.9191288784586288, + "learning_rate": 0.003, + "loss": 4.017, + "step": 40735 + }, + { + "epoch": 0.40736, + "grad_norm": 0.8385695885169739, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 40736 + }, + { + "epoch": 0.40737, + "grad_norm": 0.668497939721325, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 40737 + }, + { + "epoch": 0.40738, + "grad_norm": 0.6365504255345851, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 40738 + }, + { + "epoch": 0.40739, + "grad_norm": 0.5275048548363681, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 40739 + }, + { + "epoch": 0.4074, + "grad_norm": 0.5860874631266679, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 40740 + }, + { + "epoch": 0.40741, + "grad_norm": 0.4993861740497285, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 40741 + }, + { + "epoch": 0.40742, + "grad_norm": 0.5168757970109263, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 40742 + }, + { + "epoch": 0.40743, + "grad_norm": 0.5310003708972711, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 40743 + }, + { + "epoch": 0.40744, + "grad_norm": 0.6327567426440562, + "learning_rate": 0.003, + "loss": 3.96, + "step": 40744 + }, + { + "epoch": 0.40745, + "grad_norm": 0.7084821719702846, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 40745 + }, + { + "epoch": 0.40746, + "grad_norm": 0.8545218820422356, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 40746 + }, + { + "epoch": 0.40747, + "grad_norm": 0.925349693379952, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 40747 + }, + { + "epoch": 0.40748, + "grad_norm": 0.8714746912847167, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 40748 + }, + { + "epoch": 0.40749, + "grad_norm": 0.8168146665729482, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 40749 + }, + { + "epoch": 0.4075, + "grad_norm": 0.7486012719380061, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 40750 + }, + { + "epoch": 0.40751, + "grad_norm": 0.6820477992305445, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 40751 + }, + { + "epoch": 0.40752, + "grad_norm": 0.7178639407461462, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 40752 + }, + { + "epoch": 0.40753, + "grad_norm": 0.7368204105133058, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 40753 + }, + { + "epoch": 0.40754, + "grad_norm": 0.638140906563384, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 40754 + }, + { + "epoch": 0.40755, + "grad_norm": 0.640512671614913, + "learning_rate": 0.003, + "loss": 3.978, + "step": 40755 + }, + { + "epoch": 0.40756, + "grad_norm": 0.6969501726027615, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 40756 + }, + { + "epoch": 0.40757, + "grad_norm": 0.7491258693646755, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 40757 + }, + { + "epoch": 0.40758, + "grad_norm": 0.8508675007397781, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 40758 + }, + { + "epoch": 0.40759, + "grad_norm": 0.8750888440188719, + "learning_rate": 0.003, + "loss": 3.988, + "step": 40759 + }, + { + "epoch": 0.4076, + "grad_norm": 0.9228234055338539, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 40760 + }, + { + "epoch": 0.40761, + "grad_norm": 0.9410404808262867, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 40761 + }, + { + "epoch": 0.40762, + "grad_norm": 0.9748802420735272, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 40762 + }, + { + "epoch": 0.40763, + "grad_norm": 0.9950752637050758, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 40763 + }, + { + "epoch": 0.40764, + "grad_norm": 1.1362432403454379, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 40764 + }, + { + "epoch": 0.40765, + "grad_norm": 0.8570555586349886, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 40765 + }, + { + "epoch": 0.40766, + "grad_norm": 0.7985875831331628, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 40766 + }, + { + "epoch": 0.40767, + "grad_norm": 0.7266284037456855, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 40767 + }, + { + "epoch": 0.40768, + "grad_norm": 0.781003998410348, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 40768 + }, + { + "epoch": 0.40769, + "grad_norm": 0.8287204294086129, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 40769 + }, + { + "epoch": 0.4077, + "grad_norm": 1.0417966429236207, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 40770 + }, + { + "epoch": 0.40771, + "grad_norm": 1.1699535924593147, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 40771 + }, + { + "epoch": 0.40772, + "grad_norm": 0.95173180488117, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 40772 + }, + { + "epoch": 0.40773, + "grad_norm": 0.9370045465883272, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 40773 + }, + { + "epoch": 0.40774, + "grad_norm": 0.8875685280363874, + "learning_rate": 0.003, + "loss": 4.002, + "step": 40774 + }, + { + "epoch": 0.40775, + "grad_norm": 0.959838305608335, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 40775 + }, + { + "epoch": 0.40776, + "grad_norm": 0.880793490061605, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 40776 + }, + { + "epoch": 0.40777, + "grad_norm": 0.7798312660495158, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 40777 + }, + { + "epoch": 0.40778, + "grad_norm": 0.6990932030626196, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 40778 + }, + { + "epoch": 0.40779, + "grad_norm": 0.722039864091706, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 40779 + }, + { + "epoch": 0.4078, + "grad_norm": 0.8574994840487812, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 40780 + }, + { + "epoch": 0.40781, + "grad_norm": 0.9235783788557387, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 40781 + }, + { + "epoch": 0.40782, + "grad_norm": 0.9045401784567269, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 40782 + }, + { + "epoch": 0.40783, + "grad_norm": 0.8016679191299912, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 40783 + }, + { + "epoch": 0.40784, + "grad_norm": 0.8318914099134345, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 40784 + }, + { + "epoch": 0.40785, + "grad_norm": 0.7712225256694608, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 40785 + }, + { + "epoch": 0.40786, + "grad_norm": 0.7323106893969101, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 40786 + }, + { + "epoch": 0.40787, + "grad_norm": 0.6106235302254287, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 40787 + }, + { + "epoch": 0.40788, + "grad_norm": 0.595721567847841, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 40788 + }, + { + "epoch": 0.40789, + "grad_norm": 0.7108075543627306, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 40789 + }, + { + "epoch": 0.4079, + "grad_norm": 0.677337568700202, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 40790 + }, + { + "epoch": 0.40791, + "grad_norm": 0.6985927091167046, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 40791 + }, + { + "epoch": 0.40792, + "grad_norm": 0.6789251195510472, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 40792 + }, + { + "epoch": 0.40793, + "grad_norm": 0.6594182777695266, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 40793 + }, + { + "epoch": 0.40794, + "grad_norm": 0.8111818574615308, + "learning_rate": 0.003, + "loss": 4.013, + "step": 40794 + }, + { + "epoch": 0.40795, + "grad_norm": 0.8613605605199096, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 40795 + }, + { + "epoch": 0.40796, + "grad_norm": 0.9368663417890972, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 40796 + }, + { + "epoch": 0.40797, + "grad_norm": 1.0626710203881267, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 40797 + }, + { + "epoch": 0.40798, + "grad_norm": 0.8563565748211941, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 40798 + }, + { + "epoch": 0.40799, + "grad_norm": 0.7502941551003179, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 40799 + }, + { + "epoch": 0.408, + "grad_norm": 0.615226726226231, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 40800 + }, + { + "epoch": 0.40801, + "grad_norm": 0.6286663415849665, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 40801 + }, + { + "epoch": 0.40802, + "grad_norm": 0.6068335635999124, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 40802 + }, + { + "epoch": 0.40803, + "grad_norm": 0.6580024602454453, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 40803 + }, + { + "epoch": 0.40804, + "grad_norm": 0.6664539409353862, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 40804 + }, + { + "epoch": 0.40805, + "grad_norm": 0.7649404703352528, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 40805 + }, + { + "epoch": 0.40806, + "grad_norm": 0.7884199716915855, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 40806 + }, + { + "epoch": 0.40807, + "grad_norm": 0.7342994469807393, + "learning_rate": 0.003, + "loss": 3.9285, + "step": 40807 + }, + { + "epoch": 0.40808, + "grad_norm": 0.7886271539030928, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 40808 + }, + { + "epoch": 0.40809, + "grad_norm": 0.7141716812334229, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 40809 + }, + { + "epoch": 0.4081, + "grad_norm": 0.6161890793472167, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 40810 + }, + { + "epoch": 0.40811, + "grad_norm": 0.6575312225698583, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 40811 + }, + { + "epoch": 0.40812, + "grad_norm": 0.6954669697462028, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 40812 + }, + { + "epoch": 0.40813, + "grad_norm": 0.708181013289021, + "learning_rate": 0.003, + "loss": 3.981, + "step": 40813 + }, + { + "epoch": 0.40814, + "grad_norm": 0.7993696191572861, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 40814 + }, + { + "epoch": 0.40815, + "grad_norm": 0.8793748550129707, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 40815 + }, + { + "epoch": 0.40816, + "grad_norm": 1.0865759774613517, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 40816 + }, + { + "epoch": 0.40817, + "grad_norm": 1.0329234797338382, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 40817 + }, + { + "epoch": 0.40818, + "grad_norm": 0.9302967028743886, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 40818 + }, + { + "epoch": 0.40819, + "grad_norm": 0.9446414358923614, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 40819 + }, + { + "epoch": 0.4082, + "grad_norm": 0.8407371180112295, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 40820 + }, + { + "epoch": 0.40821, + "grad_norm": 0.9361272813690426, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 40821 + }, + { + "epoch": 0.40822, + "grad_norm": 0.8460403432546261, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 40822 + }, + { + "epoch": 0.40823, + "grad_norm": 0.9110558015272155, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 40823 + }, + { + "epoch": 0.40824, + "grad_norm": 1.0015531492768974, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 40824 + }, + { + "epoch": 0.40825, + "grad_norm": 1.0303578554669484, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 40825 + }, + { + "epoch": 0.40826, + "grad_norm": 1.0505891982547273, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 40826 + }, + { + "epoch": 0.40827, + "grad_norm": 0.8707168060468128, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 40827 + }, + { + "epoch": 0.40828, + "grad_norm": 0.8615634417913534, + "learning_rate": 0.003, + "loss": 4.0336, + "step": 40828 + }, + { + "epoch": 0.40829, + "grad_norm": 0.8102652317232374, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 40829 + }, + { + "epoch": 0.4083, + "grad_norm": 0.7651682501621339, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 40830 + }, + { + "epoch": 0.40831, + "grad_norm": 0.7339795403575906, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 40831 + }, + { + "epoch": 0.40832, + "grad_norm": 0.7360213623586075, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 40832 + }, + { + "epoch": 0.40833, + "grad_norm": 0.7273396051492216, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 40833 + }, + { + "epoch": 0.40834, + "grad_norm": 0.686325830672605, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 40834 + }, + { + "epoch": 0.40835, + "grad_norm": 0.7151305186745306, + "learning_rate": 0.003, + "loss": 3.98, + "step": 40835 + }, + { + "epoch": 0.40836, + "grad_norm": 0.6759221501010418, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 40836 + }, + { + "epoch": 0.40837, + "grad_norm": 0.707668534627995, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 40837 + }, + { + "epoch": 0.40838, + "grad_norm": 0.6587054960389196, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 40838 + }, + { + "epoch": 0.40839, + "grad_norm": 0.6061319817267222, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 40839 + }, + { + "epoch": 0.4084, + "grad_norm": 0.6256196564278874, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 40840 + }, + { + "epoch": 0.40841, + "grad_norm": 0.5831410262030439, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 40841 + }, + { + "epoch": 0.40842, + "grad_norm": 0.48628105321421633, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 40842 + }, + { + "epoch": 0.40843, + "grad_norm": 0.5249875886753642, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 40843 + }, + { + "epoch": 0.40844, + "grad_norm": 0.5581428891032575, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 40844 + }, + { + "epoch": 0.40845, + "grad_norm": 0.5842638020488827, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 40845 + }, + { + "epoch": 0.40846, + "grad_norm": 0.6321261416551208, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 40846 + }, + { + "epoch": 0.40847, + "grad_norm": 0.6853901948475152, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 40847 + }, + { + "epoch": 0.40848, + "grad_norm": 0.7860987660595293, + "learning_rate": 0.003, + "loss": 3.977, + "step": 40848 + }, + { + "epoch": 0.40849, + "grad_norm": 0.9748411007610394, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 40849 + }, + { + "epoch": 0.4085, + "grad_norm": 1.1283487567698556, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 40850 + }, + { + "epoch": 0.40851, + "grad_norm": 0.91928732676344, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 40851 + }, + { + "epoch": 0.40852, + "grad_norm": 0.9646535460414052, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 40852 + }, + { + "epoch": 0.40853, + "grad_norm": 0.8275790281093787, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 40853 + }, + { + "epoch": 0.40854, + "grad_norm": 0.7583914405716992, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 40854 + }, + { + "epoch": 0.40855, + "grad_norm": 0.6583018028796997, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 40855 + }, + { + "epoch": 0.40856, + "grad_norm": 0.6728433901404254, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 40856 + }, + { + "epoch": 0.40857, + "grad_norm": 0.6794777649816578, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 40857 + }, + { + "epoch": 0.40858, + "grad_norm": 0.6616247274233643, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 40858 + }, + { + "epoch": 0.40859, + "grad_norm": 0.6339799972116097, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 40859 + }, + { + "epoch": 0.4086, + "grad_norm": 0.6354707361128216, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 40860 + }, + { + "epoch": 0.40861, + "grad_norm": 0.6960413263817247, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 40861 + }, + { + "epoch": 0.40862, + "grad_norm": 0.708899121774969, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 40862 + }, + { + "epoch": 0.40863, + "grad_norm": 0.8474266445697042, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 40863 + }, + { + "epoch": 0.40864, + "grad_norm": 0.8660720703939762, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 40864 + }, + { + "epoch": 0.40865, + "grad_norm": 0.7214603070548452, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 40865 + }, + { + "epoch": 0.40866, + "grad_norm": 0.6712408276869992, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 40866 + }, + { + "epoch": 0.40867, + "grad_norm": 0.733628208459614, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 40867 + }, + { + "epoch": 0.40868, + "grad_norm": 0.7933473980731337, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 40868 + }, + { + "epoch": 0.40869, + "grad_norm": 0.8334200341807986, + "learning_rate": 0.003, + "loss": 3.989, + "step": 40869 + }, + { + "epoch": 0.4087, + "grad_norm": 0.7964529787542662, + "learning_rate": 0.003, + "loss": 4.024, + "step": 40870 + }, + { + "epoch": 0.40871, + "grad_norm": 0.7542208085406646, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 40871 + }, + { + "epoch": 0.40872, + "grad_norm": 0.6941276094161821, + "learning_rate": 0.003, + "loss": 3.997, + "step": 40872 + }, + { + "epoch": 0.40873, + "grad_norm": 0.8569390115019706, + "learning_rate": 0.003, + "loss": 3.993, + "step": 40873 + }, + { + "epoch": 0.40874, + "grad_norm": 0.9404256358445129, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 40874 + }, + { + "epoch": 0.40875, + "grad_norm": 0.9650340092824422, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 40875 + }, + { + "epoch": 0.40876, + "grad_norm": 1.003110914057566, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 40876 + }, + { + "epoch": 0.40877, + "grad_norm": 1.0569430375410709, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 40877 + }, + { + "epoch": 0.40878, + "grad_norm": 0.8926600273078726, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 40878 + }, + { + "epoch": 0.40879, + "grad_norm": 0.768751876375285, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 40879 + }, + { + "epoch": 0.4088, + "grad_norm": 0.7871584107397476, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 40880 + }, + { + "epoch": 0.40881, + "grad_norm": 0.783691089523909, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 40881 + }, + { + "epoch": 0.40882, + "grad_norm": 0.7518975503695062, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 40882 + }, + { + "epoch": 0.40883, + "grad_norm": 0.7418685373775971, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 40883 + }, + { + "epoch": 0.40884, + "grad_norm": 0.841552445376247, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 40884 + }, + { + "epoch": 0.40885, + "grad_norm": 0.8662925890021629, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 40885 + }, + { + "epoch": 0.40886, + "grad_norm": 0.9390123058020674, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 40886 + }, + { + "epoch": 0.40887, + "grad_norm": 1.0017772669767984, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 40887 + }, + { + "epoch": 0.40888, + "grad_norm": 1.0084406926581304, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 40888 + }, + { + "epoch": 0.40889, + "grad_norm": 0.9027556511606437, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 40889 + }, + { + "epoch": 0.4089, + "grad_norm": 0.8101366307806825, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 40890 + }, + { + "epoch": 0.40891, + "grad_norm": 0.7984899553571623, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 40891 + }, + { + "epoch": 0.40892, + "grad_norm": 0.688603260223175, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 40892 + }, + { + "epoch": 0.40893, + "grad_norm": 0.6458658003794352, + "learning_rate": 0.003, + "loss": 3.9427, + "step": 40893 + }, + { + "epoch": 0.40894, + "grad_norm": 0.6035110644618565, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 40894 + }, + { + "epoch": 0.40895, + "grad_norm": 0.533136346878219, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 40895 + }, + { + "epoch": 0.40896, + "grad_norm": 0.5162580724454519, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 40896 + }, + { + "epoch": 0.40897, + "grad_norm": 0.4714046015457305, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 40897 + }, + { + "epoch": 0.40898, + "grad_norm": 0.49880303061962383, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 40898 + }, + { + "epoch": 0.40899, + "grad_norm": 0.5214409675274523, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 40899 + }, + { + "epoch": 0.409, + "grad_norm": 0.5736624591830559, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 40900 + }, + { + "epoch": 0.40901, + "grad_norm": 0.6005293380547477, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 40901 + }, + { + "epoch": 0.40902, + "grad_norm": 0.67314067668213, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 40902 + }, + { + "epoch": 0.40903, + "grad_norm": 0.7646135307269818, + "learning_rate": 0.003, + "loss": 3.97, + "step": 40903 + }, + { + "epoch": 0.40904, + "grad_norm": 0.91715798626352, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 40904 + }, + { + "epoch": 0.40905, + "grad_norm": 1.0765914978870497, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 40905 + }, + { + "epoch": 0.40906, + "grad_norm": 1.1248302321236547, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 40906 + }, + { + "epoch": 0.40907, + "grad_norm": 0.821836524447724, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 40907 + }, + { + "epoch": 0.40908, + "grad_norm": 0.8175255386719689, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 40908 + }, + { + "epoch": 0.40909, + "grad_norm": 0.7879940338346694, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 40909 + }, + { + "epoch": 0.4091, + "grad_norm": 0.7210649343613844, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 40910 + }, + { + "epoch": 0.40911, + "grad_norm": 0.7176446561845063, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 40911 + }, + { + "epoch": 0.40912, + "grad_norm": 0.7381712046565565, + "learning_rate": 0.003, + "loss": 3.961, + "step": 40912 + }, + { + "epoch": 0.40913, + "grad_norm": 0.6572012408730096, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 40913 + }, + { + "epoch": 0.40914, + "grad_norm": 0.6228049592298813, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 40914 + }, + { + "epoch": 0.40915, + "grad_norm": 0.7139609929480406, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 40915 + }, + { + "epoch": 0.40916, + "grad_norm": 0.8885460622524309, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 40916 + }, + { + "epoch": 0.40917, + "grad_norm": 1.0835787015516427, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 40917 + }, + { + "epoch": 0.40918, + "grad_norm": 0.8615342392594829, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 40918 + }, + { + "epoch": 0.40919, + "grad_norm": 0.7859035597884577, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 40919 + }, + { + "epoch": 0.4092, + "grad_norm": 0.7594804182841493, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 40920 + }, + { + "epoch": 0.40921, + "grad_norm": 0.8803655362028138, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 40921 + }, + { + "epoch": 0.40922, + "grad_norm": 0.7352993044835102, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 40922 + }, + { + "epoch": 0.40923, + "grad_norm": 0.7972331250405045, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 40923 + }, + { + "epoch": 0.40924, + "grad_norm": 0.9581455247098598, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 40924 + }, + { + "epoch": 0.40925, + "grad_norm": 1.1150221314865945, + "learning_rate": 0.003, + "loss": 4.021, + "step": 40925 + }, + { + "epoch": 0.40926, + "grad_norm": 1.031929670650492, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 40926 + }, + { + "epoch": 0.40927, + "grad_norm": 0.8488519464488694, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 40927 + }, + { + "epoch": 0.40928, + "grad_norm": 0.8091592989369253, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 40928 + }, + { + "epoch": 0.40929, + "grad_norm": 0.7981520086470048, + "learning_rate": 0.003, + "loss": 3.9422, + "step": 40929 + }, + { + "epoch": 0.4093, + "grad_norm": 0.844339226044227, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 40930 + }, + { + "epoch": 0.40931, + "grad_norm": 1.001774024509494, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 40931 + }, + { + "epoch": 0.40932, + "grad_norm": 1.00982703178051, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 40932 + }, + { + "epoch": 0.40933, + "grad_norm": 0.9710422598513225, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 40933 + }, + { + "epoch": 0.40934, + "grad_norm": 0.8422698968043013, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 40934 + }, + { + "epoch": 0.40935, + "grad_norm": 0.7344100421735118, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 40935 + }, + { + "epoch": 0.40936, + "grad_norm": 0.679185119376394, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 40936 + }, + { + "epoch": 0.40937, + "grad_norm": 0.7596720878440504, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 40937 + }, + { + "epoch": 0.40938, + "grad_norm": 0.7516104693580825, + "learning_rate": 0.003, + "loss": 3.985, + "step": 40938 + }, + { + "epoch": 0.40939, + "grad_norm": 0.8555446453114712, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 40939 + }, + { + "epoch": 0.4094, + "grad_norm": 0.8626386226616419, + "learning_rate": 0.003, + "loss": 4.0, + "step": 40940 + }, + { + "epoch": 0.40941, + "grad_norm": 0.8569735702392518, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 40941 + }, + { + "epoch": 0.40942, + "grad_norm": 0.8678006838586687, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 40942 + }, + { + "epoch": 0.40943, + "grad_norm": 0.8679255634077234, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 40943 + }, + { + "epoch": 0.40944, + "grad_norm": 0.8400168045883979, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 40944 + }, + { + "epoch": 0.40945, + "grad_norm": 0.7881071986025298, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 40945 + }, + { + "epoch": 0.40946, + "grad_norm": 0.7712877883376521, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 40946 + }, + { + "epoch": 0.40947, + "grad_norm": 0.6939729579439108, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 40947 + }, + { + "epoch": 0.40948, + "grad_norm": 0.6376183500898581, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 40948 + }, + { + "epoch": 0.40949, + "grad_norm": 0.6755823471928508, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 40949 + }, + { + "epoch": 0.4095, + "grad_norm": 0.7160652827781274, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 40950 + }, + { + "epoch": 0.40951, + "grad_norm": 0.7658305653660575, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 40951 + }, + { + "epoch": 0.40952, + "grad_norm": 0.6984468544400072, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 40952 + }, + { + "epoch": 0.40953, + "grad_norm": 0.6653654796865791, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 40953 + }, + { + "epoch": 0.40954, + "grad_norm": 0.70641683257236, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 40954 + }, + { + "epoch": 0.40955, + "grad_norm": 0.6841596989933133, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 40955 + }, + { + "epoch": 0.40956, + "grad_norm": 0.5654357842149953, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 40956 + }, + { + "epoch": 0.40957, + "grad_norm": 0.590388050611476, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 40957 + }, + { + "epoch": 0.40958, + "grad_norm": 0.5966720234132931, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 40958 + }, + { + "epoch": 0.40959, + "grad_norm": 0.6106230102304182, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 40959 + }, + { + "epoch": 0.4096, + "grad_norm": 0.7450847541845241, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 40960 + }, + { + "epoch": 0.40961, + "grad_norm": 0.934961942164851, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 40961 + }, + { + "epoch": 0.40962, + "grad_norm": 0.9829460931867406, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 40962 + }, + { + "epoch": 0.40963, + "grad_norm": 0.9166618692157937, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 40963 + }, + { + "epoch": 0.40964, + "grad_norm": 0.8488637030442836, + "learning_rate": 0.003, + "loss": 3.987, + "step": 40964 + }, + { + "epoch": 0.40965, + "grad_norm": 0.8536540607226735, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 40965 + }, + { + "epoch": 0.40966, + "grad_norm": 0.9665279161940074, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 40966 + }, + { + "epoch": 0.40967, + "grad_norm": 0.910326426742192, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 40967 + }, + { + "epoch": 0.40968, + "grad_norm": 0.8883882086787686, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 40968 + }, + { + "epoch": 0.40969, + "grad_norm": 0.9098277624969825, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 40969 + }, + { + "epoch": 0.4097, + "grad_norm": 1.0854852885400879, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 40970 + }, + { + "epoch": 0.40971, + "grad_norm": 0.6892645497189482, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 40971 + }, + { + "epoch": 0.40972, + "grad_norm": 0.7652372625278672, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 40972 + }, + { + "epoch": 0.40973, + "grad_norm": 0.7394907891474504, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 40973 + }, + { + "epoch": 0.40974, + "grad_norm": 0.7681400967350398, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 40974 + }, + { + "epoch": 0.40975, + "grad_norm": 0.7868966338064802, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 40975 + }, + { + "epoch": 0.40976, + "grad_norm": 0.8131636429337186, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 40976 + }, + { + "epoch": 0.40977, + "grad_norm": 0.7709690723275756, + "learning_rate": 0.003, + "loss": 3.985, + "step": 40977 + }, + { + "epoch": 0.40978, + "grad_norm": 0.7119488927794122, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 40978 + }, + { + "epoch": 0.40979, + "grad_norm": 0.7325103123177653, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 40979 + }, + { + "epoch": 0.4098, + "grad_norm": 0.8328833729813767, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 40980 + }, + { + "epoch": 0.40981, + "grad_norm": 0.8860192803110373, + "learning_rate": 0.003, + "loss": 3.982, + "step": 40981 + }, + { + "epoch": 0.40982, + "grad_norm": 0.9296108700579809, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 40982 + }, + { + "epoch": 0.40983, + "grad_norm": 0.86763438533029, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 40983 + }, + { + "epoch": 0.40984, + "grad_norm": 0.7811018288770988, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 40984 + }, + { + "epoch": 0.40985, + "grad_norm": 0.6754463041849096, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 40985 + }, + { + "epoch": 0.40986, + "grad_norm": 0.6486828313242782, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 40986 + }, + { + "epoch": 0.40987, + "grad_norm": 0.7112378684831385, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 40987 + }, + { + "epoch": 0.40988, + "grad_norm": 0.7758509675959969, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 40988 + }, + { + "epoch": 0.40989, + "grad_norm": 0.795851470290018, + "learning_rate": 0.003, + "loss": 3.98, + "step": 40989 + }, + { + "epoch": 0.4099, + "grad_norm": 0.8505512431104744, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 40990 + }, + { + "epoch": 0.40991, + "grad_norm": 0.6774520367086854, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 40991 + }, + { + "epoch": 0.40992, + "grad_norm": 0.6674980676879682, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 40992 + }, + { + "epoch": 0.40993, + "grad_norm": 0.6005895057320847, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 40993 + }, + { + "epoch": 0.40994, + "grad_norm": 0.691201141757856, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 40994 + }, + { + "epoch": 0.40995, + "grad_norm": 0.7674564406393319, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 40995 + }, + { + "epoch": 0.40996, + "grad_norm": 0.918518243680427, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 40996 + }, + { + "epoch": 0.40997, + "grad_norm": 1.0601746390265945, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 40997 + }, + { + "epoch": 0.40998, + "grad_norm": 1.0633738464829494, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 40998 + }, + { + "epoch": 0.40999, + "grad_norm": 0.909878001325168, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 40999 + }, + { + "epoch": 0.41, + "grad_norm": 0.884254202879315, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 41000 + }, + { + "epoch": 0.41001, + "grad_norm": 0.7105440157918108, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 41001 + }, + { + "epoch": 0.41002, + "grad_norm": 0.6828491675984278, + "learning_rate": 0.003, + "loss": 3.9483, + "step": 41002 + }, + { + "epoch": 0.41003, + "grad_norm": 0.5854521794777359, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 41003 + }, + { + "epoch": 0.41004, + "grad_norm": 0.6572407708452713, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 41004 + }, + { + "epoch": 0.41005, + "grad_norm": 0.7242626964049811, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 41005 + }, + { + "epoch": 0.41006, + "grad_norm": 0.8021278608464701, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 41006 + }, + { + "epoch": 0.41007, + "grad_norm": 0.7760083541190036, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 41007 + }, + { + "epoch": 0.41008, + "grad_norm": 0.7094396604164376, + "learning_rate": 0.003, + "loss": 4.0, + "step": 41008 + }, + { + "epoch": 0.41009, + "grad_norm": 0.8411057562890079, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 41009 + }, + { + "epoch": 0.4101, + "grad_norm": 0.8928756396946267, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 41010 + }, + { + "epoch": 0.41011, + "grad_norm": 0.7933482694652093, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 41011 + }, + { + "epoch": 0.41012, + "grad_norm": 0.7434142150408348, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 41012 + }, + { + "epoch": 0.41013, + "grad_norm": 0.7504692297026316, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 41013 + }, + { + "epoch": 0.41014, + "grad_norm": 0.8183917712981864, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 41014 + }, + { + "epoch": 0.41015, + "grad_norm": 0.7856705914663317, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 41015 + }, + { + "epoch": 0.41016, + "grad_norm": 0.7763568175166506, + "learning_rate": 0.003, + "loss": 3.981, + "step": 41016 + }, + { + "epoch": 0.41017, + "grad_norm": 0.8415370335137822, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 41017 + }, + { + "epoch": 0.41018, + "grad_norm": 1.003273422232033, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 41018 + }, + { + "epoch": 0.41019, + "grad_norm": 1.1127012422132758, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 41019 + }, + { + "epoch": 0.4102, + "grad_norm": 0.8403599915500233, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 41020 + }, + { + "epoch": 0.41021, + "grad_norm": 0.7516471348862394, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 41021 + }, + { + "epoch": 0.41022, + "grad_norm": 0.7390031068194612, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 41022 + }, + { + "epoch": 0.41023, + "grad_norm": 0.6784650741793113, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 41023 + }, + { + "epoch": 0.41024, + "grad_norm": 0.6935752967914429, + "learning_rate": 0.003, + "loss": 3.98, + "step": 41024 + }, + { + "epoch": 0.41025, + "grad_norm": 0.7186716235870201, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 41025 + }, + { + "epoch": 0.41026, + "grad_norm": 0.6559348051239441, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 41026 + }, + { + "epoch": 0.41027, + "grad_norm": 0.6291954351145411, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 41027 + }, + { + "epoch": 0.41028, + "grad_norm": 0.6278305936612696, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 41028 + }, + { + "epoch": 0.41029, + "grad_norm": 0.6566368192791827, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 41029 + }, + { + "epoch": 0.4103, + "grad_norm": 0.6049814323985717, + "learning_rate": 0.003, + "loss": 3.966, + "step": 41030 + }, + { + "epoch": 0.41031, + "grad_norm": 0.5252502046396139, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 41031 + }, + { + "epoch": 0.41032, + "grad_norm": 0.5622128022404649, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 41032 + }, + { + "epoch": 0.41033, + "grad_norm": 0.6373998403559892, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 41033 + }, + { + "epoch": 0.41034, + "grad_norm": 0.7301239763823845, + "learning_rate": 0.003, + "loss": 3.984, + "step": 41034 + }, + { + "epoch": 0.41035, + "grad_norm": 0.9329140754449079, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 41035 + }, + { + "epoch": 0.41036, + "grad_norm": 1.126047329915157, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 41036 + }, + { + "epoch": 0.41037, + "grad_norm": 0.9155563664888504, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 41037 + }, + { + "epoch": 0.41038, + "grad_norm": 0.9120199586293295, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 41038 + }, + { + "epoch": 0.41039, + "grad_norm": 0.8748369317228658, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 41039 + }, + { + "epoch": 0.4104, + "grad_norm": 0.8156780098102601, + "learning_rate": 0.003, + "loss": 4.032, + "step": 41040 + }, + { + "epoch": 0.41041, + "grad_norm": 0.7436868292681457, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 41041 + }, + { + "epoch": 0.41042, + "grad_norm": 0.7362855113475476, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 41042 + }, + { + "epoch": 0.41043, + "grad_norm": 0.7354743675492492, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 41043 + }, + { + "epoch": 0.41044, + "grad_norm": 0.8045071148169394, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 41044 + }, + { + "epoch": 0.41045, + "grad_norm": 0.9260402924674107, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 41045 + }, + { + "epoch": 0.41046, + "grad_norm": 1.005848665389195, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 41046 + }, + { + "epoch": 0.41047, + "grad_norm": 1.040514037780205, + "learning_rate": 0.003, + "loss": 3.998, + "step": 41047 + }, + { + "epoch": 0.41048, + "grad_norm": 0.860603413453318, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 41048 + }, + { + "epoch": 0.41049, + "grad_norm": 0.780330035950817, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 41049 + }, + { + "epoch": 0.4105, + "grad_norm": 0.7741378270454162, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 41050 + }, + { + "epoch": 0.41051, + "grad_norm": 0.769033581653955, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 41051 + }, + { + "epoch": 0.41052, + "grad_norm": 0.6807472793554337, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 41052 + }, + { + "epoch": 0.41053, + "grad_norm": 0.6920859820682878, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 41053 + }, + { + "epoch": 0.41054, + "grad_norm": 0.7742073932408602, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 41054 + }, + { + "epoch": 0.41055, + "grad_norm": 0.7621665908527245, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 41055 + }, + { + "epoch": 0.41056, + "grad_norm": 0.7637686475667752, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 41056 + }, + { + "epoch": 0.41057, + "grad_norm": 0.7012896127831679, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 41057 + }, + { + "epoch": 0.41058, + "grad_norm": 0.6318751671153525, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 41058 + }, + { + "epoch": 0.41059, + "grad_norm": 0.5645842029764623, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 41059 + }, + { + "epoch": 0.4106, + "grad_norm": 0.571735247525978, + "learning_rate": 0.003, + "loss": 3.9436, + "step": 41060 + }, + { + "epoch": 0.41061, + "grad_norm": 0.5998295751141671, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 41061 + }, + { + "epoch": 0.41062, + "grad_norm": 0.7652390180590255, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 41062 + }, + { + "epoch": 0.41063, + "grad_norm": 0.8737637783146985, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 41063 + }, + { + "epoch": 0.41064, + "grad_norm": 0.9051056630936475, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 41064 + }, + { + "epoch": 0.41065, + "grad_norm": 0.8652756582577117, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 41065 + }, + { + "epoch": 0.41066, + "grad_norm": 0.9620274415455949, + "learning_rate": 0.003, + "loss": 4.001, + "step": 41066 + }, + { + "epoch": 0.41067, + "grad_norm": 1.061280578340127, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 41067 + }, + { + "epoch": 0.41068, + "grad_norm": 0.820963817695247, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 41068 + }, + { + "epoch": 0.41069, + "grad_norm": 1.0260435595627018, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 41069 + }, + { + "epoch": 0.4107, + "grad_norm": 1.1418405668012892, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 41070 + }, + { + "epoch": 0.41071, + "grad_norm": 0.8610569055410549, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 41071 + }, + { + "epoch": 0.41072, + "grad_norm": 0.907649706908446, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 41072 + }, + { + "epoch": 0.41073, + "grad_norm": 1.1084498349837673, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 41073 + }, + { + "epoch": 0.41074, + "grad_norm": 1.0312511497714565, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 41074 + }, + { + "epoch": 0.41075, + "grad_norm": 1.1908806253001594, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 41075 + }, + { + "epoch": 0.41076, + "grad_norm": 1.3636444562833994, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 41076 + }, + { + "epoch": 0.41077, + "grad_norm": 1.2756386863646403, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 41077 + }, + { + "epoch": 0.41078, + "grad_norm": 0.9468173940304998, + "learning_rate": 0.003, + "loss": 3.993, + "step": 41078 + }, + { + "epoch": 0.41079, + "grad_norm": 0.9941109258843502, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 41079 + }, + { + "epoch": 0.4108, + "grad_norm": 1.044753386840827, + "learning_rate": 0.003, + "loss": 3.986, + "step": 41080 + }, + { + "epoch": 0.41081, + "grad_norm": 0.9672436519473516, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 41081 + }, + { + "epoch": 0.41082, + "grad_norm": 0.8656582874515664, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 41082 + }, + { + "epoch": 0.41083, + "grad_norm": 0.92449516002832, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 41083 + }, + { + "epoch": 0.41084, + "grad_norm": 1.3534460075798007, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 41084 + }, + { + "epoch": 0.41085, + "grad_norm": 1.09010619746639, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 41085 + }, + { + "epoch": 0.41086, + "grad_norm": 1.078280295843465, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 41086 + }, + { + "epoch": 0.41087, + "grad_norm": 0.8219908230987548, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 41087 + }, + { + "epoch": 0.41088, + "grad_norm": 0.7187155244676782, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 41088 + }, + { + "epoch": 0.41089, + "grad_norm": 0.8493940530800971, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 41089 + }, + { + "epoch": 0.4109, + "grad_norm": 0.8554191500101949, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 41090 + }, + { + "epoch": 0.41091, + "grad_norm": 0.7926476688003765, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 41091 + }, + { + "epoch": 0.41092, + "grad_norm": 0.8020202441982726, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 41092 + }, + { + "epoch": 0.41093, + "grad_norm": 0.7363426062991195, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 41093 + }, + { + "epoch": 0.41094, + "grad_norm": 0.6864237728808095, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 41094 + }, + { + "epoch": 0.41095, + "grad_norm": 0.7466041452080064, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 41095 + }, + { + "epoch": 0.41096, + "grad_norm": 0.7464372229213568, + "learning_rate": 0.003, + "loss": 3.996, + "step": 41096 + }, + { + "epoch": 0.41097, + "grad_norm": 0.7852477852617341, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 41097 + }, + { + "epoch": 0.41098, + "grad_norm": 0.775269690200319, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 41098 + }, + { + "epoch": 0.41099, + "grad_norm": 0.766173882791954, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 41099 + }, + { + "epoch": 0.411, + "grad_norm": 0.7967363122912479, + "learning_rate": 0.003, + "loss": 3.995, + "step": 41100 + }, + { + "epoch": 0.41101, + "grad_norm": 0.8342738627077637, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 41101 + }, + { + "epoch": 0.41102, + "grad_norm": 0.7900370913043944, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 41102 + }, + { + "epoch": 0.41103, + "grad_norm": 0.7056722257193059, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41103 + }, + { + "epoch": 0.41104, + "grad_norm": 0.6688988665290582, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 41104 + }, + { + "epoch": 0.41105, + "grad_norm": 0.7538414269310979, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 41105 + }, + { + "epoch": 0.41106, + "grad_norm": 0.9011562163479795, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 41106 + }, + { + "epoch": 0.41107, + "grad_norm": 1.0349100953390846, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 41107 + }, + { + "epoch": 0.41108, + "grad_norm": 0.9910906567740636, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 41108 + }, + { + "epoch": 0.41109, + "grad_norm": 0.9734023196439223, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 41109 + }, + { + "epoch": 0.4111, + "grad_norm": 1.0007827651676484, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 41110 + }, + { + "epoch": 0.41111, + "grad_norm": 0.9379613833910666, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 41111 + }, + { + "epoch": 0.41112, + "grad_norm": 0.883347739805082, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 41112 + }, + { + "epoch": 0.41113, + "grad_norm": 0.7866913546180365, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41113 + }, + { + "epoch": 0.41114, + "grad_norm": 0.6922943475901046, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 41114 + }, + { + "epoch": 0.41115, + "grad_norm": 0.7619108732991483, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 41115 + }, + { + "epoch": 0.41116, + "grad_norm": 0.7509775727442733, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 41116 + }, + { + "epoch": 0.41117, + "grad_norm": 0.7104019653176624, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 41117 + }, + { + "epoch": 0.41118, + "grad_norm": 0.7656687338327984, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 41118 + }, + { + "epoch": 0.41119, + "grad_norm": 0.8209010296015681, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 41119 + }, + { + "epoch": 0.4112, + "grad_norm": 0.8591550085999762, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 41120 + }, + { + "epoch": 0.41121, + "grad_norm": 0.7615822094984821, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 41121 + }, + { + "epoch": 0.41122, + "grad_norm": 0.622099510593609, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 41122 + }, + { + "epoch": 0.41123, + "grad_norm": 0.6653030221680858, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 41123 + }, + { + "epoch": 0.41124, + "grad_norm": 0.6006403639270581, + "learning_rate": 0.003, + "loss": 3.973, + "step": 41124 + }, + { + "epoch": 0.41125, + "grad_norm": 0.619971298571943, + "learning_rate": 0.003, + "loss": 3.982, + "step": 41125 + }, + { + "epoch": 0.41126, + "grad_norm": 0.6271343127339101, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 41126 + }, + { + "epoch": 0.41127, + "grad_norm": 0.7138373987067652, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 41127 + }, + { + "epoch": 0.41128, + "grad_norm": 0.8751302359864774, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 41128 + }, + { + "epoch": 0.41129, + "grad_norm": 0.9693647391237815, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 41129 + }, + { + "epoch": 0.4113, + "grad_norm": 0.8397546513009349, + "learning_rate": 0.003, + "loss": 3.981, + "step": 41130 + }, + { + "epoch": 0.41131, + "grad_norm": 0.8045904429774797, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 41131 + }, + { + "epoch": 0.41132, + "grad_norm": 0.7140941510352143, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 41132 + }, + { + "epoch": 0.41133, + "grad_norm": 0.77077044956994, + "learning_rate": 0.003, + "loss": 3.9488, + "step": 41133 + }, + { + "epoch": 0.41134, + "grad_norm": 0.7942079267166043, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 41134 + }, + { + "epoch": 0.41135, + "grad_norm": 0.7681774873975437, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 41135 + }, + { + "epoch": 0.41136, + "grad_norm": 0.7043143434098943, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 41136 + }, + { + "epoch": 0.41137, + "grad_norm": 0.7112663010687849, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 41137 + }, + { + "epoch": 0.41138, + "grad_norm": 0.661822792521502, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 41138 + }, + { + "epoch": 0.41139, + "grad_norm": 0.5292891233964961, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 41139 + }, + { + "epoch": 0.4114, + "grad_norm": 0.6139255161121816, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 41140 + }, + { + "epoch": 0.41141, + "grad_norm": 0.7345554971207983, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 41141 + }, + { + "epoch": 0.41142, + "grad_norm": 0.9537163903185099, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 41142 + }, + { + "epoch": 0.41143, + "grad_norm": 1.0504649857929675, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 41143 + }, + { + "epoch": 0.41144, + "grad_norm": 0.9203956149898725, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 41144 + }, + { + "epoch": 0.41145, + "grad_norm": 0.7002684943739789, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 41145 + }, + { + "epoch": 0.41146, + "grad_norm": 0.7435762252222112, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 41146 + }, + { + "epoch": 0.41147, + "grad_norm": 0.9038758964572384, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 41147 + }, + { + "epoch": 0.41148, + "grad_norm": 0.9533366389295066, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 41148 + }, + { + "epoch": 0.41149, + "grad_norm": 0.9051290847259664, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 41149 + }, + { + "epoch": 0.4115, + "grad_norm": 0.9012091879514088, + "learning_rate": 0.003, + "loss": 3.999, + "step": 41150 + }, + { + "epoch": 0.41151, + "grad_norm": 0.7449897189628467, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 41151 + }, + { + "epoch": 0.41152, + "grad_norm": 0.7171958099200683, + "learning_rate": 0.003, + "loss": 4.017, + "step": 41152 + }, + { + "epoch": 0.41153, + "grad_norm": 0.7670524795287406, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 41153 + }, + { + "epoch": 0.41154, + "grad_norm": 0.8348605972157608, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 41154 + }, + { + "epoch": 0.41155, + "grad_norm": 0.863039663136941, + "learning_rate": 0.003, + "loss": 3.9399, + "step": 41155 + }, + { + "epoch": 0.41156, + "grad_norm": 0.9507966618810596, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 41156 + }, + { + "epoch": 0.41157, + "grad_norm": 0.8273198514347906, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 41157 + }, + { + "epoch": 0.41158, + "grad_norm": 0.6774055941767152, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 41158 + }, + { + "epoch": 0.41159, + "grad_norm": 0.6846454791284173, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 41159 + }, + { + "epoch": 0.4116, + "grad_norm": 0.6725513351546194, + "learning_rate": 0.003, + "loss": 3.9473, + "step": 41160 + }, + { + "epoch": 0.41161, + "grad_norm": 0.7593229925625721, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 41161 + }, + { + "epoch": 0.41162, + "grad_norm": 0.8294744685086881, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 41162 + }, + { + "epoch": 0.41163, + "grad_norm": 0.9007003568504919, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 41163 + }, + { + "epoch": 0.41164, + "grad_norm": 1.0152121262683103, + "learning_rate": 0.003, + "loss": 3.973, + "step": 41164 + }, + { + "epoch": 0.41165, + "grad_norm": 1.0716076928979024, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 41165 + }, + { + "epoch": 0.41166, + "grad_norm": 0.7799015931489317, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 41166 + }, + { + "epoch": 0.41167, + "grad_norm": 0.7218895655668396, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 41167 + }, + { + "epoch": 0.41168, + "grad_norm": 0.6861942057846013, + "learning_rate": 0.003, + "loss": 3.997, + "step": 41168 + }, + { + "epoch": 0.41169, + "grad_norm": 0.7163417807891589, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 41169 + }, + { + "epoch": 0.4117, + "grad_norm": 0.731246776825747, + "learning_rate": 0.003, + "loss": 3.988, + "step": 41170 + }, + { + "epoch": 0.41171, + "grad_norm": 0.862578453627493, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 41171 + }, + { + "epoch": 0.41172, + "grad_norm": 0.8458115118231092, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 41172 + }, + { + "epoch": 0.41173, + "grad_norm": 0.8930301878794766, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 41173 + }, + { + "epoch": 0.41174, + "grad_norm": 1.0203589704489224, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 41174 + }, + { + "epoch": 0.41175, + "grad_norm": 0.8218390426532209, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 41175 + }, + { + "epoch": 0.41176, + "grad_norm": 0.6914972256627754, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41176 + }, + { + "epoch": 0.41177, + "grad_norm": 0.6538864150889475, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 41177 + }, + { + "epoch": 0.41178, + "grad_norm": 0.6576686618824763, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 41178 + }, + { + "epoch": 0.41179, + "grad_norm": 0.5966534035178501, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 41179 + }, + { + "epoch": 0.4118, + "grad_norm": 0.5316252040353334, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 41180 + }, + { + "epoch": 0.41181, + "grad_norm": 0.5814919739127264, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 41181 + }, + { + "epoch": 0.41182, + "grad_norm": 0.5897150016040054, + "learning_rate": 0.003, + "loss": 4.002, + "step": 41182 + }, + { + "epoch": 0.41183, + "grad_norm": 0.6103301242960145, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 41183 + }, + { + "epoch": 0.41184, + "grad_norm": 0.572039919127553, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 41184 + }, + { + "epoch": 0.41185, + "grad_norm": 0.45606691404596444, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 41185 + }, + { + "epoch": 0.41186, + "grad_norm": 0.6086769248578655, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 41186 + }, + { + "epoch": 0.41187, + "grad_norm": 0.7742061567077234, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 41187 + }, + { + "epoch": 0.41188, + "grad_norm": 0.7958048578145583, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 41188 + }, + { + "epoch": 0.41189, + "grad_norm": 0.798772114431248, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 41189 + }, + { + "epoch": 0.4119, + "grad_norm": 0.7760443384181891, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 41190 + }, + { + "epoch": 0.41191, + "grad_norm": 0.7949779082098293, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 41191 + }, + { + "epoch": 0.41192, + "grad_norm": 0.873635394276386, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 41192 + }, + { + "epoch": 0.41193, + "grad_norm": 0.9183577599917729, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 41193 + }, + { + "epoch": 0.41194, + "grad_norm": 0.9424102083148064, + "learning_rate": 0.003, + "loss": 4.015, + "step": 41194 + }, + { + "epoch": 0.41195, + "grad_norm": 0.7990242256203399, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 41195 + }, + { + "epoch": 0.41196, + "grad_norm": 0.723541918527369, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 41196 + }, + { + "epoch": 0.41197, + "grad_norm": 0.7143943814427282, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 41197 + }, + { + "epoch": 0.41198, + "grad_norm": 0.8459136809365532, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 41198 + }, + { + "epoch": 0.41199, + "grad_norm": 1.0795074328680943, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 41199 + }, + { + "epoch": 0.412, + "grad_norm": 1.0746031665700357, + "learning_rate": 0.003, + "loss": 3.96, + "step": 41200 + }, + { + "epoch": 0.41201, + "grad_norm": 0.6962163414534029, + "learning_rate": 0.003, + "loss": 3.993, + "step": 41201 + }, + { + "epoch": 0.41202, + "grad_norm": 0.6341868866885435, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41202 + }, + { + "epoch": 0.41203, + "grad_norm": 0.7975943336710882, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 41203 + }, + { + "epoch": 0.41204, + "grad_norm": 0.9644568485759965, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41204 + }, + { + "epoch": 0.41205, + "grad_norm": 0.9706115558983409, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 41205 + }, + { + "epoch": 0.41206, + "grad_norm": 0.887812442531271, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 41206 + }, + { + "epoch": 0.41207, + "grad_norm": 0.7716027723893879, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 41207 + }, + { + "epoch": 0.41208, + "grad_norm": 0.7012450349568528, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 41208 + }, + { + "epoch": 0.41209, + "grad_norm": 0.6653288042882666, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 41209 + }, + { + "epoch": 0.4121, + "grad_norm": 0.5897728573431367, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 41210 + }, + { + "epoch": 0.41211, + "grad_norm": 0.6191563278818834, + "learning_rate": 0.003, + "loss": 3.9407, + "step": 41211 + }, + { + "epoch": 0.41212, + "grad_norm": 0.778907738514019, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 41212 + }, + { + "epoch": 0.41213, + "grad_norm": 0.8595450950692997, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 41213 + }, + { + "epoch": 0.41214, + "grad_norm": 0.8551961152200085, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 41214 + }, + { + "epoch": 0.41215, + "grad_norm": 0.7212599903093392, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 41215 + }, + { + "epoch": 0.41216, + "grad_norm": 0.6960495746792321, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 41216 + }, + { + "epoch": 0.41217, + "grad_norm": 0.7478894931215143, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 41217 + }, + { + "epoch": 0.41218, + "grad_norm": 0.8470520039748437, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 41218 + }, + { + "epoch": 0.41219, + "grad_norm": 0.8953708822905435, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 41219 + }, + { + "epoch": 0.4122, + "grad_norm": 0.8854607479634464, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 41220 + }, + { + "epoch": 0.41221, + "grad_norm": 0.7870061868654318, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 41221 + }, + { + "epoch": 0.41222, + "grad_norm": 0.811610995915471, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 41222 + }, + { + "epoch": 0.41223, + "grad_norm": 0.7987523744620892, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 41223 + }, + { + "epoch": 0.41224, + "grad_norm": 0.7589093524747396, + "learning_rate": 0.003, + "loss": 3.981, + "step": 41224 + }, + { + "epoch": 0.41225, + "grad_norm": 0.6723798975291901, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 41225 + }, + { + "epoch": 0.41226, + "grad_norm": 0.6680122537025159, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 41226 + }, + { + "epoch": 0.41227, + "grad_norm": 0.7384030115013619, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 41227 + }, + { + "epoch": 0.41228, + "grad_norm": 0.7881415987625952, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 41228 + }, + { + "epoch": 0.41229, + "grad_norm": 0.8513632305886475, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 41229 + }, + { + "epoch": 0.4123, + "grad_norm": 0.8126483198519457, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 41230 + }, + { + "epoch": 0.41231, + "grad_norm": 0.8027274012741046, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 41231 + }, + { + "epoch": 0.41232, + "grad_norm": 0.8652913610870497, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 41232 + }, + { + "epoch": 0.41233, + "grad_norm": 0.8003589381417595, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 41233 + }, + { + "epoch": 0.41234, + "grad_norm": 0.723299663705009, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 41234 + }, + { + "epoch": 0.41235, + "grad_norm": 0.8160361473029099, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 41235 + }, + { + "epoch": 0.41236, + "grad_norm": 0.7930032201983667, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 41236 + }, + { + "epoch": 0.41237, + "grad_norm": 0.8353456136119667, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 41237 + }, + { + "epoch": 0.41238, + "grad_norm": 0.8886468878864661, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 41238 + }, + { + "epoch": 0.41239, + "grad_norm": 0.7906258653995992, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 41239 + }, + { + "epoch": 0.4124, + "grad_norm": 0.876956003173558, + "learning_rate": 0.003, + "loss": 4.006, + "step": 41240 + }, + { + "epoch": 0.41241, + "grad_norm": 0.9033362537171236, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 41241 + }, + { + "epoch": 0.41242, + "grad_norm": 0.8965770235820733, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 41242 + }, + { + "epoch": 0.41243, + "grad_norm": 0.7885047681576192, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 41243 + }, + { + "epoch": 0.41244, + "grad_norm": 0.6793484652263555, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 41244 + }, + { + "epoch": 0.41245, + "grad_norm": 0.7073358222112536, + "learning_rate": 0.003, + "loss": 3.98, + "step": 41245 + }, + { + "epoch": 0.41246, + "grad_norm": 0.7497228714179931, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 41246 + }, + { + "epoch": 0.41247, + "grad_norm": 0.8187384625693689, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 41247 + }, + { + "epoch": 0.41248, + "grad_norm": 0.8404447367912533, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 41248 + }, + { + "epoch": 0.41249, + "grad_norm": 0.8779220330333464, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 41249 + }, + { + "epoch": 0.4125, + "grad_norm": 0.9227930820775689, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 41250 + }, + { + "epoch": 0.41251, + "grad_norm": 0.9294516003684461, + "learning_rate": 0.003, + "loss": 4.015, + "step": 41251 + }, + { + "epoch": 0.41252, + "grad_norm": 0.9655875487907454, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 41252 + }, + { + "epoch": 0.41253, + "grad_norm": 0.8338326426732694, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 41253 + }, + { + "epoch": 0.41254, + "grad_norm": 0.7370309223292664, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 41254 + }, + { + "epoch": 0.41255, + "grad_norm": 0.7857906978561502, + "learning_rate": 0.003, + "loss": 4.002, + "step": 41255 + }, + { + "epoch": 0.41256, + "grad_norm": 0.88878721634023, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 41256 + }, + { + "epoch": 0.41257, + "grad_norm": 0.8598589156292337, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 41257 + }, + { + "epoch": 0.41258, + "grad_norm": 0.822950461780033, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 41258 + }, + { + "epoch": 0.41259, + "grad_norm": 0.944714806891876, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 41259 + }, + { + "epoch": 0.4126, + "grad_norm": 0.8786214164880223, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 41260 + }, + { + "epoch": 0.41261, + "grad_norm": 0.6364984157373427, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 41261 + }, + { + "epoch": 0.41262, + "grad_norm": 0.6171497772157278, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 41262 + }, + { + "epoch": 0.41263, + "grad_norm": 0.6337757499262424, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 41263 + }, + { + "epoch": 0.41264, + "grad_norm": 0.735466892395417, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 41264 + }, + { + "epoch": 0.41265, + "grad_norm": 0.7007815165155511, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 41265 + }, + { + "epoch": 0.41266, + "grad_norm": 0.6471734914399582, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 41266 + }, + { + "epoch": 0.41267, + "grad_norm": 0.6738052824527992, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 41267 + }, + { + "epoch": 0.41268, + "grad_norm": 0.697515104975136, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 41268 + }, + { + "epoch": 0.41269, + "grad_norm": 0.6745806610608731, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 41269 + }, + { + "epoch": 0.4127, + "grad_norm": 0.6689524721151205, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 41270 + }, + { + "epoch": 0.41271, + "grad_norm": 0.5873817104838424, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 41271 + }, + { + "epoch": 0.41272, + "grad_norm": 0.6320652964524431, + "learning_rate": 0.003, + "loss": 3.982, + "step": 41272 + }, + { + "epoch": 0.41273, + "grad_norm": 0.6410316227044157, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 41273 + }, + { + "epoch": 0.41274, + "grad_norm": 0.6476884203969064, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 41274 + }, + { + "epoch": 0.41275, + "grad_norm": 0.612433559827304, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 41275 + }, + { + "epoch": 0.41276, + "grad_norm": 0.638481752719147, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 41276 + }, + { + "epoch": 0.41277, + "grad_norm": 0.671388774760477, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 41277 + }, + { + "epoch": 0.41278, + "grad_norm": 0.6888992254280251, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 41278 + }, + { + "epoch": 0.41279, + "grad_norm": 0.6979208217843151, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 41279 + }, + { + "epoch": 0.4128, + "grad_norm": 0.7541990564235731, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 41280 + }, + { + "epoch": 0.41281, + "grad_norm": 0.8086833170172977, + "learning_rate": 0.003, + "loss": 3.973, + "step": 41281 + }, + { + "epoch": 0.41282, + "grad_norm": 0.8398602130346755, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 41282 + }, + { + "epoch": 0.41283, + "grad_norm": 0.8036067195461536, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 41283 + }, + { + "epoch": 0.41284, + "grad_norm": 0.8127814981066201, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 41284 + }, + { + "epoch": 0.41285, + "grad_norm": 0.979573097940948, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 41285 + }, + { + "epoch": 0.41286, + "grad_norm": 1.2415046792382167, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 41286 + }, + { + "epoch": 0.41287, + "grad_norm": 0.9076806974514471, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 41287 + }, + { + "epoch": 0.41288, + "grad_norm": 0.8075613961300935, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 41288 + }, + { + "epoch": 0.41289, + "grad_norm": 0.8215376312598911, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 41289 + }, + { + "epoch": 0.4129, + "grad_norm": 0.8115492381779645, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 41290 + }, + { + "epoch": 0.41291, + "grad_norm": 0.833674141395511, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 41291 + }, + { + "epoch": 0.41292, + "grad_norm": 0.7172051195608182, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 41292 + }, + { + "epoch": 0.41293, + "grad_norm": 0.8142969283115116, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 41293 + }, + { + "epoch": 0.41294, + "grad_norm": 0.8650871106579576, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 41294 + }, + { + "epoch": 0.41295, + "grad_norm": 0.8704735187216157, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 41295 + }, + { + "epoch": 0.41296, + "grad_norm": 0.7195072675898383, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 41296 + }, + { + "epoch": 0.41297, + "grad_norm": 0.7326414695347058, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 41297 + }, + { + "epoch": 0.41298, + "grad_norm": 0.7373053893263944, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 41298 + }, + { + "epoch": 0.41299, + "grad_norm": 0.7292246074514165, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 41299 + }, + { + "epoch": 0.413, + "grad_norm": 0.7488749117356035, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 41300 + }, + { + "epoch": 0.41301, + "grad_norm": 0.7287966346596466, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 41301 + }, + { + "epoch": 0.41302, + "grad_norm": 0.7366513171230867, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 41302 + }, + { + "epoch": 0.41303, + "grad_norm": 0.7318315945054578, + "learning_rate": 0.003, + "loss": 3.974, + "step": 41303 + }, + { + "epoch": 0.41304, + "grad_norm": 0.7791620768417454, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 41304 + }, + { + "epoch": 0.41305, + "grad_norm": 0.7875597862790388, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41305 + }, + { + "epoch": 0.41306, + "grad_norm": 0.8435679514221194, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 41306 + }, + { + "epoch": 0.41307, + "grad_norm": 0.7598907672510751, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 41307 + }, + { + "epoch": 0.41308, + "grad_norm": 0.7771510367913047, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 41308 + }, + { + "epoch": 0.41309, + "grad_norm": 0.8195842976391445, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 41309 + }, + { + "epoch": 0.4131, + "grad_norm": 0.8088169003356908, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 41310 + }, + { + "epoch": 0.41311, + "grad_norm": 0.9591340332648628, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 41311 + }, + { + "epoch": 0.41312, + "grad_norm": 0.9681534175759418, + "learning_rate": 0.003, + "loss": 3.996, + "step": 41312 + }, + { + "epoch": 0.41313, + "grad_norm": 1.0284024549612782, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 41313 + }, + { + "epoch": 0.41314, + "grad_norm": 1.099695992917106, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 41314 + }, + { + "epoch": 0.41315, + "grad_norm": 0.7587871985043565, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 41315 + }, + { + "epoch": 0.41316, + "grad_norm": 0.7124595527074999, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 41316 + }, + { + "epoch": 0.41317, + "grad_norm": 0.7852859341940541, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 41317 + }, + { + "epoch": 0.41318, + "grad_norm": 0.922485394671438, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 41318 + }, + { + "epoch": 0.41319, + "grad_norm": 1.0803479655351986, + "learning_rate": 0.003, + "loss": 4.0327, + "step": 41319 + }, + { + "epoch": 0.4132, + "grad_norm": 0.9023546714325296, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 41320 + }, + { + "epoch": 0.41321, + "grad_norm": 1.0862370120058709, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 41321 + }, + { + "epoch": 0.41322, + "grad_norm": 1.2221280096282, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 41322 + }, + { + "epoch": 0.41323, + "grad_norm": 0.8516047528588635, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 41323 + }, + { + "epoch": 0.41324, + "grad_norm": 0.6998081773421674, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 41324 + }, + { + "epoch": 0.41325, + "grad_norm": 0.7104966873677308, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 41325 + }, + { + "epoch": 0.41326, + "grad_norm": 0.7293310655145234, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 41326 + }, + { + "epoch": 0.41327, + "grad_norm": 0.6334910298050875, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 41327 + }, + { + "epoch": 0.41328, + "grad_norm": 0.60766919049141, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 41328 + }, + { + "epoch": 0.41329, + "grad_norm": 0.6631006876280224, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 41329 + }, + { + "epoch": 0.4133, + "grad_norm": 0.6124446556666167, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 41330 + }, + { + "epoch": 0.41331, + "grad_norm": 0.593156174564646, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 41331 + }, + { + "epoch": 0.41332, + "grad_norm": 0.5823607715852316, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 41332 + }, + { + "epoch": 0.41333, + "grad_norm": 0.5992369939385316, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 41333 + }, + { + "epoch": 0.41334, + "grad_norm": 0.6988020675385033, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 41334 + }, + { + "epoch": 0.41335, + "grad_norm": 0.8272803953541022, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 41335 + }, + { + "epoch": 0.41336, + "grad_norm": 0.9126736195719864, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 41336 + }, + { + "epoch": 0.41337, + "grad_norm": 0.8343766250878873, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 41337 + }, + { + "epoch": 0.41338, + "grad_norm": 0.8533727043856396, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 41338 + }, + { + "epoch": 0.41339, + "grad_norm": 0.8860903850821422, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 41339 + }, + { + "epoch": 0.4134, + "grad_norm": 0.8315207915395625, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 41340 + }, + { + "epoch": 0.41341, + "grad_norm": 0.7472274976510385, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 41341 + }, + { + "epoch": 0.41342, + "grad_norm": 0.7258674537364908, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 41342 + }, + { + "epoch": 0.41343, + "grad_norm": 0.75323082418054, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 41343 + }, + { + "epoch": 0.41344, + "grad_norm": 0.7650810180798321, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 41344 + }, + { + "epoch": 0.41345, + "grad_norm": 0.5945054615510965, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 41345 + }, + { + "epoch": 0.41346, + "grad_norm": 0.6501147563881504, + "learning_rate": 0.003, + "loss": 3.9447, + "step": 41346 + }, + { + "epoch": 0.41347, + "grad_norm": 0.6460670649085896, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 41347 + }, + { + "epoch": 0.41348, + "grad_norm": 0.641542043609043, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 41348 + }, + { + "epoch": 0.41349, + "grad_norm": 0.7254593156684225, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 41349 + }, + { + "epoch": 0.4135, + "grad_norm": 0.7853188944415811, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 41350 + }, + { + "epoch": 0.41351, + "grad_norm": 0.8317346732104033, + "learning_rate": 0.003, + "loss": 4.001, + "step": 41351 + }, + { + "epoch": 0.41352, + "grad_norm": 0.9623075173151255, + "learning_rate": 0.003, + "loss": 3.975, + "step": 41352 + }, + { + "epoch": 0.41353, + "grad_norm": 1.0156527578575667, + "learning_rate": 0.003, + "loss": 3.994, + "step": 41353 + }, + { + "epoch": 0.41354, + "grad_norm": 1.030828515043789, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 41354 + }, + { + "epoch": 0.41355, + "grad_norm": 1.0499637916716846, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 41355 + }, + { + "epoch": 0.41356, + "grad_norm": 0.9578112683483758, + "learning_rate": 0.003, + "loss": 3.999, + "step": 41356 + }, + { + "epoch": 0.41357, + "grad_norm": 0.9428771338272489, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 41357 + }, + { + "epoch": 0.41358, + "grad_norm": 0.9702396299618441, + "learning_rate": 0.003, + "loss": 3.985, + "step": 41358 + }, + { + "epoch": 0.41359, + "grad_norm": 0.7887734744096074, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 41359 + }, + { + "epoch": 0.4136, + "grad_norm": 0.7641453665827589, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 41360 + }, + { + "epoch": 0.41361, + "grad_norm": 0.7601613110402655, + "learning_rate": 0.003, + "loss": 3.966, + "step": 41361 + }, + { + "epoch": 0.41362, + "grad_norm": 0.7339469481571292, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 41362 + }, + { + "epoch": 0.41363, + "grad_norm": 0.7818145006913655, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 41363 + }, + { + "epoch": 0.41364, + "grad_norm": 0.7432310163146622, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 41364 + }, + { + "epoch": 0.41365, + "grad_norm": 0.6602505494553595, + "learning_rate": 0.003, + "loss": 3.991, + "step": 41365 + }, + { + "epoch": 0.41366, + "grad_norm": 0.6059856579994298, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 41366 + }, + { + "epoch": 0.41367, + "grad_norm": 0.6702209296277968, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 41367 + }, + { + "epoch": 0.41368, + "grad_norm": 0.6051183063388261, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 41368 + }, + { + "epoch": 0.41369, + "grad_norm": 0.6712862489531444, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 41369 + }, + { + "epoch": 0.4137, + "grad_norm": 0.7527047744427798, + "learning_rate": 0.003, + "loss": 3.975, + "step": 41370 + }, + { + "epoch": 0.41371, + "grad_norm": 0.9370756543872506, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 41371 + }, + { + "epoch": 0.41372, + "grad_norm": 0.7962676114174172, + "learning_rate": 0.003, + "loss": 3.994, + "step": 41372 + }, + { + "epoch": 0.41373, + "grad_norm": 0.8259693308362615, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 41373 + }, + { + "epoch": 0.41374, + "grad_norm": 0.9888631384333777, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 41374 + }, + { + "epoch": 0.41375, + "grad_norm": 0.9235961364751737, + "learning_rate": 0.003, + "loss": 3.983, + "step": 41375 + }, + { + "epoch": 0.41376, + "grad_norm": 0.8283951868881817, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 41376 + }, + { + "epoch": 0.41377, + "grad_norm": 0.9685799998021883, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 41377 + }, + { + "epoch": 0.41378, + "grad_norm": 1.0480920082937955, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 41378 + }, + { + "epoch": 0.41379, + "grad_norm": 1.039366342375711, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 41379 + }, + { + "epoch": 0.4138, + "grad_norm": 0.9423680671037862, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 41380 + }, + { + "epoch": 0.41381, + "grad_norm": 0.8528207467870503, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 41381 + }, + { + "epoch": 0.41382, + "grad_norm": 0.7249597460648947, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 41382 + }, + { + "epoch": 0.41383, + "grad_norm": 0.6701773735178396, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 41383 + }, + { + "epoch": 0.41384, + "grad_norm": 0.6828741924232501, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 41384 + }, + { + "epoch": 0.41385, + "grad_norm": 0.6930056786858295, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 41385 + }, + { + "epoch": 0.41386, + "grad_norm": 0.698297051842655, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 41386 + }, + { + "epoch": 0.41387, + "grad_norm": 0.6979806180234602, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 41387 + }, + { + "epoch": 0.41388, + "grad_norm": 0.6026810208736688, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 41388 + }, + { + "epoch": 0.41389, + "grad_norm": 0.5402322925160136, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 41389 + }, + { + "epoch": 0.4139, + "grad_norm": 0.6406422479071553, + "learning_rate": 0.003, + "loss": 3.9433, + "step": 41390 + }, + { + "epoch": 0.41391, + "grad_norm": 0.6766518428741239, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 41391 + }, + { + "epoch": 0.41392, + "grad_norm": 0.7707174093704672, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 41392 + }, + { + "epoch": 0.41393, + "grad_norm": 0.8459830485623039, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 41393 + }, + { + "epoch": 0.41394, + "grad_norm": 1.0008042305524476, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 41394 + }, + { + "epoch": 0.41395, + "grad_norm": 1.0411779853849483, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 41395 + }, + { + "epoch": 0.41396, + "grad_norm": 0.7996264862840013, + "learning_rate": 0.003, + "loss": 3.994, + "step": 41396 + }, + { + "epoch": 0.41397, + "grad_norm": 0.6251207488578239, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 41397 + }, + { + "epoch": 0.41398, + "grad_norm": 0.635146969865529, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 41398 + }, + { + "epoch": 0.41399, + "grad_norm": 0.6531571419522157, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 41399 + }, + { + "epoch": 0.414, + "grad_norm": 0.8332141850325123, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 41400 + }, + { + "epoch": 0.41401, + "grad_norm": 0.8529132910554331, + "learning_rate": 0.003, + "loss": 3.997, + "step": 41401 + }, + { + "epoch": 0.41402, + "grad_norm": 0.7431127207448892, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 41402 + }, + { + "epoch": 0.41403, + "grad_norm": 0.6282050841477738, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 41403 + }, + { + "epoch": 0.41404, + "grad_norm": 0.6277996670041275, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 41404 + }, + { + "epoch": 0.41405, + "grad_norm": 0.6208868670713091, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 41405 + }, + { + "epoch": 0.41406, + "grad_norm": 0.6613149347357568, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 41406 + }, + { + "epoch": 0.41407, + "grad_norm": 0.7527192266722585, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 41407 + }, + { + "epoch": 0.41408, + "grad_norm": 0.8802692811909456, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 41408 + }, + { + "epoch": 0.41409, + "grad_norm": 0.835086915679407, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 41409 + }, + { + "epoch": 0.4141, + "grad_norm": 0.7737104657033356, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 41410 + }, + { + "epoch": 0.41411, + "grad_norm": 0.7236802181530065, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 41411 + }, + { + "epoch": 0.41412, + "grad_norm": 0.6894600100815943, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 41412 + }, + { + "epoch": 0.41413, + "grad_norm": 0.7747393581049413, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 41413 + }, + { + "epoch": 0.41414, + "grad_norm": 0.7900937035046804, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 41414 + }, + { + "epoch": 0.41415, + "grad_norm": 0.7838106741113594, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 41415 + }, + { + "epoch": 0.41416, + "grad_norm": 0.7652594847417266, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 41416 + }, + { + "epoch": 0.41417, + "grad_norm": 0.8494362091038242, + "learning_rate": 0.003, + "loss": 3.946, + "step": 41417 + }, + { + "epoch": 0.41418, + "grad_norm": 0.8628600572438169, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 41418 + }, + { + "epoch": 0.41419, + "grad_norm": 1.009149404268423, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 41419 + }, + { + "epoch": 0.4142, + "grad_norm": 0.9163078020708227, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 41420 + }, + { + "epoch": 0.41421, + "grad_norm": 0.6784238775145051, + "learning_rate": 0.003, + "loss": 3.999, + "step": 41421 + }, + { + "epoch": 0.41422, + "grad_norm": 0.7305905544376441, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 41422 + }, + { + "epoch": 0.41423, + "grad_norm": 0.7674663927352212, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 41423 + }, + { + "epoch": 0.41424, + "grad_norm": 0.8669708895791263, + "learning_rate": 0.003, + "loss": 4.006, + "step": 41424 + }, + { + "epoch": 0.41425, + "grad_norm": 0.8645343597415006, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 41425 + }, + { + "epoch": 0.41426, + "grad_norm": 0.8762496975732393, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 41426 + }, + { + "epoch": 0.41427, + "grad_norm": 1.0219917653388497, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 41427 + }, + { + "epoch": 0.41428, + "grad_norm": 1.0189504693067042, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 41428 + }, + { + "epoch": 0.41429, + "grad_norm": 0.9943279604363914, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 41429 + }, + { + "epoch": 0.4143, + "grad_norm": 1.0357781032407893, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 41430 + }, + { + "epoch": 0.41431, + "grad_norm": 0.8535858598284726, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 41431 + }, + { + "epoch": 0.41432, + "grad_norm": 0.9028099964928811, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 41432 + }, + { + "epoch": 0.41433, + "grad_norm": 1.0420747923035514, + "learning_rate": 0.003, + "loss": 4.0374, + "step": 41433 + }, + { + "epoch": 0.41434, + "grad_norm": 0.9692212615061329, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 41434 + }, + { + "epoch": 0.41435, + "grad_norm": 0.9014520450514555, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 41435 + }, + { + "epoch": 0.41436, + "grad_norm": 0.9145149416628184, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 41436 + }, + { + "epoch": 0.41437, + "grad_norm": 0.7281796746607235, + "learning_rate": 0.003, + "loss": 3.9419, + "step": 41437 + }, + { + "epoch": 0.41438, + "grad_norm": 0.6444815051679468, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 41438 + }, + { + "epoch": 0.41439, + "grad_norm": 0.5898816190728885, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 41439 + }, + { + "epoch": 0.4144, + "grad_norm": 0.6098466487360229, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 41440 + }, + { + "epoch": 0.41441, + "grad_norm": 0.6845024984361061, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 41441 + }, + { + "epoch": 0.41442, + "grad_norm": 0.6533333175070739, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 41442 + }, + { + "epoch": 0.41443, + "grad_norm": 0.5952864875368791, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 41443 + }, + { + "epoch": 0.41444, + "grad_norm": 0.5442065459794407, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 41444 + }, + { + "epoch": 0.41445, + "grad_norm": 0.5427445231760032, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 41445 + }, + { + "epoch": 0.41446, + "grad_norm": 0.5069365485831882, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 41446 + }, + { + "epoch": 0.41447, + "grad_norm": 0.5242579855933178, + "learning_rate": 0.003, + "loss": 3.9438, + "step": 41447 + }, + { + "epoch": 0.41448, + "grad_norm": 0.5653718180414783, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 41448 + }, + { + "epoch": 0.41449, + "grad_norm": 0.6257899409120254, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 41449 + }, + { + "epoch": 0.4145, + "grad_norm": 0.6298377347042226, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 41450 + }, + { + "epoch": 0.41451, + "grad_norm": 0.6207753418812544, + "learning_rate": 0.003, + "loss": 3.96, + "step": 41451 + }, + { + "epoch": 0.41452, + "grad_norm": 0.7062221205845727, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 41452 + }, + { + "epoch": 0.41453, + "grad_norm": 0.7366899401878029, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 41453 + }, + { + "epoch": 0.41454, + "grad_norm": 0.7148408516890113, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 41454 + }, + { + "epoch": 0.41455, + "grad_norm": 0.7051075337814593, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 41455 + }, + { + "epoch": 0.41456, + "grad_norm": 0.7907370936482919, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 41456 + }, + { + "epoch": 0.41457, + "grad_norm": 0.8617281818402031, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 41457 + }, + { + "epoch": 0.41458, + "grad_norm": 0.9185431207407528, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 41458 + }, + { + "epoch": 0.41459, + "grad_norm": 0.9812293632814802, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 41459 + }, + { + "epoch": 0.4146, + "grad_norm": 1.0070644300850513, + "learning_rate": 0.003, + "loss": 3.975, + "step": 41460 + }, + { + "epoch": 0.41461, + "grad_norm": 1.0995036351960825, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 41461 + }, + { + "epoch": 0.41462, + "grad_norm": 1.0308752516430693, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 41462 + }, + { + "epoch": 0.41463, + "grad_norm": 0.9126908657803433, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 41463 + }, + { + "epoch": 0.41464, + "grad_norm": 0.921176939801133, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 41464 + }, + { + "epoch": 0.41465, + "grad_norm": 0.8518971954341702, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 41465 + }, + { + "epoch": 0.41466, + "grad_norm": 0.8744402656780984, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 41466 + }, + { + "epoch": 0.41467, + "grad_norm": 0.9310055396782213, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 41467 + }, + { + "epoch": 0.41468, + "grad_norm": 0.8836234749968634, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 41468 + }, + { + "epoch": 0.41469, + "grad_norm": 0.7413586442941053, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 41469 + }, + { + "epoch": 0.4147, + "grad_norm": 0.8189208648297213, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 41470 + }, + { + "epoch": 0.41471, + "grad_norm": 0.8047342756627777, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 41471 + }, + { + "epoch": 0.41472, + "grad_norm": 0.6890725417658514, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 41472 + }, + { + "epoch": 0.41473, + "grad_norm": 0.630668981432774, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 41473 + }, + { + "epoch": 0.41474, + "grad_norm": 0.6205073089724341, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 41474 + }, + { + "epoch": 0.41475, + "grad_norm": 0.5632222741970022, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 41475 + }, + { + "epoch": 0.41476, + "grad_norm": 0.5628405191133306, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 41476 + }, + { + "epoch": 0.41477, + "grad_norm": 0.6140596716522552, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 41477 + }, + { + "epoch": 0.41478, + "grad_norm": 0.5937136028702785, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 41478 + }, + { + "epoch": 0.41479, + "grad_norm": 0.6505651709810151, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 41479 + }, + { + "epoch": 0.4148, + "grad_norm": 0.6153074969746479, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 41480 + }, + { + "epoch": 0.41481, + "grad_norm": 0.5957269571275061, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 41481 + }, + { + "epoch": 0.41482, + "grad_norm": 0.8332477528913141, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 41482 + }, + { + "epoch": 0.41483, + "grad_norm": 1.038346773898753, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 41483 + }, + { + "epoch": 0.41484, + "grad_norm": 1.1224701057296629, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 41484 + }, + { + "epoch": 0.41485, + "grad_norm": 0.7128753931015077, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 41485 + }, + { + "epoch": 0.41486, + "grad_norm": 0.6195433811933428, + "learning_rate": 0.003, + "loss": 3.985, + "step": 41486 + }, + { + "epoch": 0.41487, + "grad_norm": 0.6161457669708494, + "learning_rate": 0.003, + "loss": 3.961, + "step": 41487 + }, + { + "epoch": 0.41488, + "grad_norm": 0.6538360963504741, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 41488 + }, + { + "epoch": 0.41489, + "grad_norm": 0.7608220008681958, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41489 + }, + { + "epoch": 0.4149, + "grad_norm": 0.7708190137931578, + "learning_rate": 0.003, + "loss": 3.937, + "step": 41490 + }, + { + "epoch": 0.41491, + "grad_norm": 0.773956192072262, + "learning_rate": 0.003, + "loss": 4.0288, + "step": 41491 + }, + { + "epoch": 0.41492, + "grad_norm": 0.7704289333452786, + "learning_rate": 0.003, + "loss": 3.959, + "step": 41492 + }, + { + "epoch": 0.41493, + "grad_norm": 0.7084938732880204, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 41493 + }, + { + "epoch": 0.41494, + "grad_norm": 0.7124946264677505, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 41494 + }, + { + "epoch": 0.41495, + "grad_norm": 0.7688091420050825, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 41495 + }, + { + "epoch": 0.41496, + "grad_norm": 0.7973095513557651, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 41496 + }, + { + "epoch": 0.41497, + "grad_norm": 0.8575583879630421, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 41497 + }, + { + "epoch": 0.41498, + "grad_norm": 0.8328943102397701, + "learning_rate": 0.003, + "loss": 4.0127, + "step": 41498 + }, + { + "epoch": 0.41499, + "grad_norm": 0.9528864408054986, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 41499 + }, + { + "epoch": 0.415, + "grad_norm": 0.847584035411189, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 41500 + }, + { + "epoch": 0.41501, + "grad_norm": 0.8348163371296413, + "learning_rate": 0.003, + "loss": 3.973, + "step": 41501 + }, + { + "epoch": 0.41502, + "grad_norm": 0.894738565289983, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 41502 + }, + { + "epoch": 0.41503, + "grad_norm": 0.9736311802137496, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 41503 + }, + { + "epoch": 0.41504, + "grad_norm": 0.8791701135412201, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 41504 + }, + { + "epoch": 0.41505, + "grad_norm": 0.7614019710128721, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 41505 + }, + { + "epoch": 0.41506, + "grad_norm": 0.7239167097017879, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 41506 + }, + { + "epoch": 0.41507, + "grad_norm": 0.6619480394423957, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 41507 + }, + { + "epoch": 0.41508, + "grad_norm": 0.7256093059045563, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 41508 + }, + { + "epoch": 0.41509, + "grad_norm": 0.9146836135996934, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 41509 + }, + { + "epoch": 0.4151, + "grad_norm": 1.097593436978205, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 41510 + }, + { + "epoch": 0.41511, + "grad_norm": 0.8453362484988801, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 41511 + }, + { + "epoch": 0.41512, + "grad_norm": 0.6998248271760423, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 41512 + }, + { + "epoch": 0.41513, + "grad_norm": 0.692172224961604, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 41513 + }, + { + "epoch": 0.41514, + "grad_norm": 0.7688483090099816, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 41514 + }, + { + "epoch": 0.41515, + "grad_norm": 0.8300871192645672, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 41515 + }, + { + "epoch": 0.41516, + "grad_norm": 0.8686911484468376, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 41516 + }, + { + "epoch": 0.41517, + "grad_norm": 1.0240616651115444, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 41517 + }, + { + "epoch": 0.41518, + "grad_norm": 0.9293103176351856, + "learning_rate": 0.003, + "loss": 3.974, + "step": 41518 + }, + { + "epoch": 0.41519, + "grad_norm": 0.9370401670455986, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 41519 + }, + { + "epoch": 0.4152, + "grad_norm": 0.9227058709732145, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 41520 + }, + { + "epoch": 0.41521, + "grad_norm": 0.9730984051722689, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 41521 + }, + { + "epoch": 0.41522, + "grad_norm": 0.7797015024642353, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 41522 + }, + { + "epoch": 0.41523, + "grad_norm": 0.5978347984110711, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 41523 + }, + { + "epoch": 0.41524, + "grad_norm": 0.6469727876096517, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 41524 + }, + { + "epoch": 0.41525, + "grad_norm": 0.6929353170701942, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 41525 + }, + { + "epoch": 0.41526, + "grad_norm": 0.8001063536723811, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 41526 + }, + { + "epoch": 0.41527, + "grad_norm": 0.8513900396323049, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 41527 + }, + { + "epoch": 0.41528, + "grad_norm": 0.8222127671040058, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 41528 + }, + { + "epoch": 0.41529, + "grad_norm": 0.8525717466908067, + "learning_rate": 0.003, + "loss": 3.9454, + "step": 41529 + }, + { + "epoch": 0.4153, + "grad_norm": 0.7742268048317427, + "learning_rate": 0.003, + "loss": 3.99, + "step": 41530 + }, + { + "epoch": 0.41531, + "grad_norm": 0.7224619612029969, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 41531 + }, + { + "epoch": 0.41532, + "grad_norm": 0.6619264025284667, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 41532 + }, + { + "epoch": 0.41533, + "grad_norm": 0.6635409195209339, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 41533 + }, + { + "epoch": 0.41534, + "grad_norm": 0.6067390539681936, + "learning_rate": 0.003, + "loss": 4.005, + "step": 41534 + }, + { + "epoch": 0.41535, + "grad_norm": 0.6475333328474883, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 41535 + }, + { + "epoch": 0.41536, + "grad_norm": 0.6675159793586134, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 41536 + }, + { + "epoch": 0.41537, + "grad_norm": 0.5995158726414227, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 41537 + }, + { + "epoch": 0.41538, + "grad_norm": 0.5509502234939253, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 41538 + }, + { + "epoch": 0.41539, + "grad_norm": 0.6703181928389564, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 41539 + }, + { + "epoch": 0.4154, + "grad_norm": 0.6670603480903641, + "learning_rate": 0.003, + "loss": 3.9435, + "step": 41540 + }, + { + "epoch": 0.41541, + "grad_norm": 0.7704844251683036, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 41541 + }, + { + "epoch": 0.41542, + "grad_norm": 0.8346303622439016, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 41542 + }, + { + "epoch": 0.41543, + "grad_norm": 0.9678957279287612, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 41543 + }, + { + "epoch": 0.41544, + "grad_norm": 1.0518305955571177, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 41544 + }, + { + "epoch": 0.41545, + "grad_norm": 1.0034622835832718, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 41545 + }, + { + "epoch": 0.41546, + "grad_norm": 0.9329316404753033, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 41546 + }, + { + "epoch": 0.41547, + "grad_norm": 0.9642655879265105, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 41547 + }, + { + "epoch": 0.41548, + "grad_norm": 1.052586454927663, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 41548 + }, + { + "epoch": 0.41549, + "grad_norm": 0.8340980235611102, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 41549 + }, + { + "epoch": 0.4155, + "grad_norm": 0.78051617156275, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 41550 + }, + { + "epoch": 0.41551, + "grad_norm": 0.8966750246055785, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 41551 + }, + { + "epoch": 0.41552, + "grad_norm": 0.9837119529165879, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 41552 + }, + { + "epoch": 0.41553, + "grad_norm": 0.9164171024838818, + "learning_rate": 0.003, + "loss": 4.004, + "step": 41553 + }, + { + "epoch": 0.41554, + "grad_norm": 0.8600390020516664, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 41554 + }, + { + "epoch": 0.41555, + "grad_norm": 0.6977781342083705, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 41555 + }, + { + "epoch": 0.41556, + "grad_norm": 0.74691930904346, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 41556 + }, + { + "epoch": 0.41557, + "grad_norm": 0.8141451620944676, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 41557 + }, + { + "epoch": 0.41558, + "grad_norm": 0.8739963925239862, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 41558 + }, + { + "epoch": 0.41559, + "grad_norm": 0.8576984397257967, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 41559 + }, + { + "epoch": 0.4156, + "grad_norm": 0.898293826329344, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 41560 + }, + { + "epoch": 0.41561, + "grad_norm": 0.9643396441590562, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 41561 + }, + { + "epoch": 0.41562, + "grad_norm": 1.020470657533241, + "learning_rate": 0.003, + "loss": 4.015, + "step": 41562 + }, + { + "epoch": 0.41563, + "grad_norm": 0.815125299353147, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 41563 + }, + { + "epoch": 0.41564, + "grad_norm": 0.7272368040378323, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 41564 + }, + { + "epoch": 0.41565, + "grad_norm": 0.7281888458253328, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 41565 + }, + { + "epoch": 0.41566, + "grad_norm": 0.7516099923743399, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 41566 + }, + { + "epoch": 0.41567, + "grad_norm": 0.7832383939689713, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 41567 + }, + { + "epoch": 0.41568, + "grad_norm": 0.752191006096955, + "learning_rate": 0.003, + "loss": 3.98, + "step": 41568 + }, + { + "epoch": 0.41569, + "grad_norm": 0.6895168791471586, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 41569 + }, + { + "epoch": 0.4157, + "grad_norm": 0.6071539708976469, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 41570 + }, + { + "epoch": 0.41571, + "grad_norm": 0.6142736699617123, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 41571 + }, + { + "epoch": 0.41572, + "grad_norm": 0.6384174643572081, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 41572 + }, + { + "epoch": 0.41573, + "grad_norm": 0.6900745664353044, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 41573 + }, + { + "epoch": 0.41574, + "grad_norm": 0.738657726867638, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 41574 + }, + { + "epoch": 0.41575, + "grad_norm": 0.7611893325706306, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 41575 + }, + { + "epoch": 0.41576, + "grad_norm": 0.8297384352967567, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 41576 + }, + { + "epoch": 0.41577, + "grad_norm": 0.7673265551297569, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 41577 + }, + { + "epoch": 0.41578, + "grad_norm": 0.743966701949248, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 41578 + }, + { + "epoch": 0.41579, + "grad_norm": 0.7328009851902475, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 41579 + }, + { + "epoch": 0.4158, + "grad_norm": 0.6958987133287313, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 41580 + }, + { + "epoch": 0.41581, + "grad_norm": 0.6565892856200988, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 41581 + }, + { + "epoch": 0.41582, + "grad_norm": 0.6431833215302696, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 41582 + }, + { + "epoch": 0.41583, + "grad_norm": 0.6074600361727229, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 41583 + }, + { + "epoch": 0.41584, + "grad_norm": 0.7571850268466251, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 41584 + }, + { + "epoch": 0.41585, + "grad_norm": 0.9530322429633886, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 41585 + }, + { + "epoch": 0.41586, + "grad_norm": 1.098190762667968, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 41586 + }, + { + "epoch": 0.41587, + "grad_norm": 0.9418523804785197, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 41587 + }, + { + "epoch": 0.41588, + "grad_norm": 0.8888138449737846, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 41588 + }, + { + "epoch": 0.41589, + "grad_norm": 0.8586934025775019, + "learning_rate": 0.003, + "loss": 3.981, + "step": 41589 + }, + { + "epoch": 0.4159, + "grad_norm": 0.7794610501932925, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 41590 + }, + { + "epoch": 0.41591, + "grad_norm": 0.7470324016426247, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 41591 + }, + { + "epoch": 0.41592, + "grad_norm": 0.6916348222992701, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 41592 + }, + { + "epoch": 0.41593, + "grad_norm": 0.686202986192533, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 41593 + }, + { + "epoch": 0.41594, + "grad_norm": 0.63229828838622, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 41594 + }, + { + "epoch": 0.41595, + "grad_norm": 0.612371796638194, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 41595 + }, + { + "epoch": 0.41596, + "grad_norm": 0.5623557819753661, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 41596 + }, + { + "epoch": 0.41597, + "grad_norm": 0.6244811223144469, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 41597 + }, + { + "epoch": 0.41598, + "grad_norm": 0.6907020669063391, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 41598 + }, + { + "epoch": 0.41599, + "grad_norm": 0.8185533877793433, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 41599 + }, + { + "epoch": 0.416, + "grad_norm": 0.8299117243713351, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 41600 + }, + { + "epoch": 0.41601, + "grad_norm": 0.8571035482029394, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 41601 + }, + { + "epoch": 0.41602, + "grad_norm": 0.8812267495426465, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 41602 + }, + { + "epoch": 0.41603, + "grad_norm": 0.8587273882565705, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 41603 + }, + { + "epoch": 0.41604, + "grad_norm": 0.983998094940643, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 41604 + }, + { + "epoch": 0.41605, + "grad_norm": 1.1723962507991326, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 41605 + }, + { + "epoch": 0.41606, + "grad_norm": 0.8427621807433363, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 41606 + }, + { + "epoch": 0.41607, + "grad_norm": 0.7529115007180351, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 41607 + }, + { + "epoch": 0.41608, + "grad_norm": 0.7508948017798202, + "learning_rate": 0.003, + "loss": 4.021, + "step": 41608 + }, + { + "epoch": 0.41609, + "grad_norm": 0.8903999260319937, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 41609 + }, + { + "epoch": 0.4161, + "grad_norm": 0.8777992291192416, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 41610 + }, + { + "epoch": 0.41611, + "grad_norm": 0.9359772420232229, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 41611 + }, + { + "epoch": 0.41612, + "grad_norm": 0.967559824995334, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 41612 + }, + { + "epoch": 0.41613, + "grad_norm": 1.0801923397034485, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 41613 + }, + { + "epoch": 0.41614, + "grad_norm": 1.0323891614198606, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 41614 + }, + { + "epoch": 0.41615, + "grad_norm": 0.8125853664939692, + "learning_rate": 0.003, + "loss": 3.963, + "step": 41615 + }, + { + "epoch": 0.41616, + "grad_norm": 0.8363523343078517, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 41616 + }, + { + "epoch": 0.41617, + "grad_norm": 0.7575938313019559, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 41617 + }, + { + "epoch": 0.41618, + "grad_norm": 0.6980148936971796, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 41618 + }, + { + "epoch": 0.41619, + "grad_norm": 0.6362975622640458, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 41619 + }, + { + "epoch": 0.4162, + "grad_norm": 0.6129118072179505, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 41620 + }, + { + "epoch": 0.41621, + "grad_norm": 0.5742993728305493, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 41621 + }, + { + "epoch": 0.41622, + "grad_norm": 0.5875200145791653, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 41622 + }, + { + "epoch": 0.41623, + "grad_norm": 0.6081253538463016, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 41623 + }, + { + "epoch": 0.41624, + "grad_norm": 0.6994788979773109, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 41624 + }, + { + "epoch": 0.41625, + "grad_norm": 0.8072282908146244, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 41625 + }, + { + "epoch": 0.41626, + "grad_norm": 0.9290478149358846, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 41626 + }, + { + "epoch": 0.41627, + "grad_norm": 0.9067053720737068, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 41627 + }, + { + "epoch": 0.41628, + "grad_norm": 0.806954897121237, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 41628 + }, + { + "epoch": 0.41629, + "grad_norm": 0.7402216232478944, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 41629 + }, + { + "epoch": 0.4163, + "grad_norm": 0.7682996406992537, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 41630 + }, + { + "epoch": 0.41631, + "grad_norm": 0.7695277904924464, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 41631 + }, + { + "epoch": 0.41632, + "grad_norm": 0.6879262035110829, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 41632 + }, + { + "epoch": 0.41633, + "grad_norm": 0.7612414090747471, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 41633 + }, + { + "epoch": 0.41634, + "grad_norm": 0.8075182788196499, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 41634 + }, + { + "epoch": 0.41635, + "grad_norm": 0.9370388567077577, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 41635 + }, + { + "epoch": 0.41636, + "grad_norm": 0.9557343064224515, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 41636 + }, + { + "epoch": 0.41637, + "grad_norm": 0.8385415621488663, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 41637 + }, + { + "epoch": 0.41638, + "grad_norm": 0.9218260059236784, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 41638 + }, + { + "epoch": 0.41639, + "grad_norm": 0.85711098897038, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 41639 + }, + { + "epoch": 0.4164, + "grad_norm": 0.7524834259936318, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 41640 + }, + { + "epoch": 0.41641, + "grad_norm": 0.7421640852839597, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 41641 + }, + { + "epoch": 0.41642, + "grad_norm": 0.8706262323444595, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 41642 + }, + { + "epoch": 0.41643, + "grad_norm": 0.9967728055501666, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 41643 + }, + { + "epoch": 0.41644, + "grad_norm": 0.914602775163566, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 41644 + }, + { + "epoch": 0.41645, + "grad_norm": 0.7689699346911467, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 41645 + }, + { + "epoch": 0.41646, + "grad_norm": 0.885600544377438, + "learning_rate": 0.003, + "loss": 3.96, + "step": 41646 + }, + { + "epoch": 0.41647, + "grad_norm": 0.8111967779186914, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 41647 + }, + { + "epoch": 0.41648, + "grad_norm": 0.8590199404299671, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 41648 + }, + { + "epoch": 0.41649, + "grad_norm": 0.802844188812144, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 41649 + }, + { + "epoch": 0.4165, + "grad_norm": 0.7615740138746722, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 41650 + }, + { + "epoch": 0.41651, + "grad_norm": 0.8202255381578067, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 41651 + }, + { + "epoch": 0.41652, + "grad_norm": 0.8071833157628863, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 41652 + }, + { + "epoch": 0.41653, + "grad_norm": 0.7532366482550277, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 41653 + }, + { + "epoch": 0.41654, + "grad_norm": 0.7776785150882324, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 41654 + }, + { + "epoch": 0.41655, + "grad_norm": 0.7222285200175566, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 41655 + }, + { + "epoch": 0.41656, + "grad_norm": 0.6410513234436293, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 41656 + }, + { + "epoch": 0.41657, + "grad_norm": 0.6585454633961021, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 41657 + }, + { + "epoch": 0.41658, + "grad_norm": 0.6148775371526788, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 41658 + }, + { + "epoch": 0.41659, + "grad_norm": 0.7342271893325523, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 41659 + }, + { + "epoch": 0.4166, + "grad_norm": 0.8330614353336059, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 41660 + }, + { + "epoch": 0.41661, + "grad_norm": 0.8174211463847996, + "learning_rate": 0.003, + "loss": 3.9276, + "step": 41661 + }, + { + "epoch": 0.41662, + "grad_norm": 0.8052412150411145, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 41662 + }, + { + "epoch": 0.41663, + "grad_norm": 0.8993835631979938, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 41663 + }, + { + "epoch": 0.41664, + "grad_norm": 0.9800684269759352, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 41664 + }, + { + "epoch": 0.41665, + "grad_norm": 0.9643913050643341, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 41665 + }, + { + "epoch": 0.41666, + "grad_norm": 0.832500517324253, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41666 + }, + { + "epoch": 0.41667, + "grad_norm": 0.8244659074573351, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 41667 + }, + { + "epoch": 0.41668, + "grad_norm": 0.7218763287760213, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 41668 + }, + { + "epoch": 0.41669, + "grad_norm": 0.7207973351626538, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 41669 + }, + { + "epoch": 0.4167, + "grad_norm": 0.8083756465388653, + "learning_rate": 0.003, + "loss": 3.999, + "step": 41670 + }, + { + "epoch": 0.41671, + "grad_norm": 0.809180660714045, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 41671 + }, + { + "epoch": 0.41672, + "grad_norm": 0.8092147353819071, + "learning_rate": 0.003, + "loss": 3.965, + "step": 41672 + }, + { + "epoch": 0.41673, + "grad_norm": 0.8158457813572529, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 41673 + }, + { + "epoch": 0.41674, + "grad_norm": 0.9834286077249168, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 41674 + }, + { + "epoch": 0.41675, + "grad_norm": 1.0629354638467223, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 41675 + }, + { + "epoch": 0.41676, + "grad_norm": 0.8533248655401253, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 41676 + }, + { + "epoch": 0.41677, + "grad_norm": 0.7812939736474656, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 41677 + }, + { + "epoch": 0.41678, + "grad_norm": 0.7589626747067894, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41678 + }, + { + "epoch": 0.41679, + "grad_norm": 0.6794048903698242, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 41679 + }, + { + "epoch": 0.4168, + "grad_norm": 0.59382431138946, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 41680 + }, + { + "epoch": 0.41681, + "grad_norm": 0.6831338717945881, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 41681 + }, + { + "epoch": 0.41682, + "grad_norm": 0.8841282355890101, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 41682 + }, + { + "epoch": 0.41683, + "grad_norm": 0.9928122666212211, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 41683 + }, + { + "epoch": 0.41684, + "grad_norm": 1.011468982165733, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 41684 + }, + { + "epoch": 0.41685, + "grad_norm": 0.7859716595229144, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 41685 + }, + { + "epoch": 0.41686, + "grad_norm": 0.6798612050415763, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 41686 + }, + { + "epoch": 0.41687, + "grad_norm": 0.7793272509537432, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 41687 + }, + { + "epoch": 0.41688, + "grad_norm": 0.8197382419928432, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 41688 + }, + { + "epoch": 0.41689, + "grad_norm": 0.8206349082546543, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 41689 + }, + { + "epoch": 0.4169, + "grad_norm": 0.7228278261488499, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 41690 + }, + { + "epoch": 0.41691, + "grad_norm": 0.6045782412186131, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 41691 + }, + { + "epoch": 0.41692, + "grad_norm": 0.5480422929485737, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 41692 + }, + { + "epoch": 0.41693, + "grad_norm": 0.4694565283670828, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 41693 + }, + { + "epoch": 0.41694, + "grad_norm": 0.49459711978423776, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 41694 + }, + { + "epoch": 0.41695, + "grad_norm": 0.5128516449297994, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 41695 + }, + { + "epoch": 0.41696, + "grad_norm": 0.54331182668578, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 41696 + }, + { + "epoch": 0.41697, + "grad_norm": 0.6261809493568008, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 41697 + }, + { + "epoch": 0.41698, + "grad_norm": 0.6630868007625567, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 41698 + }, + { + "epoch": 0.41699, + "grad_norm": 0.4877978028562624, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 41699 + }, + { + "epoch": 0.417, + "grad_norm": 0.4524053878061106, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 41700 + }, + { + "epoch": 0.41701, + "grad_norm": 0.4774986643317372, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 41701 + }, + { + "epoch": 0.41702, + "grad_norm": 0.4940706459776065, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 41702 + }, + { + "epoch": 0.41703, + "grad_norm": 0.5727950600960764, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 41703 + }, + { + "epoch": 0.41704, + "grad_norm": 0.6382147625492275, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 41704 + }, + { + "epoch": 0.41705, + "grad_norm": 0.7117640609353286, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 41705 + }, + { + "epoch": 0.41706, + "grad_norm": 0.7638479604522452, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 41706 + }, + { + "epoch": 0.41707, + "grad_norm": 0.8699314487846362, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 41707 + }, + { + "epoch": 0.41708, + "grad_norm": 1.0608557981657893, + "learning_rate": 0.003, + "loss": 3.982, + "step": 41708 + }, + { + "epoch": 0.41709, + "grad_norm": 0.8777015491588231, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 41709 + }, + { + "epoch": 0.4171, + "grad_norm": 0.8947629991047424, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 41710 + }, + { + "epoch": 0.41711, + "grad_norm": 0.8399728892485426, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 41711 + }, + { + "epoch": 0.41712, + "grad_norm": 0.8229149860066607, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 41712 + }, + { + "epoch": 0.41713, + "grad_norm": 0.8607784906567627, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 41713 + }, + { + "epoch": 0.41714, + "grad_norm": 0.9134688456815644, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 41714 + }, + { + "epoch": 0.41715, + "grad_norm": 0.8573476243070286, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 41715 + }, + { + "epoch": 0.41716, + "grad_norm": 0.9113422780171772, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 41716 + }, + { + "epoch": 0.41717, + "grad_norm": 0.9617675255963657, + "learning_rate": 0.003, + "loss": 3.987, + "step": 41717 + }, + { + "epoch": 0.41718, + "grad_norm": 1.008330363561094, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 41718 + }, + { + "epoch": 0.41719, + "grad_norm": 1.25154364624705, + "learning_rate": 0.003, + "loss": 4.0308, + "step": 41719 + }, + { + "epoch": 0.4172, + "grad_norm": 0.7340137263217871, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 41720 + }, + { + "epoch": 0.41721, + "grad_norm": 0.7461497841453317, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 41721 + }, + { + "epoch": 0.41722, + "grad_norm": 0.8727968970034475, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 41722 + }, + { + "epoch": 0.41723, + "grad_norm": 1.0382646528567092, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 41723 + }, + { + "epoch": 0.41724, + "grad_norm": 1.1447571936682959, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 41724 + }, + { + "epoch": 0.41725, + "grad_norm": 0.8118766886583182, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 41725 + }, + { + "epoch": 0.41726, + "grad_norm": 0.8046177201815415, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 41726 + }, + { + "epoch": 0.41727, + "grad_norm": 0.797078743606111, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 41727 + }, + { + "epoch": 0.41728, + "grad_norm": 0.7847152093970283, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 41728 + }, + { + "epoch": 0.41729, + "grad_norm": 0.7467978326210719, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 41729 + }, + { + "epoch": 0.4173, + "grad_norm": 0.7383150579029872, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 41730 + }, + { + "epoch": 0.41731, + "grad_norm": 0.769475161434144, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 41731 + }, + { + "epoch": 0.41732, + "grad_norm": 0.810792656055412, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 41732 + }, + { + "epoch": 0.41733, + "grad_norm": 0.7295937634544089, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 41733 + }, + { + "epoch": 0.41734, + "grad_norm": 0.6400745651983997, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 41734 + }, + { + "epoch": 0.41735, + "grad_norm": 0.6493505287229672, + "learning_rate": 0.003, + "loss": 3.966, + "step": 41735 + }, + { + "epoch": 0.41736, + "grad_norm": 0.7511771439917724, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 41736 + }, + { + "epoch": 0.41737, + "grad_norm": 0.7011348874792654, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 41737 + }, + { + "epoch": 0.41738, + "grad_norm": 0.6428340481475869, + "learning_rate": 0.003, + "loss": 3.988, + "step": 41738 + }, + { + "epoch": 0.41739, + "grad_norm": 0.6716622532422191, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 41739 + }, + { + "epoch": 0.4174, + "grad_norm": 0.535654275287453, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 41740 + }, + { + "epoch": 0.41741, + "grad_norm": 0.5589614162520434, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 41741 + }, + { + "epoch": 0.41742, + "grad_norm": 0.5379458450434903, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 41742 + }, + { + "epoch": 0.41743, + "grad_norm": 0.5571563997185419, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 41743 + }, + { + "epoch": 0.41744, + "grad_norm": 0.6199210163931459, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 41744 + }, + { + "epoch": 0.41745, + "grad_norm": 0.6396195317185615, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 41745 + }, + { + "epoch": 0.41746, + "grad_norm": 0.8859796899868846, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 41746 + }, + { + "epoch": 0.41747, + "grad_norm": 1.3314088257805392, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 41747 + }, + { + "epoch": 0.41748, + "grad_norm": 0.6132040653760777, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 41748 + }, + { + "epoch": 0.41749, + "grad_norm": 0.7628644446196244, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 41749 + }, + { + "epoch": 0.4175, + "grad_norm": 0.8285835420288479, + "learning_rate": 0.003, + "loss": 4.01, + "step": 41750 + }, + { + "epoch": 0.41751, + "grad_norm": 1.0045209798939774, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 41751 + }, + { + "epoch": 0.41752, + "grad_norm": 1.14741856779824, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 41752 + }, + { + "epoch": 0.41753, + "grad_norm": 0.7649554984621683, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 41753 + }, + { + "epoch": 0.41754, + "grad_norm": 0.8524271472273626, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 41754 + }, + { + "epoch": 0.41755, + "grad_norm": 0.9620932455840802, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 41755 + }, + { + "epoch": 0.41756, + "grad_norm": 0.9597181628243988, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 41756 + }, + { + "epoch": 0.41757, + "grad_norm": 0.8947258431095785, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 41757 + }, + { + "epoch": 0.41758, + "grad_norm": 0.913604897345639, + "learning_rate": 0.003, + "loss": 3.968, + "step": 41758 + }, + { + "epoch": 0.41759, + "grad_norm": 0.9462766699848615, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 41759 + }, + { + "epoch": 0.4176, + "grad_norm": 0.8737677661659028, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 41760 + }, + { + "epoch": 0.41761, + "grad_norm": 0.7913696524897516, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 41761 + }, + { + "epoch": 0.41762, + "grad_norm": 0.5805559968389442, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 41762 + }, + { + "epoch": 0.41763, + "grad_norm": 0.6096559151456681, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 41763 + }, + { + "epoch": 0.41764, + "grad_norm": 0.5102280668905532, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 41764 + }, + { + "epoch": 0.41765, + "grad_norm": 0.5118309306688603, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 41765 + }, + { + "epoch": 0.41766, + "grad_norm": 0.5951014698470166, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 41766 + }, + { + "epoch": 0.41767, + "grad_norm": 0.6314195309369206, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 41767 + }, + { + "epoch": 0.41768, + "grad_norm": 0.5651773455989818, + "learning_rate": 0.003, + "loss": 3.9383, + "step": 41768 + }, + { + "epoch": 0.41769, + "grad_norm": 0.6420618807081128, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 41769 + }, + { + "epoch": 0.4177, + "grad_norm": 0.6806146597424035, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 41770 + }, + { + "epoch": 0.41771, + "grad_norm": 0.7806205792235921, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 41771 + }, + { + "epoch": 0.41772, + "grad_norm": 0.8178306982314675, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 41772 + }, + { + "epoch": 0.41773, + "grad_norm": 0.973938043807744, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 41773 + }, + { + "epoch": 0.41774, + "grad_norm": 1.039101245783335, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 41774 + }, + { + "epoch": 0.41775, + "grad_norm": 0.8030488009124651, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 41775 + }, + { + "epoch": 0.41776, + "grad_norm": 0.803981286590682, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 41776 + }, + { + "epoch": 0.41777, + "grad_norm": 0.7522430645236227, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 41777 + }, + { + "epoch": 0.41778, + "grad_norm": 0.7487300857212688, + "learning_rate": 0.003, + "loss": 4.031, + "step": 41778 + }, + { + "epoch": 0.41779, + "grad_norm": 0.8552931809509264, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 41779 + }, + { + "epoch": 0.4178, + "grad_norm": 0.9286512246807567, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 41780 + }, + { + "epoch": 0.41781, + "grad_norm": 1.0817586380645001, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 41781 + }, + { + "epoch": 0.41782, + "grad_norm": 0.8600801737980415, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 41782 + }, + { + "epoch": 0.41783, + "grad_norm": 0.8100776827944842, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 41783 + }, + { + "epoch": 0.41784, + "grad_norm": 0.8774205269936327, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 41784 + }, + { + "epoch": 0.41785, + "grad_norm": 0.9218196655300509, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 41785 + }, + { + "epoch": 0.41786, + "grad_norm": 0.9240136691278787, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 41786 + }, + { + "epoch": 0.41787, + "grad_norm": 0.9546561138810936, + "learning_rate": 0.003, + "loss": 3.992, + "step": 41787 + }, + { + "epoch": 0.41788, + "grad_norm": 0.9412307387715476, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 41788 + }, + { + "epoch": 0.41789, + "grad_norm": 0.9784709581465451, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 41789 + }, + { + "epoch": 0.4179, + "grad_norm": 1.0501194059391215, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 41790 + }, + { + "epoch": 0.41791, + "grad_norm": 0.9068927142223107, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 41791 + }, + { + "epoch": 0.41792, + "grad_norm": 0.6956835013484287, + "learning_rate": 0.003, + "loss": 4.016, + "step": 41792 + }, + { + "epoch": 0.41793, + "grad_norm": 0.6462425528722899, + "learning_rate": 0.003, + "loss": 3.977, + "step": 41793 + }, + { + "epoch": 0.41794, + "grad_norm": 0.6458013363922718, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 41794 + }, + { + "epoch": 0.41795, + "grad_norm": 0.620354428723712, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 41795 + }, + { + "epoch": 0.41796, + "grad_norm": 0.5647993839204732, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 41796 + }, + { + "epoch": 0.41797, + "grad_norm": 0.5478710415988373, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 41797 + }, + { + "epoch": 0.41798, + "grad_norm": 0.5489141085735123, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 41798 + }, + { + "epoch": 0.41799, + "grad_norm": 0.6326954602368488, + "learning_rate": 0.003, + "loss": 4.013, + "step": 41799 + }, + { + "epoch": 0.418, + "grad_norm": 0.7312610023144992, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 41800 + }, + { + "epoch": 0.41801, + "grad_norm": 0.9095871668452623, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 41801 + }, + { + "epoch": 0.41802, + "grad_norm": 1.1212475322191415, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 41802 + }, + { + "epoch": 0.41803, + "grad_norm": 0.8038974002169309, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 41803 + }, + { + "epoch": 0.41804, + "grad_norm": 0.6925831970456799, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 41804 + }, + { + "epoch": 0.41805, + "grad_norm": 0.6252331664799343, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 41805 + }, + { + "epoch": 0.41806, + "grad_norm": 0.6468193587165094, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 41806 + }, + { + "epoch": 0.41807, + "grad_norm": 0.5884661924136876, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 41807 + }, + { + "epoch": 0.41808, + "grad_norm": 0.6125026411905213, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 41808 + }, + { + "epoch": 0.41809, + "grad_norm": 0.6141470674676044, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 41809 + }, + { + "epoch": 0.4181, + "grad_norm": 0.6730799439461189, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 41810 + }, + { + "epoch": 0.41811, + "grad_norm": 0.7041328962312691, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 41811 + }, + { + "epoch": 0.41812, + "grad_norm": 0.8872567672965165, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 41812 + }, + { + "epoch": 0.41813, + "grad_norm": 1.0328107060642233, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 41813 + }, + { + "epoch": 0.41814, + "grad_norm": 1.0836717353994023, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 41814 + }, + { + "epoch": 0.41815, + "grad_norm": 0.6790721033396776, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 41815 + }, + { + "epoch": 0.41816, + "grad_norm": 0.7866645239256363, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 41816 + }, + { + "epoch": 0.41817, + "grad_norm": 0.8923858122553929, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 41817 + }, + { + "epoch": 0.41818, + "grad_norm": 0.7725170625053607, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 41818 + }, + { + "epoch": 0.41819, + "grad_norm": 0.7107711513121369, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 41819 + }, + { + "epoch": 0.4182, + "grad_norm": 0.7547593997722213, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 41820 + }, + { + "epoch": 0.41821, + "grad_norm": 0.6923831091793283, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 41821 + }, + { + "epoch": 0.41822, + "grad_norm": 0.6178005384571765, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 41822 + }, + { + "epoch": 0.41823, + "grad_norm": 0.5997628213977513, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 41823 + }, + { + "epoch": 0.41824, + "grad_norm": 0.5950488319638721, + "learning_rate": 0.003, + "loss": 3.974, + "step": 41824 + }, + { + "epoch": 0.41825, + "grad_norm": 0.5560374983476507, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 41825 + }, + { + "epoch": 0.41826, + "grad_norm": 0.4778918561900625, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 41826 + }, + { + "epoch": 0.41827, + "grad_norm": 0.4776047490004483, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 41827 + }, + { + "epoch": 0.41828, + "grad_norm": 0.5639448550178855, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 41828 + }, + { + "epoch": 0.41829, + "grad_norm": 0.6890598995986866, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 41829 + }, + { + "epoch": 0.4183, + "grad_norm": 0.791451029921012, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 41830 + }, + { + "epoch": 0.41831, + "grad_norm": 0.9213745090146713, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 41831 + }, + { + "epoch": 0.41832, + "grad_norm": 1.0465967779930836, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 41832 + }, + { + "epoch": 0.41833, + "grad_norm": 1.0698640499191874, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 41833 + }, + { + "epoch": 0.41834, + "grad_norm": 1.0652411774885482, + "learning_rate": 0.003, + "loss": 4.005, + "step": 41834 + }, + { + "epoch": 0.41835, + "grad_norm": 0.8819247854685807, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 41835 + }, + { + "epoch": 0.41836, + "grad_norm": 0.781348621799732, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 41836 + }, + { + "epoch": 0.41837, + "grad_norm": 0.875966978234018, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 41837 + }, + { + "epoch": 0.41838, + "grad_norm": 0.8919046072241925, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 41838 + }, + { + "epoch": 0.41839, + "grad_norm": 0.9621376074477149, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 41839 + }, + { + "epoch": 0.4184, + "grad_norm": 0.9969896575893135, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 41840 + }, + { + "epoch": 0.41841, + "grad_norm": 0.9622339417871921, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 41841 + }, + { + "epoch": 0.41842, + "grad_norm": 1.086332930818965, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 41842 + }, + { + "epoch": 0.41843, + "grad_norm": 1.0766427527337636, + "learning_rate": 0.003, + "loss": 4.015, + "step": 41843 + }, + { + "epoch": 0.41844, + "grad_norm": 0.8627533343492161, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 41844 + }, + { + "epoch": 0.41845, + "grad_norm": 0.9176919308263057, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 41845 + }, + { + "epoch": 0.41846, + "grad_norm": 0.8547247399872007, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 41846 + }, + { + "epoch": 0.41847, + "grad_norm": 0.9199025412143735, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 41847 + }, + { + "epoch": 0.41848, + "grad_norm": 0.9720234909097097, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 41848 + }, + { + "epoch": 0.41849, + "grad_norm": 0.8129470741153046, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 41849 + }, + { + "epoch": 0.4185, + "grad_norm": 0.8001804907239779, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 41850 + }, + { + "epoch": 0.41851, + "grad_norm": 0.7201793372086298, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 41851 + }, + { + "epoch": 0.41852, + "grad_norm": 0.7582276725352963, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 41852 + }, + { + "epoch": 0.41853, + "grad_norm": 0.7994058687091616, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 41853 + }, + { + "epoch": 0.41854, + "grad_norm": 0.7784784588171375, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 41854 + }, + { + "epoch": 0.41855, + "grad_norm": 0.707937375287179, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 41855 + }, + { + "epoch": 0.41856, + "grad_norm": 0.8140727846222642, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 41856 + }, + { + "epoch": 0.41857, + "grad_norm": 0.8320915288962584, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 41857 + }, + { + "epoch": 0.41858, + "grad_norm": 0.6970685974121216, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 41858 + }, + { + "epoch": 0.41859, + "grad_norm": 0.6247825514724219, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 41859 + }, + { + "epoch": 0.4186, + "grad_norm": 0.5977328283041824, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 41860 + }, + { + "epoch": 0.41861, + "grad_norm": 0.5518696283966432, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 41861 + }, + { + "epoch": 0.41862, + "grad_norm": 0.6047320739787495, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 41862 + }, + { + "epoch": 0.41863, + "grad_norm": 0.6526882289312931, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 41863 + }, + { + "epoch": 0.41864, + "grad_norm": 0.6741530816088199, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 41864 + }, + { + "epoch": 0.41865, + "grad_norm": 0.7860358055958911, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 41865 + }, + { + "epoch": 0.41866, + "grad_norm": 0.8500710580956873, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 41866 + }, + { + "epoch": 0.41867, + "grad_norm": 0.9489711282659624, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 41867 + }, + { + "epoch": 0.41868, + "grad_norm": 0.9836576388965215, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 41868 + }, + { + "epoch": 0.41869, + "grad_norm": 0.7810782354086202, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 41869 + }, + { + "epoch": 0.4187, + "grad_norm": 0.6857438307201964, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 41870 + }, + { + "epoch": 0.41871, + "grad_norm": 0.6687199599870867, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 41871 + }, + { + "epoch": 0.41872, + "grad_norm": 0.6639284920015842, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 41872 + }, + { + "epoch": 0.41873, + "grad_norm": 0.6036515832159797, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 41873 + }, + { + "epoch": 0.41874, + "grad_norm": 0.6486387481289777, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 41874 + }, + { + "epoch": 0.41875, + "grad_norm": 0.6423886735131251, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 41875 + }, + { + "epoch": 0.41876, + "grad_norm": 0.6589043224827017, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 41876 + }, + { + "epoch": 0.41877, + "grad_norm": 0.8172291373608128, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 41877 + }, + { + "epoch": 0.41878, + "grad_norm": 0.86815576554338, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 41878 + }, + { + "epoch": 0.41879, + "grad_norm": 0.8100280229922653, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 41879 + }, + { + "epoch": 0.4188, + "grad_norm": 0.8089925555936316, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 41880 + }, + { + "epoch": 0.41881, + "grad_norm": 0.8834884431159508, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 41881 + }, + { + "epoch": 0.41882, + "grad_norm": 0.927502056720258, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 41882 + }, + { + "epoch": 0.41883, + "grad_norm": 0.8659178083426137, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 41883 + }, + { + "epoch": 0.41884, + "grad_norm": 0.7885038344144168, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 41884 + }, + { + "epoch": 0.41885, + "grad_norm": 0.9291687505834962, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 41885 + }, + { + "epoch": 0.41886, + "grad_norm": 0.9368138362071936, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 41886 + }, + { + "epoch": 0.41887, + "grad_norm": 1.0803035072792049, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 41887 + }, + { + "epoch": 0.41888, + "grad_norm": 0.9971466398157348, + "learning_rate": 0.003, + "loss": 4.0217, + "step": 41888 + }, + { + "epoch": 0.41889, + "grad_norm": 1.0665015656789465, + "learning_rate": 0.003, + "loss": 4.011, + "step": 41889 + }, + { + "epoch": 0.4189, + "grad_norm": 0.761256726405341, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 41890 + }, + { + "epoch": 0.41891, + "grad_norm": 0.7847995104036232, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 41891 + }, + { + "epoch": 0.41892, + "grad_norm": 0.8377513422975452, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 41892 + }, + { + "epoch": 0.41893, + "grad_norm": 0.8327419877160555, + "learning_rate": 0.003, + "loss": 4.006, + "step": 41893 + }, + { + "epoch": 0.41894, + "grad_norm": 0.7954412770384678, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 41894 + }, + { + "epoch": 0.41895, + "grad_norm": 0.8486437581191847, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 41895 + }, + { + "epoch": 0.41896, + "grad_norm": 0.9805054556540465, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 41896 + }, + { + "epoch": 0.41897, + "grad_norm": 1.0104758546615173, + "learning_rate": 0.003, + "loss": 4.0203, + "step": 41897 + }, + { + "epoch": 0.41898, + "grad_norm": 0.8615061801328384, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 41898 + }, + { + "epoch": 0.41899, + "grad_norm": 0.6925835142983877, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 41899 + }, + { + "epoch": 0.419, + "grad_norm": 0.6920905337780716, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 41900 + }, + { + "epoch": 0.41901, + "grad_norm": 0.78980618965922, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 41901 + }, + { + "epoch": 0.41902, + "grad_norm": 0.8266487725457072, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 41902 + }, + { + "epoch": 0.41903, + "grad_norm": 0.8768528346749648, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 41903 + }, + { + "epoch": 0.41904, + "grad_norm": 0.832984626257423, + "learning_rate": 0.003, + "loss": 3.976, + "step": 41904 + }, + { + "epoch": 0.41905, + "grad_norm": 0.7882531590041576, + "learning_rate": 0.003, + "loss": 4.004, + "step": 41905 + }, + { + "epoch": 0.41906, + "grad_norm": 0.6927353877320209, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 41906 + }, + { + "epoch": 0.41907, + "grad_norm": 0.6888799441125649, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 41907 + }, + { + "epoch": 0.41908, + "grad_norm": 0.6017336269183309, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 41908 + }, + { + "epoch": 0.41909, + "grad_norm": 0.6096284051802153, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 41909 + }, + { + "epoch": 0.4191, + "grad_norm": 0.5546723918882948, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 41910 + }, + { + "epoch": 0.41911, + "grad_norm": 0.5326255272853361, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 41911 + }, + { + "epoch": 0.41912, + "grad_norm": 0.5936215639451461, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 41912 + }, + { + "epoch": 0.41913, + "grad_norm": 0.6381002509429154, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 41913 + }, + { + "epoch": 0.41914, + "grad_norm": 0.6212457178187184, + "learning_rate": 0.003, + "loss": 3.985, + "step": 41914 + }, + { + "epoch": 0.41915, + "grad_norm": 0.593072552009219, + "learning_rate": 0.003, + "loss": 3.9456, + "step": 41915 + }, + { + "epoch": 0.41916, + "grad_norm": 0.5669032485032325, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 41916 + }, + { + "epoch": 0.41917, + "grad_norm": 0.6071980069868531, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 41917 + }, + { + "epoch": 0.41918, + "grad_norm": 0.6791913315825382, + "learning_rate": 0.003, + "loss": 3.993, + "step": 41918 + }, + { + "epoch": 0.41919, + "grad_norm": 0.8109998252155907, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 41919 + }, + { + "epoch": 0.4192, + "grad_norm": 0.7867788419355088, + "learning_rate": 0.003, + "loss": 3.972, + "step": 41920 + }, + { + "epoch": 0.41921, + "grad_norm": 0.7354931704355753, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 41921 + }, + { + "epoch": 0.41922, + "grad_norm": 0.8198450425667627, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 41922 + }, + { + "epoch": 0.41923, + "grad_norm": 0.8718984144882289, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 41923 + }, + { + "epoch": 0.41924, + "grad_norm": 1.0061098004986384, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 41924 + }, + { + "epoch": 0.41925, + "grad_norm": 1.1983782960022873, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 41925 + }, + { + "epoch": 0.41926, + "grad_norm": 0.7667822687375989, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 41926 + }, + { + "epoch": 0.41927, + "grad_norm": 0.6748402750354808, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 41927 + }, + { + "epoch": 0.41928, + "grad_norm": 0.6500882496472248, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 41928 + }, + { + "epoch": 0.41929, + "grad_norm": 0.7059224923356181, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 41929 + }, + { + "epoch": 0.4193, + "grad_norm": 0.6936182300677988, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 41930 + }, + { + "epoch": 0.41931, + "grad_norm": 0.7596668859311014, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 41931 + }, + { + "epoch": 0.41932, + "grad_norm": 0.8345518976576617, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 41932 + }, + { + "epoch": 0.41933, + "grad_norm": 0.7671911484101839, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 41933 + }, + { + "epoch": 0.41934, + "grad_norm": 0.7951548188935627, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 41934 + }, + { + "epoch": 0.41935, + "grad_norm": 0.7350269717978432, + "learning_rate": 0.003, + "loss": 3.971, + "step": 41935 + }, + { + "epoch": 0.41936, + "grad_norm": 0.7624091606154194, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 41936 + }, + { + "epoch": 0.41937, + "grad_norm": 0.8491137634907068, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 41937 + }, + { + "epoch": 0.41938, + "grad_norm": 0.8255737150085565, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 41938 + }, + { + "epoch": 0.41939, + "grad_norm": 0.8941188664586573, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 41939 + }, + { + "epoch": 0.4194, + "grad_norm": 1.0533835003701766, + "learning_rate": 0.003, + "loss": 3.98, + "step": 41940 + }, + { + "epoch": 0.41941, + "grad_norm": 0.9849205971011679, + "learning_rate": 0.003, + "loss": 3.988, + "step": 41941 + }, + { + "epoch": 0.41942, + "grad_norm": 0.8145647832617344, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 41942 + }, + { + "epoch": 0.41943, + "grad_norm": 0.8219845402328837, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 41943 + }, + { + "epoch": 0.41944, + "grad_norm": 0.726767603265875, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 41944 + }, + { + "epoch": 0.41945, + "grad_norm": 0.7651781952407172, + "learning_rate": 0.003, + "loss": 3.9449, + "step": 41945 + }, + { + "epoch": 0.41946, + "grad_norm": 0.9653417297429536, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 41946 + }, + { + "epoch": 0.41947, + "grad_norm": 1.0936406361769473, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 41947 + }, + { + "epoch": 0.41948, + "grad_norm": 0.909578340846541, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 41948 + }, + { + "epoch": 0.41949, + "grad_norm": 0.6542830698109319, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 41949 + }, + { + "epoch": 0.4195, + "grad_norm": 0.677229190858614, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 41950 + }, + { + "epoch": 0.41951, + "grad_norm": 0.7193415598570989, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 41951 + }, + { + "epoch": 0.41952, + "grad_norm": 0.8405345432606648, + "learning_rate": 0.003, + "loss": 3.981, + "step": 41952 + }, + { + "epoch": 0.41953, + "grad_norm": 0.9833714036868811, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 41953 + }, + { + "epoch": 0.41954, + "grad_norm": 0.9422591620911117, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 41954 + }, + { + "epoch": 0.41955, + "grad_norm": 1.0579056127514797, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 41955 + }, + { + "epoch": 0.41956, + "grad_norm": 0.9754794657451703, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 41956 + }, + { + "epoch": 0.41957, + "grad_norm": 0.965608561377673, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 41957 + }, + { + "epoch": 0.41958, + "grad_norm": 0.9587291206896383, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 41958 + }, + { + "epoch": 0.41959, + "grad_norm": 0.9461309457566242, + "learning_rate": 0.003, + "loss": 3.98, + "step": 41959 + }, + { + "epoch": 0.4196, + "grad_norm": 1.0008029133666294, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 41960 + }, + { + "epoch": 0.41961, + "grad_norm": 1.0886153387083022, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 41961 + }, + { + "epoch": 0.41962, + "grad_norm": 0.8144493618982258, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 41962 + }, + { + "epoch": 0.41963, + "grad_norm": 0.7607085968878199, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 41963 + }, + { + "epoch": 0.41964, + "grad_norm": 0.6771474593502588, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 41964 + }, + { + "epoch": 0.41965, + "grad_norm": 0.6817355331545898, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 41965 + }, + { + "epoch": 0.41966, + "grad_norm": 0.746547284335349, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 41966 + }, + { + "epoch": 0.41967, + "grad_norm": 0.7165913236430669, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 41967 + }, + { + "epoch": 0.41968, + "grad_norm": 0.7740037880538796, + "learning_rate": 0.003, + "loss": 4.002, + "step": 41968 + }, + { + "epoch": 0.41969, + "grad_norm": 0.824001958751157, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 41969 + }, + { + "epoch": 0.4197, + "grad_norm": 0.7522513185949162, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 41970 + }, + { + "epoch": 0.41971, + "grad_norm": 0.6826012559279989, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 41971 + }, + { + "epoch": 0.41972, + "grad_norm": 0.6838293436568609, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 41972 + }, + { + "epoch": 0.41973, + "grad_norm": 0.6165957115611385, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 41973 + }, + { + "epoch": 0.41974, + "grad_norm": 0.6871386701638009, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 41974 + }, + { + "epoch": 0.41975, + "grad_norm": 0.7165335977609181, + "learning_rate": 0.003, + "loss": 3.953, + "step": 41975 + }, + { + "epoch": 0.41976, + "grad_norm": 0.7577721982007987, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 41976 + }, + { + "epoch": 0.41977, + "grad_norm": 0.7869153385712382, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 41977 + }, + { + "epoch": 0.41978, + "grad_norm": 0.8319311768559724, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 41978 + }, + { + "epoch": 0.41979, + "grad_norm": 0.7912356292694015, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 41979 + }, + { + "epoch": 0.4198, + "grad_norm": 0.7079689803379914, + "learning_rate": 0.003, + "loss": 3.976, + "step": 41980 + }, + { + "epoch": 0.41981, + "grad_norm": 0.6422195138182832, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 41981 + }, + { + "epoch": 0.41982, + "grad_norm": 0.5508673962413821, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 41982 + }, + { + "epoch": 0.41983, + "grad_norm": 0.5893450421266824, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 41983 + }, + { + "epoch": 0.41984, + "grad_norm": 0.6681058406242801, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 41984 + }, + { + "epoch": 0.41985, + "grad_norm": 0.7213617813143486, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 41985 + }, + { + "epoch": 0.41986, + "grad_norm": 0.727459958136992, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 41986 + }, + { + "epoch": 0.41987, + "grad_norm": 0.8628906799892667, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 41987 + }, + { + "epoch": 0.41988, + "grad_norm": 0.955121506468913, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 41988 + }, + { + "epoch": 0.41989, + "grad_norm": 0.9415115919148089, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 41989 + }, + { + "epoch": 0.4199, + "grad_norm": 0.6967995233267422, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 41990 + }, + { + "epoch": 0.41991, + "grad_norm": 0.6412783396863034, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 41991 + }, + { + "epoch": 0.41992, + "grad_norm": 0.8021938040176956, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 41992 + }, + { + "epoch": 0.41993, + "grad_norm": 0.820562450399777, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 41993 + }, + { + "epoch": 0.41994, + "grad_norm": 0.8454279559184348, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 41994 + }, + { + "epoch": 0.41995, + "grad_norm": 0.8727241291646615, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 41995 + }, + { + "epoch": 0.41996, + "grad_norm": 0.8413007349297558, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 41996 + }, + { + "epoch": 0.41997, + "grad_norm": 0.7456273559791041, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 41997 + }, + { + "epoch": 0.41998, + "grad_norm": 0.7181966055188891, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 41998 + }, + { + "epoch": 0.41999, + "grad_norm": 0.7164848726568838, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 41999 + }, + { + "epoch": 0.42, + "grad_norm": 0.7185294462086383, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 42000 + }, + { + "epoch": 0.42001, + "grad_norm": 0.7820872967549217, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 42001 + }, + { + "epoch": 0.42002, + "grad_norm": 0.7819321895221645, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 42002 + }, + { + "epoch": 0.42003, + "grad_norm": 0.7930310502705922, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 42003 + }, + { + "epoch": 0.42004, + "grad_norm": 0.7665791637583274, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 42004 + }, + { + "epoch": 0.42005, + "grad_norm": 0.7309318259454065, + "learning_rate": 0.003, + "loss": 3.953, + "step": 42005 + }, + { + "epoch": 0.42006, + "grad_norm": 0.6648622872584911, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 42006 + }, + { + "epoch": 0.42007, + "grad_norm": 0.5989808214164782, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 42007 + }, + { + "epoch": 0.42008, + "grad_norm": 0.5966804104155897, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 42008 + }, + { + "epoch": 0.42009, + "grad_norm": 0.5876106201208225, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 42009 + }, + { + "epoch": 0.4201, + "grad_norm": 0.56712355635161, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 42010 + }, + { + "epoch": 0.42011, + "grad_norm": 0.633549125778434, + "learning_rate": 0.003, + "loss": 3.991, + "step": 42011 + }, + { + "epoch": 0.42012, + "grad_norm": 0.6604520640400557, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 42012 + }, + { + "epoch": 0.42013, + "grad_norm": 0.6928306797006204, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 42013 + }, + { + "epoch": 0.42014, + "grad_norm": 0.7992564967917885, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 42014 + }, + { + "epoch": 0.42015, + "grad_norm": 0.9438833094675131, + "learning_rate": 0.003, + "loss": 3.998, + "step": 42015 + }, + { + "epoch": 0.42016, + "grad_norm": 1.1443732641863718, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 42016 + }, + { + "epoch": 0.42017, + "grad_norm": 0.9697090247452467, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 42017 + }, + { + "epoch": 0.42018, + "grad_norm": 0.9495996104943828, + "learning_rate": 0.003, + "loss": 4.0229, + "step": 42018 + }, + { + "epoch": 0.42019, + "grad_norm": 0.8258578226387444, + "learning_rate": 0.003, + "loss": 3.997, + "step": 42019 + }, + { + "epoch": 0.4202, + "grad_norm": 0.7763886205378426, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 42020 + }, + { + "epoch": 0.42021, + "grad_norm": 0.718612355559085, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 42021 + }, + { + "epoch": 0.42022, + "grad_norm": 0.7684143117795869, + "learning_rate": 0.003, + "loss": 4.01, + "step": 42022 + }, + { + "epoch": 0.42023, + "grad_norm": 0.9090213700017024, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 42023 + }, + { + "epoch": 0.42024, + "grad_norm": 0.9448541341960033, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 42024 + }, + { + "epoch": 0.42025, + "grad_norm": 0.7734279134056984, + "learning_rate": 0.003, + "loss": 3.958, + "step": 42025 + }, + { + "epoch": 0.42026, + "grad_norm": 0.694797401789906, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 42026 + }, + { + "epoch": 0.42027, + "grad_norm": 0.6738833182473632, + "learning_rate": 0.003, + "loss": 3.98, + "step": 42027 + }, + { + "epoch": 0.42028, + "grad_norm": 0.6541822528153255, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 42028 + }, + { + "epoch": 0.42029, + "grad_norm": 0.6827839149758693, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 42029 + }, + { + "epoch": 0.4203, + "grad_norm": 0.8333380284094472, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 42030 + }, + { + "epoch": 0.42031, + "grad_norm": 1.1261522482999327, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 42031 + }, + { + "epoch": 0.42032, + "grad_norm": 1.0069351262595803, + "learning_rate": 0.003, + "loss": 4.02, + "step": 42032 + }, + { + "epoch": 0.42033, + "grad_norm": 1.090308627637641, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 42033 + }, + { + "epoch": 0.42034, + "grad_norm": 0.8321146611563499, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 42034 + }, + { + "epoch": 0.42035, + "grad_norm": 0.7563853344969694, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 42035 + }, + { + "epoch": 0.42036, + "grad_norm": 0.847949774174638, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 42036 + }, + { + "epoch": 0.42037, + "grad_norm": 0.8674801277322873, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 42037 + }, + { + "epoch": 0.42038, + "grad_norm": 0.771513334613887, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 42038 + }, + { + "epoch": 0.42039, + "grad_norm": 0.8287923687987411, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 42039 + }, + { + "epoch": 0.4204, + "grad_norm": 0.7853923500389473, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 42040 + }, + { + "epoch": 0.42041, + "grad_norm": 0.7866800278320365, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 42041 + }, + { + "epoch": 0.42042, + "grad_norm": 0.7807461833581981, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 42042 + }, + { + "epoch": 0.42043, + "grad_norm": 0.6972908159684114, + "learning_rate": 0.003, + "loss": 3.985, + "step": 42043 + }, + { + "epoch": 0.42044, + "grad_norm": 0.6905748897852535, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 42044 + }, + { + "epoch": 0.42045, + "grad_norm": 0.8142474950970433, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 42045 + }, + { + "epoch": 0.42046, + "grad_norm": 0.864023224761583, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 42046 + }, + { + "epoch": 0.42047, + "grad_norm": 0.9031531625165822, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 42047 + }, + { + "epoch": 0.42048, + "grad_norm": 0.860802343514414, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 42048 + }, + { + "epoch": 0.42049, + "grad_norm": 0.8290151257811229, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 42049 + }, + { + "epoch": 0.4205, + "grad_norm": 0.7443867779039204, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 42050 + }, + { + "epoch": 0.42051, + "grad_norm": 0.6325449223458388, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 42051 + }, + { + "epoch": 0.42052, + "grad_norm": 0.6112561108728504, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 42052 + }, + { + "epoch": 0.42053, + "grad_norm": 0.6144885164809476, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 42053 + }, + { + "epoch": 0.42054, + "grad_norm": 0.5794966380328009, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 42054 + }, + { + "epoch": 0.42055, + "grad_norm": 0.6792850325030338, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 42055 + }, + { + "epoch": 0.42056, + "grad_norm": 0.7978742571020399, + "learning_rate": 0.003, + "loss": 3.955, + "step": 42056 + }, + { + "epoch": 0.42057, + "grad_norm": 0.8870732571561578, + "learning_rate": 0.003, + "loss": 3.958, + "step": 42057 + }, + { + "epoch": 0.42058, + "grad_norm": 1.1285818757762194, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 42058 + }, + { + "epoch": 0.42059, + "grad_norm": 0.9425213653809078, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 42059 + }, + { + "epoch": 0.4206, + "grad_norm": 0.8784144990523781, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 42060 + }, + { + "epoch": 0.42061, + "grad_norm": 0.9631461768872472, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 42061 + }, + { + "epoch": 0.42062, + "grad_norm": 0.901384104369306, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 42062 + }, + { + "epoch": 0.42063, + "grad_norm": 0.8396931230200669, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 42063 + }, + { + "epoch": 0.42064, + "grad_norm": 0.9282567235842084, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 42064 + }, + { + "epoch": 0.42065, + "grad_norm": 0.8889099200181004, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 42065 + }, + { + "epoch": 0.42066, + "grad_norm": 0.6538580317227765, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 42066 + }, + { + "epoch": 0.42067, + "grad_norm": 0.6592564211713114, + "learning_rate": 0.003, + "loss": 4.008, + "step": 42067 + }, + { + "epoch": 0.42068, + "grad_norm": 0.6921084439108746, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 42068 + }, + { + "epoch": 0.42069, + "grad_norm": 0.6497147364447894, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 42069 + }, + { + "epoch": 0.4207, + "grad_norm": 0.5703650752809423, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 42070 + }, + { + "epoch": 0.42071, + "grad_norm": 0.5394484425939748, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 42071 + }, + { + "epoch": 0.42072, + "grad_norm": 0.5189481314036192, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 42072 + }, + { + "epoch": 0.42073, + "grad_norm": 0.5272724246466376, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 42073 + }, + { + "epoch": 0.42074, + "grad_norm": 0.5572595113028652, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 42074 + }, + { + "epoch": 0.42075, + "grad_norm": 0.660302318234785, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 42075 + }, + { + "epoch": 0.42076, + "grad_norm": 0.7575685842625687, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 42076 + }, + { + "epoch": 0.42077, + "grad_norm": 0.8543543726276535, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 42077 + }, + { + "epoch": 0.42078, + "grad_norm": 0.9474422394807812, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 42078 + }, + { + "epoch": 0.42079, + "grad_norm": 0.9761436716108501, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 42079 + }, + { + "epoch": 0.4208, + "grad_norm": 0.8624389016831102, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 42080 + }, + { + "epoch": 0.42081, + "grad_norm": 0.7733615485393518, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 42081 + }, + { + "epoch": 0.42082, + "grad_norm": 0.8623932011122716, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 42082 + }, + { + "epoch": 0.42083, + "grad_norm": 0.9300077406742057, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 42083 + }, + { + "epoch": 0.42084, + "grad_norm": 1.0236249843973484, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 42084 + }, + { + "epoch": 0.42085, + "grad_norm": 0.904610616283565, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 42085 + }, + { + "epoch": 0.42086, + "grad_norm": 0.783136597641865, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 42086 + }, + { + "epoch": 0.42087, + "grad_norm": 0.6631737948214232, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 42087 + }, + { + "epoch": 0.42088, + "grad_norm": 0.718656892236632, + "learning_rate": 0.003, + "loss": 4.014, + "step": 42088 + }, + { + "epoch": 0.42089, + "grad_norm": 0.7368307239187901, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 42089 + }, + { + "epoch": 0.4209, + "grad_norm": 0.668679618751608, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 42090 + }, + { + "epoch": 0.42091, + "grad_norm": 0.6688325802779232, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 42091 + }, + { + "epoch": 0.42092, + "grad_norm": 0.7925574083630379, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 42092 + }, + { + "epoch": 0.42093, + "grad_norm": 0.8331854506055015, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 42093 + }, + { + "epoch": 0.42094, + "grad_norm": 0.7104707688982564, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 42094 + }, + { + "epoch": 0.42095, + "grad_norm": 0.7551125379093914, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 42095 + }, + { + "epoch": 0.42096, + "grad_norm": 0.7471496645161575, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 42096 + }, + { + "epoch": 0.42097, + "grad_norm": 0.8151810030268795, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 42097 + }, + { + "epoch": 0.42098, + "grad_norm": 1.103826304040152, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 42098 + }, + { + "epoch": 0.42099, + "grad_norm": 0.928980849678639, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 42099 + }, + { + "epoch": 0.421, + "grad_norm": 0.9155076385219, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 42100 + }, + { + "epoch": 0.42101, + "grad_norm": 0.85066239568644, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 42101 + }, + { + "epoch": 0.42102, + "grad_norm": 0.842306902970256, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 42102 + }, + { + "epoch": 0.42103, + "grad_norm": 0.8294174717123636, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 42103 + }, + { + "epoch": 0.42104, + "grad_norm": 0.7824068090638385, + "learning_rate": 0.003, + "loss": 3.969, + "step": 42104 + }, + { + "epoch": 0.42105, + "grad_norm": 0.7577462328766584, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 42105 + }, + { + "epoch": 0.42106, + "grad_norm": 0.8708663449316917, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 42106 + }, + { + "epoch": 0.42107, + "grad_norm": 0.7580439160015384, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 42107 + }, + { + "epoch": 0.42108, + "grad_norm": 0.7784049551247263, + "learning_rate": 0.003, + "loss": 3.992, + "step": 42108 + }, + { + "epoch": 0.42109, + "grad_norm": 0.807412401688705, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 42109 + }, + { + "epoch": 0.4211, + "grad_norm": 0.7692651785103267, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 42110 + }, + { + "epoch": 0.42111, + "grad_norm": 0.7263063850675912, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 42111 + }, + { + "epoch": 0.42112, + "grad_norm": 0.6172674188028398, + "learning_rate": 0.003, + "loss": 3.961, + "step": 42112 + }, + { + "epoch": 0.42113, + "grad_norm": 0.5923669392050115, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 42113 + }, + { + "epoch": 0.42114, + "grad_norm": 0.6319060009931589, + "learning_rate": 0.003, + "loss": 3.992, + "step": 42114 + }, + { + "epoch": 0.42115, + "grad_norm": 0.6805823214499453, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 42115 + }, + { + "epoch": 0.42116, + "grad_norm": 0.7431722441767186, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 42116 + }, + { + "epoch": 0.42117, + "grad_norm": 0.7680348934380035, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 42117 + }, + { + "epoch": 0.42118, + "grad_norm": 0.7931952112953234, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 42118 + }, + { + "epoch": 0.42119, + "grad_norm": 0.7726341302997938, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 42119 + }, + { + "epoch": 0.4212, + "grad_norm": 0.669483857714548, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 42120 + }, + { + "epoch": 0.42121, + "grad_norm": 0.6826180593690946, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 42121 + }, + { + "epoch": 0.42122, + "grad_norm": 0.7419754389720512, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 42122 + }, + { + "epoch": 0.42123, + "grad_norm": 0.8692576313870947, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 42123 + }, + { + "epoch": 0.42124, + "grad_norm": 0.9564996384356563, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 42124 + }, + { + "epoch": 0.42125, + "grad_norm": 0.9460159435480944, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 42125 + }, + { + "epoch": 0.42126, + "grad_norm": 0.9069998166902462, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 42126 + }, + { + "epoch": 0.42127, + "grad_norm": 0.8198326212412531, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 42127 + }, + { + "epoch": 0.42128, + "grad_norm": 0.803158413445528, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 42128 + }, + { + "epoch": 0.42129, + "grad_norm": 0.8018228647837096, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 42129 + }, + { + "epoch": 0.4213, + "grad_norm": 0.834151225248857, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 42130 + }, + { + "epoch": 0.42131, + "grad_norm": 0.9472793433377986, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 42131 + }, + { + "epoch": 0.42132, + "grad_norm": 0.9511096977821738, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 42132 + }, + { + "epoch": 0.42133, + "grad_norm": 1.029541720696203, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 42133 + }, + { + "epoch": 0.42134, + "grad_norm": 0.8800066455109644, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 42134 + }, + { + "epoch": 0.42135, + "grad_norm": 0.7491340049711184, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 42135 + }, + { + "epoch": 0.42136, + "grad_norm": 0.7469665626876085, + "learning_rate": 0.003, + "loss": 3.98, + "step": 42136 + }, + { + "epoch": 0.42137, + "grad_norm": 0.7429212222079571, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 42137 + }, + { + "epoch": 0.42138, + "grad_norm": 0.7796582854276947, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 42138 + }, + { + "epoch": 0.42139, + "grad_norm": 0.9149932133561041, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 42139 + }, + { + "epoch": 0.4214, + "grad_norm": 1.0304459821118386, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 42140 + }, + { + "epoch": 0.42141, + "grad_norm": 0.8879407812777492, + "learning_rate": 0.003, + "loss": 3.952, + "step": 42141 + }, + { + "epoch": 0.42142, + "grad_norm": 0.7944976703467831, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 42142 + }, + { + "epoch": 0.42143, + "grad_norm": 0.6226579990424486, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 42143 + }, + { + "epoch": 0.42144, + "grad_norm": 0.6537303958780227, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 42144 + }, + { + "epoch": 0.42145, + "grad_norm": 0.6965022979793954, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 42145 + }, + { + "epoch": 0.42146, + "grad_norm": 0.7237386898842374, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 42146 + }, + { + "epoch": 0.42147, + "grad_norm": 0.7449832942945722, + "learning_rate": 0.003, + "loss": 3.973, + "step": 42147 + }, + { + "epoch": 0.42148, + "grad_norm": 0.8111870545717107, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 42148 + }, + { + "epoch": 0.42149, + "grad_norm": 0.8897681850162632, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 42149 + }, + { + "epoch": 0.4215, + "grad_norm": 0.9900156061228574, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 42150 + }, + { + "epoch": 0.42151, + "grad_norm": 1.0294153048123291, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 42151 + }, + { + "epoch": 0.42152, + "grad_norm": 1.0762550476307748, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 42152 + }, + { + "epoch": 0.42153, + "grad_norm": 1.0355908066499544, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 42153 + }, + { + "epoch": 0.42154, + "grad_norm": 0.8298392377326791, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 42154 + }, + { + "epoch": 0.42155, + "grad_norm": 0.8583560386124515, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 42155 + }, + { + "epoch": 0.42156, + "grad_norm": 0.759083407587666, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 42156 + }, + { + "epoch": 0.42157, + "grad_norm": 0.8498576708603592, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 42157 + }, + { + "epoch": 0.42158, + "grad_norm": 0.8193285127362288, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 42158 + }, + { + "epoch": 0.42159, + "grad_norm": 0.8247808757688101, + "learning_rate": 0.003, + "loss": 3.989, + "step": 42159 + }, + { + "epoch": 0.4216, + "grad_norm": 0.9192327359824488, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 42160 + }, + { + "epoch": 0.42161, + "grad_norm": 0.9773596001124157, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 42161 + }, + { + "epoch": 0.42162, + "grad_norm": 1.0133885458425602, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 42162 + }, + { + "epoch": 0.42163, + "grad_norm": 0.8346233402852978, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 42163 + }, + { + "epoch": 0.42164, + "grad_norm": 0.689936878564394, + "learning_rate": 0.003, + "loss": 3.977, + "step": 42164 + }, + { + "epoch": 0.42165, + "grad_norm": 0.65374608378523, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 42165 + }, + { + "epoch": 0.42166, + "grad_norm": 0.5825382219753446, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 42166 + }, + { + "epoch": 0.42167, + "grad_norm": 0.5857382548234095, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 42167 + }, + { + "epoch": 0.42168, + "grad_norm": 0.5451584275554073, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 42168 + }, + { + "epoch": 0.42169, + "grad_norm": 0.5173548177409508, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 42169 + }, + { + "epoch": 0.4217, + "grad_norm": 0.5662561650050274, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 42170 + }, + { + "epoch": 0.42171, + "grad_norm": 0.6146332591260982, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 42171 + }, + { + "epoch": 0.42172, + "grad_norm": 0.6889216560614817, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 42172 + }, + { + "epoch": 0.42173, + "grad_norm": 0.8551977025151024, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 42173 + }, + { + "epoch": 0.42174, + "grad_norm": 0.9066624337456979, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 42174 + }, + { + "epoch": 0.42175, + "grad_norm": 0.8444451611304716, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 42175 + }, + { + "epoch": 0.42176, + "grad_norm": 0.6926987171696634, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 42176 + }, + { + "epoch": 0.42177, + "grad_norm": 0.6515476993663637, + "learning_rate": 0.003, + "loss": 3.971, + "step": 42177 + }, + { + "epoch": 0.42178, + "grad_norm": 0.6926792861108887, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 42178 + }, + { + "epoch": 0.42179, + "grad_norm": 0.6432189702987542, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 42179 + }, + { + "epoch": 0.4218, + "grad_norm": 0.6133142013604238, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 42180 + }, + { + "epoch": 0.42181, + "grad_norm": 0.604441056317074, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 42181 + }, + { + "epoch": 0.42182, + "grad_norm": 0.5864712047599215, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 42182 + }, + { + "epoch": 0.42183, + "grad_norm": 0.6026831752791537, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 42183 + }, + { + "epoch": 0.42184, + "grad_norm": 0.5909096250539287, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 42184 + }, + { + "epoch": 0.42185, + "grad_norm": 0.5582233807958907, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 42185 + }, + { + "epoch": 0.42186, + "grad_norm": 0.5487145473919349, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 42186 + }, + { + "epoch": 0.42187, + "grad_norm": 0.573128511696416, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 42187 + }, + { + "epoch": 0.42188, + "grad_norm": 0.6916287695831506, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 42188 + }, + { + "epoch": 0.42189, + "grad_norm": 0.6841410252431828, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 42189 + }, + { + "epoch": 0.4219, + "grad_norm": 0.6538038781418667, + "learning_rate": 0.003, + "loss": 3.983, + "step": 42190 + }, + { + "epoch": 0.42191, + "grad_norm": 0.6869959237133728, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 42191 + }, + { + "epoch": 0.42192, + "grad_norm": 0.8399392543841896, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 42192 + }, + { + "epoch": 0.42193, + "grad_norm": 1.001139299677521, + "learning_rate": 0.003, + "loss": 3.9444, + "step": 42193 + }, + { + "epoch": 0.42194, + "grad_norm": 1.1487018121296206, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 42194 + }, + { + "epoch": 0.42195, + "grad_norm": 0.7180933556519578, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 42195 + }, + { + "epoch": 0.42196, + "grad_norm": 0.8432068566580492, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 42196 + }, + { + "epoch": 0.42197, + "grad_norm": 1.1397553942212055, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 42197 + }, + { + "epoch": 0.42198, + "grad_norm": 0.8885848149551598, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 42198 + }, + { + "epoch": 0.42199, + "grad_norm": 0.7533706024655972, + "learning_rate": 0.003, + "loss": 3.977, + "step": 42199 + }, + { + "epoch": 0.422, + "grad_norm": 0.8012110695270929, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 42200 + }, + { + "epoch": 0.42201, + "grad_norm": 0.7135057721408068, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 42201 + }, + { + "epoch": 0.42202, + "grad_norm": 0.7198389805906475, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 42202 + }, + { + "epoch": 0.42203, + "grad_norm": 0.8240966543953432, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 42203 + }, + { + "epoch": 0.42204, + "grad_norm": 0.8446726268906302, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 42204 + }, + { + "epoch": 0.42205, + "grad_norm": 0.8994940241037357, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 42205 + }, + { + "epoch": 0.42206, + "grad_norm": 0.8412240247275221, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 42206 + }, + { + "epoch": 0.42207, + "grad_norm": 0.7052395698951565, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 42207 + }, + { + "epoch": 0.42208, + "grad_norm": 0.6729364418252934, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 42208 + }, + { + "epoch": 0.42209, + "grad_norm": 0.7104016298177243, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 42209 + }, + { + "epoch": 0.4221, + "grad_norm": 0.7386586101626212, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 42210 + }, + { + "epoch": 0.42211, + "grad_norm": 0.7923565389432353, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 42211 + }, + { + "epoch": 0.42212, + "grad_norm": 0.8342478512398822, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 42212 + }, + { + "epoch": 0.42213, + "grad_norm": 0.8417719759306833, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 42213 + }, + { + "epoch": 0.42214, + "grad_norm": 0.909934104637122, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 42214 + }, + { + "epoch": 0.42215, + "grad_norm": 0.8158127530913356, + "learning_rate": 0.003, + "loss": 3.971, + "step": 42215 + }, + { + "epoch": 0.42216, + "grad_norm": 0.7425814349247386, + "learning_rate": 0.003, + "loss": 3.944, + "step": 42216 + }, + { + "epoch": 0.42217, + "grad_norm": 0.6797512103960492, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 42217 + }, + { + "epoch": 0.42218, + "grad_norm": 0.7811146246196893, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 42218 + }, + { + "epoch": 0.42219, + "grad_norm": 0.9448371854589589, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 42219 + }, + { + "epoch": 0.4222, + "grad_norm": 0.9989067856460518, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 42220 + }, + { + "epoch": 0.42221, + "grad_norm": 1.053170684337444, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 42221 + }, + { + "epoch": 0.42222, + "grad_norm": 1.11909707797068, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 42222 + }, + { + "epoch": 0.42223, + "grad_norm": 0.9830656769910107, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 42223 + }, + { + "epoch": 0.42224, + "grad_norm": 0.9402022488239207, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 42224 + }, + { + "epoch": 0.42225, + "grad_norm": 0.8737187362855477, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 42225 + }, + { + "epoch": 0.42226, + "grad_norm": 0.8315640319161941, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 42226 + }, + { + "epoch": 0.42227, + "grad_norm": 0.9051767008593584, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 42227 + }, + { + "epoch": 0.42228, + "grad_norm": 1.151809630918756, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 42228 + }, + { + "epoch": 0.42229, + "grad_norm": 0.7770955852196413, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 42229 + }, + { + "epoch": 0.4223, + "grad_norm": 0.7873472940151144, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 42230 + }, + { + "epoch": 0.42231, + "grad_norm": 0.763540797174986, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 42231 + }, + { + "epoch": 0.42232, + "grad_norm": 0.6931741297855122, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 42232 + }, + { + "epoch": 0.42233, + "grad_norm": 0.7209618037173432, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 42233 + }, + { + "epoch": 0.42234, + "grad_norm": 0.7378963478874212, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 42234 + }, + { + "epoch": 0.42235, + "grad_norm": 0.6917118358836879, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 42235 + }, + { + "epoch": 0.42236, + "grad_norm": 0.5938464966845328, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 42236 + }, + { + "epoch": 0.42237, + "grad_norm": 0.6184929523607728, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 42237 + }, + { + "epoch": 0.42238, + "grad_norm": 0.7281111778455152, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 42238 + }, + { + "epoch": 0.42239, + "grad_norm": 0.7268020138722846, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 42239 + }, + { + "epoch": 0.4224, + "grad_norm": 0.780604033943062, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 42240 + }, + { + "epoch": 0.42241, + "grad_norm": 0.7963116834585408, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 42241 + }, + { + "epoch": 0.42242, + "grad_norm": 0.844638253560713, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 42242 + }, + { + "epoch": 0.42243, + "grad_norm": 0.8447925934130249, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 42243 + }, + { + "epoch": 0.42244, + "grad_norm": 0.8539704256981984, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 42244 + }, + { + "epoch": 0.42245, + "grad_norm": 0.708562032823319, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 42245 + }, + { + "epoch": 0.42246, + "grad_norm": 0.6815269058891488, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 42246 + }, + { + "epoch": 0.42247, + "grad_norm": 0.7061593071958271, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 42247 + }, + { + "epoch": 0.42248, + "grad_norm": 0.8706716152013354, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 42248 + }, + { + "epoch": 0.42249, + "grad_norm": 0.938776806686849, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 42249 + }, + { + "epoch": 0.4225, + "grad_norm": 0.836670091099124, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 42250 + }, + { + "epoch": 0.42251, + "grad_norm": 0.8951864876543644, + "learning_rate": 0.003, + "loss": 3.972, + "step": 42251 + }, + { + "epoch": 0.42252, + "grad_norm": 0.7826628259726187, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 42252 + }, + { + "epoch": 0.42253, + "grad_norm": 0.8021248817484229, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 42253 + }, + { + "epoch": 0.42254, + "grad_norm": 0.8600001007857909, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 42254 + }, + { + "epoch": 0.42255, + "grad_norm": 0.8631440142887071, + "learning_rate": 0.003, + "loss": 3.995, + "step": 42255 + }, + { + "epoch": 0.42256, + "grad_norm": 0.9571255893657006, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 42256 + }, + { + "epoch": 0.42257, + "grad_norm": 1.1956896335181488, + "learning_rate": 0.003, + "loss": 4.004, + "step": 42257 + }, + { + "epoch": 0.42258, + "grad_norm": 0.8946499717671379, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 42258 + }, + { + "epoch": 0.42259, + "grad_norm": 0.7364633843847217, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 42259 + }, + { + "epoch": 0.4226, + "grad_norm": 0.6692360114134109, + "learning_rate": 0.003, + "loss": 3.9446, + "step": 42260 + }, + { + "epoch": 0.42261, + "grad_norm": 0.6008754614073945, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 42261 + }, + { + "epoch": 0.42262, + "grad_norm": 0.6130687355614798, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 42262 + }, + { + "epoch": 0.42263, + "grad_norm": 0.6286257144063977, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 42263 + }, + { + "epoch": 0.42264, + "grad_norm": 0.6763837034042299, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 42264 + }, + { + "epoch": 0.42265, + "grad_norm": 0.709743364266265, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 42265 + }, + { + "epoch": 0.42266, + "grad_norm": 0.7064779238059605, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 42266 + }, + { + "epoch": 0.42267, + "grad_norm": 0.6916431295470328, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 42267 + }, + { + "epoch": 0.42268, + "grad_norm": 0.6293575048448511, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 42268 + }, + { + "epoch": 0.42269, + "grad_norm": 0.6082793783382532, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 42269 + }, + { + "epoch": 0.4227, + "grad_norm": 0.5998885485087877, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 42270 + }, + { + "epoch": 0.42271, + "grad_norm": 0.5814619478416319, + "learning_rate": 0.003, + "loss": 3.957, + "step": 42271 + }, + { + "epoch": 0.42272, + "grad_norm": 0.5768815646041265, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 42272 + }, + { + "epoch": 0.42273, + "grad_norm": 0.6786582105128299, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 42273 + }, + { + "epoch": 0.42274, + "grad_norm": 0.7923568259904596, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 42274 + }, + { + "epoch": 0.42275, + "grad_norm": 0.8978090546681514, + "learning_rate": 0.003, + "loss": 3.971, + "step": 42275 + }, + { + "epoch": 0.42276, + "grad_norm": 1.012492616153127, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 42276 + }, + { + "epoch": 0.42277, + "grad_norm": 1.0678440006108891, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 42277 + }, + { + "epoch": 0.42278, + "grad_norm": 0.9853012727970015, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 42278 + }, + { + "epoch": 0.42279, + "grad_norm": 0.8628310746119268, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 42279 + }, + { + "epoch": 0.4228, + "grad_norm": 0.829401037224788, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 42280 + }, + { + "epoch": 0.42281, + "grad_norm": 0.7788268132896687, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 42281 + }, + { + "epoch": 0.42282, + "grad_norm": 0.6802920909166129, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 42282 + }, + { + "epoch": 0.42283, + "grad_norm": 0.6915608607136571, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 42283 + }, + { + "epoch": 0.42284, + "grad_norm": 0.7305639846386532, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 42284 + }, + { + "epoch": 0.42285, + "grad_norm": 0.7068514605445437, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 42285 + }, + { + "epoch": 0.42286, + "grad_norm": 0.7419871171666057, + "learning_rate": 0.003, + "loss": 3.983, + "step": 42286 + }, + { + "epoch": 0.42287, + "grad_norm": 0.7846488183555486, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 42287 + }, + { + "epoch": 0.42288, + "grad_norm": 0.6955897433614489, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 42288 + }, + { + "epoch": 0.42289, + "grad_norm": 0.7111330741283204, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 42289 + }, + { + "epoch": 0.4229, + "grad_norm": 0.7420416704793341, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 42290 + }, + { + "epoch": 0.42291, + "grad_norm": 0.8016943427001856, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 42291 + }, + { + "epoch": 0.42292, + "grad_norm": 0.8658936219255773, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 42292 + }, + { + "epoch": 0.42293, + "grad_norm": 0.943082302830232, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 42293 + }, + { + "epoch": 0.42294, + "grad_norm": 1.0027467433270831, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 42294 + }, + { + "epoch": 0.42295, + "grad_norm": 1.053927107536458, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 42295 + }, + { + "epoch": 0.42296, + "grad_norm": 1.0029976612300548, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 42296 + }, + { + "epoch": 0.42297, + "grad_norm": 0.8990318039985126, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 42297 + }, + { + "epoch": 0.42298, + "grad_norm": 0.9336525513867217, + "learning_rate": 0.003, + "loss": 4.0231, + "step": 42298 + }, + { + "epoch": 0.42299, + "grad_norm": 1.0865351671619856, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 42299 + }, + { + "epoch": 0.423, + "grad_norm": 0.8880433514672255, + "learning_rate": 0.003, + "loss": 3.993, + "step": 42300 + }, + { + "epoch": 0.42301, + "grad_norm": 0.7938822454796437, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 42301 + }, + { + "epoch": 0.42302, + "grad_norm": 0.8228508461174463, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 42302 + }, + { + "epoch": 0.42303, + "grad_norm": 0.7809093098098188, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 42303 + }, + { + "epoch": 0.42304, + "grad_norm": 0.7209341769187837, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 42304 + }, + { + "epoch": 0.42305, + "grad_norm": 0.7065667658731067, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 42305 + }, + { + "epoch": 0.42306, + "grad_norm": 0.6162998077384233, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 42306 + }, + { + "epoch": 0.42307, + "grad_norm": 0.6657823832997894, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 42307 + }, + { + "epoch": 0.42308, + "grad_norm": 0.7336238070314844, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 42308 + }, + { + "epoch": 0.42309, + "grad_norm": 0.6967190401699114, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 42309 + }, + { + "epoch": 0.4231, + "grad_norm": 0.7343962548211583, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 42310 + }, + { + "epoch": 0.42311, + "grad_norm": 0.857498703360974, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 42311 + }, + { + "epoch": 0.42312, + "grad_norm": 0.9695032182515466, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 42312 + }, + { + "epoch": 0.42313, + "grad_norm": 1.0761641791167078, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 42313 + }, + { + "epoch": 0.42314, + "grad_norm": 0.9078144753497213, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 42314 + }, + { + "epoch": 0.42315, + "grad_norm": 0.793434696673026, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 42315 + }, + { + "epoch": 0.42316, + "grad_norm": 0.6786341765237351, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 42316 + }, + { + "epoch": 0.42317, + "grad_norm": 0.7276308360615096, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 42317 + }, + { + "epoch": 0.42318, + "grad_norm": 0.9081067081409504, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 42318 + }, + { + "epoch": 0.42319, + "grad_norm": 0.9295418627885654, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 42319 + }, + { + "epoch": 0.4232, + "grad_norm": 0.8426474056294116, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 42320 + }, + { + "epoch": 0.42321, + "grad_norm": 0.8756431336705137, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 42321 + }, + { + "epoch": 0.42322, + "grad_norm": 0.778096215477784, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 42322 + }, + { + "epoch": 0.42323, + "grad_norm": 0.6777030769452086, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 42323 + }, + { + "epoch": 0.42324, + "grad_norm": 0.582356106679338, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 42324 + }, + { + "epoch": 0.42325, + "grad_norm": 0.5462836607449629, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 42325 + }, + { + "epoch": 0.42326, + "grad_norm": 0.5532862306008792, + "learning_rate": 0.003, + "loss": 3.9429, + "step": 42326 + }, + { + "epoch": 0.42327, + "grad_norm": 0.6381218882683396, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 42327 + }, + { + "epoch": 0.42328, + "grad_norm": 0.6314004912184891, + "learning_rate": 0.003, + "loss": 3.993, + "step": 42328 + }, + { + "epoch": 0.42329, + "grad_norm": 0.6594492151869444, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 42329 + }, + { + "epoch": 0.4233, + "grad_norm": 0.6778733261111562, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 42330 + }, + { + "epoch": 0.42331, + "grad_norm": 0.8674092894756066, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 42331 + }, + { + "epoch": 0.42332, + "grad_norm": 1.085798977205325, + "learning_rate": 0.003, + "loss": 3.973, + "step": 42332 + }, + { + "epoch": 0.42333, + "grad_norm": 0.9090952576367621, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 42333 + }, + { + "epoch": 0.42334, + "grad_norm": 0.7558364368667303, + "learning_rate": 0.003, + "loss": 3.987, + "step": 42334 + }, + { + "epoch": 0.42335, + "grad_norm": 0.6366564382179777, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 42335 + }, + { + "epoch": 0.42336, + "grad_norm": 0.7728790523830877, + "learning_rate": 0.003, + "loss": 3.977, + "step": 42336 + }, + { + "epoch": 0.42337, + "grad_norm": 0.8104878221554388, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 42337 + }, + { + "epoch": 0.42338, + "grad_norm": 0.7584071833563125, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 42338 + }, + { + "epoch": 0.42339, + "grad_norm": 0.8232944375998155, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 42339 + }, + { + "epoch": 0.4234, + "grad_norm": 0.9597693624476936, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 42340 + }, + { + "epoch": 0.42341, + "grad_norm": 0.8305889759755313, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 42341 + }, + { + "epoch": 0.42342, + "grad_norm": 0.7326855971520976, + "learning_rate": 0.003, + "loss": 3.974, + "step": 42342 + }, + { + "epoch": 0.42343, + "grad_norm": 0.7431939978299577, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 42343 + }, + { + "epoch": 0.42344, + "grad_norm": 0.7362361488696256, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 42344 + }, + { + "epoch": 0.42345, + "grad_norm": 0.7607002435165237, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 42345 + }, + { + "epoch": 0.42346, + "grad_norm": 0.7818780700668119, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 42346 + }, + { + "epoch": 0.42347, + "grad_norm": 0.7958343061170771, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 42347 + }, + { + "epoch": 0.42348, + "grad_norm": 0.8043558297857234, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 42348 + }, + { + "epoch": 0.42349, + "grad_norm": 0.8889509428776446, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 42349 + }, + { + "epoch": 0.4235, + "grad_norm": 0.8851222713223736, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 42350 + }, + { + "epoch": 0.42351, + "grad_norm": 0.8519518492423176, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 42351 + }, + { + "epoch": 0.42352, + "grad_norm": 0.9517921484318915, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 42352 + }, + { + "epoch": 0.42353, + "grad_norm": 0.9595232115524808, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 42353 + }, + { + "epoch": 0.42354, + "grad_norm": 0.9151889779998853, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 42354 + }, + { + "epoch": 0.42355, + "grad_norm": 0.9119359486435552, + "learning_rate": 0.003, + "loss": 3.987, + "step": 42355 + }, + { + "epoch": 0.42356, + "grad_norm": 0.8131134818873442, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 42356 + }, + { + "epoch": 0.42357, + "grad_norm": 0.687266944416025, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 42357 + }, + { + "epoch": 0.42358, + "grad_norm": 0.8168208710495906, + "learning_rate": 0.003, + "loss": 4.005, + "step": 42358 + }, + { + "epoch": 0.42359, + "grad_norm": 0.8121190092890983, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 42359 + }, + { + "epoch": 0.4236, + "grad_norm": 0.8657557294798148, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 42360 + }, + { + "epoch": 0.42361, + "grad_norm": 0.9369228401331253, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 42361 + }, + { + "epoch": 0.42362, + "grad_norm": 0.8272277280088733, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 42362 + }, + { + "epoch": 0.42363, + "grad_norm": 0.6445798022523521, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 42363 + }, + { + "epoch": 0.42364, + "grad_norm": 0.6410557023024054, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 42364 + }, + { + "epoch": 0.42365, + "grad_norm": 0.6316016510273049, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 42365 + }, + { + "epoch": 0.42366, + "grad_norm": 0.6020460640093146, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 42366 + }, + { + "epoch": 0.42367, + "grad_norm": 0.7087131509704644, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 42367 + }, + { + "epoch": 0.42368, + "grad_norm": 0.7407982006269376, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 42368 + }, + { + "epoch": 0.42369, + "grad_norm": 0.8483667719774944, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 42369 + }, + { + "epoch": 0.4237, + "grad_norm": 0.8697132176520771, + "learning_rate": 0.003, + "loss": 3.9475, + "step": 42370 + }, + { + "epoch": 0.42371, + "grad_norm": 0.6289521870925951, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 42371 + }, + { + "epoch": 0.42372, + "grad_norm": 0.4999442607693992, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 42372 + }, + { + "epoch": 0.42373, + "grad_norm": 0.5746334946071194, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 42373 + }, + { + "epoch": 0.42374, + "grad_norm": 0.5390091808327618, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 42374 + }, + { + "epoch": 0.42375, + "grad_norm": 0.5632658778851161, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 42375 + }, + { + "epoch": 0.42376, + "grad_norm": 0.640986556174193, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 42376 + }, + { + "epoch": 0.42377, + "grad_norm": 0.8827852639272781, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 42377 + }, + { + "epoch": 0.42378, + "grad_norm": 0.953283705591244, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 42378 + }, + { + "epoch": 0.42379, + "grad_norm": 0.7332267124941467, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 42379 + }, + { + "epoch": 0.4238, + "grad_norm": 0.6505640201346744, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 42380 + }, + { + "epoch": 0.42381, + "grad_norm": 0.722871224036211, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 42381 + }, + { + "epoch": 0.42382, + "grad_norm": 0.7600913539579898, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 42382 + }, + { + "epoch": 0.42383, + "grad_norm": 0.7760430704806033, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 42383 + }, + { + "epoch": 0.42384, + "grad_norm": 0.7713252481847909, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 42384 + }, + { + "epoch": 0.42385, + "grad_norm": 0.8003237053786116, + "learning_rate": 0.003, + "loss": 3.9443, + "step": 42385 + }, + { + "epoch": 0.42386, + "grad_norm": 0.8422828974929568, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 42386 + }, + { + "epoch": 0.42387, + "grad_norm": 0.817430102618988, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 42387 + }, + { + "epoch": 0.42388, + "grad_norm": 0.7889417325965679, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 42388 + }, + { + "epoch": 0.42389, + "grad_norm": 0.8420563657373121, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 42389 + }, + { + "epoch": 0.4239, + "grad_norm": 0.9412597967815521, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 42390 + }, + { + "epoch": 0.42391, + "grad_norm": 1.0156750938741428, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 42391 + }, + { + "epoch": 0.42392, + "grad_norm": 0.98952195016486, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 42392 + }, + { + "epoch": 0.42393, + "grad_norm": 1.1237331740156937, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 42393 + }, + { + "epoch": 0.42394, + "grad_norm": 0.9265965774262861, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 42394 + }, + { + "epoch": 0.42395, + "grad_norm": 0.7772938260464447, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 42395 + }, + { + "epoch": 0.42396, + "grad_norm": 0.6782808035738601, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 42396 + }, + { + "epoch": 0.42397, + "grad_norm": 0.648627819682107, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 42397 + }, + { + "epoch": 0.42398, + "grad_norm": 0.5859213194468672, + "learning_rate": 0.003, + "loss": 3.965, + "step": 42398 + }, + { + "epoch": 0.42399, + "grad_norm": 0.7027047440882591, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 42399 + }, + { + "epoch": 0.424, + "grad_norm": 0.8095339890625751, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 42400 + }, + { + "epoch": 0.42401, + "grad_norm": 0.9401982232111844, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 42401 + }, + { + "epoch": 0.42402, + "grad_norm": 0.9892604104850415, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 42402 + }, + { + "epoch": 0.42403, + "grad_norm": 0.9236458230657976, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 42403 + }, + { + "epoch": 0.42404, + "grad_norm": 0.8174262800053333, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 42404 + }, + { + "epoch": 0.42405, + "grad_norm": 0.6887262803812753, + "learning_rate": 0.003, + "loss": 3.985, + "step": 42405 + }, + { + "epoch": 0.42406, + "grad_norm": 0.5500549295544238, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 42406 + }, + { + "epoch": 0.42407, + "grad_norm": 0.6149717499917856, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 42407 + }, + { + "epoch": 0.42408, + "grad_norm": 0.6956820779570373, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 42408 + }, + { + "epoch": 0.42409, + "grad_norm": 0.7341408778221734, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 42409 + }, + { + "epoch": 0.4241, + "grad_norm": 0.7633802635220633, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 42410 + }, + { + "epoch": 0.42411, + "grad_norm": 0.8415065068301613, + "learning_rate": 0.003, + "loss": 3.976, + "step": 42411 + }, + { + "epoch": 0.42412, + "grad_norm": 0.8939088975882217, + "learning_rate": 0.003, + "loss": 4.0223, + "step": 42412 + }, + { + "epoch": 0.42413, + "grad_norm": 0.9097281227238517, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 42413 + }, + { + "epoch": 0.42414, + "grad_norm": 0.8451826968002686, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 42414 + }, + { + "epoch": 0.42415, + "grad_norm": 0.7861240903297967, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 42415 + }, + { + "epoch": 0.42416, + "grad_norm": 0.8287319625968799, + "learning_rate": 0.003, + "loss": 3.941, + "step": 42416 + }, + { + "epoch": 0.42417, + "grad_norm": 0.9152657307619972, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 42417 + }, + { + "epoch": 0.42418, + "grad_norm": 0.804036248285328, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 42418 + }, + { + "epoch": 0.42419, + "grad_norm": 0.8209006841379696, + "learning_rate": 0.003, + "loss": 3.984, + "step": 42419 + }, + { + "epoch": 0.4242, + "grad_norm": 0.857645768282743, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 42420 + }, + { + "epoch": 0.42421, + "grad_norm": 0.7805041740481989, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 42421 + }, + { + "epoch": 0.42422, + "grad_norm": 0.7337084332504005, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 42422 + }, + { + "epoch": 0.42423, + "grad_norm": 0.6966049536672969, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 42423 + }, + { + "epoch": 0.42424, + "grad_norm": 0.7138801482580708, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 42424 + }, + { + "epoch": 0.42425, + "grad_norm": 0.7301145081509806, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 42425 + }, + { + "epoch": 0.42426, + "grad_norm": 0.7064441767540947, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 42426 + }, + { + "epoch": 0.42427, + "grad_norm": 0.6293186583357329, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 42427 + }, + { + "epoch": 0.42428, + "grad_norm": 0.6684325713068022, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 42428 + }, + { + "epoch": 0.42429, + "grad_norm": 0.741043247878992, + "learning_rate": 0.003, + "loss": 4.001, + "step": 42429 + }, + { + "epoch": 0.4243, + "grad_norm": 0.786653166866696, + "learning_rate": 0.003, + "loss": 3.951, + "step": 42430 + }, + { + "epoch": 0.42431, + "grad_norm": 0.8432831643364354, + "learning_rate": 0.003, + "loss": 3.954, + "step": 42431 + }, + { + "epoch": 0.42432, + "grad_norm": 0.7667228859034355, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 42432 + }, + { + "epoch": 0.42433, + "grad_norm": 0.8007753849046014, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 42433 + }, + { + "epoch": 0.42434, + "grad_norm": 0.9186249230595286, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 42434 + }, + { + "epoch": 0.42435, + "grad_norm": 0.9621960189662452, + "learning_rate": 0.003, + "loss": 3.978, + "step": 42435 + }, + { + "epoch": 0.42436, + "grad_norm": 0.9722942766142866, + "learning_rate": 0.003, + "loss": 3.98, + "step": 42436 + }, + { + "epoch": 0.42437, + "grad_norm": 0.9332452243425718, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 42437 + }, + { + "epoch": 0.42438, + "grad_norm": 0.9359239501461525, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 42438 + }, + { + "epoch": 0.42439, + "grad_norm": 0.9269299191085456, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 42439 + }, + { + "epoch": 0.4244, + "grad_norm": 0.8410409677239546, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 42440 + }, + { + "epoch": 0.42441, + "grad_norm": 0.8407025321024871, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 42441 + }, + { + "epoch": 0.42442, + "grad_norm": 0.772091560530285, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 42442 + }, + { + "epoch": 0.42443, + "grad_norm": 0.7430980008565633, + "learning_rate": 0.003, + "loss": 4.006, + "step": 42443 + }, + { + "epoch": 0.42444, + "grad_norm": 0.7646935637026657, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 42444 + }, + { + "epoch": 0.42445, + "grad_norm": 0.8478358317725307, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 42445 + }, + { + "epoch": 0.42446, + "grad_norm": 0.9373953076497147, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 42446 + }, + { + "epoch": 0.42447, + "grad_norm": 0.8624226716782036, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 42447 + }, + { + "epoch": 0.42448, + "grad_norm": 0.8037608368038786, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 42448 + }, + { + "epoch": 0.42449, + "grad_norm": 0.7044762865740759, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 42449 + }, + { + "epoch": 0.4245, + "grad_norm": 0.762107985052262, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 42450 + }, + { + "epoch": 0.42451, + "grad_norm": 0.9167717660753062, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 42451 + }, + { + "epoch": 0.42452, + "grad_norm": 0.9758087652141731, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 42452 + }, + { + "epoch": 0.42453, + "grad_norm": 1.1301794757985268, + "learning_rate": 0.003, + "loss": 4.033, + "step": 42453 + }, + { + "epoch": 0.42454, + "grad_norm": 0.8084519572540292, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 42454 + }, + { + "epoch": 0.42455, + "grad_norm": 0.7506211316538876, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 42455 + }, + { + "epoch": 0.42456, + "grad_norm": 0.6195245568815738, + "learning_rate": 0.003, + "loss": 3.993, + "step": 42456 + }, + { + "epoch": 0.42457, + "grad_norm": 0.5103350844476496, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 42457 + }, + { + "epoch": 0.42458, + "grad_norm": 0.49595363963028727, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 42458 + }, + { + "epoch": 0.42459, + "grad_norm": 0.5246603277471376, + "learning_rate": 0.003, + "loss": 3.9378, + "step": 42459 + }, + { + "epoch": 0.4246, + "grad_norm": 0.6976842619240109, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 42460 + }, + { + "epoch": 0.42461, + "grad_norm": 0.9111316580732743, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 42461 + }, + { + "epoch": 0.42462, + "grad_norm": 1.0246521456460151, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 42462 + }, + { + "epoch": 0.42463, + "grad_norm": 0.7347117108502286, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 42463 + }, + { + "epoch": 0.42464, + "grad_norm": 0.6553765996508627, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 42464 + }, + { + "epoch": 0.42465, + "grad_norm": 0.7848613316823799, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 42465 + }, + { + "epoch": 0.42466, + "grad_norm": 0.786041287798123, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 42466 + }, + { + "epoch": 0.42467, + "grad_norm": 0.6130441866650372, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 42467 + }, + { + "epoch": 0.42468, + "grad_norm": 0.6172737639210549, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 42468 + }, + { + "epoch": 0.42469, + "grad_norm": 0.6009151854943434, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 42469 + }, + { + "epoch": 0.4247, + "grad_norm": 0.5746143220319857, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 42470 + }, + { + "epoch": 0.42471, + "grad_norm": 0.6109365564493158, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 42471 + }, + { + "epoch": 0.42472, + "grad_norm": 0.7136122866648992, + "learning_rate": 0.003, + "loss": 3.965, + "step": 42472 + }, + { + "epoch": 0.42473, + "grad_norm": 0.7830499781342735, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 42473 + }, + { + "epoch": 0.42474, + "grad_norm": 0.8606042358900506, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 42474 + }, + { + "epoch": 0.42475, + "grad_norm": 0.892449421659301, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 42475 + }, + { + "epoch": 0.42476, + "grad_norm": 0.7682277774209747, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 42476 + }, + { + "epoch": 0.42477, + "grad_norm": 0.7307610944874952, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 42477 + }, + { + "epoch": 0.42478, + "grad_norm": 0.6798724585560032, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 42478 + }, + { + "epoch": 0.42479, + "grad_norm": 0.7457880803240594, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 42479 + }, + { + "epoch": 0.4248, + "grad_norm": 0.8625817395345607, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 42480 + }, + { + "epoch": 0.42481, + "grad_norm": 0.9021157069620481, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 42481 + }, + { + "epoch": 0.42482, + "grad_norm": 0.8715344736275378, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 42482 + }, + { + "epoch": 0.42483, + "grad_norm": 1.0669698685426394, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 42483 + }, + { + "epoch": 0.42484, + "grad_norm": 1.1180882239277272, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 42484 + }, + { + "epoch": 0.42485, + "grad_norm": 0.8527104846549186, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 42485 + }, + { + "epoch": 0.42486, + "grad_norm": 0.8332796415098219, + "learning_rate": 0.003, + "loss": 3.95, + "step": 42486 + }, + { + "epoch": 0.42487, + "grad_norm": 0.8280728978326526, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 42487 + }, + { + "epoch": 0.42488, + "grad_norm": 0.7420849111968096, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 42488 + }, + { + "epoch": 0.42489, + "grad_norm": 0.7091803672323449, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 42489 + }, + { + "epoch": 0.4249, + "grad_norm": 0.7659082331545217, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 42490 + }, + { + "epoch": 0.42491, + "grad_norm": 0.7391116927266453, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 42491 + }, + { + "epoch": 0.42492, + "grad_norm": 0.7795960209273436, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 42492 + }, + { + "epoch": 0.42493, + "grad_norm": 0.9388127645587638, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 42493 + }, + { + "epoch": 0.42494, + "grad_norm": 0.9984758419099776, + "learning_rate": 0.003, + "loss": 3.976, + "step": 42494 + }, + { + "epoch": 0.42495, + "grad_norm": 0.90556250702167, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 42495 + }, + { + "epoch": 0.42496, + "grad_norm": 0.8506861596282516, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 42496 + }, + { + "epoch": 0.42497, + "grad_norm": 0.7031989259434538, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 42497 + }, + { + "epoch": 0.42498, + "grad_norm": 0.6827111779333643, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 42498 + }, + { + "epoch": 0.42499, + "grad_norm": 0.7144555267394496, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 42499 + }, + { + "epoch": 0.425, + "grad_norm": 0.8188900808074717, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 42500 + }, + { + "epoch": 0.42501, + "grad_norm": 0.8306879092832969, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 42501 + }, + { + "epoch": 0.42502, + "grad_norm": 0.8729512035763007, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 42502 + }, + { + "epoch": 0.42503, + "grad_norm": 0.8988446860952477, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 42503 + }, + { + "epoch": 0.42504, + "grad_norm": 1.1460313194650535, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 42504 + }, + { + "epoch": 0.42505, + "grad_norm": 0.9691130023339196, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 42505 + }, + { + "epoch": 0.42506, + "grad_norm": 0.8802783541106932, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 42506 + }, + { + "epoch": 0.42507, + "grad_norm": 0.8118293668562103, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 42507 + }, + { + "epoch": 0.42508, + "grad_norm": 0.7274414762095767, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 42508 + }, + { + "epoch": 0.42509, + "grad_norm": 0.650492781949363, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 42509 + }, + { + "epoch": 0.4251, + "grad_norm": 0.6677978108601283, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 42510 + }, + { + "epoch": 0.42511, + "grad_norm": 0.6868333463916712, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 42511 + }, + { + "epoch": 0.42512, + "grad_norm": 0.6474758079882559, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 42512 + }, + { + "epoch": 0.42513, + "grad_norm": 0.7196694893058408, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 42513 + }, + { + "epoch": 0.42514, + "grad_norm": 0.771090123502608, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 42514 + }, + { + "epoch": 0.42515, + "grad_norm": 0.7348700587740324, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 42515 + }, + { + "epoch": 0.42516, + "grad_norm": 0.7863473173255517, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 42516 + }, + { + "epoch": 0.42517, + "grad_norm": 1.0296160724510506, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 42517 + }, + { + "epoch": 0.42518, + "grad_norm": 1.0517935172737443, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 42518 + }, + { + "epoch": 0.42519, + "grad_norm": 0.7029655649595317, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 42519 + }, + { + "epoch": 0.4252, + "grad_norm": 0.6134703873968489, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 42520 + }, + { + "epoch": 0.42521, + "grad_norm": 0.7467322536406902, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 42521 + }, + { + "epoch": 0.42522, + "grad_norm": 0.7907128817166178, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 42522 + }, + { + "epoch": 0.42523, + "grad_norm": 0.7333082509479846, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 42523 + }, + { + "epoch": 0.42524, + "grad_norm": 0.6419650695948698, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 42524 + }, + { + "epoch": 0.42525, + "grad_norm": 0.6595342085662184, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 42525 + }, + { + "epoch": 0.42526, + "grad_norm": 0.8763554568738893, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 42526 + }, + { + "epoch": 0.42527, + "grad_norm": 0.9666378072097948, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 42527 + }, + { + "epoch": 0.42528, + "grad_norm": 0.8412112571181284, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 42528 + }, + { + "epoch": 0.42529, + "grad_norm": 0.7219105884598107, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 42529 + }, + { + "epoch": 0.4253, + "grad_norm": 0.844877345493254, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 42530 + }, + { + "epoch": 0.42531, + "grad_norm": 0.7725652447580611, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 42531 + }, + { + "epoch": 0.42532, + "grad_norm": 0.764540903286893, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 42532 + }, + { + "epoch": 0.42533, + "grad_norm": 0.6883002057436329, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 42533 + }, + { + "epoch": 0.42534, + "grad_norm": 0.7352840892237076, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 42534 + }, + { + "epoch": 0.42535, + "grad_norm": 0.7488529287321397, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 42535 + }, + { + "epoch": 0.42536, + "grad_norm": 0.7926507715261876, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 42536 + }, + { + "epoch": 0.42537, + "grad_norm": 0.8534967293236405, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 42537 + }, + { + "epoch": 0.42538, + "grad_norm": 0.7671115621603634, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 42538 + }, + { + "epoch": 0.42539, + "grad_norm": 0.7252629711258985, + "learning_rate": 0.003, + "loss": 3.996, + "step": 42539 + }, + { + "epoch": 0.4254, + "grad_norm": 0.7622696028957083, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 42540 + }, + { + "epoch": 0.42541, + "grad_norm": 0.9059553286511567, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 42541 + }, + { + "epoch": 0.42542, + "grad_norm": 1.1385139213181548, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 42542 + }, + { + "epoch": 0.42543, + "grad_norm": 0.9267123126647135, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 42543 + }, + { + "epoch": 0.42544, + "grad_norm": 0.8945652216845246, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 42544 + }, + { + "epoch": 0.42545, + "grad_norm": 0.9318403270491287, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 42545 + }, + { + "epoch": 0.42546, + "grad_norm": 1.0244570572547185, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 42546 + }, + { + "epoch": 0.42547, + "grad_norm": 0.9410713469951928, + "learning_rate": 0.003, + "loss": 4.024, + "step": 42547 + }, + { + "epoch": 0.42548, + "grad_norm": 0.9440301877356185, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 42548 + }, + { + "epoch": 0.42549, + "grad_norm": 0.8829783219915096, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 42549 + }, + { + "epoch": 0.4255, + "grad_norm": 0.8784155781389398, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 42550 + }, + { + "epoch": 0.42551, + "grad_norm": 0.9004048402676499, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 42551 + }, + { + "epoch": 0.42552, + "grad_norm": 0.9566053166846742, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 42552 + }, + { + "epoch": 0.42553, + "grad_norm": 1.1215923952516345, + "learning_rate": 0.003, + "loss": 4.0448, + "step": 42553 + }, + { + "epoch": 0.42554, + "grad_norm": 0.8732720572092524, + "learning_rate": 0.003, + "loss": 3.99, + "step": 42554 + }, + { + "epoch": 0.42555, + "grad_norm": 0.684837827850813, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 42555 + }, + { + "epoch": 0.42556, + "grad_norm": 0.6461676680294149, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 42556 + }, + { + "epoch": 0.42557, + "grad_norm": 0.6247067428810142, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 42557 + }, + { + "epoch": 0.42558, + "grad_norm": 0.5950820352882693, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 42558 + }, + { + "epoch": 0.42559, + "grad_norm": 0.5852714215785475, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 42559 + }, + { + "epoch": 0.4256, + "grad_norm": 0.625565854629765, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 42560 + }, + { + "epoch": 0.42561, + "grad_norm": 0.6460517922039077, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 42561 + }, + { + "epoch": 0.42562, + "grad_norm": 0.6830000127728799, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 42562 + }, + { + "epoch": 0.42563, + "grad_norm": 0.625024480211221, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 42563 + }, + { + "epoch": 0.42564, + "grad_norm": 0.5684847974573307, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 42564 + }, + { + "epoch": 0.42565, + "grad_norm": 0.5719431337650387, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 42565 + }, + { + "epoch": 0.42566, + "grad_norm": 0.6433678691084787, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 42566 + }, + { + "epoch": 0.42567, + "grad_norm": 0.67839358651239, + "learning_rate": 0.003, + "loss": 3.975, + "step": 42567 + }, + { + "epoch": 0.42568, + "grad_norm": 0.7076162728850753, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 42568 + }, + { + "epoch": 0.42569, + "grad_norm": 0.7843412554341122, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 42569 + }, + { + "epoch": 0.4257, + "grad_norm": 0.9201938209229191, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 42570 + }, + { + "epoch": 0.42571, + "grad_norm": 1.0200740488602678, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 42571 + }, + { + "epoch": 0.42572, + "grad_norm": 0.8950284299534103, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 42572 + }, + { + "epoch": 0.42573, + "grad_norm": 0.8052820566588507, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 42573 + }, + { + "epoch": 0.42574, + "grad_norm": 0.7465774993048804, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 42574 + }, + { + "epoch": 0.42575, + "grad_norm": 0.786008404662586, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 42575 + }, + { + "epoch": 0.42576, + "grad_norm": 0.8383220531819079, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 42576 + }, + { + "epoch": 0.42577, + "grad_norm": 0.9353712346501821, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 42577 + }, + { + "epoch": 0.42578, + "grad_norm": 1.0543611805273574, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 42578 + }, + { + "epoch": 0.42579, + "grad_norm": 0.8782908977973785, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 42579 + }, + { + "epoch": 0.4258, + "grad_norm": 0.6858102270872947, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 42580 + }, + { + "epoch": 0.42581, + "grad_norm": 0.7654503793691375, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 42581 + }, + { + "epoch": 0.42582, + "grad_norm": 0.721682597852002, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 42582 + }, + { + "epoch": 0.42583, + "grad_norm": 0.7728554333461349, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 42583 + }, + { + "epoch": 0.42584, + "grad_norm": 0.8445341895475494, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 42584 + }, + { + "epoch": 0.42585, + "grad_norm": 0.7520815175155802, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 42585 + }, + { + "epoch": 0.42586, + "grad_norm": 0.7113416535656653, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 42586 + }, + { + "epoch": 0.42587, + "grad_norm": 0.8116013978448298, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 42587 + }, + { + "epoch": 0.42588, + "grad_norm": 0.81609064969917, + "learning_rate": 0.003, + "loss": 3.993, + "step": 42588 + }, + { + "epoch": 0.42589, + "grad_norm": 0.9705272354677037, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 42589 + }, + { + "epoch": 0.4259, + "grad_norm": 1.0020587126063154, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 42590 + }, + { + "epoch": 0.42591, + "grad_norm": 0.9070296758554457, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 42591 + }, + { + "epoch": 0.42592, + "grad_norm": 0.8619108180974108, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 42592 + }, + { + "epoch": 0.42593, + "grad_norm": 0.7270832835624365, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 42593 + }, + { + "epoch": 0.42594, + "grad_norm": 0.6634412668650039, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 42594 + }, + { + "epoch": 0.42595, + "grad_norm": 0.691334488055538, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 42595 + }, + { + "epoch": 0.42596, + "grad_norm": 0.7094228521208846, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 42596 + }, + { + "epoch": 0.42597, + "grad_norm": 0.6633424299810974, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 42597 + }, + { + "epoch": 0.42598, + "grad_norm": 0.6850473491205825, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 42598 + }, + { + "epoch": 0.42599, + "grad_norm": 0.8125385488209911, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 42599 + }, + { + "epoch": 0.426, + "grad_norm": 0.7221771672914049, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 42600 + }, + { + "epoch": 0.42601, + "grad_norm": 0.6713295867654976, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 42601 + }, + { + "epoch": 0.42602, + "grad_norm": 0.6585116176250859, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 42602 + }, + { + "epoch": 0.42603, + "grad_norm": 0.7489703844652837, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 42603 + }, + { + "epoch": 0.42604, + "grad_norm": 0.8649971698432581, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 42604 + }, + { + "epoch": 0.42605, + "grad_norm": 0.9560405116007533, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 42605 + }, + { + "epoch": 0.42606, + "grad_norm": 0.9261387361035959, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 42606 + }, + { + "epoch": 0.42607, + "grad_norm": 0.9594338024616749, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 42607 + }, + { + "epoch": 0.42608, + "grad_norm": 0.8505740974237962, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 42608 + }, + { + "epoch": 0.42609, + "grad_norm": 0.7766405162831811, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 42609 + }, + { + "epoch": 0.4261, + "grad_norm": 0.7295546967143224, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 42610 + }, + { + "epoch": 0.42611, + "grad_norm": 0.7451831159565881, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 42611 + }, + { + "epoch": 0.42612, + "grad_norm": 0.7859647570848219, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 42612 + }, + { + "epoch": 0.42613, + "grad_norm": 0.8214953424620892, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 42613 + }, + { + "epoch": 0.42614, + "grad_norm": 0.7836886097739341, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 42614 + }, + { + "epoch": 0.42615, + "grad_norm": 0.8496729508787619, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 42615 + }, + { + "epoch": 0.42616, + "grad_norm": 0.9570190085983092, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 42616 + }, + { + "epoch": 0.42617, + "grad_norm": 0.8593864358206508, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 42617 + }, + { + "epoch": 0.42618, + "grad_norm": 0.7461098878092837, + "learning_rate": 0.003, + "loss": 3.978, + "step": 42618 + }, + { + "epoch": 0.42619, + "grad_norm": 0.7794987978658303, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 42619 + }, + { + "epoch": 0.4262, + "grad_norm": 0.6818667851823702, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 42620 + }, + { + "epoch": 0.42621, + "grad_norm": 0.7312823855015842, + "learning_rate": 0.003, + "loss": 3.96, + "step": 42621 + }, + { + "epoch": 0.42622, + "grad_norm": 0.697914435886324, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 42622 + }, + { + "epoch": 0.42623, + "grad_norm": 0.718498322706181, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 42623 + }, + { + "epoch": 0.42624, + "grad_norm": 0.7888793537604787, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 42624 + }, + { + "epoch": 0.42625, + "grad_norm": 0.8148629665719238, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 42625 + }, + { + "epoch": 0.42626, + "grad_norm": 0.7237785020176853, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 42626 + }, + { + "epoch": 0.42627, + "grad_norm": 0.6871264995775844, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 42627 + }, + { + "epoch": 0.42628, + "grad_norm": 0.5993469833556889, + "learning_rate": 0.003, + "loss": 3.967, + "step": 42628 + }, + { + "epoch": 0.42629, + "grad_norm": 0.5952628860706459, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 42629 + }, + { + "epoch": 0.4263, + "grad_norm": 0.6644256069751139, + "learning_rate": 0.003, + "loss": 3.961, + "step": 42630 + }, + { + "epoch": 0.42631, + "grad_norm": 0.7196313665000091, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 42631 + }, + { + "epoch": 0.42632, + "grad_norm": 0.9125497400496395, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 42632 + }, + { + "epoch": 0.42633, + "grad_norm": 1.0286299148502636, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 42633 + }, + { + "epoch": 0.42634, + "grad_norm": 1.0033617910782475, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 42634 + }, + { + "epoch": 0.42635, + "grad_norm": 1.1970606227762353, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 42635 + }, + { + "epoch": 0.42636, + "grad_norm": 0.7813949338537783, + "learning_rate": 0.003, + "loss": 3.969, + "step": 42636 + }, + { + "epoch": 0.42637, + "grad_norm": 0.6416816764435429, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 42637 + }, + { + "epoch": 0.42638, + "grad_norm": 0.6354808584888972, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 42638 + }, + { + "epoch": 0.42639, + "grad_norm": 0.6440596493052546, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 42639 + }, + { + "epoch": 0.4264, + "grad_norm": 0.7119952642356786, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 42640 + }, + { + "epoch": 0.42641, + "grad_norm": 0.8845503457099078, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 42641 + }, + { + "epoch": 0.42642, + "grad_norm": 0.9936934156207026, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 42642 + }, + { + "epoch": 0.42643, + "grad_norm": 1.1315555777944986, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 42643 + }, + { + "epoch": 0.42644, + "grad_norm": 0.9691397065991801, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 42644 + }, + { + "epoch": 0.42645, + "grad_norm": 0.7989666734594788, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 42645 + }, + { + "epoch": 0.42646, + "grad_norm": 0.676927635674392, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 42646 + }, + { + "epoch": 0.42647, + "grad_norm": 0.6314873693789688, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 42647 + }, + { + "epoch": 0.42648, + "grad_norm": 0.6479155088082973, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 42648 + }, + { + "epoch": 0.42649, + "grad_norm": 0.7812190141191159, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 42649 + }, + { + "epoch": 0.4265, + "grad_norm": 0.9673262542399415, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 42650 + }, + { + "epoch": 0.42651, + "grad_norm": 1.1155226875881383, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 42651 + }, + { + "epoch": 0.42652, + "grad_norm": 0.7297650068615059, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 42652 + }, + { + "epoch": 0.42653, + "grad_norm": 0.618959480153069, + "learning_rate": 0.003, + "loss": 3.978, + "step": 42653 + }, + { + "epoch": 0.42654, + "grad_norm": 0.6709599184324018, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 42654 + }, + { + "epoch": 0.42655, + "grad_norm": 0.7452990933285212, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 42655 + }, + { + "epoch": 0.42656, + "grad_norm": 0.7264609400126104, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 42656 + }, + { + "epoch": 0.42657, + "grad_norm": 0.7214383528367926, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 42657 + }, + { + "epoch": 0.42658, + "grad_norm": 0.6953135555269596, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 42658 + }, + { + "epoch": 0.42659, + "grad_norm": 0.7394036873293671, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 42659 + }, + { + "epoch": 0.4266, + "grad_norm": 0.8180461623291241, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 42660 + }, + { + "epoch": 0.42661, + "grad_norm": 0.9069171086431519, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 42661 + }, + { + "epoch": 0.42662, + "grad_norm": 0.8509622586392952, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 42662 + }, + { + "epoch": 0.42663, + "grad_norm": 0.7816330628875097, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 42663 + }, + { + "epoch": 0.42664, + "grad_norm": 0.7413540110682635, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 42664 + }, + { + "epoch": 0.42665, + "grad_norm": 0.757920016798565, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 42665 + }, + { + "epoch": 0.42666, + "grad_norm": 0.698071487950747, + "learning_rate": 0.003, + "loss": 3.993, + "step": 42666 + }, + { + "epoch": 0.42667, + "grad_norm": 0.676456368648001, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 42667 + }, + { + "epoch": 0.42668, + "grad_norm": 0.79489276790591, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 42668 + }, + { + "epoch": 0.42669, + "grad_norm": 0.8056313551955594, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 42669 + }, + { + "epoch": 0.4267, + "grad_norm": 0.8149516301000826, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 42670 + }, + { + "epoch": 0.42671, + "grad_norm": 0.998976032028455, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 42671 + }, + { + "epoch": 0.42672, + "grad_norm": 0.9554081561591571, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 42672 + }, + { + "epoch": 0.42673, + "grad_norm": 0.7656507653023077, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 42673 + }, + { + "epoch": 0.42674, + "grad_norm": 0.9154940585471827, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 42674 + }, + { + "epoch": 0.42675, + "grad_norm": 0.9862299847465339, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 42675 + }, + { + "epoch": 0.42676, + "grad_norm": 1.025138690753768, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 42676 + }, + { + "epoch": 0.42677, + "grad_norm": 1.1213971042344297, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 42677 + }, + { + "epoch": 0.42678, + "grad_norm": 0.8935818865079526, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 42678 + }, + { + "epoch": 0.42679, + "grad_norm": 0.8941875771981367, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 42679 + }, + { + "epoch": 0.4268, + "grad_norm": 0.8237844080340813, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 42680 + }, + { + "epoch": 0.42681, + "grad_norm": 0.822610551846449, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 42681 + }, + { + "epoch": 0.42682, + "grad_norm": 0.6830118095550289, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 42682 + }, + { + "epoch": 0.42683, + "grad_norm": 0.814261379217865, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 42683 + }, + { + "epoch": 0.42684, + "grad_norm": 0.9178545131067096, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 42684 + }, + { + "epoch": 0.42685, + "grad_norm": 0.8732388490571587, + "learning_rate": 0.003, + "loss": 4.0201, + "step": 42685 + }, + { + "epoch": 0.42686, + "grad_norm": 0.8187798935314441, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 42686 + }, + { + "epoch": 0.42687, + "grad_norm": 0.8453690736549025, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 42687 + }, + { + "epoch": 0.42688, + "grad_norm": 0.8875832426836088, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 42688 + }, + { + "epoch": 0.42689, + "grad_norm": 0.8900052542603174, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 42689 + }, + { + "epoch": 0.4269, + "grad_norm": 0.8666483850652361, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 42690 + }, + { + "epoch": 0.42691, + "grad_norm": 0.8590238463104234, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 42691 + }, + { + "epoch": 0.42692, + "grad_norm": 0.7653732010885451, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 42692 + }, + { + "epoch": 0.42693, + "grad_norm": 0.708946883739381, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 42693 + }, + { + "epoch": 0.42694, + "grad_norm": 0.6878239060532485, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 42694 + }, + { + "epoch": 0.42695, + "grad_norm": 0.6961616230511782, + "learning_rate": 0.003, + "loss": 4.021, + "step": 42695 + }, + { + "epoch": 0.42696, + "grad_norm": 0.7944182865079538, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 42696 + }, + { + "epoch": 0.42697, + "grad_norm": 0.9856575291049704, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 42697 + }, + { + "epoch": 0.42698, + "grad_norm": 0.9434967737907938, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 42698 + }, + { + "epoch": 0.42699, + "grad_norm": 0.8299166935807271, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 42699 + }, + { + "epoch": 0.427, + "grad_norm": 0.9176964068046108, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 42700 + }, + { + "epoch": 0.42701, + "grad_norm": 0.8071041992057587, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 42701 + }, + { + "epoch": 0.42702, + "grad_norm": 0.7087297415088597, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 42702 + }, + { + "epoch": 0.42703, + "grad_norm": 0.5021515518718955, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 42703 + }, + { + "epoch": 0.42704, + "grad_norm": 0.5358054666482001, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 42704 + }, + { + "epoch": 0.42705, + "grad_norm": 0.49904448173370364, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 42705 + }, + { + "epoch": 0.42706, + "grad_norm": 0.5057053672758149, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 42706 + }, + { + "epoch": 0.42707, + "grad_norm": 0.5801370795446805, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 42707 + }, + { + "epoch": 0.42708, + "grad_norm": 0.667407036442952, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 42708 + }, + { + "epoch": 0.42709, + "grad_norm": 0.7539305083640256, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 42709 + }, + { + "epoch": 0.4271, + "grad_norm": 0.7701817895901225, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 42710 + }, + { + "epoch": 0.42711, + "grad_norm": 0.6721695312791031, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 42711 + }, + { + "epoch": 0.42712, + "grad_norm": 0.6084230080028997, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 42712 + }, + { + "epoch": 0.42713, + "grad_norm": 0.598780869976102, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 42713 + }, + { + "epoch": 0.42714, + "grad_norm": 0.5337702269342527, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 42714 + }, + { + "epoch": 0.42715, + "grad_norm": 0.6643656992535156, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 42715 + }, + { + "epoch": 0.42716, + "grad_norm": 0.7918951338114482, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 42716 + }, + { + "epoch": 0.42717, + "grad_norm": 0.6794153921570899, + "learning_rate": 0.003, + "loss": 3.944, + "step": 42717 + }, + { + "epoch": 0.42718, + "grad_norm": 0.5115989865245864, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 42718 + }, + { + "epoch": 0.42719, + "grad_norm": 0.43149827462238055, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 42719 + }, + { + "epoch": 0.4272, + "grad_norm": 0.549569720043869, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 42720 + }, + { + "epoch": 0.42721, + "grad_norm": 0.6294463322486805, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 42721 + }, + { + "epoch": 0.42722, + "grad_norm": 0.593980336158829, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 42722 + }, + { + "epoch": 0.42723, + "grad_norm": 0.5933456915793304, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 42723 + }, + { + "epoch": 0.42724, + "grad_norm": 0.6628308791339234, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 42724 + }, + { + "epoch": 0.42725, + "grad_norm": 0.7043632830419991, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 42725 + }, + { + "epoch": 0.42726, + "grad_norm": 0.8266508755319035, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 42726 + }, + { + "epoch": 0.42727, + "grad_norm": 0.8209599577152508, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 42727 + }, + { + "epoch": 0.42728, + "grad_norm": 0.7745052002164653, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 42728 + }, + { + "epoch": 0.42729, + "grad_norm": 0.7586867865364799, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 42729 + }, + { + "epoch": 0.4273, + "grad_norm": 0.8538899056957759, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 42730 + }, + { + "epoch": 0.42731, + "grad_norm": 0.8690337707169333, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 42731 + }, + { + "epoch": 0.42732, + "grad_norm": 0.980099363374212, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 42732 + }, + { + "epoch": 0.42733, + "grad_norm": 1.007963542045994, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 42733 + }, + { + "epoch": 0.42734, + "grad_norm": 1.201052480097814, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 42734 + }, + { + "epoch": 0.42735, + "grad_norm": 0.7920404415169247, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 42735 + }, + { + "epoch": 0.42736, + "grad_norm": 0.8165562740362647, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 42736 + }, + { + "epoch": 0.42737, + "grad_norm": 0.879310993661854, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 42737 + }, + { + "epoch": 0.42738, + "grad_norm": 0.9696153124862534, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 42738 + }, + { + "epoch": 0.42739, + "grad_norm": 0.9696873210843406, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 42739 + }, + { + "epoch": 0.4274, + "grad_norm": 1.0136076721096081, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 42740 + }, + { + "epoch": 0.42741, + "grad_norm": 0.9077760211618835, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 42741 + }, + { + "epoch": 0.42742, + "grad_norm": 0.9635914878452546, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 42742 + }, + { + "epoch": 0.42743, + "grad_norm": 0.987152630695319, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 42743 + }, + { + "epoch": 0.42744, + "grad_norm": 0.8304261711094663, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 42744 + }, + { + "epoch": 0.42745, + "grad_norm": 0.8273793616215697, + "learning_rate": 0.003, + "loss": 4.0576, + "step": 42745 + }, + { + "epoch": 0.42746, + "grad_norm": 0.754509252721255, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 42746 + }, + { + "epoch": 0.42747, + "grad_norm": 0.8354784925932838, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 42747 + }, + { + "epoch": 0.42748, + "grad_norm": 0.7941823111152335, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 42748 + }, + { + "epoch": 0.42749, + "grad_norm": 0.8293815187997812, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 42749 + }, + { + "epoch": 0.4275, + "grad_norm": 0.8320307077542652, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 42750 + }, + { + "epoch": 0.42751, + "grad_norm": 0.9560196611475151, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 42751 + }, + { + "epoch": 0.42752, + "grad_norm": 0.9581715085592056, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 42752 + }, + { + "epoch": 0.42753, + "grad_norm": 0.8441447171226584, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 42753 + }, + { + "epoch": 0.42754, + "grad_norm": 0.7415282953439553, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 42754 + }, + { + "epoch": 0.42755, + "grad_norm": 0.6838187010229235, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 42755 + }, + { + "epoch": 0.42756, + "grad_norm": 0.7052528953904855, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 42756 + }, + { + "epoch": 0.42757, + "grad_norm": 0.6261883031557597, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 42757 + }, + { + "epoch": 0.42758, + "grad_norm": 0.6475396210050478, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 42758 + }, + { + "epoch": 0.42759, + "grad_norm": 0.70237741242566, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 42759 + }, + { + "epoch": 0.4276, + "grad_norm": 0.7287184929302823, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 42760 + }, + { + "epoch": 0.42761, + "grad_norm": 0.7522412820101075, + "learning_rate": 0.003, + "loss": 3.991, + "step": 42761 + }, + { + "epoch": 0.42762, + "grad_norm": 0.7573612526414355, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 42762 + }, + { + "epoch": 0.42763, + "grad_norm": 0.7341821136683053, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 42763 + }, + { + "epoch": 0.42764, + "grad_norm": 0.7001711742032772, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 42764 + }, + { + "epoch": 0.42765, + "grad_norm": 0.8201408695120176, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 42765 + }, + { + "epoch": 0.42766, + "grad_norm": 0.7189532267613675, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 42766 + }, + { + "epoch": 0.42767, + "grad_norm": 0.635505276397799, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 42767 + }, + { + "epoch": 0.42768, + "grad_norm": 0.6420540271370846, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 42768 + }, + { + "epoch": 0.42769, + "grad_norm": 0.619266359947602, + "learning_rate": 0.003, + "loss": 3.969, + "step": 42769 + }, + { + "epoch": 0.4277, + "grad_norm": 0.6416337688374434, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 42770 + }, + { + "epoch": 0.42771, + "grad_norm": 0.6356106428710038, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 42771 + }, + { + "epoch": 0.42772, + "grad_norm": 0.731645580045126, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 42772 + }, + { + "epoch": 0.42773, + "grad_norm": 0.9058543186941138, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 42773 + }, + { + "epoch": 0.42774, + "grad_norm": 0.998153940335113, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 42774 + }, + { + "epoch": 0.42775, + "grad_norm": 0.9470394372035247, + "learning_rate": 0.003, + "loss": 3.98, + "step": 42775 + }, + { + "epoch": 0.42776, + "grad_norm": 0.8850467911087699, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 42776 + }, + { + "epoch": 0.42777, + "grad_norm": 0.8305593073685438, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 42777 + }, + { + "epoch": 0.42778, + "grad_norm": 0.9926674008293672, + "learning_rate": 0.003, + "loss": 3.965, + "step": 42778 + }, + { + "epoch": 0.42779, + "grad_norm": 1.040480278244432, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 42779 + }, + { + "epoch": 0.4278, + "grad_norm": 0.9189674136820634, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 42780 + }, + { + "epoch": 0.42781, + "grad_norm": 0.8118337611984766, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 42781 + }, + { + "epoch": 0.42782, + "grad_norm": 0.7700796137833485, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 42782 + }, + { + "epoch": 0.42783, + "grad_norm": 0.7533854547390388, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 42783 + }, + { + "epoch": 0.42784, + "grad_norm": 0.6989746811328348, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 42784 + }, + { + "epoch": 0.42785, + "grad_norm": 0.6999396386335773, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 42785 + }, + { + "epoch": 0.42786, + "grad_norm": 0.7255888298163224, + "learning_rate": 0.003, + "loss": 3.976, + "step": 42786 + }, + { + "epoch": 0.42787, + "grad_norm": 0.7895187766284701, + "learning_rate": 0.003, + "loss": 4.029, + "step": 42787 + }, + { + "epoch": 0.42788, + "grad_norm": 0.6769466272319982, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 42788 + }, + { + "epoch": 0.42789, + "grad_norm": 0.631635476000888, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 42789 + }, + { + "epoch": 0.4279, + "grad_norm": 0.6080407670354803, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 42790 + }, + { + "epoch": 0.42791, + "grad_norm": 0.6607912153244465, + "learning_rate": 0.003, + "loss": 3.9378, + "step": 42791 + }, + { + "epoch": 0.42792, + "grad_norm": 0.7053535003648778, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 42792 + }, + { + "epoch": 0.42793, + "grad_norm": 0.762044007608085, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 42793 + }, + { + "epoch": 0.42794, + "grad_norm": 0.8304503299065835, + "learning_rate": 0.003, + "loss": 3.986, + "step": 42794 + }, + { + "epoch": 0.42795, + "grad_norm": 1.0460405930825465, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 42795 + }, + { + "epoch": 0.42796, + "grad_norm": 1.0148159117993685, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 42796 + }, + { + "epoch": 0.42797, + "grad_norm": 0.9197103012275518, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 42797 + }, + { + "epoch": 0.42798, + "grad_norm": 0.9152681885503708, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 42798 + }, + { + "epoch": 0.42799, + "grad_norm": 0.8861487709160514, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 42799 + }, + { + "epoch": 0.428, + "grad_norm": 0.7973884759233782, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 42800 + }, + { + "epoch": 0.42801, + "grad_norm": 0.7914980328431599, + "learning_rate": 0.003, + "loss": 3.984, + "step": 42801 + }, + { + "epoch": 0.42802, + "grad_norm": 0.9062119344368044, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 42802 + }, + { + "epoch": 0.42803, + "grad_norm": 0.9071776006983292, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 42803 + }, + { + "epoch": 0.42804, + "grad_norm": 0.783984077024931, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 42804 + }, + { + "epoch": 0.42805, + "grad_norm": 0.6785467849047424, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 42805 + }, + { + "epoch": 0.42806, + "grad_norm": 0.6554615234775707, + "learning_rate": 0.003, + "loss": 3.967, + "step": 42806 + }, + { + "epoch": 0.42807, + "grad_norm": 0.6989353257415849, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 42807 + }, + { + "epoch": 0.42808, + "grad_norm": 0.6579918701466614, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 42808 + }, + { + "epoch": 0.42809, + "grad_norm": 0.6168989170351361, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 42809 + }, + { + "epoch": 0.4281, + "grad_norm": 0.6995695412975479, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 42810 + }, + { + "epoch": 0.42811, + "grad_norm": 0.8599477420136374, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 42811 + }, + { + "epoch": 0.42812, + "grad_norm": 0.919989328541213, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 42812 + }, + { + "epoch": 0.42813, + "grad_norm": 0.9773046125132449, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 42813 + }, + { + "epoch": 0.42814, + "grad_norm": 0.9971449200820642, + "learning_rate": 0.003, + "loss": 3.983, + "step": 42814 + }, + { + "epoch": 0.42815, + "grad_norm": 0.8282322800343689, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 42815 + }, + { + "epoch": 0.42816, + "grad_norm": 0.7260239189319574, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 42816 + }, + { + "epoch": 0.42817, + "grad_norm": 0.715060645625218, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 42817 + }, + { + "epoch": 0.42818, + "grad_norm": 0.6462563332826843, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 42818 + }, + { + "epoch": 0.42819, + "grad_norm": 0.6378110869687517, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 42819 + }, + { + "epoch": 0.4282, + "grad_norm": 0.6291827670661425, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 42820 + }, + { + "epoch": 0.42821, + "grad_norm": 0.6415356624818955, + "learning_rate": 0.003, + "loss": 3.9293, + "step": 42821 + }, + { + "epoch": 0.42822, + "grad_norm": 0.5913153163624782, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 42822 + }, + { + "epoch": 0.42823, + "grad_norm": 0.49216494867470384, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 42823 + }, + { + "epoch": 0.42824, + "grad_norm": 0.4754111588489071, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 42824 + }, + { + "epoch": 0.42825, + "grad_norm": 0.5255115665842945, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 42825 + }, + { + "epoch": 0.42826, + "grad_norm": 0.5747331357475316, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 42826 + }, + { + "epoch": 0.42827, + "grad_norm": 0.6107140859734728, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 42827 + }, + { + "epoch": 0.42828, + "grad_norm": 0.660977668047099, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 42828 + }, + { + "epoch": 0.42829, + "grad_norm": 0.7476807387633143, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 42829 + }, + { + "epoch": 0.4283, + "grad_norm": 0.7684964990899364, + "learning_rate": 0.003, + "loss": 3.983, + "step": 42830 + }, + { + "epoch": 0.42831, + "grad_norm": 0.7270226830890879, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 42831 + }, + { + "epoch": 0.42832, + "grad_norm": 0.7452111302573404, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 42832 + }, + { + "epoch": 0.42833, + "grad_norm": 0.7493088559910985, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 42833 + }, + { + "epoch": 0.42834, + "grad_norm": 0.7799877929960549, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 42834 + }, + { + "epoch": 0.42835, + "grad_norm": 0.9611099289887605, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 42835 + }, + { + "epoch": 0.42836, + "grad_norm": 1.1860101867243567, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 42836 + }, + { + "epoch": 0.42837, + "grad_norm": 0.8889999320964127, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 42837 + }, + { + "epoch": 0.42838, + "grad_norm": 0.9778051102735675, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 42838 + }, + { + "epoch": 0.42839, + "grad_norm": 1.180455382633059, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 42839 + }, + { + "epoch": 0.4284, + "grad_norm": 1.0194372841826769, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 42840 + }, + { + "epoch": 0.42841, + "grad_norm": 1.0156457254840932, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 42841 + }, + { + "epoch": 0.42842, + "grad_norm": 0.9675622839020986, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 42842 + }, + { + "epoch": 0.42843, + "grad_norm": 0.9361539079020595, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 42843 + }, + { + "epoch": 0.42844, + "grad_norm": 0.8734358736355838, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 42844 + }, + { + "epoch": 0.42845, + "grad_norm": 0.811923841583702, + "learning_rate": 0.003, + "loss": 3.977, + "step": 42845 + }, + { + "epoch": 0.42846, + "grad_norm": 0.7634005900596106, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 42846 + }, + { + "epoch": 0.42847, + "grad_norm": 0.7390171994449142, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 42847 + }, + { + "epoch": 0.42848, + "grad_norm": 0.7485231362176933, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 42848 + }, + { + "epoch": 0.42849, + "grad_norm": 0.765397880858599, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 42849 + }, + { + "epoch": 0.4285, + "grad_norm": 0.6420970940724465, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 42850 + }, + { + "epoch": 0.42851, + "grad_norm": 0.6850729139590471, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 42851 + }, + { + "epoch": 0.42852, + "grad_norm": 0.6817266162328748, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 42852 + }, + { + "epoch": 0.42853, + "grad_norm": 0.6661049759036053, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 42853 + }, + { + "epoch": 0.42854, + "grad_norm": 0.6922664185416234, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 42854 + }, + { + "epoch": 0.42855, + "grad_norm": 0.7918327322126233, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 42855 + }, + { + "epoch": 0.42856, + "grad_norm": 0.8653833109386552, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 42856 + }, + { + "epoch": 0.42857, + "grad_norm": 0.8635593545616043, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 42857 + }, + { + "epoch": 0.42858, + "grad_norm": 0.7508857461667954, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 42858 + }, + { + "epoch": 0.42859, + "grad_norm": 0.7366560938671987, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 42859 + }, + { + "epoch": 0.4286, + "grad_norm": 0.8099679572104591, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 42860 + }, + { + "epoch": 0.42861, + "grad_norm": 0.7553703827009233, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 42861 + }, + { + "epoch": 0.42862, + "grad_norm": 0.7183331891422347, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 42862 + }, + { + "epoch": 0.42863, + "grad_norm": 0.9305609087755996, + "learning_rate": 0.003, + "loss": 3.9439, + "step": 42863 + }, + { + "epoch": 0.42864, + "grad_norm": 1.1281908624511499, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 42864 + }, + { + "epoch": 0.42865, + "grad_norm": 0.8360811314023825, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 42865 + }, + { + "epoch": 0.42866, + "grad_norm": 0.7615659369256383, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 42866 + }, + { + "epoch": 0.42867, + "grad_norm": 0.8129490112576213, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 42867 + }, + { + "epoch": 0.42868, + "grad_norm": 0.7847967629887411, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 42868 + }, + { + "epoch": 0.42869, + "grad_norm": 0.8091148794312303, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 42869 + }, + { + "epoch": 0.4287, + "grad_norm": 0.8614574602934576, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 42870 + }, + { + "epoch": 0.42871, + "grad_norm": 0.9154816401941912, + "learning_rate": 0.003, + "loss": 3.982, + "step": 42871 + }, + { + "epoch": 0.42872, + "grad_norm": 0.8645973678424442, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 42872 + }, + { + "epoch": 0.42873, + "grad_norm": 0.7826622760201244, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 42873 + }, + { + "epoch": 0.42874, + "grad_norm": 0.7175090799082323, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 42874 + }, + { + "epoch": 0.42875, + "grad_norm": 0.7115668288426009, + "learning_rate": 0.003, + "loss": 3.986, + "step": 42875 + }, + { + "epoch": 0.42876, + "grad_norm": 0.7124720938710062, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 42876 + }, + { + "epoch": 0.42877, + "grad_norm": 0.8799595248556952, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 42877 + }, + { + "epoch": 0.42878, + "grad_norm": 0.9109712367678197, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 42878 + }, + { + "epoch": 0.42879, + "grad_norm": 0.9662285524978358, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 42879 + }, + { + "epoch": 0.4288, + "grad_norm": 0.9994386995624165, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 42880 + }, + { + "epoch": 0.42881, + "grad_norm": 0.8670727754165379, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 42881 + }, + { + "epoch": 0.42882, + "grad_norm": 0.8525096392393635, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 42882 + }, + { + "epoch": 0.42883, + "grad_norm": 0.8183757528460336, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 42883 + }, + { + "epoch": 0.42884, + "grad_norm": 0.8082843230175876, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 42884 + }, + { + "epoch": 0.42885, + "grad_norm": 0.8624491152794209, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 42885 + }, + { + "epoch": 0.42886, + "grad_norm": 0.867148918962055, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 42886 + }, + { + "epoch": 0.42887, + "grad_norm": 0.6660239116935168, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 42887 + }, + { + "epoch": 0.42888, + "grad_norm": 0.6250584918411266, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 42888 + }, + { + "epoch": 0.42889, + "grad_norm": 0.5471489834839749, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 42889 + }, + { + "epoch": 0.4289, + "grad_norm": 0.5794073161586356, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 42890 + }, + { + "epoch": 0.42891, + "grad_norm": 0.5814567800872356, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 42891 + }, + { + "epoch": 0.42892, + "grad_norm": 0.629782198951241, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 42892 + }, + { + "epoch": 0.42893, + "grad_norm": 0.6430156781943175, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 42893 + }, + { + "epoch": 0.42894, + "grad_norm": 0.8383307795754422, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 42894 + }, + { + "epoch": 0.42895, + "grad_norm": 1.1094076534438582, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 42895 + }, + { + "epoch": 0.42896, + "grad_norm": 0.8996879835838454, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 42896 + }, + { + "epoch": 0.42897, + "grad_norm": 0.7706801012373712, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 42897 + }, + { + "epoch": 0.42898, + "grad_norm": 0.7447311481905298, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 42898 + }, + { + "epoch": 0.42899, + "grad_norm": 0.6504507351101996, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 42899 + }, + { + "epoch": 0.429, + "grad_norm": 0.634088021892595, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 42900 + }, + { + "epoch": 0.42901, + "grad_norm": 0.7202376760101546, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 42901 + }, + { + "epoch": 0.42902, + "grad_norm": 0.815871826390948, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 42902 + }, + { + "epoch": 0.42903, + "grad_norm": 0.7372469617854877, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 42903 + }, + { + "epoch": 0.42904, + "grad_norm": 0.9339703857996616, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 42904 + }, + { + "epoch": 0.42905, + "grad_norm": 1.0581530070673528, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 42905 + }, + { + "epoch": 0.42906, + "grad_norm": 0.8564634329213189, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 42906 + }, + { + "epoch": 0.42907, + "grad_norm": 0.7812951955904747, + "learning_rate": 0.003, + "loss": 3.985, + "step": 42907 + }, + { + "epoch": 0.42908, + "grad_norm": 0.821128353423221, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 42908 + }, + { + "epoch": 0.42909, + "grad_norm": 0.8526799224021231, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 42909 + }, + { + "epoch": 0.4291, + "grad_norm": 0.7793911505724277, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 42910 + }, + { + "epoch": 0.42911, + "grad_norm": 0.6925192213217933, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 42911 + }, + { + "epoch": 0.42912, + "grad_norm": 0.6640692132650328, + "learning_rate": 0.003, + "loss": 3.986, + "step": 42912 + }, + { + "epoch": 0.42913, + "grad_norm": 0.6370535144698807, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 42913 + }, + { + "epoch": 0.42914, + "grad_norm": 0.6288352881948951, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 42914 + }, + { + "epoch": 0.42915, + "grad_norm": 0.5626219748449287, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 42915 + }, + { + "epoch": 0.42916, + "grad_norm": 0.5532086781767371, + "learning_rate": 0.003, + "loss": 3.988, + "step": 42916 + }, + { + "epoch": 0.42917, + "grad_norm": 0.5954183081580543, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 42917 + }, + { + "epoch": 0.42918, + "grad_norm": 0.7100503525264157, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 42918 + }, + { + "epoch": 0.42919, + "grad_norm": 0.8018049969388844, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 42919 + }, + { + "epoch": 0.4292, + "grad_norm": 0.7636937263074925, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 42920 + }, + { + "epoch": 0.42921, + "grad_norm": 0.880542186495424, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 42921 + }, + { + "epoch": 0.42922, + "grad_norm": 0.8727896141181102, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 42922 + }, + { + "epoch": 0.42923, + "grad_norm": 0.8962977644098337, + "learning_rate": 0.003, + "loss": 3.994, + "step": 42923 + }, + { + "epoch": 0.42924, + "grad_norm": 1.0062215340212104, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 42924 + }, + { + "epoch": 0.42925, + "grad_norm": 1.0015589952921078, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 42925 + }, + { + "epoch": 0.42926, + "grad_norm": 0.9516334582698632, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 42926 + }, + { + "epoch": 0.42927, + "grad_norm": 1.0715102107987375, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 42927 + }, + { + "epoch": 0.42928, + "grad_norm": 1.0123649890876303, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 42928 + }, + { + "epoch": 0.42929, + "grad_norm": 1.0121774583176402, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 42929 + }, + { + "epoch": 0.4293, + "grad_norm": 1.0327074637882656, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 42930 + }, + { + "epoch": 0.42931, + "grad_norm": 1.0241423458831767, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 42931 + }, + { + "epoch": 0.42932, + "grad_norm": 1.0474896479619802, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 42932 + }, + { + "epoch": 0.42933, + "grad_norm": 0.794236375401764, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 42933 + }, + { + "epoch": 0.42934, + "grad_norm": 0.6355810106520963, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 42934 + }, + { + "epoch": 0.42935, + "grad_norm": 0.615766456165104, + "learning_rate": 0.003, + "loss": 3.975, + "step": 42935 + }, + { + "epoch": 0.42936, + "grad_norm": 0.6478719867056177, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 42936 + }, + { + "epoch": 0.42937, + "grad_norm": 0.7497593519122967, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 42937 + }, + { + "epoch": 0.42938, + "grad_norm": 0.7885732919376703, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 42938 + }, + { + "epoch": 0.42939, + "grad_norm": 0.7541641505762925, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 42939 + }, + { + "epoch": 0.4294, + "grad_norm": 0.6276472317820979, + "learning_rate": 0.003, + "loss": 3.9435, + "step": 42940 + }, + { + "epoch": 0.42941, + "grad_norm": 0.70706908487142, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 42941 + }, + { + "epoch": 0.42942, + "grad_norm": 0.9098096110412901, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 42942 + }, + { + "epoch": 0.42943, + "grad_norm": 0.8707987341498534, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 42943 + }, + { + "epoch": 0.42944, + "grad_norm": 0.7434261029334293, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 42944 + }, + { + "epoch": 0.42945, + "grad_norm": 0.646104604574662, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 42945 + }, + { + "epoch": 0.42946, + "grad_norm": 0.6276010548124656, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 42946 + }, + { + "epoch": 0.42947, + "grad_norm": 0.5704105031628747, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 42947 + }, + { + "epoch": 0.42948, + "grad_norm": 0.507067387587632, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 42948 + }, + { + "epoch": 0.42949, + "grad_norm": 0.4946987714532369, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 42949 + }, + { + "epoch": 0.4295, + "grad_norm": 0.5510107092801038, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 42950 + }, + { + "epoch": 0.42951, + "grad_norm": 0.635157548514759, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 42951 + }, + { + "epoch": 0.42952, + "grad_norm": 0.8200623242236432, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 42952 + }, + { + "epoch": 0.42953, + "grad_norm": 0.9362783540211335, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 42953 + }, + { + "epoch": 0.42954, + "grad_norm": 0.911191381524156, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 42954 + }, + { + "epoch": 0.42955, + "grad_norm": 0.8786659120996105, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 42955 + }, + { + "epoch": 0.42956, + "grad_norm": 0.7976629741177248, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 42956 + }, + { + "epoch": 0.42957, + "grad_norm": 0.8229775049373902, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 42957 + }, + { + "epoch": 0.42958, + "grad_norm": 0.8048374643712183, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 42958 + }, + { + "epoch": 0.42959, + "grad_norm": 0.8602028933674765, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 42959 + }, + { + "epoch": 0.4296, + "grad_norm": 0.8389061912180233, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 42960 + }, + { + "epoch": 0.42961, + "grad_norm": 0.752734293392802, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 42961 + }, + { + "epoch": 0.42962, + "grad_norm": 0.861385473003517, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 42962 + }, + { + "epoch": 0.42963, + "grad_norm": 0.9904235730847613, + "learning_rate": 0.003, + "loss": 3.995, + "step": 42963 + }, + { + "epoch": 0.42964, + "grad_norm": 1.1191633839690658, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 42964 + }, + { + "epoch": 0.42965, + "grad_norm": 0.851211315601911, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 42965 + }, + { + "epoch": 0.42966, + "grad_norm": 0.7650343661979027, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 42966 + }, + { + "epoch": 0.42967, + "grad_norm": 0.6736554727951809, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 42967 + }, + { + "epoch": 0.42968, + "grad_norm": 0.6805319542444611, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 42968 + }, + { + "epoch": 0.42969, + "grad_norm": 0.6648744263529444, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 42969 + }, + { + "epoch": 0.4297, + "grad_norm": 0.7568519944316789, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 42970 + }, + { + "epoch": 0.42971, + "grad_norm": 0.8105392058631388, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 42971 + }, + { + "epoch": 0.42972, + "grad_norm": 0.7832475232439093, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 42972 + }, + { + "epoch": 0.42973, + "grad_norm": 0.7927352849269338, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 42973 + }, + { + "epoch": 0.42974, + "grad_norm": 0.7944544301389135, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 42974 + }, + { + "epoch": 0.42975, + "grad_norm": 0.7725813780063157, + "learning_rate": 0.003, + "loss": 3.977, + "step": 42975 + }, + { + "epoch": 0.42976, + "grad_norm": 0.6947759561613115, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 42976 + }, + { + "epoch": 0.42977, + "grad_norm": 0.6920499861461223, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 42977 + }, + { + "epoch": 0.42978, + "grad_norm": 0.829794726872972, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 42978 + }, + { + "epoch": 0.42979, + "grad_norm": 0.955129423822428, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 42979 + }, + { + "epoch": 0.4298, + "grad_norm": 0.8581269421716715, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 42980 + }, + { + "epoch": 0.42981, + "grad_norm": 0.9774351082603511, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 42981 + }, + { + "epoch": 0.42982, + "grad_norm": 1.0166124080844456, + "learning_rate": 0.003, + "loss": 4.003, + "step": 42982 + }, + { + "epoch": 0.42983, + "grad_norm": 1.1164799456843377, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 42983 + }, + { + "epoch": 0.42984, + "grad_norm": 0.8048850431742803, + "learning_rate": 0.003, + "loss": 4.0, + "step": 42984 + }, + { + "epoch": 0.42985, + "grad_norm": 0.6830898800844079, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 42985 + }, + { + "epoch": 0.42986, + "grad_norm": 0.7730214535264321, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 42986 + }, + { + "epoch": 0.42987, + "grad_norm": 0.7126619397372693, + "learning_rate": 0.003, + "loss": 3.976, + "step": 42987 + }, + { + "epoch": 0.42988, + "grad_norm": 0.7803290410397475, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 42988 + }, + { + "epoch": 0.42989, + "grad_norm": 0.8011843980145632, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 42989 + }, + { + "epoch": 0.4299, + "grad_norm": 0.7913495083291376, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 42990 + }, + { + "epoch": 0.42991, + "grad_norm": 0.7688930938997439, + "learning_rate": 0.003, + "loss": 3.9338, + "step": 42991 + }, + { + "epoch": 0.42992, + "grad_norm": 0.6675347834136658, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 42992 + }, + { + "epoch": 0.42993, + "grad_norm": 0.6961716132217388, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 42993 + }, + { + "epoch": 0.42994, + "grad_norm": 0.7278104280283176, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 42994 + }, + { + "epoch": 0.42995, + "grad_norm": 0.6700977377623458, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 42995 + }, + { + "epoch": 0.42996, + "grad_norm": 0.6187305999214324, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 42996 + }, + { + "epoch": 0.42997, + "grad_norm": 0.6397285952738004, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 42997 + }, + { + "epoch": 0.42998, + "grad_norm": 0.7230792959514348, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 42998 + }, + { + "epoch": 0.42999, + "grad_norm": 0.7251421507072868, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 42999 + }, + { + "epoch": 0.43, + "grad_norm": 0.8034571261197059, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 43000 + }, + { + "epoch": 0.43001, + "grad_norm": 0.8293863190106378, + "learning_rate": 0.003, + "loss": 3.984, + "step": 43001 + }, + { + "epoch": 0.43002, + "grad_norm": 0.8859727262755622, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 43002 + }, + { + "epoch": 0.43003, + "grad_norm": 0.9163239295841934, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 43003 + }, + { + "epoch": 0.43004, + "grad_norm": 0.9376524298854403, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 43004 + }, + { + "epoch": 0.43005, + "grad_norm": 0.8844604355097421, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 43005 + }, + { + "epoch": 0.43006, + "grad_norm": 0.8254388212478327, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 43006 + }, + { + "epoch": 0.43007, + "grad_norm": 0.9283074455668122, + "learning_rate": 0.003, + "loss": 3.979, + "step": 43007 + }, + { + "epoch": 0.43008, + "grad_norm": 1.1352265124200211, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 43008 + }, + { + "epoch": 0.43009, + "grad_norm": 0.8874727938965106, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 43009 + }, + { + "epoch": 0.4301, + "grad_norm": 0.8263268998459952, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 43010 + }, + { + "epoch": 0.43011, + "grad_norm": 0.8591838913232108, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 43011 + }, + { + "epoch": 0.43012, + "grad_norm": 0.9018791193435436, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 43012 + }, + { + "epoch": 0.43013, + "grad_norm": 1.0119431379451893, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 43013 + }, + { + "epoch": 0.43014, + "grad_norm": 0.930920123005953, + "learning_rate": 0.003, + "loss": 3.99, + "step": 43014 + }, + { + "epoch": 0.43015, + "grad_norm": 0.9993193675507704, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 43015 + }, + { + "epoch": 0.43016, + "grad_norm": 0.7254311934485531, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 43016 + }, + { + "epoch": 0.43017, + "grad_norm": 0.6795966958449446, + "learning_rate": 0.003, + "loss": 3.996, + "step": 43017 + }, + { + "epoch": 0.43018, + "grad_norm": 0.6576646324707297, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 43018 + }, + { + "epoch": 0.43019, + "grad_norm": 0.7755782375078683, + "learning_rate": 0.003, + "loss": 3.988, + "step": 43019 + }, + { + "epoch": 0.4302, + "grad_norm": 0.8358463457095796, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 43020 + }, + { + "epoch": 0.43021, + "grad_norm": 0.8527074417738841, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 43021 + }, + { + "epoch": 0.43022, + "grad_norm": 0.7408956510643776, + "learning_rate": 0.003, + "loss": 3.983, + "step": 43022 + }, + { + "epoch": 0.43023, + "grad_norm": 0.6944864383655597, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 43023 + }, + { + "epoch": 0.43024, + "grad_norm": 0.7889098120725182, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 43024 + }, + { + "epoch": 0.43025, + "grad_norm": 0.8720585322624763, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 43025 + }, + { + "epoch": 0.43026, + "grad_norm": 0.9456433073809545, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 43026 + }, + { + "epoch": 0.43027, + "grad_norm": 0.8661534673895224, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 43027 + }, + { + "epoch": 0.43028, + "grad_norm": 0.7408007118882123, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 43028 + }, + { + "epoch": 0.43029, + "grad_norm": 0.5956765029389607, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 43029 + }, + { + "epoch": 0.4303, + "grad_norm": 0.5990639077859943, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 43030 + }, + { + "epoch": 0.43031, + "grad_norm": 0.5845844986243132, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 43031 + }, + { + "epoch": 0.43032, + "grad_norm": 0.5669838483799643, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 43032 + }, + { + "epoch": 0.43033, + "grad_norm": 0.6019781846906472, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 43033 + }, + { + "epoch": 0.43034, + "grad_norm": 0.5413588278735368, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 43034 + }, + { + "epoch": 0.43035, + "grad_norm": 0.5823718654223391, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 43035 + }, + { + "epoch": 0.43036, + "grad_norm": 0.5953790159092107, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 43036 + }, + { + "epoch": 0.43037, + "grad_norm": 0.6312030973669105, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 43037 + }, + { + "epoch": 0.43038, + "grad_norm": 0.7233513572234083, + "learning_rate": 0.003, + "loss": 3.95, + "step": 43038 + }, + { + "epoch": 0.43039, + "grad_norm": 0.8146238395627764, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 43039 + }, + { + "epoch": 0.4304, + "grad_norm": 0.8886461346683813, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 43040 + }, + { + "epoch": 0.43041, + "grad_norm": 1.044066887751673, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 43041 + }, + { + "epoch": 0.43042, + "grad_norm": 0.9063170043386094, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 43042 + }, + { + "epoch": 0.43043, + "grad_norm": 0.9244865847850016, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 43043 + }, + { + "epoch": 0.43044, + "grad_norm": 0.9767785447065968, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 43044 + }, + { + "epoch": 0.43045, + "grad_norm": 1.18757540475907, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 43045 + }, + { + "epoch": 0.43046, + "grad_norm": 0.8792651235449999, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 43046 + }, + { + "epoch": 0.43047, + "grad_norm": 0.8418367742525815, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 43047 + }, + { + "epoch": 0.43048, + "grad_norm": 0.8055846496390335, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 43048 + }, + { + "epoch": 0.43049, + "grad_norm": 0.7756788076935119, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 43049 + }, + { + "epoch": 0.4305, + "grad_norm": 0.7748992032820918, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 43050 + }, + { + "epoch": 0.43051, + "grad_norm": 0.6489516388959522, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 43051 + }, + { + "epoch": 0.43052, + "grad_norm": 0.7790867193477427, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 43052 + }, + { + "epoch": 0.43053, + "grad_norm": 0.7971767203522468, + "learning_rate": 0.003, + "loss": 4.002, + "step": 43053 + }, + { + "epoch": 0.43054, + "grad_norm": 0.8461445876882029, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 43054 + }, + { + "epoch": 0.43055, + "grad_norm": 0.7420250462977535, + "learning_rate": 0.003, + "loss": 3.974, + "step": 43055 + }, + { + "epoch": 0.43056, + "grad_norm": 0.6760005638960015, + "learning_rate": 0.003, + "loss": 3.997, + "step": 43056 + }, + { + "epoch": 0.43057, + "grad_norm": 0.7326973686821957, + "learning_rate": 0.003, + "loss": 3.9421, + "step": 43057 + }, + { + "epoch": 0.43058, + "grad_norm": 0.7029975671461136, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 43058 + }, + { + "epoch": 0.43059, + "grad_norm": 0.7977088544582732, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 43059 + }, + { + "epoch": 0.4306, + "grad_norm": 0.9296181649315239, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 43060 + }, + { + "epoch": 0.43061, + "grad_norm": 0.9680080163184775, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 43061 + }, + { + "epoch": 0.43062, + "grad_norm": 0.9876181342746472, + "learning_rate": 0.003, + "loss": 3.99, + "step": 43062 + }, + { + "epoch": 0.43063, + "grad_norm": 0.9501293594702396, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 43063 + }, + { + "epoch": 0.43064, + "grad_norm": 0.8008838589118601, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 43064 + }, + { + "epoch": 0.43065, + "grad_norm": 0.6558071541383405, + "learning_rate": 0.003, + "loss": 3.978, + "step": 43065 + }, + { + "epoch": 0.43066, + "grad_norm": 0.6127617545163739, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 43066 + }, + { + "epoch": 0.43067, + "grad_norm": 0.5755755523267729, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 43067 + }, + { + "epoch": 0.43068, + "grad_norm": 0.713363369681702, + "learning_rate": 0.003, + "loss": 4.008, + "step": 43068 + }, + { + "epoch": 0.43069, + "grad_norm": 0.8098902860308114, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 43069 + }, + { + "epoch": 0.4307, + "grad_norm": 0.8860604680220123, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 43070 + }, + { + "epoch": 0.43071, + "grad_norm": 0.803320373758162, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 43071 + }, + { + "epoch": 0.43072, + "grad_norm": 0.6830196638865913, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 43072 + }, + { + "epoch": 0.43073, + "grad_norm": 0.7112163049299332, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 43073 + }, + { + "epoch": 0.43074, + "grad_norm": 0.8247160142959669, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 43074 + }, + { + "epoch": 0.43075, + "grad_norm": 0.8369427886736763, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 43075 + }, + { + "epoch": 0.43076, + "grad_norm": 0.7531001007749052, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 43076 + }, + { + "epoch": 0.43077, + "grad_norm": 0.6898532254681089, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 43077 + }, + { + "epoch": 0.43078, + "grad_norm": 0.6369888224967127, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 43078 + }, + { + "epoch": 0.43079, + "grad_norm": 0.5533531890750084, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 43079 + }, + { + "epoch": 0.4308, + "grad_norm": 0.5973323404118309, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 43080 + }, + { + "epoch": 0.43081, + "grad_norm": 0.5888884913337065, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 43081 + }, + { + "epoch": 0.43082, + "grad_norm": 0.5199627485970286, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 43082 + }, + { + "epoch": 0.43083, + "grad_norm": 0.5526852458865058, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 43083 + }, + { + "epoch": 0.43084, + "grad_norm": 0.5655281717655993, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 43084 + }, + { + "epoch": 0.43085, + "grad_norm": 0.5935490527126694, + "learning_rate": 0.003, + "loss": 3.997, + "step": 43085 + }, + { + "epoch": 0.43086, + "grad_norm": 0.6540818770705763, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 43086 + }, + { + "epoch": 0.43087, + "grad_norm": 0.6828575779911497, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 43087 + }, + { + "epoch": 0.43088, + "grad_norm": 0.7469690045804755, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 43088 + }, + { + "epoch": 0.43089, + "grad_norm": 0.834859565560693, + "learning_rate": 0.003, + "loss": 3.9376, + "step": 43089 + }, + { + "epoch": 0.4309, + "grad_norm": 0.9909529727526496, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 43090 + }, + { + "epoch": 0.43091, + "grad_norm": 1.1293301568234104, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 43091 + }, + { + "epoch": 0.43092, + "grad_norm": 0.8819798972889686, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 43092 + }, + { + "epoch": 0.43093, + "grad_norm": 0.7918597964238993, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 43093 + }, + { + "epoch": 0.43094, + "grad_norm": 0.7102947833827001, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 43094 + }, + { + "epoch": 0.43095, + "grad_norm": 0.8149273371199439, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 43095 + }, + { + "epoch": 0.43096, + "grad_norm": 0.9293658312280985, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 43096 + }, + { + "epoch": 0.43097, + "grad_norm": 1.0700099715629752, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 43097 + }, + { + "epoch": 0.43098, + "grad_norm": 0.8531200487810259, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 43098 + }, + { + "epoch": 0.43099, + "grad_norm": 0.6836344386353579, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 43099 + }, + { + "epoch": 0.431, + "grad_norm": 0.7311555110124958, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 43100 + }, + { + "epoch": 0.43101, + "grad_norm": 0.7334674617765577, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 43101 + }, + { + "epoch": 0.43102, + "grad_norm": 0.7530933099589429, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 43102 + }, + { + "epoch": 0.43103, + "grad_norm": 0.7504972653709854, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 43103 + }, + { + "epoch": 0.43104, + "grad_norm": 0.7439687916893158, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 43104 + }, + { + "epoch": 0.43105, + "grad_norm": 0.7173963541386473, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 43105 + }, + { + "epoch": 0.43106, + "grad_norm": 0.6038726745014258, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 43106 + }, + { + "epoch": 0.43107, + "grad_norm": 0.6446692051147712, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 43107 + }, + { + "epoch": 0.43108, + "grad_norm": 0.7333869740951318, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 43108 + }, + { + "epoch": 0.43109, + "grad_norm": 0.8281233301834561, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 43109 + }, + { + "epoch": 0.4311, + "grad_norm": 1.0036591642160981, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 43110 + }, + { + "epoch": 0.43111, + "grad_norm": 1.1313238248213369, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 43111 + }, + { + "epoch": 0.43112, + "grad_norm": 0.8950518792948805, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 43112 + }, + { + "epoch": 0.43113, + "grad_norm": 0.9300478468780029, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 43113 + }, + { + "epoch": 0.43114, + "grad_norm": 0.9693961770201437, + "learning_rate": 0.003, + "loss": 3.971, + "step": 43114 + }, + { + "epoch": 0.43115, + "grad_norm": 0.9857667405071333, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 43115 + }, + { + "epoch": 0.43116, + "grad_norm": 0.9851690097588262, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 43116 + }, + { + "epoch": 0.43117, + "grad_norm": 0.8987571492531955, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 43117 + }, + { + "epoch": 0.43118, + "grad_norm": 0.9234269385760332, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 43118 + }, + { + "epoch": 0.43119, + "grad_norm": 1.0020672189049367, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 43119 + }, + { + "epoch": 0.4312, + "grad_norm": 1.0469882626546834, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 43120 + }, + { + "epoch": 0.43121, + "grad_norm": 0.9024821257874663, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 43121 + }, + { + "epoch": 0.43122, + "grad_norm": 0.8540463575010997, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 43122 + }, + { + "epoch": 0.43123, + "grad_norm": 0.9626863275910058, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 43123 + }, + { + "epoch": 0.43124, + "grad_norm": 0.8575845786817253, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 43124 + }, + { + "epoch": 0.43125, + "grad_norm": 0.7900430185781261, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 43125 + }, + { + "epoch": 0.43126, + "grad_norm": 0.8591073334299987, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 43126 + }, + { + "epoch": 0.43127, + "grad_norm": 0.7538021549603346, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 43127 + }, + { + "epoch": 0.43128, + "grad_norm": 0.745303883445532, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 43128 + }, + { + "epoch": 0.43129, + "grad_norm": 0.6348515804571497, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 43129 + }, + { + "epoch": 0.4313, + "grad_norm": 0.6986905503173838, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 43130 + }, + { + "epoch": 0.43131, + "grad_norm": 0.6949663917731257, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 43131 + }, + { + "epoch": 0.43132, + "grad_norm": 0.6432652987794906, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 43132 + }, + { + "epoch": 0.43133, + "grad_norm": 0.6140544204314525, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 43133 + }, + { + "epoch": 0.43134, + "grad_norm": 0.6017956655071304, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 43134 + }, + { + "epoch": 0.43135, + "grad_norm": 0.6361073762015106, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 43135 + }, + { + "epoch": 0.43136, + "grad_norm": 0.6641735419486688, + "learning_rate": 0.003, + "loss": 3.9397, + "step": 43136 + }, + { + "epoch": 0.43137, + "grad_norm": 0.6766033995842927, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 43137 + }, + { + "epoch": 0.43138, + "grad_norm": 0.6577919228324158, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 43138 + }, + { + "epoch": 0.43139, + "grad_norm": 0.6819362480576496, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 43139 + }, + { + "epoch": 0.4314, + "grad_norm": 0.8164273714799094, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 43140 + }, + { + "epoch": 0.43141, + "grad_norm": 0.9612668739649964, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 43141 + }, + { + "epoch": 0.43142, + "grad_norm": 1.095461592672935, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 43142 + }, + { + "epoch": 0.43143, + "grad_norm": 0.7036130916160781, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 43143 + }, + { + "epoch": 0.43144, + "grad_norm": 0.6947955979746302, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 43144 + }, + { + "epoch": 0.43145, + "grad_norm": 0.891233908647761, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 43145 + }, + { + "epoch": 0.43146, + "grad_norm": 0.8366754941928825, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 43146 + }, + { + "epoch": 0.43147, + "grad_norm": 0.7999044742232425, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 43147 + }, + { + "epoch": 0.43148, + "grad_norm": 0.8517855741219957, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 43148 + }, + { + "epoch": 0.43149, + "grad_norm": 0.6926224246895295, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 43149 + }, + { + "epoch": 0.4315, + "grad_norm": 0.6659651723619566, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 43150 + }, + { + "epoch": 0.43151, + "grad_norm": 0.7012129163758193, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 43151 + }, + { + "epoch": 0.43152, + "grad_norm": 0.645772749112144, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 43152 + }, + { + "epoch": 0.43153, + "grad_norm": 0.6970237142968715, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 43153 + }, + { + "epoch": 0.43154, + "grad_norm": 0.6807662514263947, + "learning_rate": 0.003, + "loss": 3.967, + "step": 43154 + }, + { + "epoch": 0.43155, + "grad_norm": 0.6210752902600611, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 43155 + }, + { + "epoch": 0.43156, + "grad_norm": 0.6757095436372043, + "learning_rate": 0.003, + "loss": 3.98, + "step": 43156 + }, + { + "epoch": 0.43157, + "grad_norm": 0.7293326910423157, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 43157 + }, + { + "epoch": 0.43158, + "grad_norm": 0.6440810780072168, + "learning_rate": 0.003, + "loss": 4.0, + "step": 43158 + }, + { + "epoch": 0.43159, + "grad_norm": 0.6091108739928344, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 43159 + }, + { + "epoch": 0.4316, + "grad_norm": 0.6619714163885599, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 43160 + }, + { + "epoch": 0.43161, + "grad_norm": 0.7362975655411932, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 43161 + }, + { + "epoch": 0.43162, + "grad_norm": 0.830319343325416, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 43162 + }, + { + "epoch": 0.43163, + "grad_norm": 0.9289611298129088, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 43163 + }, + { + "epoch": 0.43164, + "grad_norm": 1.021037130538735, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 43164 + }, + { + "epoch": 0.43165, + "grad_norm": 1.0125048991566121, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 43165 + }, + { + "epoch": 0.43166, + "grad_norm": 0.8844914927148512, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 43166 + }, + { + "epoch": 0.43167, + "grad_norm": 0.648271465176242, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 43167 + }, + { + "epoch": 0.43168, + "grad_norm": 0.6701371610071999, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 43168 + }, + { + "epoch": 0.43169, + "grad_norm": 0.5855066084622741, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 43169 + }, + { + "epoch": 0.4317, + "grad_norm": 0.6053292029095877, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 43170 + }, + { + "epoch": 0.43171, + "grad_norm": 0.6329151438083149, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 43171 + }, + { + "epoch": 0.43172, + "grad_norm": 0.6897656154240174, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 43172 + }, + { + "epoch": 0.43173, + "grad_norm": 0.7870625855417067, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 43173 + }, + { + "epoch": 0.43174, + "grad_norm": 0.7999388129458206, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 43174 + }, + { + "epoch": 0.43175, + "grad_norm": 0.7603015783699805, + "learning_rate": 0.003, + "loss": 3.976, + "step": 43175 + }, + { + "epoch": 0.43176, + "grad_norm": 0.782175525094012, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 43176 + }, + { + "epoch": 0.43177, + "grad_norm": 0.8086913118930682, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 43177 + }, + { + "epoch": 0.43178, + "grad_norm": 0.719686415043592, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 43178 + }, + { + "epoch": 0.43179, + "grad_norm": 0.7104153095584382, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 43179 + }, + { + "epoch": 0.4318, + "grad_norm": 0.7683132613187239, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 43180 + }, + { + "epoch": 0.43181, + "grad_norm": 0.9104799287966391, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 43181 + }, + { + "epoch": 0.43182, + "grad_norm": 0.9965039322640835, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 43182 + }, + { + "epoch": 0.43183, + "grad_norm": 0.9815302998293997, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 43183 + }, + { + "epoch": 0.43184, + "grad_norm": 0.880418400584105, + "learning_rate": 0.003, + "loss": 4.011, + "step": 43184 + }, + { + "epoch": 0.43185, + "grad_norm": 0.9040978768646017, + "learning_rate": 0.003, + "loss": 3.98, + "step": 43185 + }, + { + "epoch": 0.43186, + "grad_norm": 0.8639574397355609, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 43186 + }, + { + "epoch": 0.43187, + "grad_norm": 1.0000429319046225, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 43187 + }, + { + "epoch": 0.43188, + "grad_norm": 1.0353743056090803, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 43188 + }, + { + "epoch": 0.43189, + "grad_norm": 0.880082274826556, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 43189 + }, + { + "epoch": 0.4319, + "grad_norm": 0.9468211019097236, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 43190 + }, + { + "epoch": 0.43191, + "grad_norm": 1.1124868360167501, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 43191 + }, + { + "epoch": 0.43192, + "grad_norm": 0.9220519057342899, + "learning_rate": 0.003, + "loss": 4.011, + "step": 43192 + }, + { + "epoch": 0.43193, + "grad_norm": 0.8823184979479788, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 43193 + }, + { + "epoch": 0.43194, + "grad_norm": 0.9268576629594175, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 43194 + }, + { + "epoch": 0.43195, + "grad_norm": 0.8706733089290907, + "learning_rate": 0.003, + "loss": 4.003, + "step": 43195 + }, + { + "epoch": 0.43196, + "grad_norm": 0.955906599691598, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 43196 + }, + { + "epoch": 0.43197, + "grad_norm": 0.9931101674214303, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 43197 + }, + { + "epoch": 0.43198, + "grad_norm": 0.9688733013816307, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 43198 + }, + { + "epoch": 0.43199, + "grad_norm": 0.9079085989626626, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 43199 + }, + { + "epoch": 0.432, + "grad_norm": 0.8931141049921901, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 43200 + }, + { + "epoch": 0.43201, + "grad_norm": 0.7864678611165954, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 43201 + }, + { + "epoch": 0.43202, + "grad_norm": 0.7531555048614201, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 43202 + }, + { + "epoch": 0.43203, + "grad_norm": 0.7095774328239233, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 43203 + }, + { + "epoch": 0.43204, + "grad_norm": 0.6902401096188683, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 43204 + }, + { + "epoch": 0.43205, + "grad_norm": 0.6668607074149868, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 43205 + }, + { + "epoch": 0.43206, + "grad_norm": 0.6279399283317547, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 43206 + }, + { + "epoch": 0.43207, + "grad_norm": 0.5521001554927187, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 43207 + }, + { + "epoch": 0.43208, + "grad_norm": 0.5738529788967043, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 43208 + }, + { + "epoch": 0.43209, + "grad_norm": 0.5668777437923401, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 43209 + }, + { + "epoch": 0.4321, + "grad_norm": 0.549575644318125, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 43210 + }, + { + "epoch": 0.43211, + "grad_norm": 0.6373073546490104, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 43211 + }, + { + "epoch": 0.43212, + "grad_norm": 0.6371776659821714, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 43212 + }, + { + "epoch": 0.43213, + "grad_norm": 0.7284301372475395, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 43213 + }, + { + "epoch": 0.43214, + "grad_norm": 0.8179735755969655, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 43214 + }, + { + "epoch": 0.43215, + "grad_norm": 0.9820440227890963, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 43215 + }, + { + "epoch": 0.43216, + "grad_norm": 0.855123110830178, + "learning_rate": 0.003, + "loss": 3.956, + "step": 43216 + }, + { + "epoch": 0.43217, + "grad_norm": 0.6329428384501661, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 43217 + }, + { + "epoch": 0.43218, + "grad_norm": 0.61470684670075, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 43218 + }, + { + "epoch": 0.43219, + "grad_norm": 0.6654230650717359, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 43219 + }, + { + "epoch": 0.4322, + "grad_norm": 0.8071302327608009, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 43220 + }, + { + "epoch": 0.43221, + "grad_norm": 0.8548308250508228, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 43221 + }, + { + "epoch": 0.43222, + "grad_norm": 0.7597695553740328, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 43222 + }, + { + "epoch": 0.43223, + "grad_norm": 0.6433644946019969, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 43223 + }, + { + "epoch": 0.43224, + "grad_norm": 0.6159181467353465, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 43224 + }, + { + "epoch": 0.43225, + "grad_norm": 0.5563238908225163, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 43225 + }, + { + "epoch": 0.43226, + "grad_norm": 0.5219839958778246, + "learning_rate": 0.003, + "loss": 3.9464, + "step": 43226 + }, + { + "epoch": 0.43227, + "grad_norm": 0.5853869454508118, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 43227 + }, + { + "epoch": 0.43228, + "grad_norm": 0.6433935472201091, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 43228 + }, + { + "epoch": 0.43229, + "grad_norm": 0.6200328258317219, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 43229 + }, + { + "epoch": 0.4323, + "grad_norm": 0.6645763815512418, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 43230 + }, + { + "epoch": 0.43231, + "grad_norm": 0.7594177000061768, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 43231 + }, + { + "epoch": 0.43232, + "grad_norm": 0.8861042249317433, + "learning_rate": 0.003, + "loss": 3.977, + "step": 43232 + }, + { + "epoch": 0.43233, + "grad_norm": 0.9343598420335733, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 43233 + }, + { + "epoch": 0.43234, + "grad_norm": 1.1002255627156106, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 43234 + }, + { + "epoch": 0.43235, + "grad_norm": 1.010012699626753, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 43235 + }, + { + "epoch": 0.43236, + "grad_norm": 1.0875504564145477, + "learning_rate": 0.003, + "loss": 4.002, + "step": 43236 + }, + { + "epoch": 0.43237, + "grad_norm": 0.9232549136726126, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 43237 + }, + { + "epoch": 0.43238, + "grad_norm": 0.814249256741327, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 43238 + }, + { + "epoch": 0.43239, + "grad_norm": 0.7672440805560805, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 43239 + }, + { + "epoch": 0.4324, + "grad_norm": 0.7289076674535576, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 43240 + }, + { + "epoch": 0.43241, + "grad_norm": 0.7489560653062077, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 43241 + }, + { + "epoch": 0.43242, + "grad_norm": 0.8621619657780502, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 43242 + }, + { + "epoch": 0.43243, + "grad_norm": 0.950266483972658, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 43243 + }, + { + "epoch": 0.43244, + "grad_norm": 0.9314781402539818, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 43244 + }, + { + "epoch": 0.43245, + "grad_norm": 0.7845167209151446, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 43245 + }, + { + "epoch": 0.43246, + "grad_norm": 0.7664763437837929, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 43246 + }, + { + "epoch": 0.43247, + "grad_norm": 0.7711538668992038, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 43247 + }, + { + "epoch": 0.43248, + "grad_norm": 0.8092059051279998, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 43248 + }, + { + "epoch": 0.43249, + "grad_norm": 0.8232046579706247, + "learning_rate": 0.003, + "loss": 3.975, + "step": 43249 + }, + { + "epoch": 0.4325, + "grad_norm": 0.8695147584341018, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 43250 + }, + { + "epoch": 0.43251, + "grad_norm": 0.816054153879278, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 43251 + }, + { + "epoch": 0.43252, + "grad_norm": 0.7643985277507735, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 43252 + }, + { + "epoch": 0.43253, + "grad_norm": 0.808312706139097, + "learning_rate": 0.003, + "loss": 3.986, + "step": 43253 + }, + { + "epoch": 0.43254, + "grad_norm": 0.758824684573195, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 43254 + }, + { + "epoch": 0.43255, + "grad_norm": 0.631425261881137, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 43255 + }, + { + "epoch": 0.43256, + "grad_norm": 0.6905281595012662, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 43256 + }, + { + "epoch": 0.43257, + "grad_norm": 0.7650227446915534, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 43257 + }, + { + "epoch": 0.43258, + "grad_norm": 0.7006756339188184, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 43258 + }, + { + "epoch": 0.43259, + "grad_norm": 0.7291037870304244, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 43259 + }, + { + "epoch": 0.4326, + "grad_norm": 0.7131545704620065, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 43260 + }, + { + "epoch": 0.43261, + "grad_norm": 0.6988810115821977, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 43261 + }, + { + "epoch": 0.43262, + "grad_norm": 0.7016829050677116, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 43262 + }, + { + "epoch": 0.43263, + "grad_norm": 0.6788248530876663, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 43263 + }, + { + "epoch": 0.43264, + "grad_norm": 0.8091974507210135, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 43264 + }, + { + "epoch": 0.43265, + "grad_norm": 0.9078017859311094, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 43265 + }, + { + "epoch": 0.43266, + "grad_norm": 0.9547320162829113, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 43266 + }, + { + "epoch": 0.43267, + "grad_norm": 0.9898356109866657, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 43267 + }, + { + "epoch": 0.43268, + "grad_norm": 1.0168868464079557, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 43268 + }, + { + "epoch": 0.43269, + "grad_norm": 0.8865434518081738, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 43269 + }, + { + "epoch": 0.4327, + "grad_norm": 0.7669259318680047, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 43270 + }, + { + "epoch": 0.43271, + "grad_norm": 0.7152142499807335, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 43271 + }, + { + "epoch": 0.43272, + "grad_norm": 0.6136965555336451, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 43272 + }, + { + "epoch": 0.43273, + "grad_norm": 0.6142005037709952, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 43273 + }, + { + "epoch": 0.43274, + "grad_norm": 0.748336934769258, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 43274 + }, + { + "epoch": 0.43275, + "grad_norm": 0.853870444974362, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 43275 + }, + { + "epoch": 0.43276, + "grad_norm": 0.8914694407421192, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 43276 + }, + { + "epoch": 0.43277, + "grad_norm": 0.7931138962668342, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 43277 + }, + { + "epoch": 0.43278, + "grad_norm": 0.8078904556354187, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 43278 + }, + { + "epoch": 0.43279, + "grad_norm": 0.8024671646087858, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 43279 + }, + { + "epoch": 0.4328, + "grad_norm": 0.7587535015571385, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 43280 + }, + { + "epoch": 0.43281, + "grad_norm": 0.6711481469908009, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 43281 + }, + { + "epoch": 0.43282, + "grad_norm": 0.6691231149519036, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 43282 + }, + { + "epoch": 0.43283, + "grad_norm": 0.6136355322617021, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 43283 + }, + { + "epoch": 0.43284, + "grad_norm": 0.6651627841046448, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 43284 + }, + { + "epoch": 0.43285, + "grad_norm": 0.6841077175692057, + "learning_rate": 0.003, + "loss": 4.013, + "step": 43285 + }, + { + "epoch": 0.43286, + "grad_norm": 0.7135502151613928, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 43286 + }, + { + "epoch": 0.43287, + "grad_norm": 0.6732983079589339, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 43287 + }, + { + "epoch": 0.43288, + "grad_norm": 0.6319857706875375, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 43288 + }, + { + "epoch": 0.43289, + "grad_norm": 0.6150031591311307, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 43289 + }, + { + "epoch": 0.4329, + "grad_norm": 0.64969419196901, + "learning_rate": 0.003, + "loss": 3.9465, + "step": 43290 + }, + { + "epoch": 0.43291, + "grad_norm": 0.7345837721187765, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 43291 + }, + { + "epoch": 0.43292, + "grad_norm": 0.8104285451384448, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 43292 + }, + { + "epoch": 0.43293, + "grad_norm": 0.8456529709812661, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 43293 + }, + { + "epoch": 0.43294, + "grad_norm": 0.9716952653288143, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 43294 + }, + { + "epoch": 0.43295, + "grad_norm": 1.112877600520348, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43295 + }, + { + "epoch": 0.43296, + "grad_norm": 0.8470432671065546, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 43296 + }, + { + "epoch": 0.43297, + "grad_norm": 0.8559303383867456, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 43297 + }, + { + "epoch": 0.43298, + "grad_norm": 1.0178056228393624, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 43298 + }, + { + "epoch": 0.43299, + "grad_norm": 1.0435017078464055, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 43299 + }, + { + "epoch": 0.433, + "grad_norm": 0.9351378412217833, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 43300 + }, + { + "epoch": 0.43301, + "grad_norm": 0.9251522193907397, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 43301 + }, + { + "epoch": 0.43302, + "grad_norm": 1.0387183781279488, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 43302 + }, + { + "epoch": 0.43303, + "grad_norm": 0.90945862078298, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 43303 + }, + { + "epoch": 0.43304, + "grad_norm": 0.7263641005717245, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 43304 + }, + { + "epoch": 0.43305, + "grad_norm": 0.6724815155791, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 43305 + }, + { + "epoch": 0.43306, + "grad_norm": 0.7299641033717397, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 43306 + }, + { + "epoch": 0.43307, + "grad_norm": 0.6803420320925684, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 43307 + }, + { + "epoch": 0.43308, + "grad_norm": 0.6218287698397336, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 43308 + }, + { + "epoch": 0.43309, + "grad_norm": 0.6025616950368624, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 43309 + }, + { + "epoch": 0.4331, + "grad_norm": 0.6999647551514248, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 43310 + }, + { + "epoch": 0.43311, + "grad_norm": 0.6255943846974075, + "learning_rate": 0.003, + "loss": 3.99, + "step": 43311 + }, + { + "epoch": 0.43312, + "grad_norm": 0.5296669009236749, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 43312 + }, + { + "epoch": 0.43313, + "grad_norm": 0.5982185350768816, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 43313 + }, + { + "epoch": 0.43314, + "grad_norm": 0.6154243162682173, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 43314 + }, + { + "epoch": 0.43315, + "grad_norm": 0.5806475030047682, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 43315 + }, + { + "epoch": 0.43316, + "grad_norm": 0.48666596998405365, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 43316 + }, + { + "epoch": 0.43317, + "grad_norm": 0.4879268016375523, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 43317 + }, + { + "epoch": 0.43318, + "grad_norm": 0.5332199790962101, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 43318 + }, + { + "epoch": 0.43319, + "grad_norm": 0.5432925791339039, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 43319 + }, + { + "epoch": 0.4332, + "grad_norm": 0.5809548028292184, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 43320 + }, + { + "epoch": 0.43321, + "grad_norm": 0.6202959662141351, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 43321 + }, + { + "epoch": 0.43322, + "grad_norm": 0.6740610425598932, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 43322 + }, + { + "epoch": 0.43323, + "grad_norm": 0.7481006957937766, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 43323 + }, + { + "epoch": 0.43324, + "grad_norm": 0.943838121613539, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 43324 + }, + { + "epoch": 0.43325, + "grad_norm": 1.127547879609285, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 43325 + }, + { + "epoch": 0.43326, + "grad_norm": 0.9713431841661541, + "learning_rate": 0.003, + "loss": 3.976, + "step": 43326 + }, + { + "epoch": 0.43327, + "grad_norm": 0.9289946076541219, + "learning_rate": 0.003, + "loss": 3.985, + "step": 43327 + }, + { + "epoch": 0.43328, + "grad_norm": 0.970260235276846, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 43328 + }, + { + "epoch": 0.43329, + "grad_norm": 0.9907582945355793, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 43329 + }, + { + "epoch": 0.4333, + "grad_norm": 0.9190275732294683, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 43330 + }, + { + "epoch": 0.43331, + "grad_norm": 0.8679342240097097, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 43331 + }, + { + "epoch": 0.43332, + "grad_norm": 0.9701757209151153, + "learning_rate": 0.003, + "loss": 4.0, + "step": 43332 + }, + { + "epoch": 0.43333, + "grad_norm": 0.9730934033687824, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 43333 + }, + { + "epoch": 0.43334, + "grad_norm": 1.011594914210081, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 43334 + }, + { + "epoch": 0.43335, + "grad_norm": 0.8970525818374832, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 43335 + }, + { + "epoch": 0.43336, + "grad_norm": 0.8481932478925015, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 43336 + }, + { + "epoch": 0.43337, + "grad_norm": 0.967315999603055, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 43337 + }, + { + "epoch": 0.43338, + "grad_norm": 0.9020426293148918, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 43338 + }, + { + "epoch": 0.43339, + "grad_norm": 0.7596321646539638, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 43339 + }, + { + "epoch": 0.4334, + "grad_norm": 0.7367355389085922, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 43340 + }, + { + "epoch": 0.43341, + "grad_norm": 0.7048156904837553, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 43341 + }, + { + "epoch": 0.43342, + "grad_norm": 0.80572213714997, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 43342 + }, + { + "epoch": 0.43343, + "grad_norm": 0.8613341000715827, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 43343 + }, + { + "epoch": 0.43344, + "grad_norm": 0.8305106527735248, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 43344 + }, + { + "epoch": 0.43345, + "grad_norm": 0.9050501101347369, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 43345 + }, + { + "epoch": 0.43346, + "grad_norm": 0.925119798062499, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 43346 + }, + { + "epoch": 0.43347, + "grad_norm": 0.9881919302311843, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 43347 + }, + { + "epoch": 0.43348, + "grad_norm": 0.8923063148775416, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 43348 + }, + { + "epoch": 0.43349, + "grad_norm": 0.8199008987180338, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 43349 + }, + { + "epoch": 0.4335, + "grad_norm": 0.8437602110651536, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 43350 + }, + { + "epoch": 0.43351, + "grad_norm": 0.8349264732799903, + "learning_rate": 0.003, + "loss": 4.0194, + "step": 43351 + }, + { + "epoch": 0.43352, + "grad_norm": 0.6166330380021483, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 43352 + }, + { + "epoch": 0.43353, + "grad_norm": 0.529574348652089, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 43353 + }, + { + "epoch": 0.43354, + "grad_norm": 0.5319699385478966, + "learning_rate": 0.003, + "loss": 3.933, + "step": 43354 + }, + { + "epoch": 0.43355, + "grad_norm": 0.5774082872276294, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 43355 + }, + { + "epoch": 0.43356, + "grad_norm": 0.5568652490264814, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 43356 + }, + { + "epoch": 0.43357, + "grad_norm": 0.5643471432504993, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 43357 + }, + { + "epoch": 0.43358, + "grad_norm": 0.6006240808907324, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 43358 + }, + { + "epoch": 0.43359, + "grad_norm": 0.731163455824827, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 43359 + }, + { + "epoch": 0.4336, + "grad_norm": 0.8520289867494072, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 43360 + }, + { + "epoch": 0.43361, + "grad_norm": 0.7894078941683649, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 43361 + }, + { + "epoch": 0.43362, + "grad_norm": 0.7002377769383663, + "learning_rate": 0.003, + "loss": 4.017, + "step": 43362 + }, + { + "epoch": 0.43363, + "grad_norm": 0.7220019941338726, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 43363 + }, + { + "epoch": 0.43364, + "grad_norm": 0.7337626052514118, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 43364 + }, + { + "epoch": 0.43365, + "grad_norm": 0.7235170265631533, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 43365 + }, + { + "epoch": 0.43366, + "grad_norm": 0.7115122780433392, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 43366 + }, + { + "epoch": 0.43367, + "grad_norm": 0.7050644116092956, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 43367 + }, + { + "epoch": 0.43368, + "grad_norm": 0.7306181627005974, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 43368 + }, + { + "epoch": 0.43369, + "grad_norm": 0.7411393535168334, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 43369 + }, + { + "epoch": 0.4337, + "grad_norm": 0.7403422497801847, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 43370 + }, + { + "epoch": 0.43371, + "grad_norm": 0.8564130737433788, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 43371 + }, + { + "epoch": 0.43372, + "grad_norm": 0.9488102626900528, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 43372 + }, + { + "epoch": 0.43373, + "grad_norm": 0.995721436724551, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 43373 + }, + { + "epoch": 0.43374, + "grad_norm": 1.0411485030143273, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 43374 + }, + { + "epoch": 0.43375, + "grad_norm": 0.8882784731101904, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 43375 + }, + { + "epoch": 0.43376, + "grad_norm": 0.8704658337155637, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 43376 + }, + { + "epoch": 0.43377, + "grad_norm": 0.7942308235281701, + "learning_rate": 0.003, + "loss": 3.9426, + "step": 43377 + }, + { + "epoch": 0.43378, + "grad_norm": 0.7039386028424014, + "learning_rate": 0.003, + "loss": 3.973, + "step": 43378 + }, + { + "epoch": 0.43379, + "grad_norm": 0.6755334080353118, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 43379 + }, + { + "epoch": 0.4338, + "grad_norm": 0.6576328321234789, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 43380 + }, + { + "epoch": 0.43381, + "grad_norm": 0.6765656129972754, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 43381 + }, + { + "epoch": 0.43382, + "grad_norm": 0.7378097422387243, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 43382 + }, + { + "epoch": 0.43383, + "grad_norm": 0.752471408571655, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 43383 + }, + { + "epoch": 0.43384, + "grad_norm": 0.9120242098519017, + "learning_rate": 0.003, + "loss": 3.9457, + "step": 43384 + }, + { + "epoch": 0.43385, + "grad_norm": 1.0345205868293064, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 43385 + }, + { + "epoch": 0.43386, + "grad_norm": 0.9812958328527093, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 43386 + }, + { + "epoch": 0.43387, + "grad_norm": 0.9297756979158793, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 43387 + }, + { + "epoch": 0.43388, + "grad_norm": 0.7834433474687306, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 43388 + }, + { + "epoch": 0.43389, + "grad_norm": 0.7318246144795243, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 43389 + }, + { + "epoch": 0.4339, + "grad_norm": 0.6803964781639105, + "learning_rate": 0.003, + "loss": 3.985, + "step": 43390 + }, + { + "epoch": 0.43391, + "grad_norm": 0.6972343627685312, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 43391 + }, + { + "epoch": 0.43392, + "grad_norm": 0.5991049225925649, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 43392 + }, + { + "epoch": 0.43393, + "grad_norm": 0.5432950183187022, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 43393 + }, + { + "epoch": 0.43394, + "grad_norm": 0.5652816349019586, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 43394 + }, + { + "epoch": 0.43395, + "grad_norm": 0.6137561266198243, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 43395 + }, + { + "epoch": 0.43396, + "grad_norm": 0.7955037394409477, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 43396 + }, + { + "epoch": 0.43397, + "grad_norm": 1.0914616866217757, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 43397 + }, + { + "epoch": 0.43398, + "grad_norm": 1.0735779635130143, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 43398 + }, + { + "epoch": 0.43399, + "grad_norm": 1.1345466317179995, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 43399 + }, + { + "epoch": 0.434, + "grad_norm": 0.7338445443851539, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 43400 + }, + { + "epoch": 0.43401, + "grad_norm": 0.6837874487385642, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 43401 + }, + { + "epoch": 0.43402, + "grad_norm": 0.6398248342000127, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 43402 + }, + { + "epoch": 0.43403, + "grad_norm": 0.6516501261906098, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 43403 + }, + { + "epoch": 0.43404, + "grad_norm": 0.7030555201855035, + "learning_rate": 0.003, + "loss": 3.996, + "step": 43404 + }, + { + "epoch": 0.43405, + "grad_norm": 0.6183466415682815, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 43405 + }, + { + "epoch": 0.43406, + "grad_norm": 0.6399516341289458, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 43406 + }, + { + "epoch": 0.43407, + "grad_norm": 0.6687945978582363, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 43407 + }, + { + "epoch": 0.43408, + "grad_norm": 0.8021357973070088, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 43408 + }, + { + "epoch": 0.43409, + "grad_norm": 0.9088081470872988, + "learning_rate": 0.003, + "loss": 3.991, + "step": 43409 + }, + { + "epoch": 0.4341, + "grad_norm": 0.806916012892191, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 43410 + }, + { + "epoch": 0.43411, + "grad_norm": 0.7582364547951759, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 43411 + }, + { + "epoch": 0.43412, + "grad_norm": 0.80857362275993, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 43412 + }, + { + "epoch": 0.43413, + "grad_norm": 0.8839375114567358, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 43413 + }, + { + "epoch": 0.43414, + "grad_norm": 0.8877866392478364, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 43414 + }, + { + "epoch": 0.43415, + "grad_norm": 1.0237937963852828, + "learning_rate": 0.003, + "loss": 4.009, + "step": 43415 + }, + { + "epoch": 0.43416, + "grad_norm": 1.0969791389864165, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 43416 + }, + { + "epoch": 0.43417, + "grad_norm": 1.0755456810020727, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 43417 + }, + { + "epoch": 0.43418, + "grad_norm": 1.1008020607811793, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 43418 + }, + { + "epoch": 0.43419, + "grad_norm": 0.8781025159452678, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 43419 + }, + { + "epoch": 0.4342, + "grad_norm": 0.6982189670383672, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 43420 + }, + { + "epoch": 0.43421, + "grad_norm": 0.6551052734522769, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 43421 + }, + { + "epoch": 0.43422, + "grad_norm": 0.6827167348019906, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 43422 + }, + { + "epoch": 0.43423, + "grad_norm": 0.674083687542495, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 43423 + }, + { + "epoch": 0.43424, + "grad_norm": 0.7406975733084166, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 43424 + }, + { + "epoch": 0.43425, + "grad_norm": 0.8710487287913199, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 43425 + }, + { + "epoch": 0.43426, + "grad_norm": 1.0416910170684661, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 43426 + }, + { + "epoch": 0.43427, + "grad_norm": 0.8882461846722429, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 43427 + }, + { + "epoch": 0.43428, + "grad_norm": 0.7709371105841651, + "learning_rate": 0.003, + "loss": 3.993, + "step": 43428 + }, + { + "epoch": 0.43429, + "grad_norm": 0.7839737433350806, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 43429 + }, + { + "epoch": 0.4343, + "grad_norm": 0.8870432756442443, + "learning_rate": 0.003, + "loss": 3.9439, + "step": 43430 + }, + { + "epoch": 0.43431, + "grad_norm": 0.991960843079435, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 43431 + }, + { + "epoch": 0.43432, + "grad_norm": 1.03463522464142, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 43432 + }, + { + "epoch": 0.43433, + "grad_norm": 0.8835298167199234, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 43433 + }, + { + "epoch": 0.43434, + "grad_norm": 0.8545139608049598, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 43434 + }, + { + "epoch": 0.43435, + "grad_norm": 0.9342240157663834, + "learning_rate": 0.003, + "loss": 4.001, + "step": 43435 + }, + { + "epoch": 0.43436, + "grad_norm": 0.791276159982671, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 43436 + }, + { + "epoch": 0.43437, + "grad_norm": 0.7675740807880633, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 43437 + }, + { + "epoch": 0.43438, + "grad_norm": 0.7925633507380624, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 43438 + }, + { + "epoch": 0.43439, + "grad_norm": 0.81962250102518, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 43439 + }, + { + "epoch": 0.4344, + "grad_norm": 0.7236261104773494, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 43440 + }, + { + "epoch": 0.43441, + "grad_norm": 0.7204278500497384, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 43441 + }, + { + "epoch": 0.43442, + "grad_norm": 0.8382496471016154, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 43442 + }, + { + "epoch": 0.43443, + "grad_norm": 0.9169440566909091, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 43443 + }, + { + "epoch": 0.43444, + "grad_norm": 0.929639747691754, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 43444 + }, + { + "epoch": 0.43445, + "grad_norm": 0.9257788391985325, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 43445 + }, + { + "epoch": 0.43446, + "grad_norm": 0.9673473317823401, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 43446 + }, + { + "epoch": 0.43447, + "grad_norm": 0.9300111608203682, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 43447 + }, + { + "epoch": 0.43448, + "grad_norm": 0.7882676953325632, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 43448 + }, + { + "epoch": 0.43449, + "grad_norm": 0.8017294591678681, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 43449 + }, + { + "epoch": 0.4345, + "grad_norm": 0.7207965984489655, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 43450 + }, + { + "epoch": 0.43451, + "grad_norm": 0.7173345388741622, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 43451 + }, + { + "epoch": 0.43452, + "grad_norm": 0.7920718473790851, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 43452 + }, + { + "epoch": 0.43453, + "grad_norm": 0.7687487733422743, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 43453 + }, + { + "epoch": 0.43454, + "grad_norm": 0.7512671341272666, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 43454 + }, + { + "epoch": 0.43455, + "grad_norm": 0.7012231798353166, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 43455 + }, + { + "epoch": 0.43456, + "grad_norm": 0.6219023070514754, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 43456 + }, + { + "epoch": 0.43457, + "grad_norm": 0.5641593333930149, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 43457 + }, + { + "epoch": 0.43458, + "grad_norm": 0.5215136965328349, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 43458 + }, + { + "epoch": 0.43459, + "grad_norm": 0.5366982687707149, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 43459 + }, + { + "epoch": 0.4346, + "grad_norm": 0.5839754725013601, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 43460 + }, + { + "epoch": 0.43461, + "grad_norm": 0.6019790710518664, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 43461 + }, + { + "epoch": 0.43462, + "grad_norm": 0.5925476767090508, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 43462 + }, + { + "epoch": 0.43463, + "grad_norm": 0.5800667009154836, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 43463 + }, + { + "epoch": 0.43464, + "grad_norm": 0.5950170816106248, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 43464 + }, + { + "epoch": 0.43465, + "grad_norm": 0.6266857998588995, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 43465 + }, + { + "epoch": 0.43466, + "grad_norm": 0.6255475065742878, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 43466 + }, + { + "epoch": 0.43467, + "grad_norm": 0.7077213400487399, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 43467 + }, + { + "epoch": 0.43468, + "grad_norm": 0.8188987369601943, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 43468 + }, + { + "epoch": 0.43469, + "grad_norm": 0.9576248263677266, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 43469 + }, + { + "epoch": 0.4347, + "grad_norm": 1.091573887176377, + "learning_rate": 0.003, + "loss": 3.987, + "step": 43470 + }, + { + "epoch": 0.43471, + "grad_norm": 0.8207685364692489, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 43471 + }, + { + "epoch": 0.43472, + "grad_norm": 0.7357067306847239, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 43472 + }, + { + "epoch": 0.43473, + "grad_norm": 0.6898968176537456, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 43473 + }, + { + "epoch": 0.43474, + "grad_norm": 0.7327631320472292, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 43474 + }, + { + "epoch": 0.43475, + "grad_norm": 0.8022449433592889, + "learning_rate": 0.003, + "loss": 3.989, + "step": 43475 + }, + { + "epoch": 0.43476, + "grad_norm": 0.8942550738172841, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 43476 + }, + { + "epoch": 0.43477, + "grad_norm": 1.0105279144527926, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 43477 + }, + { + "epoch": 0.43478, + "grad_norm": 1.1516515519752297, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 43478 + }, + { + "epoch": 0.43479, + "grad_norm": 1.035587505841156, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 43479 + }, + { + "epoch": 0.4348, + "grad_norm": 0.8825817788334571, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 43480 + }, + { + "epoch": 0.43481, + "grad_norm": 0.9290634815957892, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 43481 + }, + { + "epoch": 0.43482, + "grad_norm": 0.7803849487586252, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 43482 + }, + { + "epoch": 0.43483, + "grad_norm": 0.7831570658414548, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 43483 + }, + { + "epoch": 0.43484, + "grad_norm": 0.8030744438394669, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 43484 + }, + { + "epoch": 0.43485, + "grad_norm": 0.8486438771526679, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43485 + }, + { + "epoch": 0.43486, + "grad_norm": 0.8824895590285158, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 43486 + }, + { + "epoch": 0.43487, + "grad_norm": 0.9975722377733903, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 43487 + }, + { + "epoch": 0.43488, + "grad_norm": 1.2630028231639625, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 43488 + }, + { + "epoch": 0.43489, + "grad_norm": 0.9273932737760125, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 43489 + }, + { + "epoch": 0.4349, + "grad_norm": 0.9878471981057978, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 43490 + }, + { + "epoch": 0.43491, + "grad_norm": 1.0640951155339033, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 43491 + }, + { + "epoch": 0.43492, + "grad_norm": 0.9605897279251051, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 43492 + }, + { + "epoch": 0.43493, + "grad_norm": 0.8824256212099602, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43493 + }, + { + "epoch": 0.43494, + "grad_norm": 0.7115252879782964, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 43494 + }, + { + "epoch": 0.43495, + "grad_norm": 0.7961931253655823, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 43495 + }, + { + "epoch": 0.43496, + "grad_norm": 0.7834622158957165, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 43496 + }, + { + "epoch": 0.43497, + "grad_norm": 0.7503253377303222, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 43497 + }, + { + "epoch": 0.43498, + "grad_norm": 0.6742055919004435, + "learning_rate": 0.003, + "loss": 3.987, + "step": 43498 + }, + { + "epoch": 0.43499, + "grad_norm": 0.719686813136845, + "learning_rate": 0.003, + "loss": 3.999, + "step": 43499 + }, + { + "epoch": 0.435, + "grad_norm": 0.7406734127258212, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 43500 + }, + { + "epoch": 0.43501, + "grad_norm": 0.7070729356618095, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 43501 + }, + { + "epoch": 0.43502, + "grad_norm": 0.6477001173659532, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 43502 + }, + { + "epoch": 0.43503, + "grad_norm": 0.6204622301700021, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 43503 + }, + { + "epoch": 0.43504, + "grad_norm": 0.6539148729401985, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 43504 + }, + { + "epoch": 0.43505, + "grad_norm": 0.7409347719806635, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 43505 + }, + { + "epoch": 0.43506, + "grad_norm": 0.8089831190457984, + "learning_rate": 0.003, + "loss": 3.999, + "step": 43506 + }, + { + "epoch": 0.43507, + "grad_norm": 0.7717943852034364, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 43507 + }, + { + "epoch": 0.43508, + "grad_norm": 0.7938924742371254, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 43508 + }, + { + "epoch": 0.43509, + "grad_norm": 0.7920423883233476, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 43509 + }, + { + "epoch": 0.4351, + "grad_norm": 0.5533568870623032, + "learning_rate": 0.003, + "loss": 3.954, + "step": 43510 + }, + { + "epoch": 0.43511, + "grad_norm": 0.610642004343027, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 43511 + }, + { + "epoch": 0.43512, + "grad_norm": 0.5782771123787497, + "learning_rate": 0.003, + "loss": 3.991, + "step": 43512 + }, + { + "epoch": 0.43513, + "grad_norm": 0.6463293717988523, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 43513 + }, + { + "epoch": 0.43514, + "grad_norm": 0.7670382879425146, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 43514 + }, + { + "epoch": 0.43515, + "grad_norm": 1.0212522278421656, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 43515 + }, + { + "epoch": 0.43516, + "grad_norm": 1.1329360898892593, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 43516 + }, + { + "epoch": 0.43517, + "grad_norm": 0.7831763155606959, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 43517 + }, + { + "epoch": 0.43518, + "grad_norm": 0.6938410267422059, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 43518 + }, + { + "epoch": 0.43519, + "grad_norm": 0.6465399790633848, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 43519 + }, + { + "epoch": 0.4352, + "grad_norm": 0.5889664897226486, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 43520 + }, + { + "epoch": 0.43521, + "grad_norm": 0.576371864282027, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 43521 + }, + { + "epoch": 0.43522, + "grad_norm": 0.48480425355085116, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43522 + }, + { + "epoch": 0.43523, + "grad_norm": 0.5246647105524798, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 43523 + }, + { + "epoch": 0.43524, + "grad_norm": 0.48843852415625255, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 43524 + }, + { + "epoch": 0.43525, + "grad_norm": 0.566606043113065, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 43525 + }, + { + "epoch": 0.43526, + "grad_norm": 0.6077660013979145, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 43526 + }, + { + "epoch": 0.43527, + "grad_norm": 0.6569733109144325, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 43527 + }, + { + "epoch": 0.43528, + "grad_norm": 0.7874291024039785, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 43528 + }, + { + "epoch": 0.43529, + "grad_norm": 0.9926480744868719, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 43529 + }, + { + "epoch": 0.4353, + "grad_norm": 1.0006785969315164, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 43530 + }, + { + "epoch": 0.43531, + "grad_norm": 0.816385585202357, + "learning_rate": 0.003, + "loss": 3.98, + "step": 43531 + }, + { + "epoch": 0.43532, + "grad_norm": 0.7241391760757512, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 43532 + }, + { + "epoch": 0.43533, + "grad_norm": 0.872578953746671, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 43533 + }, + { + "epoch": 0.43534, + "grad_norm": 0.9257087925404685, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 43534 + }, + { + "epoch": 0.43535, + "grad_norm": 0.8720289349397646, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 43535 + }, + { + "epoch": 0.43536, + "grad_norm": 0.9540658726811494, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 43536 + }, + { + "epoch": 0.43537, + "grad_norm": 0.9786767118686198, + "learning_rate": 0.003, + "loss": 3.977, + "step": 43537 + }, + { + "epoch": 0.43538, + "grad_norm": 0.9934044512879697, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 43538 + }, + { + "epoch": 0.43539, + "grad_norm": 0.9350174638489543, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 43539 + }, + { + "epoch": 0.4354, + "grad_norm": 0.7695561400496735, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 43540 + }, + { + "epoch": 0.43541, + "grad_norm": 0.7556670566847632, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 43541 + }, + { + "epoch": 0.43542, + "grad_norm": 0.7788618578761912, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 43542 + }, + { + "epoch": 0.43543, + "grad_norm": 0.7745371559184379, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 43543 + }, + { + "epoch": 0.43544, + "grad_norm": 0.7346084478722412, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 43544 + }, + { + "epoch": 0.43545, + "grad_norm": 0.7249419827075152, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 43545 + }, + { + "epoch": 0.43546, + "grad_norm": 0.8935703090179182, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 43546 + }, + { + "epoch": 0.43547, + "grad_norm": 0.8998275065361037, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 43547 + }, + { + "epoch": 0.43548, + "grad_norm": 1.0329176947678285, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 43548 + }, + { + "epoch": 0.43549, + "grad_norm": 1.0406810926114898, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43549 + }, + { + "epoch": 0.4355, + "grad_norm": 0.8543796860509295, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 43550 + }, + { + "epoch": 0.43551, + "grad_norm": 0.7506529118256834, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 43551 + }, + { + "epoch": 0.43552, + "grad_norm": 0.7551500626247286, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 43552 + }, + { + "epoch": 0.43553, + "grad_norm": 0.7298756663747266, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 43553 + }, + { + "epoch": 0.43554, + "grad_norm": 0.7183011729028232, + "learning_rate": 0.003, + "loss": 3.978, + "step": 43554 + }, + { + "epoch": 0.43555, + "grad_norm": 0.7087300573415793, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 43555 + }, + { + "epoch": 0.43556, + "grad_norm": 0.7513133665191951, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 43556 + }, + { + "epoch": 0.43557, + "grad_norm": 0.73215499503607, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 43557 + }, + { + "epoch": 0.43558, + "grad_norm": 0.8365872560659777, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 43558 + }, + { + "epoch": 0.43559, + "grad_norm": 0.9289603097764567, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 43559 + }, + { + "epoch": 0.4356, + "grad_norm": 0.9271613812474807, + "learning_rate": 0.003, + "loss": 3.978, + "step": 43560 + }, + { + "epoch": 0.43561, + "grad_norm": 0.744322022296714, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 43561 + }, + { + "epoch": 0.43562, + "grad_norm": 0.8155727772685704, + "learning_rate": 0.003, + "loss": 3.9385, + "step": 43562 + }, + { + "epoch": 0.43563, + "grad_norm": 0.8007819908088536, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 43563 + }, + { + "epoch": 0.43564, + "grad_norm": 0.7484243234936745, + "learning_rate": 0.003, + "loss": 4.013, + "step": 43564 + }, + { + "epoch": 0.43565, + "grad_norm": 0.7164027492909616, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 43565 + }, + { + "epoch": 0.43566, + "grad_norm": 0.6955303858207587, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 43566 + }, + { + "epoch": 0.43567, + "grad_norm": 0.8169506558192786, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 43567 + }, + { + "epoch": 0.43568, + "grad_norm": 0.8348865187138886, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 43568 + }, + { + "epoch": 0.43569, + "grad_norm": 0.8117806619134245, + "learning_rate": 0.003, + "loss": 3.976, + "step": 43569 + }, + { + "epoch": 0.4357, + "grad_norm": 0.6943024877459095, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 43570 + }, + { + "epoch": 0.43571, + "grad_norm": 0.614508860596038, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 43571 + }, + { + "epoch": 0.43572, + "grad_norm": 0.5499402743869951, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 43572 + }, + { + "epoch": 0.43573, + "grad_norm": 0.5241001559857703, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 43573 + }, + { + "epoch": 0.43574, + "grad_norm": 0.5588699587011983, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 43574 + }, + { + "epoch": 0.43575, + "grad_norm": 0.5569052803567178, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 43575 + }, + { + "epoch": 0.43576, + "grad_norm": 0.6039940797326857, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 43576 + }, + { + "epoch": 0.43577, + "grad_norm": 0.6617175479782466, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 43577 + }, + { + "epoch": 0.43578, + "grad_norm": 0.7661865579379042, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 43578 + }, + { + "epoch": 0.43579, + "grad_norm": 0.8681434942730528, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 43579 + }, + { + "epoch": 0.4358, + "grad_norm": 0.819006142294188, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 43580 + }, + { + "epoch": 0.43581, + "grad_norm": 0.679565011753817, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 43581 + }, + { + "epoch": 0.43582, + "grad_norm": 0.6415994119518483, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 43582 + }, + { + "epoch": 0.43583, + "grad_norm": 0.7482908196294179, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 43583 + }, + { + "epoch": 0.43584, + "grad_norm": 0.8788685062084596, + "learning_rate": 0.003, + "loss": 3.986, + "step": 43584 + }, + { + "epoch": 0.43585, + "grad_norm": 0.7983899820524163, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 43585 + }, + { + "epoch": 0.43586, + "grad_norm": 0.7740333797968679, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 43586 + }, + { + "epoch": 0.43587, + "grad_norm": 0.7574142621113479, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 43587 + }, + { + "epoch": 0.43588, + "grad_norm": 0.8258351853929501, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 43588 + }, + { + "epoch": 0.43589, + "grad_norm": 0.8147801591358151, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 43589 + }, + { + "epoch": 0.4359, + "grad_norm": 0.8599889982445745, + "learning_rate": 0.003, + "loss": 3.986, + "step": 43590 + }, + { + "epoch": 0.43591, + "grad_norm": 0.8641882578125769, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 43591 + }, + { + "epoch": 0.43592, + "grad_norm": 0.8343307467643044, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 43592 + }, + { + "epoch": 0.43593, + "grad_norm": 0.7310902680892031, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 43593 + }, + { + "epoch": 0.43594, + "grad_norm": 0.7607550656028501, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 43594 + }, + { + "epoch": 0.43595, + "grad_norm": 0.8376600522898746, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 43595 + }, + { + "epoch": 0.43596, + "grad_norm": 0.8427259830673918, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 43596 + }, + { + "epoch": 0.43597, + "grad_norm": 0.765263357214399, + "learning_rate": 0.003, + "loss": 3.994, + "step": 43597 + }, + { + "epoch": 0.43598, + "grad_norm": 1.004487882987911, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 43598 + }, + { + "epoch": 0.43599, + "grad_norm": 1.2649351603670964, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 43599 + }, + { + "epoch": 0.436, + "grad_norm": 0.5628642289150523, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 43600 + }, + { + "epoch": 0.43601, + "grad_norm": 0.6462742702619702, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 43601 + }, + { + "epoch": 0.43602, + "grad_norm": 0.8210214065920639, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 43602 + }, + { + "epoch": 0.43603, + "grad_norm": 0.8460783667436318, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 43603 + }, + { + "epoch": 0.43604, + "grad_norm": 0.815847113470301, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 43604 + }, + { + "epoch": 0.43605, + "grad_norm": 0.7728779638963115, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 43605 + }, + { + "epoch": 0.43606, + "grad_norm": 0.7297223460513775, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 43606 + }, + { + "epoch": 0.43607, + "grad_norm": 0.8255513755041319, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 43607 + }, + { + "epoch": 0.43608, + "grad_norm": 0.9077549270754556, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 43608 + }, + { + "epoch": 0.43609, + "grad_norm": 0.938770986202428, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 43609 + }, + { + "epoch": 0.4361, + "grad_norm": 1.1562613769234733, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 43610 + }, + { + "epoch": 0.43611, + "grad_norm": 0.8503744271197089, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 43611 + }, + { + "epoch": 0.43612, + "grad_norm": 0.8064644710243924, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 43612 + }, + { + "epoch": 0.43613, + "grad_norm": 0.8305051857541811, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 43613 + }, + { + "epoch": 0.43614, + "grad_norm": 0.8008347994456609, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 43614 + }, + { + "epoch": 0.43615, + "grad_norm": 0.8297475381189255, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 43615 + }, + { + "epoch": 0.43616, + "grad_norm": 0.8111069601373416, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 43616 + }, + { + "epoch": 0.43617, + "grad_norm": 0.859896752757543, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 43617 + }, + { + "epoch": 0.43618, + "grad_norm": 0.9930583482262744, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43618 + }, + { + "epoch": 0.43619, + "grad_norm": 1.1064701268046482, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 43619 + }, + { + "epoch": 0.4362, + "grad_norm": 0.8754475349387287, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 43620 + }, + { + "epoch": 0.43621, + "grad_norm": 0.7941288137782109, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 43621 + }, + { + "epoch": 0.43622, + "grad_norm": 0.7225213806339758, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 43622 + }, + { + "epoch": 0.43623, + "grad_norm": 0.6775793706297063, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 43623 + }, + { + "epoch": 0.43624, + "grad_norm": 0.6171517831130352, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 43624 + }, + { + "epoch": 0.43625, + "grad_norm": 0.6585896324256251, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 43625 + }, + { + "epoch": 0.43626, + "grad_norm": 0.5784044577684248, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 43626 + }, + { + "epoch": 0.43627, + "grad_norm": 0.5840596350706586, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 43627 + }, + { + "epoch": 0.43628, + "grad_norm": 0.7294640934679588, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 43628 + }, + { + "epoch": 0.43629, + "grad_norm": 0.8671855948891286, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 43629 + }, + { + "epoch": 0.4363, + "grad_norm": 0.9517715520615688, + "learning_rate": 0.003, + "loss": 3.966, + "step": 43630 + }, + { + "epoch": 0.43631, + "grad_norm": 1.0168489043695352, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 43631 + }, + { + "epoch": 0.43632, + "grad_norm": 0.8927431923820569, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 43632 + }, + { + "epoch": 0.43633, + "grad_norm": 0.6666832688339654, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 43633 + }, + { + "epoch": 0.43634, + "grad_norm": 0.6708978860083045, + "learning_rate": 0.003, + "loss": 3.9474, + "step": 43634 + }, + { + "epoch": 0.43635, + "grad_norm": 0.7751786363378406, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 43635 + }, + { + "epoch": 0.43636, + "grad_norm": 0.6655478850086155, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 43636 + }, + { + "epoch": 0.43637, + "grad_norm": 0.7062670210494144, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 43637 + }, + { + "epoch": 0.43638, + "grad_norm": 0.733600424303989, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 43638 + }, + { + "epoch": 0.43639, + "grad_norm": 0.7723151426284637, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 43639 + }, + { + "epoch": 0.4364, + "grad_norm": 0.76656481783174, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 43640 + }, + { + "epoch": 0.43641, + "grad_norm": 0.7698561838806248, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 43641 + }, + { + "epoch": 0.43642, + "grad_norm": 0.7004391965238758, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 43642 + }, + { + "epoch": 0.43643, + "grad_norm": 0.5959961092402735, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 43643 + }, + { + "epoch": 0.43644, + "grad_norm": 0.6582016243245985, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 43644 + }, + { + "epoch": 0.43645, + "grad_norm": 0.7566008903015881, + "learning_rate": 0.003, + "loss": 3.94, + "step": 43645 + }, + { + "epoch": 0.43646, + "grad_norm": 0.9794697270692952, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 43646 + }, + { + "epoch": 0.43647, + "grad_norm": 1.0712691237889571, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 43647 + }, + { + "epoch": 0.43648, + "grad_norm": 0.9314688099515233, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 43648 + }, + { + "epoch": 0.43649, + "grad_norm": 0.8132176635935575, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 43649 + }, + { + "epoch": 0.4365, + "grad_norm": 0.7826843041117715, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 43650 + }, + { + "epoch": 0.43651, + "grad_norm": 0.711709765358227, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 43651 + }, + { + "epoch": 0.43652, + "grad_norm": 0.5826216972347716, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 43652 + }, + { + "epoch": 0.43653, + "grad_norm": 0.5454823069009742, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 43653 + }, + { + "epoch": 0.43654, + "grad_norm": 0.607990231512706, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 43654 + }, + { + "epoch": 0.43655, + "grad_norm": 0.5807766538282041, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 43655 + }, + { + "epoch": 0.43656, + "grad_norm": 0.6731697936092336, + "learning_rate": 0.003, + "loss": 3.973, + "step": 43656 + }, + { + "epoch": 0.43657, + "grad_norm": 0.7801999402476101, + "learning_rate": 0.003, + "loss": 3.9394, + "step": 43657 + }, + { + "epoch": 0.43658, + "grad_norm": 0.9455983232131602, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 43658 + }, + { + "epoch": 0.43659, + "grad_norm": 0.9663081530603496, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 43659 + }, + { + "epoch": 0.4366, + "grad_norm": 0.924006152621669, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 43660 + }, + { + "epoch": 0.43661, + "grad_norm": 0.9441713964648039, + "learning_rate": 0.003, + "loss": 3.991, + "step": 43661 + }, + { + "epoch": 0.43662, + "grad_norm": 0.9751280873871792, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 43662 + }, + { + "epoch": 0.43663, + "grad_norm": 1.0152475623685917, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 43663 + }, + { + "epoch": 0.43664, + "grad_norm": 0.9925694297956813, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 43664 + }, + { + "epoch": 0.43665, + "grad_norm": 0.9389719576080434, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 43665 + }, + { + "epoch": 0.43666, + "grad_norm": 0.9538108932009731, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 43666 + }, + { + "epoch": 0.43667, + "grad_norm": 0.9288376658371492, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 43667 + }, + { + "epoch": 0.43668, + "grad_norm": 0.8681680324491257, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 43668 + }, + { + "epoch": 0.43669, + "grad_norm": 0.7884796082268066, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 43669 + }, + { + "epoch": 0.4367, + "grad_norm": 0.8229775468292121, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 43670 + }, + { + "epoch": 0.43671, + "grad_norm": 0.861178461539055, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 43671 + }, + { + "epoch": 0.43672, + "grad_norm": 0.9708880719158169, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 43672 + }, + { + "epoch": 0.43673, + "grad_norm": 1.0739815774974084, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 43673 + }, + { + "epoch": 0.43674, + "grad_norm": 0.8577652123685026, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 43674 + }, + { + "epoch": 0.43675, + "grad_norm": 0.7490478360461017, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 43675 + }, + { + "epoch": 0.43676, + "grad_norm": 0.6609542614019573, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 43676 + }, + { + "epoch": 0.43677, + "grad_norm": 0.6363400484591839, + "learning_rate": 0.003, + "loss": 4.0, + "step": 43677 + }, + { + "epoch": 0.43678, + "grad_norm": 0.6299803911546505, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 43678 + }, + { + "epoch": 0.43679, + "grad_norm": 0.6208996314968156, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 43679 + }, + { + "epoch": 0.4368, + "grad_norm": 0.5950316656724669, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 43680 + }, + { + "epoch": 0.43681, + "grad_norm": 0.5212563702283738, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 43681 + }, + { + "epoch": 0.43682, + "grad_norm": 0.5804573445171278, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 43682 + }, + { + "epoch": 0.43683, + "grad_norm": 0.6539315516210931, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 43683 + }, + { + "epoch": 0.43684, + "grad_norm": 0.7293281805945588, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 43684 + }, + { + "epoch": 0.43685, + "grad_norm": 0.6934594524833411, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 43685 + }, + { + "epoch": 0.43686, + "grad_norm": 0.6285517441485245, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 43686 + }, + { + "epoch": 0.43687, + "grad_norm": 0.6639558371099118, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 43687 + }, + { + "epoch": 0.43688, + "grad_norm": 0.7140583711852642, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 43688 + }, + { + "epoch": 0.43689, + "grad_norm": 0.7154506800704491, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 43689 + }, + { + "epoch": 0.4369, + "grad_norm": 0.767377601036157, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 43690 + }, + { + "epoch": 0.43691, + "grad_norm": 0.866016988022179, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 43691 + }, + { + "epoch": 0.43692, + "grad_norm": 0.9283691527487136, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 43692 + }, + { + "epoch": 0.43693, + "grad_norm": 0.9604206850611765, + "learning_rate": 0.003, + "loss": 4.01, + "step": 43693 + }, + { + "epoch": 0.43694, + "grad_norm": 0.965537584378549, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 43694 + }, + { + "epoch": 0.43695, + "grad_norm": 1.142974895471867, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 43695 + }, + { + "epoch": 0.43696, + "grad_norm": 0.9750493600890355, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 43696 + }, + { + "epoch": 0.43697, + "grad_norm": 1.0241667666542653, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 43697 + }, + { + "epoch": 0.43698, + "grad_norm": 0.9444304863722345, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 43698 + }, + { + "epoch": 0.43699, + "grad_norm": 0.9143104731585993, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 43699 + }, + { + "epoch": 0.437, + "grad_norm": 0.7981314862685657, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 43700 + }, + { + "epoch": 0.43701, + "grad_norm": 0.7518410777431639, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 43701 + }, + { + "epoch": 0.43702, + "grad_norm": 0.8656312703303269, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 43702 + }, + { + "epoch": 0.43703, + "grad_norm": 0.9134417811866863, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 43703 + }, + { + "epoch": 0.43704, + "grad_norm": 0.8926910420758974, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 43704 + }, + { + "epoch": 0.43705, + "grad_norm": 0.9042959215882572, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 43705 + }, + { + "epoch": 0.43706, + "grad_norm": 0.8136932686164362, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 43706 + }, + { + "epoch": 0.43707, + "grad_norm": 0.768035229492963, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 43707 + }, + { + "epoch": 0.43708, + "grad_norm": 0.737629157264411, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 43708 + }, + { + "epoch": 0.43709, + "grad_norm": 0.6622789423576211, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 43709 + }, + { + "epoch": 0.4371, + "grad_norm": 0.5489825444192352, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 43710 + }, + { + "epoch": 0.43711, + "grad_norm": 0.4999801564562855, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 43711 + }, + { + "epoch": 0.43712, + "grad_norm": 0.5431045084891306, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 43712 + }, + { + "epoch": 0.43713, + "grad_norm": 0.5697837338055626, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 43713 + }, + { + "epoch": 0.43714, + "grad_norm": 0.5993684776261345, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 43714 + }, + { + "epoch": 0.43715, + "grad_norm": 0.6319094869005853, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 43715 + }, + { + "epoch": 0.43716, + "grad_norm": 0.704793753181374, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 43716 + }, + { + "epoch": 0.43717, + "grad_norm": 0.8523710642067934, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 43717 + }, + { + "epoch": 0.43718, + "grad_norm": 0.9298946329694623, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 43718 + }, + { + "epoch": 0.43719, + "grad_norm": 0.8973490125001298, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 43719 + }, + { + "epoch": 0.4372, + "grad_norm": 0.861898781083249, + "learning_rate": 0.003, + "loss": 3.979, + "step": 43720 + }, + { + "epoch": 0.43721, + "grad_norm": 0.8944048168250098, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 43721 + }, + { + "epoch": 0.43722, + "grad_norm": 0.779335552880413, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 43722 + }, + { + "epoch": 0.43723, + "grad_norm": 0.8560092323836758, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 43723 + }, + { + "epoch": 0.43724, + "grad_norm": 0.7269302282606036, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 43724 + }, + { + "epoch": 0.43725, + "grad_norm": 0.7536866329945536, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 43725 + }, + { + "epoch": 0.43726, + "grad_norm": 0.8180741018491616, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 43726 + }, + { + "epoch": 0.43727, + "grad_norm": 0.8868838750804283, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 43727 + }, + { + "epoch": 0.43728, + "grad_norm": 0.9478016407844834, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 43728 + }, + { + "epoch": 0.43729, + "grad_norm": 1.0644573017244845, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 43729 + }, + { + "epoch": 0.4373, + "grad_norm": 1.0026926731523447, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 43730 + }, + { + "epoch": 0.43731, + "grad_norm": 0.9204742769572983, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 43731 + }, + { + "epoch": 0.43732, + "grad_norm": 0.8756362276029249, + "learning_rate": 0.003, + "loss": 3.959, + "step": 43732 + }, + { + "epoch": 0.43733, + "grad_norm": 0.6783603400632783, + "learning_rate": 0.003, + "loss": 4.009, + "step": 43733 + }, + { + "epoch": 0.43734, + "grad_norm": 0.6991275187038684, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 43734 + }, + { + "epoch": 0.43735, + "grad_norm": 0.6350243237454634, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 43735 + }, + { + "epoch": 0.43736, + "grad_norm": 0.5819673738864156, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 43736 + }, + { + "epoch": 0.43737, + "grad_norm": 0.6308404123758916, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 43737 + }, + { + "epoch": 0.43738, + "grad_norm": 0.651294419394791, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 43738 + }, + { + "epoch": 0.43739, + "grad_norm": 0.6404270584613999, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 43739 + }, + { + "epoch": 0.4374, + "grad_norm": 0.6222243194384581, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 43740 + }, + { + "epoch": 0.43741, + "grad_norm": 0.5745214035208259, + "learning_rate": 0.003, + "loss": 3.935, + "step": 43741 + }, + { + "epoch": 0.43742, + "grad_norm": 0.5812488222548218, + "learning_rate": 0.003, + "loss": 3.98, + "step": 43742 + }, + { + "epoch": 0.43743, + "grad_norm": 0.6250059136018501, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 43743 + }, + { + "epoch": 0.43744, + "grad_norm": 0.6880177954606579, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 43744 + }, + { + "epoch": 0.43745, + "grad_norm": 0.8197220759465674, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 43745 + }, + { + "epoch": 0.43746, + "grad_norm": 0.931130125108409, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 43746 + }, + { + "epoch": 0.43747, + "grad_norm": 0.8287261589404122, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 43747 + }, + { + "epoch": 0.43748, + "grad_norm": 0.6557325139884123, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 43748 + }, + { + "epoch": 0.43749, + "grad_norm": 0.6308751548563877, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 43749 + }, + { + "epoch": 0.4375, + "grad_norm": 0.7150734792317683, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 43750 + }, + { + "epoch": 0.43751, + "grad_norm": 0.7578253380566395, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 43751 + }, + { + "epoch": 0.43752, + "grad_norm": 0.8013282813579549, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 43752 + }, + { + "epoch": 0.43753, + "grad_norm": 0.7665398258012129, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 43753 + }, + { + "epoch": 0.43754, + "grad_norm": 0.5995971711472043, + "learning_rate": 0.003, + "loss": 3.957, + "step": 43754 + }, + { + "epoch": 0.43755, + "grad_norm": 0.620301858924946, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 43755 + }, + { + "epoch": 0.43756, + "grad_norm": 0.6548215722084454, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 43756 + }, + { + "epoch": 0.43757, + "grad_norm": 0.6904326794557472, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 43757 + }, + { + "epoch": 0.43758, + "grad_norm": 0.6905722261678862, + "learning_rate": 0.003, + "loss": 3.946, + "step": 43758 + }, + { + "epoch": 0.43759, + "grad_norm": 0.6481959915998564, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 43759 + }, + { + "epoch": 0.4376, + "grad_norm": 0.7192032722075883, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 43760 + }, + { + "epoch": 0.43761, + "grad_norm": 0.9573520836741574, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 43761 + }, + { + "epoch": 0.43762, + "grad_norm": 1.1687130148400544, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 43762 + }, + { + "epoch": 0.43763, + "grad_norm": 0.7512632741658163, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 43763 + }, + { + "epoch": 0.43764, + "grad_norm": 0.7484047019895714, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 43764 + }, + { + "epoch": 0.43765, + "grad_norm": 0.7109140210266961, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 43765 + }, + { + "epoch": 0.43766, + "grad_norm": 0.7332184139808741, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 43766 + }, + { + "epoch": 0.43767, + "grad_norm": 0.8645776341226392, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 43767 + }, + { + "epoch": 0.43768, + "grad_norm": 0.8643608121483634, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 43768 + }, + { + "epoch": 0.43769, + "grad_norm": 0.9013690819478322, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 43769 + }, + { + "epoch": 0.4377, + "grad_norm": 0.8318065904359013, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 43770 + }, + { + "epoch": 0.43771, + "grad_norm": 0.7188059165265354, + "learning_rate": 0.003, + "loss": 3.943, + "step": 43771 + }, + { + "epoch": 0.43772, + "grad_norm": 0.7259408078048671, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 43772 + }, + { + "epoch": 0.43773, + "grad_norm": 0.8029413193552876, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 43773 + }, + { + "epoch": 0.43774, + "grad_norm": 0.8895765128148558, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 43774 + }, + { + "epoch": 0.43775, + "grad_norm": 1.0701884076030772, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 43775 + }, + { + "epoch": 0.43776, + "grad_norm": 1.0608481219561283, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 43776 + }, + { + "epoch": 0.43777, + "grad_norm": 1.0069587342160637, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 43777 + }, + { + "epoch": 0.43778, + "grad_norm": 1.049286716123705, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 43778 + }, + { + "epoch": 0.43779, + "grad_norm": 1.1060597783994865, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 43779 + }, + { + "epoch": 0.4378, + "grad_norm": 0.9288064315767285, + "learning_rate": 0.003, + "loss": 4.004, + "step": 43780 + }, + { + "epoch": 0.43781, + "grad_norm": 0.8825761554750685, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 43781 + }, + { + "epoch": 0.43782, + "grad_norm": 0.8667601210463156, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 43782 + }, + { + "epoch": 0.43783, + "grad_norm": 0.8469860541199725, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 43783 + }, + { + "epoch": 0.43784, + "grad_norm": 0.8133241942275984, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 43784 + }, + { + "epoch": 0.43785, + "grad_norm": 0.9452203957840187, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 43785 + }, + { + "epoch": 0.43786, + "grad_norm": 0.9963901541505804, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 43786 + }, + { + "epoch": 0.43787, + "grad_norm": 0.9751293991650625, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 43787 + }, + { + "epoch": 0.43788, + "grad_norm": 0.9280812918185087, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 43788 + }, + { + "epoch": 0.43789, + "grad_norm": 0.8886969521703048, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 43789 + }, + { + "epoch": 0.4379, + "grad_norm": 0.8441458900737815, + "learning_rate": 0.003, + "loss": 4.0386, + "step": 43790 + }, + { + "epoch": 0.43791, + "grad_norm": 0.8496871548290834, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 43791 + }, + { + "epoch": 0.43792, + "grad_norm": 0.8086041454224864, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 43792 + }, + { + "epoch": 0.43793, + "grad_norm": 0.8282197348842404, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 43793 + }, + { + "epoch": 0.43794, + "grad_norm": 0.7083753362720667, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 43794 + }, + { + "epoch": 0.43795, + "grad_norm": 0.5947680086498521, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 43795 + }, + { + "epoch": 0.43796, + "grad_norm": 0.6461026683837618, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 43796 + }, + { + "epoch": 0.43797, + "grad_norm": 0.7288082118632203, + "learning_rate": 0.003, + "loss": 4.013, + "step": 43797 + }, + { + "epoch": 0.43798, + "grad_norm": 0.7302659536580463, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 43798 + }, + { + "epoch": 0.43799, + "grad_norm": 0.7649466416183024, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 43799 + }, + { + "epoch": 0.438, + "grad_norm": 0.8153271244720683, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 43800 + }, + { + "epoch": 0.43801, + "grad_norm": 0.8071248063793494, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 43801 + }, + { + "epoch": 0.43802, + "grad_norm": 0.8008913961172578, + "learning_rate": 0.003, + "loss": 3.96, + "step": 43802 + }, + { + "epoch": 0.43803, + "grad_norm": 0.7385404416086394, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 43803 + }, + { + "epoch": 0.43804, + "grad_norm": 0.6282690001298101, + "learning_rate": 0.003, + "loss": 4.019, + "step": 43804 + }, + { + "epoch": 0.43805, + "grad_norm": 0.5395729638913861, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 43805 + }, + { + "epoch": 0.43806, + "grad_norm": 0.5489134764532917, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 43806 + }, + { + "epoch": 0.43807, + "grad_norm": 0.6872737116656298, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 43807 + }, + { + "epoch": 0.43808, + "grad_norm": 0.8422073895729935, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 43808 + }, + { + "epoch": 0.43809, + "grad_norm": 0.9044974490010813, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 43809 + }, + { + "epoch": 0.4381, + "grad_norm": 0.8552297091371781, + "learning_rate": 0.003, + "loss": 3.943, + "step": 43810 + }, + { + "epoch": 0.43811, + "grad_norm": 0.7491633918967412, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 43811 + }, + { + "epoch": 0.43812, + "grad_norm": 0.6984368112031126, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 43812 + }, + { + "epoch": 0.43813, + "grad_norm": 0.6532918876735897, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 43813 + }, + { + "epoch": 0.43814, + "grad_norm": 0.7139652165877987, + "learning_rate": 0.003, + "loss": 3.9271, + "step": 43814 + }, + { + "epoch": 0.43815, + "grad_norm": 0.7840716232036391, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 43815 + }, + { + "epoch": 0.43816, + "grad_norm": 0.757877751214302, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 43816 + }, + { + "epoch": 0.43817, + "grad_norm": 0.589526286943474, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 43817 + }, + { + "epoch": 0.43818, + "grad_norm": 0.6567385567095241, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 43818 + }, + { + "epoch": 0.43819, + "grad_norm": 0.7027986014782379, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 43819 + }, + { + "epoch": 0.4382, + "grad_norm": 0.6250777110276393, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 43820 + }, + { + "epoch": 0.43821, + "grad_norm": 0.6325636981982345, + "learning_rate": 0.003, + "loss": 3.9387, + "step": 43821 + }, + { + "epoch": 0.43822, + "grad_norm": 0.6719999998488109, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 43822 + }, + { + "epoch": 0.43823, + "grad_norm": 0.6784870793109954, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 43823 + }, + { + "epoch": 0.43824, + "grad_norm": 0.7288397746656532, + "learning_rate": 0.003, + "loss": 3.9418, + "step": 43824 + }, + { + "epoch": 0.43825, + "grad_norm": 0.7061277781072143, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 43825 + }, + { + "epoch": 0.43826, + "grad_norm": 0.8138005619910877, + "learning_rate": 0.003, + "loss": 3.955, + "step": 43826 + }, + { + "epoch": 0.43827, + "grad_norm": 0.8844177255966558, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 43827 + }, + { + "epoch": 0.43828, + "grad_norm": 0.8173412611245592, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 43828 + }, + { + "epoch": 0.43829, + "grad_norm": 0.7955706043612404, + "learning_rate": 0.003, + "loss": 3.98, + "step": 43829 + }, + { + "epoch": 0.4383, + "grad_norm": 0.8525004461676176, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 43830 + }, + { + "epoch": 0.43831, + "grad_norm": 0.9017905041486597, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 43831 + }, + { + "epoch": 0.43832, + "grad_norm": 1.0075040023662991, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 43832 + }, + { + "epoch": 0.43833, + "grad_norm": 0.8946376826773761, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 43833 + }, + { + "epoch": 0.43834, + "grad_norm": 0.9475243247888365, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 43834 + }, + { + "epoch": 0.43835, + "grad_norm": 1.1913737396609476, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 43835 + }, + { + "epoch": 0.43836, + "grad_norm": 1.0202970756360374, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 43836 + }, + { + "epoch": 0.43837, + "grad_norm": 0.8628655339922202, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 43837 + }, + { + "epoch": 0.43838, + "grad_norm": 0.807309719842553, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 43838 + }, + { + "epoch": 0.43839, + "grad_norm": 0.8195379496121821, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 43839 + }, + { + "epoch": 0.4384, + "grad_norm": 0.7650912751603565, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 43840 + }, + { + "epoch": 0.43841, + "grad_norm": 0.8301947965316445, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 43841 + }, + { + "epoch": 0.43842, + "grad_norm": 0.7637815012527978, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 43842 + }, + { + "epoch": 0.43843, + "grad_norm": 0.6681956222012073, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 43843 + }, + { + "epoch": 0.43844, + "grad_norm": 0.6678876355503891, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 43844 + }, + { + "epoch": 0.43845, + "grad_norm": 0.6636868314723856, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 43845 + }, + { + "epoch": 0.43846, + "grad_norm": 0.7025137570951427, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 43846 + }, + { + "epoch": 0.43847, + "grad_norm": 0.7785800087156474, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 43847 + }, + { + "epoch": 0.43848, + "grad_norm": 0.7397085963912111, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 43848 + }, + { + "epoch": 0.43849, + "grad_norm": 0.6826016362753059, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 43849 + }, + { + "epoch": 0.4385, + "grad_norm": 0.6313025256331849, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 43850 + }, + { + "epoch": 0.43851, + "grad_norm": 0.6990063846319218, + "learning_rate": 0.003, + "loss": 3.967, + "step": 43851 + }, + { + "epoch": 0.43852, + "grad_norm": 0.7628671892837409, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 43852 + }, + { + "epoch": 0.43853, + "grad_norm": 0.7501646856919759, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 43853 + }, + { + "epoch": 0.43854, + "grad_norm": 0.6777237588924724, + "learning_rate": 0.003, + "loss": 3.98, + "step": 43854 + }, + { + "epoch": 0.43855, + "grad_norm": 0.7045763387831278, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 43855 + }, + { + "epoch": 0.43856, + "grad_norm": 0.7583927080329064, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 43856 + }, + { + "epoch": 0.43857, + "grad_norm": 0.8240031219005363, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 43857 + }, + { + "epoch": 0.43858, + "grad_norm": 0.9745212123710137, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 43858 + }, + { + "epoch": 0.43859, + "grad_norm": 1.092277538207217, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 43859 + }, + { + "epoch": 0.4386, + "grad_norm": 0.9166699248183651, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 43860 + }, + { + "epoch": 0.43861, + "grad_norm": 0.7109157420611281, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 43861 + }, + { + "epoch": 0.43862, + "grad_norm": 0.6780007081913685, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 43862 + }, + { + "epoch": 0.43863, + "grad_norm": 0.6273732434316375, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 43863 + }, + { + "epoch": 0.43864, + "grad_norm": 0.67499107617835, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 43864 + }, + { + "epoch": 0.43865, + "grad_norm": 0.7086821283896373, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 43865 + }, + { + "epoch": 0.43866, + "grad_norm": 0.71384559389832, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 43866 + }, + { + "epoch": 0.43867, + "grad_norm": 0.7622905464260635, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 43867 + }, + { + "epoch": 0.43868, + "grad_norm": 0.7268629878874291, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 43868 + }, + { + "epoch": 0.43869, + "grad_norm": 0.7238019915017386, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 43869 + }, + { + "epoch": 0.4387, + "grad_norm": 0.7780340619681425, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 43870 + }, + { + "epoch": 0.43871, + "grad_norm": 0.7844382554171698, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 43871 + }, + { + "epoch": 0.43872, + "grad_norm": 0.8056981814644366, + "learning_rate": 0.003, + "loss": 3.938, + "step": 43872 + }, + { + "epoch": 0.43873, + "grad_norm": 0.936562541368082, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 43873 + }, + { + "epoch": 0.43874, + "grad_norm": 1.1325443676438236, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 43874 + }, + { + "epoch": 0.43875, + "grad_norm": 0.7791223042796916, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 43875 + }, + { + "epoch": 0.43876, + "grad_norm": 0.6855346297426793, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 43876 + }, + { + "epoch": 0.43877, + "grad_norm": 0.787261761764446, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 43877 + }, + { + "epoch": 0.43878, + "grad_norm": 0.7948290666460004, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 43878 + }, + { + "epoch": 0.43879, + "grad_norm": 0.7794021192492553, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 43879 + }, + { + "epoch": 0.4388, + "grad_norm": 0.6081473870917611, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 43880 + }, + { + "epoch": 0.43881, + "grad_norm": 0.5809515543882307, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 43881 + }, + { + "epoch": 0.43882, + "grad_norm": 0.5832254506541704, + "learning_rate": 0.003, + "loss": 3.9274, + "step": 43882 + }, + { + "epoch": 0.43883, + "grad_norm": 0.6039230370389711, + "learning_rate": 0.003, + "loss": 3.979, + "step": 43883 + }, + { + "epoch": 0.43884, + "grad_norm": 0.7095517620500658, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 43884 + }, + { + "epoch": 0.43885, + "grad_norm": 0.7848301632522848, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 43885 + }, + { + "epoch": 0.43886, + "grad_norm": 0.7890770119690723, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 43886 + }, + { + "epoch": 0.43887, + "grad_norm": 0.8200718999538403, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 43887 + }, + { + "epoch": 0.43888, + "grad_norm": 0.9119013518299853, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 43888 + }, + { + "epoch": 0.43889, + "grad_norm": 0.9510925306760675, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 43889 + }, + { + "epoch": 0.4389, + "grad_norm": 0.979577435472941, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 43890 + }, + { + "epoch": 0.43891, + "grad_norm": 1.019347111949277, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 43891 + }, + { + "epoch": 0.43892, + "grad_norm": 0.9736392562755719, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 43892 + }, + { + "epoch": 0.43893, + "grad_norm": 0.988195973143905, + "learning_rate": 0.003, + "loss": 3.988, + "step": 43893 + }, + { + "epoch": 0.43894, + "grad_norm": 1.0822892140903517, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 43894 + }, + { + "epoch": 0.43895, + "grad_norm": 0.8459971661354889, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 43895 + }, + { + "epoch": 0.43896, + "grad_norm": 0.7829387801700576, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 43896 + }, + { + "epoch": 0.43897, + "grad_norm": 0.682675832944956, + "learning_rate": 0.003, + "loss": 3.9319, + "step": 43897 + }, + { + "epoch": 0.43898, + "grad_norm": 0.6466724835574278, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 43898 + }, + { + "epoch": 0.43899, + "grad_norm": 0.6215936151076613, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 43899 + }, + { + "epoch": 0.439, + "grad_norm": 0.7204456001806516, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 43900 + }, + { + "epoch": 0.43901, + "grad_norm": 0.8257588505974366, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 43901 + }, + { + "epoch": 0.43902, + "grad_norm": 0.8179451656741248, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 43902 + }, + { + "epoch": 0.43903, + "grad_norm": 0.7373515539622096, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 43903 + }, + { + "epoch": 0.43904, + "grad_norm": 0.7219241104120173, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 43904 + }, + { + "epoch": 0.43905, + "grad_norm": 0.7844054063466113, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 43905 + }, + { + "epoch": 0.43906, + "grad_norm": 0.7061702415424321, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 43906 + }, + { + "epoch": 0.43907, + "grad_norm": 0.6144047391262861, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 43907 + }, + { + "epoch": 0.43908, + "grad_norm": 0.640589995663731, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 43908 + }, + { + "epoch": 0.43909, + "grad_norm": 0.616343347706755, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 43909 + }, + { + "epoch": 0.4391, + "grad_norm": 0.5755921199065142, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 43910 + }, + { + "epoch": 0.43911, + "grad_norm": 0.5867460932362248, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 43911 + }, + { + "epoch": 0.43912, + "grad_norm": 0.7573925007429608, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 43912 + }, + { + "epoch": 0.43913, + "grad_norm": 0.8281467697733284, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 43913 + }, + { + "epoch": 0.43914, + "grad_norm": 0.7549902680979756, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 43914 + }, + { + "epoch": 0.43915, + "grad_norm": 0.799159237471663, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 43915 + }, + { + "epoch": 0.43916, + "grad_norm": 0.8080945657020864, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 43916 + }, + { + "epoch": 0.43917, + "grad_norm": 0.8692901196869778, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 43917 + }, + { + "epoch": 0.43918, + "grad_norm": 0.9874412552652325, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 43918 + }, + { + "epoch": 0.43919, + "grad_norm": 1.054947287803899, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 43919 + }, + { + "epoch": 0.4392, + "grad_norm": 1.0517560903101142, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 43920 + }, + { + "epoch": 0.43921, + "grad_norm": 0.9800850226507163, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 43921 + }, + { + "epoch": 0.43922, + "grad_norm": 0.953352691010883, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 43922 + }, + { + "epoch": 0.43923, + "grad_norm": 0.9586957765257267, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 43923 + }, + { + "epoch": 0.43924, + "grad_norm": 1.004152686097587, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 43924 + }, + { + "epoch": 0.43925, + "grad_norm": 0.9890750800953844, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 43925 + }, + { + "epoch": 0.43926, + "grad_norm": 0.9762655976911984, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 43926 + }, + { + "epoch": 0.43927, + "grad_norm": 0.9375020185230537, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 43927 + }, + { + "epoch": 0.43928, + "grad_norm": 0.823767091336757, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 43928 + }, + { + "epoch": 0.43929, + "grad_norm": 0.8960339892474697, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 43929 + }, + { + "epoch": 0.4393, + "grad_norm": 1.0388151443079197, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 43930 + }, + { + "epoch": 0.43931, + "grad_norm": 1.0170266758947222, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 43931 + }, + { + "epoch": 0.43932, + "grad_norm": 0.9382483973230368, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 43932 + }, + { + "epoch": 0.43933, + "grad_norm": 1.1067182179976542, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 43933 + }, + { + "epoch": 0.43934, + "grad_norm": 0.9514760595566544, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 43934 + }, + { + "epoch": 0.43935, + "grad_norm": 0.9167455013636145, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 43935 + }, + { + "epoch": 0.43936, + "grad_norm": 0.8331513408745957, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 43936 + }, + { + "epoch": 0.43937, + "grad_norm": 0.7549442508108887, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 43937 + }, + { + "epoch": 0.43938, + "grad_norm": 0.6293753828930514, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 43938 + }, + { + "epoch": 0.43939, + "grad_norm": 0.6354862847869015, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 43939 + }, + { + "epoch": 0.4394, + "grad_norm": 0.5875920771753924, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 43940 + }, + { + "epoch": 0.43941, + "grad_norm": 0.6319120493870092, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 43941 + }, + { + "epoch": 0.43942, + "grad_norm": 0.5698947795256464, + "learning_rate": 0.003, + "loss": 3.987, + "step": 43942 + }, + { + "epoch": 0.43943, + "grad_norm": 0.5497181715149224, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 43943 + }, + { + "epoch": 0.43944, + "grad_norm": 0.572495455932644, + "learning_rate": 0.003, + "loss": 3.9453, + "step": 43944 + }, + { + "epoch": 0.43945, + "grad_norm": 0.6179291928338759, + "learning_rate": 0.003, + "loss": 3.953, + "step": 43945 + }, + { + "epoch": 0.43946, + "grad_norm": 0.7276617284359433, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 43946 + }, + { + "epoch": 0.43947, + "grad_norm": 0.8228610378064997, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 43947 + }, + { + "epoch": 0.43948, + "grad_norm": 0.9143332307793536, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 43948 + }, + { + "epoch": 0.43949, + "grad_norm": 0.9011528194179503, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 43949 + }, + { + "epoch": 0.4395, + "grad_norm": 0.8669790594735167, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 43950 + }, + { + "epoch": 0.43951, + "grad_norm": 0.9049095185222741, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 43951 + }, + { + "epoch": 0.43952, + "grad_norm": 0.7933603344183312, + "learning_rate": 0.003, + "loss": 3.989, + "step": 43952 + }, + { + "epoch": 0.43953, + "grad_norm": 0.6593905723752891, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 43953 + }, + { + "epoch": 0.43954, + "grad_norm": 0.6341146830956113, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 43954 + }, + { + "epoch": 0.43955, + "grad_norm": 0.7246043107218587, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 43955 + }, + { + "epoch": 0.43956, + "grad_norm": 0.7310185447831016, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 43956 + }, + { + "epoch": 0.43957, + "grad_norm": 0.6334990182769606, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 43957 + }, + { + "epoch": 0.43958, + "grad_norm": 0.5582252620971314, + "learning_rate": 0.003, + "loss": 3.958, + "step": 43958 + }, + { + "epoch": 0.43959, + "grad_norm": 0.5521080020333129, + "learning_rate": 0.003, + "loss": 3.996, + "step": 43959 + }, + { + "epoch": 0.4396, + "grad_norm": 0.6202841843020007, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 43960 + }, + { + "epoch": 0.43961, + "grad_norm": 0.6532139776417556, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 43961 + }, + { + "epoch": 0.43962, + "grad_norm": 0.6620334912744797, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 43962 + }, + { + "epoch": 0.43963, + "grad_norm": 0.6833054716692493, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 43963 + }, + { + "epoch": 0.43964, + "grad_norm": 0.6027026870372199, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 43964 + }, + { + "epoch": 0.43965, + "grad_norm": 0.6809116977801577, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 43965 + }, + { + "epoch": 0.43966, + "grad_norm": 0.8491000430805817, + "learning_rate": 0.003, + "loss": 3.941, + "step": 43966 + }, + { + "epoch": 0.43967, + "grad_norm": 0.896731967954526, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 43967 + }, + { + "epoch": 0.43968, + "grad_norm": 0.8105088241466782, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 43968 + }, + { + "epoch": 0.43969, + "grad_norm": 0.7259665249896249, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 43969 + }, + { + "epoch": 0.4397, + "grad_norm": 0.6731154077775169, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 43970 + }, + { + "epoch": 0.43971, + "grad_norm": 0.5561503646781533, + "learning_rate": 0.003, + "loss": 3.976, + "step": 43971 + }, + { + "epoch": 0.43972, + "grad_norm": 0.5572940555190021, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 43972 + }, + { + "epoch": 0.43973, + "grad_norm": 0.690704713507498, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 43973 + }, + { + "epoch": 0.43974, + "grad_norm": 0.7477163408389018, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 43974 + }, + { + "epoch": 0.43975, + "grad_norm": 0.7800281021276045, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 43975 + }, + { + "epoch": 0.43976, + "grad_norm": 0.7169478483308819, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 43976 + }, + { + "epoch": 0.43977, + "grad_norm": 0.8496547126342886, + "learning_rate": 0.003, + "loss": 3.99, + "step": 43977 + }, + { + "epoch": 0.43978, + "grad_norm": 0.93857699185805, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 43978 + }, + { + "epoch": 0.43979, + "grad_norm": 0.9499370803607711, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 43979 + }, + { + "epoch": 0.4398, + "grad_norm": 1.0142673654140828, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 43980 + }, + { + "epoch": 0.43981, + "grad_norm": 1.0224943894746192, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 43981 + }, + { + "epoch": 0.43982, + "grad_norm": 0.9587235623284548, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 43982 + }, + { + "epoch": 0.43983, + "grad_norm": 0.9134818941636011, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 43983 + }, + { + "epoch": 0.43984, + "grad_norm": 0.8093864301503341, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 43984 + }, + { + "epoch": 0.43985, + "grad_norm": 0.7681729806637603, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 43985 + }, + { + "epoch": 0.43986, + "grad_norm": 0.7743357816329176, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 43986 + }, + { + "epoch": 0.43987, + "grad_norm": 0.7944211504504872, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 43987 + }, + { + "epoch": 0.43988, + "grad_norm": 0.8590790291077935, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 43988 + }, + { + "epoch": 0.43989, + "grad_norm": 0.8459756149465366, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 43989 + }, + { + "epoch": 0.4399, + "grad_norm": 0.8217667202183091, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 43990 + }, + { + "epoch": 0.43991, + "grad_norm": 0.7038797422380689, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 43991 + }, + { + "epoch": 0.43992, + "grad_norm": 0.8110279477503466, + "learning_rate": 0.003, + "loss": 3.979, + "step": 43992 + }, + { + "epoch": 0.43993, + "grad_norm": 0.8298574308064722, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 43993 + }, + { + "epoch": 0.43994, + "grad_norm": 0.7849335877403117, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 43994 + }, + { + "epoch": 0.43995, + "grad_norm": 0.7528018329151296, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 43995 + }, + { + "epoch": 0.43996, + "grad_norm": 0.6460798761970276, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 43996 + }, + { + "epoch": 0.43997, + "grad_norm": 0.5841789513673783, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 43997 + }, + { + "epoch": 0.43998, + "grad_norm": 0.6287804693949193, + "learning_rate": 0.003, + "loss": 3.959, + "step": 43998 + }, + { + "epoch": 0.43999, + "grad_norm": 0.8202585427126785, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 43999 + }, + { + "epoch": 0.44, + "grad_norm": 1.1484561730223743, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 44000 + }, + { + "epoch": 0.44001, + "grad_norm": 0.9416835832909344, + "learning_rate": 0.003, + "loss": 4.026, + "step": 44001 + }, + { + "epoch": 0.44002, + "grad_norm": 0.9040133827922692, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 44002 + }, + { + "epoch": 0.44003, + "grad_norm": 0.8764040021360272, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 44003 + }, + { + "epoch": 0.44004, + "grad_norm": 0.9259594988960193, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 44004 + }, + { + "epoch": 0.44005, + "grad_norm": 0.9867975453421108, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 44005 + }, + { + "epoch": 0.44006, + "grad_norm": 0.8360173119282172, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 44006 + }, + { + "epoch": 0.44007, + "grad_norm": 0.6564800742369697, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 44007 + }, + { + "epoch": 0.44008, + "grad_norm": 0.6837153096766818, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 44008 + }, + { + "epoch": 0.44009, + "grad_norm": 0.6343889455548379, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 44009 + }, + { + "epoch": 0.4401, + "grad_norm": 0.6841463986341519, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 44010 + }, + { + "epoch": 0.44011, + "grad_norm": 0.6153559655330043, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 44011 + }, + { + "epoch": 0.44012, + "grad_norm": 0.5741256157324048, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 44012 + }, + { + "epoch": 0.44013, + "grad_norm": 0.5807573470583449, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 44013 + }, + { + "epoch": 0.44014, + "grad_norm": 0.6945158014450183, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 44014 + }, + { + "epoch": 0.44015, + "grad_norm": 0.6464451288198497, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 44015 + }, + { + "epoch": 0.44016, + "grad_norm": 0.5789712178492931, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 44016 + }, + { + "epoch": 0.44017, + "grad_norm": 0.6284935479605623, + "learning_rate": 0.003, + "loss": 3.962, + "step": 44017 + }, + { + "epoch": 0.44018, + "grad_norm": 0.6877099033993164, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 44018 + }, + { + "epoch": 0.44019, + "grad_norm": 0.7945497564211226, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 44019 + }, + { + "epoch": 0.4402, + "grad_norm": 0.8751375454240191, + "learning_rate": 0.003, + "loss": 3.964, + "step": 44020 + }, + { + "epoch": 0.44021, + "grad_norm": 0.8894970285315484, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 44021 + }, + { + "epoch": 0.44022, + "grad_norm": 0.8506252897529855, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 44022 + }, + { + "epoch": 0.44023, + "grad_norm": 0.8420390757711383, + "learning_rate": 0.003, + "loss": 3.998, + "step": 44023 + }, + { + "epoch": 0.44024, + "grad_norm": 0.7908543007451483, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 44024 + }, + { + "epoch": 0.44025, + "grad_norm": 0.906458005565148, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 44025 + }, + { + "epoch": 0.44026, + "grad_norm": 0.9016793819281874, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 44026 + }, + { + "epoch": 0.44027, + "grad_norm": 0.8752285308116536, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44027 + }, + { + "epoch": 0.44028, + "grad_norm": 0.805630861575761, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 44028 + }, + { + "epoch": 0.44029, + "grad_norm": 0.7088336975787743, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 44029 + }, + { + "epoch": 0.4403, + "grad_norm": 0.5704007150590047, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 44030 + }, + { + "epoch": 0.44031, + "grad_norm": 0.5829933115229191, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 44031 + }, + { + "epoch": 0.44032, + "grad_norm": 0.6807984280066959, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 44032 + }, + { + "epoch": 0.44033, + "grad_norm": 0.8308128290002884, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 44033 + }, + { + "epoch": 0.44034, + "grad_norm": 1.0190185367519382, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 44034 + }, + { + "epoch": 0.44035, + "grad_norm": 1.1405356225428764, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 44035 + }, + { + "epoch": 0.44036, + "grad_norm": 0.7030595497389397, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 44036 + }, + { + "epoch": 0.44037, + "grad_norm": 0.6578695811856721, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 44037 + }, + { + "epoch": 0.44038, + "grad_norm": 0.8356472842027861, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 44038 + }, + { + "epoch": 0.44039, + "grad_norm": 0.8844294236174959, + "learning_rate": 0.003, + "loss": 3.9357, + "step": 44039 + }, + { + "epoch": 0.4404, + "grad_norm": 0.9087839645872227, + "learning_rate": 0.003, + "loss": 4.001, + "step": 44040 + }, + { + "epoch": 0.44041, + "grad_norm": 0.8636617321251152, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 44041 + }, + { + "epoch": 0.44042, + "grad_norm": 0.8285502601553467, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 44042 + }, + { + "epoch": 0.44043, + "grad_norm": 0.9065793860474104, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44043 + }, + { + "epoch": 0.44044, + "grad_norm": 0.8770962459684437, + "learning_rate": 0.003, + "loss": 3.965, + "step": 44044 + }, + { + "epoch": 0.44045, + "grad_norm": 0.8657140664689139, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 44045 + }, + { + "epoch": 0.44046, + "grad_norm": 1.0277330121083057, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 44046 + }, + { + "epoch": 0.44047, + "grad_norm": 0.884553796536937, + "learning_rate": 0.003, + "loss": 3.968, + "step": 44047 + }, + { + "epoch": 0.44048, + "grad_norm": 0.7943318188486883, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 44048 + }, + { + "epoch": 0.44049, + "grad_norm": 0.7016679569444341, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 44049 + }, + { + "epoch": 0.4405, + "grad_norm": 0.6661755936121847, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 44050 + }, + { + "epoch": 0.44051, + "grad_norm": 0.7072019550106178, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 44051 + }, + { + "epoch": 0.44052, + "grad_norm": 0.8035432431054477, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 44052 + }, + { + "epoch": 0.44053, + "grad_norm": 0.7809463111380484, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 44053 + }, + { + "epoch": 0.44054, + "grad_norm": 0.7066100419840644, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 44054 + }, + { + "epoch": 0.44055, + "grad_norm": 0.7623980359090659, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 44055 + }, + { + "epoch": 0.44056, + "grad_norm": 0.677424395490742, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 44056 + }, + { + "epoch": 0.44057, + "grad_norm": 0.5524029028471831, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 44057 + }, + { + "epoch": 0.44058, + "grad_norm": 0.5927467199665403, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 44058 + }, + { + "epoch": 0.44059, + "grad_norm": 0.6492914521442849, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 44059 + }, + { + "epoch": 0.4406, + "grad_norm": 0.704233798665219, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 44060 + }, + { + "epoch": 0.44061, + "grad_norm": 0.8035863540769614, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 44061 + }, + { + "epoch": 0.44062, + "grad_norm": 0.8492342918526264, + "learning_rate": 0.003, + "loss": 3.9441, + "step": 44062 + }, + { + "epoch": 0.44063, + "grad_norm": 0.826450636333069, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 44063 + }, + { + "epoch": 0.44064, + "grad_norm": 0.8768793983246231, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 44064 + }, + { + "epoch": 0.44065, + "grad_norm": 0.9400872918347242, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 44065 + }, + { + "epoch": 0.44066, + "grad_norm": 0.8745352342215353, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 44066 + }, + { + "epoch": 0.44067, + "grad_norm": 0.755314507879027, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 44067 + }, + { + "epoch": 0.44068, + "grad_norm": 0.6984694398898484, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 44068 + }, + { + "epoch": 0.44069, + "grad_norm": 0.594497878246022, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 44069 + }, + { + "epoch": 0.4407, + "grad_norm": 0.5639802400174099, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 44070 + }, + { + "epoch": 0.44071, + "grad_norm": 0.6483608478760187, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 44071 + }, + { + "epoch": 0.44072, + "grad_norm": 0.782871452033233, + "learning_rate": 0.003, + "loss": 3.974, + "step": 44072 + }, + { + "epoch": 0.44073, + "grad_norm": 0.9904722008817781, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 44073 + }, + { + "epoch": 0.44074, + "grad_norm": 1.0154846887955453, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 44074 + }, + { + "epoch": 0.44075, + "grad_norm": 0.8291023553926932, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 44075 + }, + { + "epoch": 0.44076, + "grad_norm": 0.7135855598173088, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 44076 + }, + { + "epoch": 0.44077, + "grad_norm": 0.7383131493607593, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 44077 + }, + { + "epoch": 0.44078, + "grad_norm": 0.7060823182064166, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 44078 + }, + { + "epoch": 0.44079, + "grad_norm": 0.6878193408380735, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 44079 + }, + { + "epoch": 0.4408, + "grad_norm": 0.7785567947120173, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 44080 + }, + { + "epoch": 0.44081, + "grad_norm": 0.8332288485854836, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 44081 + }, + { + "epoch": 0.44082, + "grad_norm": 0.9202271018647985, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 44082 + }, + { + "epoch": 0.44083, + "grad_norm": 0.8228831820499828, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 44083 + }, + { + "epoch": 0.44084, + "grad_norm": 0.7036728640363981, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 44084 + }, + { + "epoch": 0.44085, + "grad_norm": 0.7105108380925366, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 44085 + }, + { + "epoch": 0.44086, + "grad_norm": 0.6930818080398106, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 44086 + }, + { + "epoch": 0.44087, + "grad_norm": 0.6845959974091135, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 44087 + }, + { + "epoch": 0.44088, + "grad_norm": 0.705883073703111, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 44088 + }, + { + "epoch": 0.44089, + "grad_norm": 0.7181728910408378, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 44089 + }, + { + "epoch": 0.4409, + "grad_norm": 0.7266047461150416, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 44090 + }, + { + "epoch": 0.44091, + "grad_norm": 0.7812820854712703, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 44091 + }, + { + "epoch": 0.44092, + "grad_norm": 0.8548212786082016, + "learning_rate": 0.003, + "loss": 3.967, + "step": 44092 + }, + { + "epoch": 0.44093, + "grad_norm": 0.8409666982780793, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 44093 + }, + { + "epoch": 0.44094, + "grad_norm": 0.8372071641401473, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 44094 + }, + { + "epoch": 0.44095, + "grad_norm": 1.0615389360324454, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 44095 + }, + { + "epoch": 0.44096, + "grad_norm": 1.2054118493370325, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 44096 + }, + { + "epoch": 0.44097, + "grad_norm": 0.9576714827032187, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 44097 + }, + { + "epoch": 0.44098, + "grad_norm": 0.7899490128046909, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 44098 + }, + { + "epoch": 0.44099, + "grad_norm": 0.6661166461649657, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 44099 + }, + { + "epoch": 0.441, + "grad_norm": 0.6442381334239727, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 44100 + }, + { + "epoch": 0.44101, + "grad_norm": 0.6202999709538991, + "learning_rate": 0.003, + "loss": 4.008, + "step": 44101 + }, + { + "epoch": 0.44102, + "grad_norm": 0.6594586962190045, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 44102 + }, + { + "epoch": 0.44103, + "grad_norm": 0.6967463817090637, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 44103 + }, + { + "epoch": 0.44104, + "grad_norm": 0.7187175627369483, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 44104 + }, + { + "epoch": 0.44105, + "grad_norm": 0.660681575808519, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 44105 + }, + { + "epoch": 0.44106, + "grad_norm": 0.6150423344804236, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 44106 + }, + { + "epoch": 0.44107, + "grad_norm": 0.5880926881196649, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 44107 + }, + { + "epoch": 0.44108, + "grad_norm": 0.6374927746983479, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 44108 + }, + { + "epoch": 0.44109, + "grad_norm": 0.7281237356839388, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 44109 + }, + { + "epoch": 0.4411, + "grad_norm": 0.8547379926818963, + "learning_rate": 0.003, + "loss": 3.991, + "step": 44110 + }, + { + "epoch": 0.44111, + "grad_norm": 0.9251560846780928, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 44111 + }, + { + "epoch": 0.44112, + "grad_norm": 0.8901072891373589, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 44112 + }, + { + "epoch": 0.44113, + "grad_norm": 0.6813452699744256, + "learning_rate": 0.003, + "loss": 3.984, + "step": 44113 + }, + { + "epoch": 0.44114, + "grad_norm": 0.6722347003953903, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 44114 + }, + { + "epoch": 0.44115, + "grad_norm": 0.6738436249840442, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 44115 + }, + { + "epoch": 0.44116, + "grad_norm": 0.7347843781980067, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 44116 + }, + { + "epoch": 0.44117, + "grad_norm": 0.7431949204699839, + "learning_rate": 0.003, + "loss": 3.983, + "step": 44117 + }, + { + "epoch": 0.44118, + "grad_norm": 0.6617839554776477, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 44118 + }, + { + "epoch": 0.44119, + "grad_norm": 0.6778765236695368, + "learning_rate": 0.003, + "loss": 3.982, + "step": 44119 + }, + { + "epoch": 0.4412, + "grad_norm": 0.7963966807068735, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44120 + }, + { + "epoch": 0.44121, + "grad_norm": 0.8754624216096788, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 44121 + }, + { + "epoch": 0.44122, + "grad_norm": 0.8719082995455215, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 44122 + }, + { + "epoch": 0.44123, + "grad_norm": 0.8517820907997657, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 44123 + }, + { + "epoch": 0.44124, + "grad_norm": 0.8737762374080946, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 44124 + }, + { + "epoch": 0.44125, + "grad_norm": 0.8898128837010432, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 44125 + }, + { + "epoch": 0.44126, + "grad_norm": 1.0189496273502225, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 44126 + }, + { + "epoch": 0.44127, + "grad_norm": 0.9670160186791805, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 44127 + }, + { + "epoch": 0.44128, + "grad_norm": 0.8790668421587984, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 44128 + }, + { + "epoch": 0.44129, + "grad_norm": 0.8638098340820166, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 44129 + }, + { + "epoch": 0.4413, + "grad_norm": 0.8726321990029822, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 44130 + }, + { + "epoch": 0.44131, + "grad_norm": 0.8846280224092512, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 44131 + }, + { + "epoch": 0.44132, + "grad_norm": 1.0027751923063632, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 44132 + }, + { + "epoch": 0.44133, + "grad_norm": 1.0516061729212853, + "learning_rate": 0.003, + "loss": 3.986, + "step": 44133 + }, + { + "epoch": 0.44134, + "grad_norm": 0.9547536354270804, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 44134 + }, + { + "epoch": 0.44135, + "grad_norm": 1.0179284463767602, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 44135 + }, + { + "epoch": 0.44136, + "grad_norm": 1.1350379235112733, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 44136 + }, + { + "epoch": 0.44137, + "grad_norm": 0.9498897745593179, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 44137 + }, + { + "epoch": 0.44138, + "grad_norm": 1.036840063248988, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 44138 + }, + { + "epoch": 0.44139, + "grad_norm": 0.8667916517474473, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 44139 + }, + { + "epoch": 0.4414, + "grad_norm": 0.8507500676631468, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 44140 + }, + { + "epoch": 0.44141, + "grad_norm": 0.8648985234892679, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 44141 + }, + { + "epoch": 0.44142, + "grad_norm": 0.8543076176521331, + "learning_rate": 0.003, + "loss": 3.997, + "step": 44142 + }, + { + "epoch": 0.44143, + "grad_norm": 0.8698300407365894, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 44143 + }, + { + "epoch": 0.44144, + "grad_norm": 0.8529645044155977, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 44144 + }, + { + "epoch": 0.44145, + "grad_norm": 0.7881918585732766, + "learning_rate": 0.003, + "loss": 3.986, + "step": 44145 + }, + { + "epoch": 0.44146, + "grad_norm": 0.8349405123124849, + "learning_rate": 0.003, + "loss": 3.981, + "step": 44146 + }, + { + "epoch": 0.44147, + "grad_norm": 0.8674867775992976, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 44147 + }, + { + "epoch": 0.44148, + "grad_norm": 0.6961258720133137, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 44148 + }, + { + "epoch": 0.44149, + "grad_norm": 0.6457826148428525, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 44149 + }, + { + "epoch": 0.4415, + "grad_norm": 0.7816089891230756, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 44150 + }, + { + "epoch": 0.44151, + "grad_norm": 0.8797219125926335, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 44151 + }, + { + "epoch": 0.44152, + "grad_norm": 0.8291832867615243, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 44152 + }, + { + "epoch": 0.44153, + "grad_norm": 0.7021332270594716, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44153 + }, + { + "epoch": 0.44154, + "grad_norm": 0.6529124803452273, + "learning_rate": 0.003, + "loss": 4.0193, + "step": 44154 + }, + { + "epoch": 0.44155, + "grad_norm": 0.6329882683569019, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 44155 + }, + { + "epoch": 0.44156, + "grad_norm": 0.6223160810602894, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 44156 + }, + { + "epoch": 0.44157, + "grad_norm": 0.6224068388823308, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 44157 + }, + { + "epoch": 0.44158, + "grad_norm": 0.5657746834633933, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 44158 + }, + { + "epoch": 0.44159, + "grad_norm": 0.5223397428170046, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44159 + }, + { + "epoch": 0.4416, + "grad_norm": 0.48170783214898816, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 44160 + }, + { + "epoch": 0.44161, + "grad_norm": 0.523864476495168, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 44161 + }, + { + "epoch": 0.44162, + "grad_norm": 0.5375601996888708, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 44162 + }, + { + "epoch": 0.44163, + "grad_norm": 0.5739674714209829, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 44163 + }, + { + "epoch": 0.44164, + "grad_norm": 0.5795815204349591, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 44164 + }, + { + "epoch": 0.44165, + "grad_norm": 0.602206994649287, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 44165 + }, + { + "epoch": 0.44166, + "grad_norm": 0.7234355882553221, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 44166 + }, + { + "epoch": 0.44167, + "grad_norm": 0.9126596634109094, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 44167 + }, + { + "epoch": 0.44168, + "grad_norm": 0.9594103999482672, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 44168 + }, + { + "epoch": 0.44169, + "grad_norm": 0.9251391073920786, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 44169 + }, + { + "epoch": 0.4417, + "grad_norm": 1.0281527113962337, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 44170 + }, + { + "epoch": 0.44171, + "grad_norm": 0.8947747547984048, + "learning_rate": 0.003, + "loss": 3.998, + "step": 44171 + }, + { + "epoch": 0.44172, + "grad_norm": 0.9567777699887758, + "learning_rate": 0.003, + "loss": 3.985, + "step": 44172 + }, + { + "epoch": 0.44173, + "grad_norm": 0.9776198489262139, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 44173 + }, + { + "epoch": 0.44174, + "grad_norm": 0.8967480482591828, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 44174 + }, + { + "epoch": 0.44175, + "grad_norm": 0.8414757916040686, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 44175 + }, + { + "epoch": 0.44176, + "grad_norm": 0.865861328657681, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 44176 + }, + { + "epoch": 0.44177, + "grad_norm": 0.9364217898586888, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 44177 + }, + { + "epoch": 0.44178, + "grad_norm": 0.9562030634929776, + "learning_rate": 0.003, + "loss": 3.965, + "step": 44178 + }, + { + "epoch": 0.44179, + "grad_norm": 1.140839388586871, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 44179 + }, + { + "epoch": 0.4418, + "grad_norm": 0.924839366987203, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 44180 + }, + { + "epoch": 0.44181, + "grad_norm": 0.7831718475656552, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 44181 + }, + { + "epoch": 0.44182, + "grad_norm": 0.712721021079328, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 44182 + }, + { + "epoch": 0.44183, + "grad_norm": 0.7112198164580625, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 44183 + }, + { + "epoch": 0.44184, + "grad_norm": 0.6600597247759069, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 44184 + }, + { + "epoch": 0.44185, + "grad_norm": 0.6412288845127562, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 44185 + }, + { + "epoch": 0.44186, + "grad_norm": 0.6215302424659533, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 44186 + }, + { + "epoch": 0.44187, + "grad_norm": 0.648667948972537, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 44187 + }, + { + "epoch": 0.44188, + "grad_norm": 0.7380676894302057, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 44188 + }, + { + "epoch": 0.44189, + "grad_norm": 0.7160278815442446, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 44189 + }, + { + "epoch": 0.4419, + "grad_norm": 0.7449562486088235, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 44190 + }, + { + "epoch": 0.44191, + "grad_norm": 0.6865840584995063, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 44191 + }, + { + "epoch": 0.44192, + "grad_norm": 0.7230913796705158, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 44192 + }, + { + "epoch": 0.44193, + "grad_norm": 0.7776301416198986, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 44193 + }, + { + "epoch": 0.44194, + "grad_norm": 0.8082129142272679, + "learning_rate": 0.003, + "loss": 3.995, + "step": 44194 + }, + { + "epoch": 0.44195, + "grad_norm": 0.8057891428928323, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 44195 + }, + { + "epoch": 0.44196, + "grad_norm": 0.7976350700948882, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 44196 + }, + { + "epoch": 0.44197, + "grad_norm": 0.8176046971177572, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 44197 + }, + { + "epoch": 0.44198, + "grad_norm": 0.7963858392845532, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 44198 + }, + { + "epoch": 0.44199, + "grad_norm": 0.7884102732607587, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 44199 + }, + { + "epoch": 0.442, + "grad_norm": 0.6456430500706233, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 44200 + }, + { + "epoch": 0.44201, + "grad_norm": 0.6328983093951063, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 44201 + }, + { + "epoch": 0.44202, + "grad_norm": 0.6275908447112063, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 44202 + }, + { + "epoch": 0.44203, + "grad_norm": 0.6975743826880854, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 44203 + }, + { + "epoch": 0.44204, + "grad_norm": 0.7974294569254198, + "learning_rate": 0.003, + "loss": 3.969, + "step": 44204 + }, + { + "epoch": 0.44205, + "grad_norm": 0.9119631650235961, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 44205 + }, + { + "epoch": 0.44206, + "grad_norm": 0.9672375803026566, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 44206 + }, + { + "epoch": 0.44207, + "grad_norm": 1.0822039943887047, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 44207 + }, + { + "epoch": 0.44208, + "grad_norm": 0.9730333004522812, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 44208 + }, + { + "epoch": 0.44209, + "grad_norm": 0.8398568833321725, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 44209 + }, + { + "epoch": 0.4421, + "grad_norm": 0.8886968518458063, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 44210 + }, + { + "epoch": 0.44211, + "grad_norm": 1.044790461939842, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44211 + }, + { + "epoch": 0.44212, + "grad_norm": 0.9406181698161902, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 44212 + }, + { + "epoch": 0.44213, + "grad_norm": 0.8753871157754528, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 44213 + }, + { + "epoch": 0.44214, + "grad_norm": 0.8240524284739219, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 44214 + }, + { + "epoch": 0.44215, + "grad_norm": 0.850923048756946, + "learning_rate": 0.003, + "loss": 3.982, + "step": 44215 + }, + { + "epoch": 0.44216, + "grad_norm": 0.792582669425812, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 44216 + }, + { + "epoch": 0.44217, + "grad_norm": 0.79487608683786, + "learning_rate": 0.003, + "loss": 3.961, + "step": 44217 + }, + { + "epoch": 0.44218, + "grad_norm": 0.7505820180731113, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 44218 + }, + { + "epoch": 0.44219, + "grad_norm": 0.6962211828090953, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 44219 + }, + { + "epoch": 0.4422, + "grad_norm": 0.6519337555874116, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 44220 + }, + { + "epoch": 0.44221, + "grad_norm": 0.6103342269837113, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 44221 + }, + { + "epoch": 0.44222, + "grad_norm": 0.5613038504649012, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 44222 + }, + { + "epoch": 0.44223, + "grad_norm": 0.6418557210834762, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 44223 + }, + { + "epoch": 0.44224, + "grad_norm": 0.7055650757720944, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 44224 + }, + { + "epoch": 0.44225, + "grad_norm": 0.7962578648614788, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 44225 + }, + { + "epoch": 0.44226, + "grad_norm": 0.7937546441591931, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 44226 + }, + { + "epoch": 0.44227, + "grad_norm": 0.7382598250996889, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 44227 + }, + { + "epoch": 0.44228, + "grad_norm": 0.7315186396618482, + "learning_rate": 0.003, + "loss": 3.97, + "step": 44228 + }, + { + "epoch": 0.44229, + "grad_norm": 0.7297095132615647, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44229 + }, + { + "epoch": 0.4423, + "grad_norm": 0.6869623257225518, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 44230 + }, + { + "epoch": 0.44231, + "grad_norm": 0.7180815575628391, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 44231 + }, + { + "epoch": 0.44232, + "grad_norm": 0.7374163738280016, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 44232 + }, + { + "epoch": 0.44233, + "grad_norm": 0.8019288013119722, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 44233 + }, + { + "epoch": 0.44234, + "grad_norm": 0.809461100574732, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 44234 + }, + { + "epoch": 0.44235, + "grad_norm": 0.7771489197592376, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 44235 + }, + { + "epoch": 0.44236, + "grad_norm": 0.7119032663363846, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 44236 + }, + { + "epoch": 0.44237, + "grad_norm": 0.9621861232594209, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 44237 + }, + { + "epoch": 0.44238, + "grad_norm": 1.1721326560069725, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 44238 + }, + { + "epoch": 0.44239, + "grad_norm": 0.7385636638314181, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 44239 + }, + { + "epoch": 0.4424, + "grad_norm": 0.6047995089688989, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 44240 + }, + { + "epoch": 0.44241, + "grad_norm": 0.7307079409936245, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 44241 + }, + { + "epoch": 0.44242, + "grad_norm": 0.6594566175224575, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 44242 + }, + { + "epoch": 0.44243, + "grad_norm": 0.6362947906797926, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 44243 + }, + { + "epoch": 0.44244, + "grad_norm": 0.6000365493529944, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 44244 + }, + { + "epoch": 0.44245, + "grad_norm": 0.5498137990105834, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 44245 + }, + { + "epoch": 0.44246, + "grad_norm": 0.5089064533352949, + "learning_rate": 0.003, + "loss": 3.97, + "step": 44246 + }, + { + "epoch": 0.44247, + "grad_norm": 0.5226428560369542, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 44247 + }, + { + "epoch": 0.44248, + "grad_norm": 0.6399384996225673, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 44248 + }, + { + "epoch": 0.44249, + "grad_norm": 0.7085411296852466, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 44249 + }, + { + "epoch": 0.4425, + "grad_norm": 0.791262779938607, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 44250 + }, + { + "epoch": 0.44251, + "grad_norm": 0.7789941720966038, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 44251 + }, + { + "epoch": 0.44252, + "grad_norm": 0.8113076198859681, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 44252 + }, + { + "epoch": 0.44253, + "grad_norm": 0.8650283322566232, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 44253 + }, + { + "epoch": 0.44254, + "grad_norm": 0.8074776052431772, + "learning_rate": 0.003, + "loss": 3.983, + "step": 44254 + }, + { + "epoch": 0.44255, + "grad_norm": 0.8487661660443093, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 44255 + }, + { + "epoch": 0.44256, + "grad_norm": 1.0336915265022748, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 44256 + }, + { + "epoch": 0.44257, + "grad_norm": 0.8392084963962241, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 44257 + }, + { + "epoch": 0.44258, + "grad_norm": 0.644356635890114, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 44258 + }, + { + "epoch": 0.44259, + "grad_norm": 0.619115998664973, + "learning_rate": 0.003, + "loss": 3.9385, + "step": 44259 + }, + { + "epoch": 0.4426, + "grad_norm": 0.7180880733304992, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 44260 + }, + { + "epoch": 0.44261, + "grad_norm": 0.8502815402192455, + "learning_rate": 0.003, + "loss": 3.9385, + "step": 44261 + }, + { + "epoch": 0.44262, + "grad_norm": 0.9362570063967891, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 44262 + }, + { + "epoch": 0.44263, + "grad_norm": 0.8833538830840445, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 44263 + }, + { + "epoch": 0.44264, + "grad_norm": 0.7783048807436534, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 44264 + }, + { + "epoch": 0.44265, + "grad_norm": 0.7472119825426561, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 44265 + }, + { + "epoch": 0.44266, + "grad_norm": 0.6987855700058703, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 44266 + }, + { + "epoch": 0.44267, + "grad_norm": 0.6796788120556698, + "learning_rate": 0.003, + "loss": 3.9404, + "step": 44267 + }, + { + "epoch": 0.44268, + "grad_norm": 0.6511681699380192, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 44268 + }, + { + "epoch": 0.44269, + "grad_norm": 0.6757130681995808, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 44269 + }, + { + "epoch": 0.4427, + "grad_norm": 0.7750071329003059, + "learning_rate": 0.003, + "loss": 3.9404, + "step": 44270 + }, + { + "epoch": 0.44271, + "grad_norm": 0.9027994564113682, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 44271 + }, + { + "epoch": 0.44272, + "grad_norm": 1.0358489072112032, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 44272 + }, + { + "epoch": 0.44273, + "grad_norm": 1.0080178540580431, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 44273 + }, + { + "epoch": 0.44274, + "grad_norm": 0.843934115743556, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 44274 + }, + { + "epoch": 0.44275, + "grad_norm": 0.9099010693961626, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 44275 + }, + { + "epoch": 0.44276, + "grad_norm": 0.9697140374330521, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 44276 + }, + { + "epoch": 0.44277, + "grad_norm": 0.9635778060153919, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 44277 + }, + { + "epoch": 0.44278, + "grad_norm": 0.9252244782367163, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 44278 + }, + { + "epoch": 0.44279, + "grad_norm": 0.836793496825614, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 44279 + }, + { + "epoch": 0.4428, + "grad_norm": 0.8450816895664083, + "learning_rate": 0.003, + "loss": 4.006, + "step": 44280 + }, + { + "epoch": 0.44281, + "grad_norm": 0.8697112912910571, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 44281 + }, + { + "epoch": 0.44282, + "grad_norm": 0.7988495104950489, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 44282 + }, + { + "epoch": 0.44283, + "grad_norm": 0.8090955060552593, + "learning_rate": 0.003, + "loss": 3.959, + "step": 44283 + }, + { + "epoch": 0.44284, + "grad_norm": 0.9222067971567149, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 44284 + }, + { + "epoch": 0.44285, + "grad_norm": 1.0030942171997879, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 44285 + }, + { + "epoch": 0.44286, + "grad_norm": 0.9753748661676324, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 44286 + }, + { + "epoch": 0.44287, + "grad_norm": 0.9343460858035205, + "learning_rate": 0.003, + "loss": 4.0235, + "step": 44287 + }, + { + "epoch": 0.44288, + "grad_norm": 0.9706217790223606, + "learning_rate": 0.003, + "loss": 3.995, + "step": 44288 + }, + { + "epoch": 0.44289, + "grad_norm": 0.8888345219350157, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 44289 + }, + { + "epoch": 0.4429, + "grad_norm": 0.8529360560128146, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 44290 + }, + { + "epoch": 0.44291, + "grad_norm": 0.7722217187952513, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 44291 + }, + { + "epoch": 0.44292, + "grad_norm": 0.7956355240225144, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 44292 + }, + { + "epoch": 0.44293, + "grad_norm": 0.8422591289940496, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44293 + }, + { + "epoch": 0.44294, + "grad_norm": 0.8466009825095396, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 44294 + }, + { + "epoch": 0.44295, + "grad_norm": 0.9182444644457636, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 44295 + }, + { + "epoch": 0.44296, + "grad_norm": 0.8540204094714181, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 44296 + }, + { + "epoch": 0.44297, + "grad_norm": 0.7270257391768886, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 44297 + }, + { + "epoch": 0.44298, + "grad_norm": 0.6976738346996708, + "learning_rate": 0.003, + "loss": 3.975, + "step": 44298 + }, + { + "epoch": 0.44299, + "grad_norm": 0.6435810253777243, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44299 + }, + { + "epoch": 0.443, + "grad_norm": 0.6222768858700601, + "learning_rate": 0.003, + "loss": 3.99, + "step": 44300 + }, + { + "epoch": 0.44301, + "grad_norm": 0.5978973678916016, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 44301 + }, + { + "epoch": 0.44302, + "grad_norm": 0.6317196029416583, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 44302 + }, + { + "epoch": 0.44303, + "grad_norm": 0.5901213693951772, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 44303 + }, + { + "epoch": 0.44304, + "grad_norm": 0.539336758870111, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 44304 + }, + { + "epoch": 0.44305, + "grad_norm": 0.4665407814742129, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 44305 + }, + { + "epoch": 0.44306, + "grad_norm": 0.42466439666756484, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 44306 + }, + { + "epoch": 0.44307, + "grad_norm": 0.45414242174212865, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44307 + }, + { + "epoch": 0.44308, + "grad_norm": 0.6028629110948241, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 44308 + }, + { + "epoch": 0.44309, + "grad_norm": 0.7661346455736391, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 44309 + }, + { + "epoch": 0.4431, + "grad_norm": 0.9093473623256495, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 44310 + }, + { + "epoch": 0.44311, + "grad_norm": 1.0007568199105734, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 44311 + }, + { + "epoch": 0.44312, + "grad_norm": 0.9296725562240129, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 44312 + }, + { + "epoch": 0.44313, + "grad_norm": 0.730500336117861, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 44313 + }, + { + "epoch": 0.44314, + "grad_norm": 0.7382635602117876, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 44314 + }, + { + "epoch": 0.44315, + "grad_norm": 0.6661190903118928, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 44315 + }, + { + "epoch": 0.44316, + "grad_norm": 0.7224609270419835, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 44316 + }, + { + "epoch": 0.44317, + "grad_norm": 0.7358820052781267, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 44317 + }, + { + "epoch": 0.44318, + "grad_norm": 0.8662922003058688, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 44318 + }, + { + "epoch": 0.44319, + "grad_norm": 1.0236513008306296, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 44319 + }, + { + "epoch": 0.4432, + "grad_norm": 1.0510356136338128, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 44320 + }, + { + "epoch": 0.44321, + "grad_norm": 0.9493405245660453, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 44321 + }, + { + "epoch": 0.44322, + "grad_norm": 0.7913886104960268, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 44322 + }, + { + "epoch": 0.44323, + "grad_norm": 0.8530863729843552, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44323 + }, + { + "epoch": 0.44324, + "grad_norm": 0.7368656852190537, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 44324 + }, + { + "epoch": 0.44325, + "grad_norm": 0.6628758268195144, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 44325 + }, + { + "epoch": 0.44326, + "grad_norm": 0.6975885755579139, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 44326 + }, + { + "epoch": 0.44327, + "grad_norm": 0.760080942348174, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 44327 + }, + { + "epoch": 0.44328, + "grad_norm": 0.7789509294781938, + "learning_rate": 0.003, + "loss": 3.9218, + "step": 44328 + }, + { + "epoch": 0.44329, + "grad_norm": 0.6998401166813357, + "learning_rate": 0.003, + "loss": 3.992, + "step": 44329 + }, + { + "epoch": 0.4433, + "grad_norm": 0.8350675487314154, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 44330 + }, + { + "epoch": 0.44331, + "grad_norm": 1.0483102013321508, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 44331 + }, + { + "epoch": 0.44332, + "grad_norm": 1.0505046590711724, + "learning_rate": 0.003, + "loss": 4.007, + "step": 44332 + }, + { + "epoch": 0.44333, + "grad_norm": 0.9632955166488603, + "learning_rate": 0.003, + "loss": 3.993, + "step": 44333 + }, + { + "epoch": 0.44334, + "grad_norm": 1.0212974544649058, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 44334 + }, + { + "epoch": 0.44335, + "grad_norm": 0.9457144377105402, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 44335 + }, + { + "epoch": 0.44336, + "grad_norm": 0.8741833306936824, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44336 + }, + { + "epoch": 0.44337, + "grad_norm": 0.8046882466353631, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 44337 + }, + { + "epoch": 0.44338, + "grad_norm": 0.7115241801913512, + "learning_rate": 0.003, + "loss": 3.9464, + "step": 44338 + }, + { + "epoch": 0.44339, + "grad_norm": 0.795009734076667, + "learning_rate": 0.003, + "loss": 3.955, + "step": 44339 + }, + { + "epoch": 0.4434, + "grad_norm": 0.7435812020924356, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 44340 + }, + { + "epoch": 0.44341, + "grad_norm": 0.7548756115580013, + "learning_rate": 0.003, + "loss": 3.982, + "step": 44341 + }, + { + "epoch": 0.44342, + "grad_norm": 0.6438725644477055, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 44342 + }, + { + "epoch": 0.44343, + "grad_norm": 0.6528458327355283, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 44343 + }, + { + "epoch": 0.44344, + "grad_norm": 0.8013580976790076, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 44344 + }, + { + "epoch": 0.44345, + "grad_norm": 0.8991762630286594, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 44345 + }, + { + "epoch": 0.44346, + "grad_norm": 0.8200719725353272, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 44346 + }, + { + "epoch": 0.44347, + "grad_norm": 0.776281612130928, + "learning_rate": 0.003, + "loss": 3.9475, + "step": 44347 + }, + { + "epoch": 0.44348, + "grad_norm": 0.8172182261615262, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 44348 + }, + { + "epoch": 0.44349, + "grad_norm": 0.9198200546639098, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 44349 + }, + { + "epoch": 0.4435, + "grad_norm": 0.8825397130642076, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 44350 + }, + { + "epoch": 0.44351, + "grad_norm": 0.8404923929643124, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 44351 + }, + { + "epoch": 0.44352, + "grad_norm": 0.8817778990280306, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 44352 + }, + { + "epoch": 0.44353, + "grad_norm": 0.8531558130930865, + "learning_rate": 0.003, + "loss": 3.98, + "step": 44353 + }, + { + "epoch": 0.44354, + "grad_norm": 0.7844105960915392, + "learning_rate": 0.003, + "loss": 3.987, + "step": 44354 + }, + { + "epoch": 0.44355, + "grad_norm": 0.7246333579678381, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 44355 + }, + { + "epoch": 0.44356, + "grad_norm": 0.6451150103177884, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 44356 + }, + { + "epoch": 0.44357, + "grad_norm": 0.6889270525478438, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 44357 + }, + { + "epoch": 0.44358, + "grad_norm": 0.5936667879803217, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 44358 + }, + { + "epoch": 0.44359, + "grad_norm": 0.722584264671611, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 44359 + }, + { + "epoch": 0.4436, + "grad_norm": 0.7576833930971811, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 44360 + }, + { + "epoch": 0.44361, + "grad_norm": 0.87541436396705, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 44361 + }, + { + "epoch": 0.44362, + "grad_norm": 1.0551583509734976, + "learning_rate": 0.003, + "loss": 4.008, + "step": 44362 + }, + { + "epoch": 0.44363, + "grad_norm": 0.8932610040061023, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 44363 + }, + { + "epoch": 0.44364, + "grad_norm": 0.7384825722891485, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 44364 + }, + { + "epoch": 0.44365, + "grad_norm": 0.7360787126269055, + "learning_rate": 0.003, + "loss": 3.9302, + "step": 44365 + }, + { + "epoch": 0.44366, + "grad_norm": 0.6249043588146092, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 44366 + }, + { + "epoch": 0.44367, + "grad_norm": 0.584593324889021, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 44367 + }, + { + "epoch": 0.44368, + "grad_norm": 0.5699344995467964, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 44368 + }, + { + "epoch": 0.44369, + "grad_norm": 0.5664278645120712, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 44369 + }, + { + "epoch": 0.4437, + "grad_norm": 0.5942314173277129, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 44370 + }, + { + "epoch": 0.44371, + "grad_norm": 0.6366712447611632, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 44371 + }, + { + "epoch": 0.44372, + "grad_norm": 0.6465664214153828, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 44372 + }, + { + "epoch": 0.44373, + "grad_norm": 0.6419405104147046, + "learning_rate": 0.003, + "loss": 3.988, + "step": 44373 + }, + { + "epoch": 0.44374, + "grad_norm": 0.7081460591861001, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 44374 + }, + { + "epoch": 0.44375, + "grad_norm": 0.896052534408632, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 44375 + }, + { + "epoch": 0.44376, + "grad_norm": 1.0007658195829618, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 44376 + }, + { + "epoch": 0.44377, + "grad_norm": 1.0825792986276552, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 44377 + }, + { + "epoch": 0.44378, + "grad_norm": 1.027624506805242, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 44378 + }, + { + "epoch": 0.44379, + "grad_norm": 0.8226701367457329, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 44379 + }, + { + "epoch": 0.4438, + "grad_norm": 0.7502328974517979, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 44380 + }, + { + "epoch": 0.44381, + "grad_norm": 0.7415924986285682, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 44381 + }, + { + "epoch": 0.44382, + "grad_norm": 0.8786489935103394, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 44382 + }, + { + "epoch": 0.44383, + "grad_norm": 1.024552391007323, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 44383 + }, + { + "epoch": 0.44384, + "grad_norm": 1.0525347544502215, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 44384 + }, + { + "epoch": 0.44385, + "grad_norm": 0.9540561673415903, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 44385 + }, + { + "epoch": 0.44386, + "grad_norm": 0.8898971134601255, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 44386 + }, + { + "epoch": 0.44387, + "grad_norm": 0.7732975052221828, + "learning_rate": 0.003, + "loss": 3.986, + "step": 44387 + }, + { + "epoch": 0.44388, + "grad_norm": 0.7047390111193784, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 44388 + }, + { + "epoch": 0.44389, + "grad_norm": 0.6522304732381693, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 44389 + }, + { + "epoch": 0.4439, + "grad_norm": 0.5981089312079934, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 44390 + }, + { + "epoch": 0.44391, + "grad_norm": 0.5601674448596979, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 44391 + }, + { + "epoch": 0.44392, + "grad_norm": 0.533261625447952, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 44392 + }, + { + "epoch": 0.44393, + "grad_norm": 0.5179075291374158, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 44393 + }, + { + "epoch": 0.44394, + "grad_norm": 0.556649303579655, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 44394 + }, + { + "epoch": 0.44395, + "grad_norm": 0.5498352748727808, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 44395 + }, + { + "epoch": 0.44396, + "grad_norm": 0.6084655775154555, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44396 + }, + { + "epoch": 0.44397, + "grad_norm": 0.6936708194875892, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 44397 + }, + { + "epoch": 0.44398, + "grad_norm": 0.7620072371370152, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 44398 + }, + { + "epoch": 0.44399, + "grad_norm": 0.8772568094532244, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 44399 + }, + { + "epoch": 0.444, + "grad_norm": 0.9743978548539411, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 44400 + }, + { + "epoch": 0.44401, + "grad_norm": 1.0921268010505372, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 44401 + }, + { + "epoch": 0.44402, + "grad_norm": 0.7419356216248655, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44402 + }, + { + "epoch": 0.44403, + "grad_norm": 0.6875558199795683, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 44403 + }, + { + "epoch": 0.44404, + "grad_norm": 0.7674695719236522, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 44404 + }, + { + "epoch": 0.44405, + "grad_norm": 0.7274429458754466, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 44405 + }, + { + "epoch": 0.44406, + "grad_norm": 0.7808475839322294, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 44406 + }, + { + "epoch": 0.44407, + "grad_norm": 0.7894944351743061, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 44407 + }, + { + "epoch": 0.44408, + "grad_norm": 0.7336581871874617, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 44408 + }, + { + "epoch": 0.44409, + "grad_norm": 0.7508289534701771, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 44409 + }, + { + "epoch": 0.4441, + "grad_norm": 0.7138977324279213, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 44410 + }, + { + "epoch": 0.44411, + "grad_norm": 0.8993749315503105, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 44411 + }, + { + "epoch": 0.44412, + "grad_norm": 0.9570951285614536, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 44412 + }, + { + "epoch": 0.44413, + "grad_norm": 0.9391878661428097, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 44413 + }, + { + "epoch": 0.44414, + "grad_norm": 0.8221104095880848, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 44414 + }, + { + "epoch": 0.44415, + "grad_norm": 0.8496932690205573, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 44415 + }, + { + "epoch": 0.44416, + "grad_norm": 0.8637656600744763, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44416 + }, + { + "epoch": 0.44417, + "grad_norm": 0.9132826703633675, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 44417 + }, + { + "epoch": 0.44418, + "grad_norm": 0.9721790413743298, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 44418 + }, + { + "epoch": 0.44419, + "grad_norm": 0.9804920655858747, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 44419 + }, + { + "epoch": 0.4442, + "grad_norm": 0.9368550621771505, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 44420 + }, + { + "epoch": 0.44421, + "grad_norm": 0.8936126471110469, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 44421 + }, + { + "epoch": 0.44422, + "grad_norm": 0.8862790645341462, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 44422 + }, + { + "epoch": 0.44423, + "grad_norm": 0.8645683942749777, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 44423 + }, + { + "epoch": 0.44424, + "grad_norm": 0.8854184245016923, + "learning_rate": 0.003, + "loss": 3.977, + "step": 44424 + }, + { + "epoch": 0.44425, + "grad_norm": 0.8850930001594611, + "learning_rate": 0.003, + "loss": 3.981, + "step": 44425 + }, + { + "epoch": 0.44426, + "grad_norm": 0.9275923378295465, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 44426 + }, + { + "epoch": 0.44427, + "grad_norm": 0.8444727074408503, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 44427 + }, + { + "epoch": 0.44428, + "grad_norm": 0.7854845188866401, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 44428 + }, + { + "epoch": 0.44429, + "grad_norm": 0.6743062244807776, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 44429 + }, + { + "epoch": 0.4443, + "grad_norm": 0.6434397799885386, + "learning_rate": 0.003, + "loss": 3.985, + "step": 44430 + }, + { + "epoch": 0.44431, + "grad_norm": 0.7321558298645664, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 44431 + }, + { + "epoch": 0.44432, + "grad_norm": 0.7573699136750205, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 44432 + }, + { + "epoch": 0.44433, + "grad_norm": 0.8111501978404103, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 44433 + }, + { + "epoch": 0.44434, + "grad_norm": 0.8415605871330221, + "learning_rate": 0.003, + "loss": 3.979, + "step": 44434 + }, + { + "epoch": 0.44435, + "grad_norm": 0.7817073797698545, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 44435 + }, + { + "epoch": 0.44436, + "grad_norm": 0.7609171281102507, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 44436 + }, + { + "epoch": 0.44437, + "grad_norm": 0.7023356766187554, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 44437 + }, + { + "epoch": 0.44438, + "grad_norm": 0.6640888337325918, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 44438 + }, + { + "epoch": 0.44439, + "grad_norm": 0.7558464927010289, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 44439 + }, + { + "epoch": 0.4444, + "grad_norm": 0.7453432425662108, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 44440 + }, + { + "epoch": 0.44441, + "grad_norm": 0.7313425035100414, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44441 + }, + { + "epoch": 0.44442, + "grad_norm": 0.7909161484339902, + "learning_rate": 0.003, + "loss": 3.998, + "step": 44442 + }, + { + "epoch": 0.44443, + "grad_norm": 0.8263380961447351, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 44443 + }, + { + "epoch": 0.44444, + "grad_norm": 0.8268770826183028, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 44444 + }, + { + "epoch": 0.44445, + "grad_norm": 0.7825076983144829, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 44445 + }, + { + "epoch": 0.44446, + "grad_norm": 0.7165390998282968, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 44446 + }, + { + "epoch": 0.44447, + "grad_norm": 0.6969742135308284, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 44447 + }, + { + "epoch": 0.44448, + "grad_norm": 0.6948946369743779, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 44448 + }, + { + "epoch": 0.44449, + "grad_norm": 0.6431025681900798, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 44449 + }, + { + "epoch": 0.4445, + "grad_norm": 0.5909499731361472, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 44450 + }, + { + "epoch": 0.44451, + "grad_norm": 0.5281773996028981, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 44451 + }, + { + "epoch": 0.44452, + "grad_norm": 0.4312226511127983, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 44452 + }, + { + "epoch": 0.44453, + "grad_norm": 0.4615856554448579, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 44453 + }, + { + "epoch": 0.44454, + "grad_norm": 0.48192852569698436, + "learning_rate": 0.003, + "loss": 3.941, + "step": 44454 + }, + { + "epoch": 0.44455, + "grad_norm": 0.5500622866709827, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 44455 + }, + { + "epoch": 0.44456, + "grad_norm": 0.6213013420517343, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 44456 + }, + { + "epoch": 0.44457, + "grad_norm": 0.7030285911720527, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 44457 + }, + { + "epoch": 0.44458, + "grad_norm": 0.7762327800775113, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 44458 + }, + { + "epoch": 0.44459, + "grad_norm": 0.7821881538907838, + "learning_rate": 0.003, + "loss": 3.9446, + "step": 44459 + }, + { + "epoch": 0.4446, + "grad_norm": 0.8745988997405096, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 44460 + }, + { + "epoch": 0.44461, + "grad_norm": 0.9107820113444716, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 44461 + }, + { + "epoch": 0.44462, + "grad_norm": 0.8758751022798602, + "learning_rate": 0.003, + "loss": 3.981, + "step": 44462 + }, + { + "epoch": 0.44463, + "grad_norm": 0.7880472288332634, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 44463 + }, + { + "epoch": 0.44464, + "grad_norm": 0.7655193171991296, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 44464 + }, + { + "epoch": 0.44465, + "grad_norm": 0.8167543467748131, + "learning_rate": 0.003, + "loss": 3.966, + "step": 44465 + }, + { + "epoch": 0.44466, + "grad_norm": 0.791702365150009, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 44466 + }, + { + "epoch": 0.44467, + "grad_norm": 0.8008647315372991, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 44467 + }, + { + "epoch": 0.44468, + "grad_norm": 0.9421589054324241, + "learning_rate": 0.003, + "loss": 3.981, + "step": 44468 + }, + { + "epoch": 0.44469, + "grad_norm": 1.0960141766959368, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 44469 + }, + { + "epoch": 0.4447, + "grad_norm": 0.8255856484038563, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 44470 + }, + { + "epoch": 0.44471, + "grad_norm": 0.7696156170691695, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 44471 + }, + { + "epoch": 0.44472, + "grad_norm": 0.7181122299883795, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 44472 + }, + { + "epoch": 0.44473, + "grad_norm": 0.8786111990512383, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 44473 + }, + { + "epoch": 0.44474, + "grad_norm": 0.8805517542252571, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 44474 + }, + { + "epoch": 0.44475, + "grad_norm": 0.889209309513131, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 44475 + }, + { + "epoch": 0.44476, + "grad_norm": 0.7520973739389473, + "learning_rate": 0.003, + "loss": 3.979, + "step": 44476 + }, + { + "epoch": 0.44477, + "grad_norm": 0.7287171631048334, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 44477 + }, + { + "epoch": 0.44478, + "grad_norm": 0.809374821501646, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 44478 + }, + { + "epoch": 0.44479, + "grad_norm": 0.9277549471062877, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 44479 + }, + { + "epoch": 0.4448, + "grad_norm": 1.0069972146540314, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 44480 + }, + { + "epoch": 0.44481, + "grad_norm": 0.8964962145886384, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 44481 + }, + { + "epoch": 0.44482, + "grad_norm": 0.9169779541092818, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 44482 + }, + { + "epoch": 0.44483, + "grad_norm": 1.0141153757549144, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 44483 + }, + { + "epoch": 0.44484, + "grad_norm": 1.0524301368888853, + "learning_rate": 0.003, + "loss": 4.03, + "step": 44484 + }, + { + "epoch": 0.44485, + "grad_norm": 1.068532752210123, + "learning_rate": 0.003, + "loss": 3.982, + "step": 44485 + }, + { + "epoch": 0.44486, + "grad_norm": 0.8647036226934963, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 44486 + }, + { + "epoch": 0.44487, + "grad_norm": 0.8253324333402919, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 44487 + }, + { + "epoch": 0.44488, + "grad_norm": 0.8058486907037801, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 44488 + }, + { + "epoch": 0.44489, + "grad_norm": 0.8352933382886736, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 44489 + }, + { + "epoch": 0.4449, + "grad_norm": 1.024900376943098, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 44490 + }, + { + "epoch": 0.44491, + "grad_norm": 1.2487842015145156, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 44491 + }, + { + "epoch": 0.44492, + "grad_norm": 0.9065374047345579, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 44492 + }, + { + "epoch": 0.44493, + "grad_norm": 0.7898728598533976, + "learning_rate": 0.003, + "loss": 3.99, + "step": 44493 + }, + { + "epoch": 0.44494, + "grad_norm": 0.72231184606667, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 44494 + }, + { + "epoch": 0.44495, + "grad_norm": 0.5921937992631708, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 44495 + }, + { + "epoch": 0.44496, + "grad_norm": 0.6306104868391169, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 44496 + }, + { + "epoch": 0.44497, + "grad_norm": 0.6462995469217673, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 44497 + }, + { + "epoch": 0.44498, + "grad_norm": 0.48647289728587445, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 44498 + }, + { + "epoch": 0.44499, + "grad_norm": 0.5553383568575758, + "learning_rate": 0.003, + "loss": 3.9475, + "step": 44499 + }, + { + "epoch": 0.445, + "grad_norm": 0.5173184652214244, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 44500 + }, + { + "epoch": 0.44501, + "grad_norm": 0.5478331918159356, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 44501 + }, + { + "epoch": 0.44502, + "grad_norm": 0.5966814388300469, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 44502 + }, + { + "epoch": 0.44503, + "grad_norm": 0.6037026756788976, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 44503 + }, + { + "epoch": 0.44504, + "grad_norm": 0.6004748591997181, + "learning_rate": 0.003, + "loss": 3.9465, + "step": 44504 + }, + { + "epoch": 0.44505, + "grad_norm": 0.7292239790321272, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 44505 + }, + { + "epoch": 0.44506, + "grad_norm": 0.8487076704536756, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 44506 + }, + { + "epoch": 0.44507, + "grad_norm": 0.8820234568174709, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 44507 + }, + { + "epoch": 0.44508, + "grad_norm": 0.8410237034194803, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 44508 + }, + { + "epoch": 0.44509, + "grad_norm": 0.7689730467103495, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44509 + }, + { + "epoch": 0.4451, + "grad_norm": 0.7159517367331768, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 44510 + }, + { + "epoch": 0.44511, + "grad_norm": 0.6593896622533314, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 44511 + }, + { + "epoch": 0.44512, + "grad_norm": 0.7366342201005106, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 44512 + }, + { + "epoch": 0.44513, + "grad_norm": 0.7323619828276765, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 44513 + }, + { + "epoch": 0.44514, + "grad_norm": 0.7594026653425358, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 44514 + }, + { + "epoch": 0.44515, + "grad_norm": 0.7994271480490809, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 44515 + }, + { + "epoch": 0.44516, + "grad_norm": 0.8880698793114383, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 44516 + }, + { + "epoch": 0.44517, + "grad_norm": 1.019602491851846, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 44517 + }, + { + "epoch": 0.44518, + "grad_norm": 1.0582440149867245, + "learning_rate": 0.003, + "loss": 4.0, + "step": 44518 + }, + { + "epoch": 0.44519, + "grad_norm": 0.7874242485888759, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 44519 + }, + { + "epoch": 0.4452, + "grad_norm": 0.6810336258870363, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 44520 + }, + { + "epoch": 0.44521, + "grad_norm": 0.7378351464165404, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 44521 + }, + { + "epoch": 0.44522, + "grad_norm": 0.8570957608943885, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 44522 + }, + { + "epoch": 0.44523, + "grad_norm": 0.959561690861344, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 44523 + }, + { + "epoch": 0.44524, + "grad_norm": 1.0057570584826645, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 44524 + }, + { + "epoch": 0.44525, + "grad_norm": 0.9708949287265174, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 44525 + }, + { + "epoch": 0.44526, + "grad_norm": 1.1583175626830646, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 44526 + }, + { + "epoch": 0.44527, + "grad_norm": 0.7200806679454504, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44527 + }, + { + "epoch": 0.44528, + "grad_norm": 0.7880288991209654, + "learning_rate": 0.003, + "loss": 3.991, + "step": 44528 + }, + { + "epoch": 0.44529, + "grad_norm": 0.8931189073107701, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 44529 + }, + { + "epoch": 0.4453, + "grad_norm": 0.8072783061890065, + "learning_rate": 0.003, + "loss": 4.021, + "step": 44530 + }, + { + "epoch": 0.44531, + "grad_norm": 0.7264180436952284, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 44531 + }, + { + "epoch": 0.44532, + "grad_norm": 0.668923898190233, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 44532 + }, + { + "epoch": 0.44533, + "grad_norm": 0.713477179201521, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 44533 + }, + { + "epoch": 0.44534, + "grad_norm": 0.6816574940058079, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 44534 + }, + { + "epoch": 0.44535, + "grad_norm": 0.6641012424080427, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 44535 + }, + { + "epoch": 0.44536, + "grad_norm": 0.6660157348567041, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 44536 + }, + { + "epoch": 0.44537, + "grad_norm": 0.7349624500267545, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 44537 + }, + { + "epoch": 0.44538, + "grad_norm": 0.7366703950104045, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 44538 + }, + { + "epoch": 0.44539, + "grad_norm": 0.7718354767077743, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 44539 + }, + { + "epoch": 0.4454, + "grad_norm": 0.753246858677835, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 44540 + }, + { + "epoch": 0.44541, + "grad_norm": 0.719795838896932, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 44541 + }, + { + "epoch": 0.44542, + "grad_norm": 0.6169914483418396, + "learning_rate": 0.003, + "loss": 3.974, + "step": 44542 + }, + { + "epoch": 0.44543, + "grad_norm": 0.6234846420377335, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44543 + }, + { + "epoch": 0.44544, + "grad_norm": 0.593977004932689, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 44544 + }, + { + "epoch": 0.44545, + "grad_norm": 0.6460442124602951, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 44545 + }, + { + "epoch": 0.44546, + "grad_norm": 0.7075588545641511, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 44546 + }, + { + "epoch": 0.44547, + "grad_norm": 0.7748597923977498, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 44547 + }, + { + "epoch": 0.44548, + "grad_norm": 0.9180838749149319, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 44548 + }, + { + "epoch": 0.44549, + "grad_norm": 1.043468310790201, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 44549 + }, + { + "epoch": 0.4455, + "grad_norm": 0.9939554208148736, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 44550 + }, + { + "epoch": 0.44551, + "grad_norm": 0.8276955236544405, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 44551 + }, + { + "epoch": 0.44552, + "grad_norm": 0.7980856240889386, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 44552 + }, + { + "epoch": 0.44553, + "grad_norm": 0.8411317781456965, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 44553 + }, + { + "epoch": 0.44554, + "grad_norm": 0.7804197469838883, + "learning_rate": 0.003, + "loss": 3.982, + "step": 44554 + }, + { + "epoch": 0.44555, + "grad_norm": 0.8193602689164042, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 44555 + }, + { + "epoch": 0.44556, + "grad_norm": 0.887213606441064, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 44556 + }, + { + "epoch": 0.44557, + "grad_norm": 0.9505772568891897, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 44557 + }, + { + "epoch": 0.44558, + "grad_norm": 0.8193320720396124, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 44558 + }, + { + "epoch": 0.44559, + "grad_norm": 0.7238202379851686, + "learning_rate": 0.003, + "loss": 3.983, + "step": 44559 + }, + { + "epoch": 0.4456, + "grad_norm": 0.7459129031173243, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 44560 + }, + { + "epoch": 0.44561, + "grad_norm": 0.7573999040143423, + "learning_rate": 0.003, + "loss": 3.977, + "step": 44561 + }, + { + "epoch": 0.44562, + "grad_norm": 0.7475695379251388, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 44562 + }, + { + "epoch": 0.44563, + "grad_norm": 0.7378488654478077, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 44563 + }, + { + "epoch": 0.44564, + "grad_norm": 0.6545784128224164, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 44564 + }, + { + "epoch": 0.44565, + "grad_norm": 0.6412192123455752, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 44565 + }, + { + "epoch": 0.44566, + "grad_norm": 0.6485438445884871, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 44566 + }, + { + "epoch": 0.44567, + "grad_norm": 0.6576802333195605, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 44567 + }, + { + "epoch": 0.44568, + "grad_norm": 0.7101157313922204, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 44568 + }, + { + "epoch": 0.44569, + "grad_norm": 0.7548875540407096, + "learning_rate": 0.003, + "loss": 3.9411, + "step": 44569 + }, + { + "epoch": 0.4457, + "grad_norm": 0.8059932654453066, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 44570 + }, + { + "epoch": 0.44571, + "grad_norm": 0.8222701331189044, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44571 + }, + { + "epoch": 0.44572, + "grad_norm": 0.8245562547737726, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 44572 + }, + { + "epoch": 0.44573, + "grad_norm": 0.8437941931118211, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 44573 + }, + { + "epoch": 0.44574, + "grad_norm": 0.9025117737082896, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 44574 + }, + { + "epoch": 0.44575, + "grad_norm": 0.7875679230386426, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 44575 + }, + { + "epoch": 0.44576, + "grad_norm": 0.8222883036630052, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 44576 + }, + { + "epoch": 0.44577, + "grad_norm": 0.8576797625990703, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 44577 + }, + { + "epoch": 0.44578, + "grad_norm": 0.8520076499695075, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 44578 + }, + { + "epoch": 0.44579, + "grad_norm": 0.8700120820379907, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 44579 + }, + { + "epoch": 0.4458, + "grad_norm": 0.895659544794789, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 44580 + }, + { + "epoch": 0.44581, + "grad_norm": 1.0067132141889612, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 44581 + }, + { + "epoch": 0.44582, + "grad_norm": 1.223440581870002, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 44582 + }, + { + "epoch": 0.44583, + "grad_norm": 0.8149692531865007, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 44583 + }, + { + "epoch": 0.44584, + "grad_norm": 0.7913108257048775, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 44584 + }, + { + "epoch": 0.44585, + "grad_norm": 0.9260376575033165, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 44585 + }, + { + "epoch": 0.44586, + "grad_norm": 1.0043940115773669, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 44586 + }, + { + "epoch": 0.44587, + "grad_norm": 1.039458322388499, + "learning_rate": 0.003, + "loss": 3.992, + "step": 44587 + }, + { + "epoch": 0.44588, + "grad_norm": 0.9710353236855371, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 44588 + }, + { + "epoch": 0.44589, + "grad_norm": 0.8775104602358392, + "learning_rate": 0.003, + "loss": 3.994, + "step": 44589 + }, + { + "epoch": 0.4459, + "grad_norm": 0.8567419875578617, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 44590 + }, + { + "epoch": 0.44591, + "grad_norm": 0.8447339693364758, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 44591 + }, + { + "epoch": 0.44592, + "grad_norm": 0.9707770020638344, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 44592 + }, + { + "epoch": 0.44593, + "grad_norm": 0.8929199885239566, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 44593 + }, + { + "epoch": 0.44594, + "grad_norm": 0.7610386233222296, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 44594 + }, + { + "epoch": 0.44595, + "grad_norm": 0.7203933891255339, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 44595 + }, + { + "epoch": 0.44596, + "grad_norm": 0.7025184495512903, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 44596 + }, + { + "epoch": 0.44597, + "grad_norm": 0.6177020159999776, + "learning_rate": 0.003, + "loss": 4.002, + "step": 44597 + }, + { + "epoch": 0.44598, + "grad_norm": 0.7316308908838463, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44598 + }, + { + "epoch": 0.44599, + "grad_norm": 0.8236844574329215, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 44599 + }, + { + "epoch": 0.446, + "grad_norm": 0.7305670490624881, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 44600 + }, + { + "epoch": 0.44601, + "grad_norm": 0.6390818152387407, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 44601 + }, + { + "epoch": 0.44602, + "grad_norm": 0.6709663838177097, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 44602 + }, + { + "epoch": 0.44603, + "grad_norm": 0.6121497631512143, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 44603 + }, + { + "epoch": 0.44604, + "grad_norm": 0.5948355252137362, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 44604 + }, + { + "epoch": 0.44605, + "grad_norm": 0.5917898351125589, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 44605 + }, + { + "epoch": 0.44606, + "grad_norm": 0.5887510037590878, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 44606 + }, + { + "epoch": 0.44607, + "grad_norm": 0.5621522509962811, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 44607 + }, + { + "epoch": 0.44608, + "grad_norm": 0.5858854590812679, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 44608 + }, + { + "epoch": 0.44609, + "grad_norm": 0.6781857457647436, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 44609 + }, + { + "epoch": 0.4461, + "grad_norm": 0.7007185548676439, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 44610 + }, + { + "epoch": 0.44611, + "grad_norm": 0.7635670617433213, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 44611 + }, + { + "epoch": 0.44612, + "grad_norm": 0.8769536021856089, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 44612 + }, + { + "epoch": 0.44613, + "grad_norm": 0.926597432918527, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 44613 + }, + { + "epoch": 0.44614, + "grad_norm": 0.8219315276714394, + "learning_rate": 0.003, + "loss": 4.003, + "step": 44614 + }, + { + "epoch": 0.44615, + "grad_norm": 0.8426447891070912, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 44615 + }, + { + "epoch": 0.44616, + "grad_norm": 0.8679908928532439, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44616 + }, + { + "epoch": 0.44617, + "grad_norm": 0.8807279086031692, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 44617 + }, + { + "epoch": 0.44618, + "grad_norm": 0.8921261504023854, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 44618 + }, + { + "epoch": 0.44619, + "grad_norm": 0.8780348703121246, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 44619 + }, + { + "epoch": 0.4462, + "grad_norm": 0.8727964087443923, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 44620 + }, + { + "epoch": 0.44621, + "grad_norm": 0.9141893244471466, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 44621 + }, + { + "epoch": 0.44622, + "grad_norm": 0.9752638945837726, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 44622 + }, + { + "epoch": 0.44623, + "grad_norm": 1.0239224159672842, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 44623 + }, + { + "epoch": 0.44624, + "grad_norm": 0.9617787418865961, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 44624 + }, + { + "epoch": 0.44625, + "grad_norm": 0.9978901480271237, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 44625 + }, + { + "epoch": 0.44626, + "grad_norm": 0.9701163490458938, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 44626 + }, + { + "epoch": 0.44627, + "grad_norm": 0.9961107751456926, + "learning_rate": 0.003, + "loss": 4.007, + "step": 44627 + }, + { + "epoch": 0.44628, + "grad_norm": 1.1197522452936972, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 44628 + }, + { + "epoch": 0.44629, + "grad_norm": 0.9108518336977357, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 44629 + }, + { + "epoch": 0.4463, + "grad_norm": 0.7819455582207108, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 44630 + }, + { + "epoch": 0.44631, + "grad_norm": 0.6609867194807659, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 44631 + }, + { + "epoch": 0.44632, + "grad_norm": 0.6048889463671839, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 44632 + }, + { + "epoch": 0.44633, + "grad_norm": 0.6809651190770789, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 44633 + }, + { + "epoch": 0.44634, + "grad_norm": 0.7359524303451178, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 44634 + }, + { + "epoch": 0.44635, + "grad_norm": 0.8943516970688221, + "learning_rate": 0.003, + "loss": 3.999, + "step": 44635 + }, + { + "epoch": 0.44636, + "grad_norm": 0.8715485802245294, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 44636 + }, + { + "epoch": 0.44637, + "grad_norm": 0.6918622210066318, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 44637 + }, + { + "epoch": 0.44638, + "grad_norm": 0.553947925703871, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 44638 + }, + { + "epoch": 0.44639, + "grad_norm": 0.5296938813670116, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 44639 + }, + { + "epoch": 0.4464, + "grad_norm": 0.5716984616502189, + "learning_rate": 0.003, + "loss": 3.959, + "step": 44640 + }, + { + "epoch": 0.44641, + "grad_norm": 0.5538047702391466, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 44641 + }, + { + "epoch": 0.44642, + "grad_norm": 0.5074529307056399, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 44642 + }, + { + "epoch": 0.44643, + "grad_norm": 0.4852806562108055, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 44643 + }, + { + "epoch": 0.44644, + "grad_norm": 0.5258101027613818, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 44644 + }, + { + "epoch": 0.44645, + "grad_norm": 0.526179256629259, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 44645 + }, + { + "epoch": 0.44646, + "grad_norm": 0.5942042964432799, + "learning_rate": 0.003, + "loss": 3.97, + "step": 44646 + }, + { + "epoch": 0.44647, + "grad_norm": 0.6893250154317158, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 44647 + }, + { + "epoch": 0.44648, + "grad_norm": 0.7857292189009026, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 44648 + }, + { + "epoch": 0.44649, + "grad_norm": 0.7528700049536369, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 44649 + }, + { + "epoch": 0.4465, + "grad_norm": 0.6517886353173372, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 44650 + }, + { + "epoch": 0.44651, + "grad_norm": 0.6087350981841813, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 44651 + }, + { + "epoch": 0.44652, + "grad_norm": 0.6388503825524846, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 44652 + }, + { + "epoch": 0.44653, + "grad_norm": 0.5826949763852003, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 44653 + }, + { + "epoch": 0.44654, + "grad_norm": 0.6021798608453313, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 44654 + }, + { + "epoch": 0.44655, + "grad_norm": 0.6111168025066614, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 44655 + }, + { + "epoch": 0.44656, + "grad_norm": 0.679005094811332, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 44656 + }, + { + "epoch": 0.44657, + "grad_norm": 0.815619167961271, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 44657 + }, + { + "epoch": 0.44658, + "grad_norm": 0.8459504436670567, + "learning_rate": 0.003, + "loss": 3.953, + "step": 44658 + }, + { + "epoch": 0.44659, + "grad_norm": 0.8386504424711881, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44659 + }, + { + "epoch": 0.4466, + "grad_norm": 1.018347324397927, + "learning_rate": 0.003, + "loss": 3.985, + "step": 44660 + }, + { + "epoch": 0.44661, + "grad_norm": 1.308210223320834, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 44661 + }, + { + "epoch": 0.44662, + "grad_norm": 0.8662429488938251, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 44662 + }, + { + "epoch": 0.44663, + "grad_norm": 0.828970969277299, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 44663 + }, + { + "epoch": 0.44664, + "grad_norm": 0.9353039584032034, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 44664 + }, + { + "epoch": 0.44665, + "grad_norm": 1.1598852822289762, + "learning_rate": 0.003, + "loss": 4.0279, + "step": 44665 + }, + { + "epoch": 0.44666, + "grad_norm": 0.898355881412346, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 44666 + }, + { + "epoch": 0.44667, + "grad_norm": 0.9744753694320778, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 44667 + }, + { + "epoch": 0.44668, + "grad_norm": 1.13712247486954, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 44668 + }, + { + "epoch": 0.44669, + "grad_norm": 0.8550226357891075, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 44669 + }, + { + "epoch": 0.4467, + "grad_norm": 0.7145802214611535, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 44670 + }, + { + "epoch": 0.44671, + "grad_norm": 0.7207933561045098, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 44671 + }, + { + "epoch": 0.44672, + "grad_norm": 0.7871497254948943, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 44672 + }, + { + "epoch": 0.44673, + "grad_norm": 0.848546785077696, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 44673 + }, + { + "epoch": 0.44674, + "grad_norm": 0.8359121846651127, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 44674 + }, + { + "epoch": 0.44675, + "grad_norm": 0.8044315975698219, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 44675 + }, + { + "epoch": 0.44676, + "grad_norm": 0.8106597152059765, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 44676 + }, + { + "epoch": 0.44677, + "grad_norm": 1.0467755505548728, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 44677 + }, + { + "epoch": 0.44678, + "grad_norm": 1.0835404560392505, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 44678 + }, + { + "epoch": 0.44679, + "grad_norm": 0.7855554211769642, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 44679 + }, + { + "epoch": 0.4468, + "grad_norm": 0.8199677676532984, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 44680 + }, + { + "epoch": 0.44681, + "grad_norm": 0.8179821764736814, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 44681 + }, + { + "epoch": 0.44682, + "grad_norm": 0.7993261927371467, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 44682 + }, + { + "epoch": 0.44683, + "grad_norm": 0.6598233097135102, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 44683 + }, + { + "epoch": 0.44684, + "grad_norm": 0.7189618770300304, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 44684 + }, + { + "epoch": 0.44685, + "grad_norm": 0.7216147849440416, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 44685 + }, + { + "epoch": 0.44686, + "grad_norm": 0.6914681904493405, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 44686 + }, + { + "epoch": 0.44687, + "grad_norm": 0.6101717714960619, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 44687 + }, + { + "epoch": 0.44688, + "grad_norm": 0.5275249469051944, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 44688 + }, + { + "epoch": 0.44689, + "grad_norm": 0.523627863919201, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 44689 + }, + { + "epoch": 0.4469, + "grad_norm": 0.48753851895569983, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 44690 + }, + { + "epoch": 0.44691, + "grad_norm": 0.5490681452604587, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 44691 + }, + { + "epoch": 0.44692, + "grad_norm": 0.606935994959527, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 44692 + }, + { + "epoch": 0.44693, + "grad_norm": 0.7112153650420382, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 44693 + }, + { + "epoch": 0.44694, + "grad_norm": 0.8702254751970817, + "learning_rate": 0.003, + "loss": 3.975, + "step": 44694 + }, + { + "epoch": 0.44695, + "grad_norm": 1.0024498920687337, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 44695 + }, + { + "epoch": 0.44696, + "grad_norm": 1.1032048517545994, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 44696 + }, + { + "epoch": 0.44697, + "grad_norm": 0.7809512435138795, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 44697 + }, + { + "epoch": 0.44698, + "grad_norm": 0.6901202018145562, + "learning_rate": 0.003, + "loss": 3.988, + "step": 44698 + }, + { + "epoch": 0.44699, + "grad_norm": 0.7126978002578124, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44699 + }, + { + "epoch": 0.447, + "grad_norm": 0.744072616022102, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 44700 + }, + { + "epoch": 0.44701, + "grad_norm": 0.7104516749028533, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 44701 + }, + { + "epoch": 0.44702, + "grad_norm": 0.6170392083722912, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 44702 + }, + { + "epoch": 0.44703, + "grad_norm": 0.6329209286568538, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 44703 + }, + { + "epoch": 0.44704, + "grad_norm": 0.6399606163382083, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 44704 + }, + { + "epoch": 0.44705, + "grad_norm": 0.6527673667727111, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 44705 + }, + { + "epoch": 0.44706, + "grad_norm": 0.7301298896617548, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 44706 + }, + { + "epoch": 0.44707, + "grad_norm": 0.7338487030084999, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 44707 + }, + { + "epoch": 0.44708, + "grad_norm": 0.6628028581240377, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 44708 + }, + { + "epoch": 0.44709, + "grad_norm": 0.773564100805598, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 44709 + }, + { + "epoch": 0.4471, + "grad_norm": 0.8518906801050143, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 44710 + }, + { + "epoch": 0.44711, + "grad_norm": 0.9194898704578447, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 44711 + }, + { + "epoch": 0.44712, + "grad_norm": 1.019605270824583, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 44712 + }, + { + "epoch": 0.44713, + "grad_norm": 1.1487148643889704, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 44713 + }, + { + "epoch": 0.44714, + "grad_norm": 0.8992250077747751, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 44714 + }, + { + "epoch": 0.44715, + "grad_norm": 0.9081689894953421, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 44715 + }, + { + "epoch": 0.44716, + "grad_norm": 0.932101482052118, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 44716 + }, + { + "epoch": 0.44717, + "grad_norm": 0.9268472616579807, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 44717 + }, + { + "epoch": 0.44718, + "grad_norm": 0.898543884532686, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 44718 + }, + { + "epoch": 0.44719, + "grad_norm": 0.8306054022058755, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 44719 + }, + { + "epoch": 0.4472, + "grad_norm": 0.7847098502268158, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 44720 + }, + { + "epoch": 0.44721, + "grad_norm": 0.859467438281963, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 44721 + }, + { + "epoch": 0.44722, + "grad_norm": 0.7677678483098734, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 44722 + }, + { + "epoch": 0.44723, + "grad_norm": 0.8959252281239581, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 44723 + }, + { + "epoch": 0.44724, + "grad_norm": 0.9032508537836696, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 44724 + }, + { + "epoch": 0.44725, + "grad_norm": 0.7892066376009294, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 44725 + }, + { + "epoch": 0.44726, + "grad_norm": 0.8857939654786811, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 44726 + }, + { + "epoch": 0.44727, + "grad_norm": 0.9766354325405927, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 44727 + }, + { + "epoch": 0.44728, + "grad_norm": 0.984773599338026, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 44728 + }, + { + "epoch": 0.44729, + "grad_norm": 0.907475926081602, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 44729 + }, + { + "epoch": 0.4473, + "grad_norm": 0.8145873091172035, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 44730 + }, + { + "epoch": 0.44731, + "grad_norm": 0.8775161184173511, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 44731 + }, + { + "epoch": 0.44732, + "grad_norm": 0.8354331157900042, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 44732 + }, + { + "epoch": 0.44733, + "grad_norm": 0.8078628276063486, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 44733 + }, + { + "epoch": 0.44734, + "grad_norm": 0.8204898020248541, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 44734 + }, + { + "epoch": 0.44735, + "grad_norm": 0.833717614060004, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 44735 + }, + { + "epoch": 0.44736, + "grad_norm": 0.7834513474291424, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 44736 + }, + { + "epoch": 0.44737, + "grad_norm": 0.6663079757899609, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 44737 + }, + { + "epoch": 0.44738, + "grad_norm": 0.6100841996693748, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 44738 + }, + { + "epoch": 0.44739, + "grad_norm": 0.6424892003518129, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 44739 + }, + { + "epoch": 0.4474, + "grad_norm": 0.6977804063689653, + "learning_rate": 0.003, + "loss": 3.977, + "step": 44740 + }, + { + "epoch": 0.44741, + "grad_norm": 0.7913817440986437, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 44741 + }, + { + "epoch": 0.44742, + "grad_norm": 0.672702765232332, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 44742 + }, + { + "epoch": 0.44743, + "grad_norm": 0.5663718684483319, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 44743 + }, + { + "epoch": 0.44744, + "grad_norm": 0.621302495542162, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 44744 + }, + { + "epoch": 0.44745, + "grad_norm": 0.5753873863750272, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 44745 + }, + { + "epoch": 0.44746, + "grad_norm": 0.6863500924747005, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 44746 + }, + { + "epoch": 0.44747, + "grad_norm": 0.7217929427806217, + "learning_rate": 0.003, + "loss": 3.9508, + "step": 44747 + }, + { + "epoch": 0.44748, + "grad_norm": 0.7538206453109213, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 44748 + }, + { + "epoch": 0.44749, + "grad_norm": 0.7214837593916902, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 44749 + }, + { + "epoch": 0.4475, + "grad_norm": 0.6264568893927344, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 44750 + }, + { + "epoch": 0.44751, + "grad_norm": 0.6294033025574263, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 44751 + }, + { + "epoch": 0.44752, + "grad_norm": 0.6424035679043979, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 44752 + }, + { + "epoch": 0.44753, + "grad_norm": 0.6895845021297153, + "learning_rate": 0.003, + "loss": 3.9447, + "step": 44753 + }, + { + "epoch": 0.44754, + "grad_norm": 0.7010995271983725, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 44754 + }, + { + "epoch": 0.44755, + "grad_norm": 0.6412075180020055, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 44755 + }, + { + "epoch": 0.44756, + "grad_norm": 0.6595551940862724, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 44756 + }, + { + "epoch": 0.44757, + "grad_norm": 0.6419936491667556, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 44757 + }, + { + "epoch": 0.44758, + "grad_norm": 0.6675491875082107, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 44758 + }, + { + "epoch": 0.44759, + "grad_norm": 0.7556572706826172, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 44759 + }, + { + "epoch": 0.4476, + "grad_norm": 0.8123859456534533, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 44760 + }, + { + "epoch": 0.44761, + "grad_norm": 0.9540343967915249, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 44761 + }, + { + "epoch": 0.44762, + "grad_norm": 1.035538645363941, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 44762 + }, + { + "epoch": 0.44763, + "grad_norm": 1.1005783656516661, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 44763 + }, + { + "epoch": 0.44764, + "grad_norm": 0.9398806440821963, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 44764 + }, + { + "epoch": 0.44765, + "grad_norm": 1.034229933024769, + "learning_rate": 0.003, + "loss": 3.996, + "step": 44765 + }, + { + "epoch": 0.44766, + "grad_norm": 0.9665255569322336, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 44766 + }, + { + "epoch": 0.44767, + "grad_norm": 1.030068692062578, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 44767 + }, + { + "epoch": 0.44768, + "grad_norm": 0.9337720668674753, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 44768 + }, + { + "epoch": 0.44769, + "grad_norm": 0.9865542509887271, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 44769 + }, + { + "epoch": 0.4477, + "grad_norm": 0.8368900203326276, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 44770 + }, + { + "epoch": 0.44771, + "grad_norm": 0.6962241633268355, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 44771 + }, + { + "epoch": 0.44772, + "grad_norm": 0.7064394036155108, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 44772 + }, + { + "epoch": 0.44773, + "grad_norm": 0.7992458147153559, + "learning_rate": 0.003, + "loss": 3.978, + "step": 44773 + }, + { + "epoch": 0.44774, + "grad_norm": 0.8508033538693993, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 44774 + }, + { + "epoch": 0.44775, + "grad_norm": 0.8691736751403698, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 44775 + }, + { + "epoch": 0.44776, + "grad_norm": 0.9080581817231089, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 44776 + }, + { + "epoch": 0.44777, + "grad_norm": 0.8930653781131996, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 44777 + }, + { + "epoch": 0.44778, + "grad_norm": 0.825896700504976, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 44778 + }, + { + "epoch": 0.44779, + "grad_norm": 0.6876959522094512, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 44779 + }, + { + "epoch": 0.4478, + "grad_norm": 0.7437888277610272, + "learning_rate": 0.003, + "loss": 3.972, + "step": 44780 + }, + { + "epoch": 0.44781, + "grad_norm": 0.6837611746010565, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 44781 + }, + { + "epoch": 0.44782, + "grad_norm": 0.7910397496395429, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 44782 + }, + { + "epoch": 0.44783, + "grad_norm": 0.8022438868327582, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 44783 + }, + { + "epoch": 0.44784, + "grad_norm": 0.7636967827324008, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 44784 + }, + { + "epoch": 0.44785, + "grad_norm": 0.8116613533570505, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 44785 + }, + { + "epoch": 0.44786, + "grad_norm": 0.8364843901073237, + "learning_rate": 0.003, + "loss": 4.004, + "step": 44786 + }, + { + "epoch": 0.44787, + "grad_norm": 0.753657688820899, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 44787 + }, + { + "epoch": 0.44788, + "grad_norm": 0.6604434364257358, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 44788 + }, + { + "epoch": 0.44789, + "grad_norm": 0.6429200422477115, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 44789 + }, + { + "epoch": 0.4479, + "grad_norm": 0.661536517562491, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 44790 + }, + { + "epoch": 0.44791, + "grad_norm": 0.6432551716504943, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 44791 + }, + { + "epoch": 0.44792, + "grad_norm": 0.7660603632553903, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 44792 + }, + { + "epoch": 0.44793, + "grad_norm": 0.8913650440216863, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 44793 + }, + { + "epoch": 0.44794, + "grad_norm": 0.8476757893274144, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 44794 + }, + { + "epoch": 0.44795, + "grad_norm": 0.7910909177574269, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 44795 + }, + { + "epoch": 0.44796, + "grad_norm": 0.7697317040178158, + "learning_rate": 0.003, + "loss": 3.976, + "step": 44796 + }, + { + "epoch": 0.44797, + "grad_norm": 0.7544306263535379, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 44797 + }, + { + "epoch": 0.44798, + "grad_norm": 0.7356952572019267, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 44798 + }, + { + "epoch": 0.44799, + "grad_norm": 0.8182968098128933, + "learning_rate": 0.003, + "loss": 4.019, + "step": 44799 + }, + { + "epoch": 0.448, + "grad_norm": 0.8210805344053401, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 44800 + }, + { + "epoch": 0.44801, + "grad_norm": 0.8100646554646111, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 44801 + }, + { + "epoch": 0.44802, + "grad_norm": 0.7038353020930813, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 44802 + }, + { + "epoch": 0.44803, + "grad_norm": 0.7573978047290967, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 44803 + }, + { + "epoch": 0.44804, + "grad_norm": 0.8161191266474619, + "learning_rate": 0.003, + "loss": 3.9274, + "step": 44804 + }, + { + "epoch": 0.44805, + "grad_norm": 0.8166190481617583, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 44805 + }, + { + "epoch": 0.44806, + "grad_norm": 0.8178642104179253, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 44806 + }, + { + "epoch": 0.44807, + "grad_norm": 0.914121396525075, + "learning_rate": 0.003, + "loss": 3.999, + "step": 44807 + }, + { + "epoch": 0.44808, + "grad_norm": 1.0265529337740094, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 44808 + }, + { + "epoch": 0.44809, + "grad_norm": 0.955701058129819, + "learning_rate": 0.003, + "loss": 3.986, + "step": 44809 + }, + { + "epoch": 0.4481, + "grad_norm": 1.0207640104790903, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 44810 + }, + { + "epoch": 0.44811, + "grad_norm": 0.9589480521706769, + "learning_rate": 0.003, + "loss": 4.0132, + "step": 44811 + }, + { + "epoch": 0.44812, + "grad_norm": 0.9953687970290477, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 44812 + }, + { + "epoch": 0.44813, + "grad_norm": 0.9963388553123566, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 44813 + }, + { + "epoch": 0.44814, + "grad_norm": 1.0423133751916855, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 44814 + }, + { + "epoch": 0.44815, + "grad_norm": 0.9388623605347501, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 44815 + }, + { + "epoch": 0.44816, + "grad_norm": 1.0665384231161739, + "learning_rate": 0.003, + "loss": 3.99, + "step": 44816 + }, + { + "epoch": 0.44817, + "grad_norm": 0.9696944559827627, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 44817 + }, + { + "epoch": 0.44818, + "grad_norm": 0.9813843922319379, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 44818 + }, + { + "epoch": 0.44819, + "grad_norm": 0.9451716006798209, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 44819 + }, + { + "epoch": 0.4482, + "grad_norm": 0.8359841832451324, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 44820 + }, + { + "epoch": 0.44821, + "grad_norm": 0.7433238250198565, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 44821 + }, + { + "epoch": 0.44822, + "grad_norm": 0.6834071803287743, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 44822 + }, + { + "epoch": 0.44823, + "grad_norm": 0.6157271895391988, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 44823 + }, + { + "epoch": 0.44824, + "grad_norm": 0.5727247792125076, + "learning_rate": 0.003, + "loss": 4.01, + "step": 44824 + }, + { + "epoch": 0.44825, + "grad_norm": 0.615582227375496, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 44825 + }, + { + "epoch": 0.44826, + "grad_norm": 0.6193209834864529, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 44826 + }, + { + "epoch": 0.44827, + "grad_norm": 0.627153649372734, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 44827 + }, + { + "epoch": 0.44828, + "grad_norm": 0.5944127848854203, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 44828 + }, + { + "epoch": 0.44829, + "grad_norm": 0.5466660629846358, + "learning_rate": 0.003, + "loss": 3.971, + "step": 44829 + }, + { + "epoch": 0.4483, + "grad_norm": 0.5622456501035868, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 44830 + }, + { + "epoch": 0.44831, + "grad_norm": 0.513112737232498, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 44831 + }, + { + "epoch": 0.44832, + "grad_norm": 0.47896941121254316, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 44832 + }, + { + "epoch": 0.44833, + "grad_norm": 0.45841529708177237, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44833 + }, + { + "epoch": 0.44834, + "grad_norm": 0.5154731256070971, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 44834 + }, + { + "epoch": 0.44835, + "grad_norm": 0.567904702534527, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 44835 + }, + { + "epoch": 0.44836, + "grad_norm": 0.5266621733443844, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 44836 + }, + { + "epoch": 0.44837, + "grad_norm": 0.5344073807085781, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 44837 + }, + { + "epoch": 0.44838, + "grad_norm": 0.5590119305844566, + "learning_rate": 0.003, + "loss": 3.963, + "step": 44838 + }, + { + "epoch": 0.44839, + "grad_norm": 0.6750760920760116, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 44839 + }, + { + "epoch": 0.4484, + "grad_norm": 0.9037125313901031, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 44840 + }, + { + "epoch": 0.44841, + "grad_norm": 1.1738682984257363, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 44841 + }, + { + "epoch": 0.44842, + "grad_norm": 0.8757687429549024, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 44842 + }, + { + "epoch": 0.44843, + "grad_norm": 0.6998537242004356, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 44843 + }, + { + "epoch": 0.44844, + "grad_norm": 0.6348374215678977, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 44844 + }, + { + "epoch": 0.44845, + "grad_norm": 0.7231547913299349, + "learning_rate": 0.003, + "loss": 3.989, + "step": 44845 + }, + { + "epoch": 0.44846, + "grad_norm": 0.6447282505725982, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 44846 + }, + { + "epoch": 0.44847, + "grad_norm": 0.6696754760211034, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 44847 + }, + { + "epoch": 0.44848, + "grad_norm": 0.6523028271881475, + "learning_rate": 0.003, + "loss": 3.9424, + "step": 44848 + }, + { + "epoch": 0.44849, + "grad_norm": 0.7834538159988667, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 44849 + }, + { + "epoch": 0.4485, + "grad_norm": 0.8410216089858766, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 44850 + }, + { + "epoch": 0.44851, + "grad_norm": 0.7898037642100862, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 44851 + }, + { + "epoch": 0.44852, + "grad_norm": 0.7276780869357179, + "learning_rate": 0.003, + "loss": 3.984, + "step": 44852 + }, + { + "epoch": 0.44853, + "grad_norm": 0.7979904865657339, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 44853 + }, + { + "epoch": 0.44854, + "grad_norm": 0.8200385535231944, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 44854 + }, + { + "epoch": 0.44855, + "grad_norm": 0.9197209153354943, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 44855 + }, + { + "epoch": 0.44856, + "grad_norm": 0.931506923023149, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 44856 + }, + { + "epoch": 0.44857, + "grad_norm": 0.8325311892527609, + "learning_rate": 0.003, + "loss": 3.976, + "step": 44857 + }, + { + "epoch": 0.44858, + "grad_norm": 0.9110109241043673, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 44858 + }, + { + "epoch": 0.44859, + "grad_norm": 0.9000012720979127, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 44859 + }, + { + "epoch": 0.4486, + "grad_norm": 0.818958826166356, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 44860 + }, + { + "epoch": 0.44861, + "grad_norm": 0.7274724689787185, + "learning_rate": 0.003, + "loss": 3.984, + "step": 44861 + }, + { + "epoch": 0.44862, + "grad_norm": 0.7274633886107874, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 44862 + }, + { + "epoch": 0.44863, + "grad_norm": 0.795927766478055, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 44863 + }, + { + "epoch": 0.44864, + "grad_norm": 0.9296158258593455, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 44864 + }, + { + "epoch": 0.44865, + "grad_norm": 1.0899376211626037, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 44865 + }, + { + "epoch": 0.44866, + "grad_norm": 1.0442169515583528, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 44866 + }, + { + "epoch": 0.44867, + "grad_norm": 0.8988954682418123, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 44867 + }, + { + "epoch": 0.44868, + "grad_norm": 0.8150966043011562, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 44868 + }, + { + "epoch": 0.44869, + "grad_norm": 0.7163826809643459, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 44869 + }, + { + "epoch": 0.4487, + "grad_norm": 0.6722575493293748, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 44870 + }, + { + "epoch": 0.44871, + "grad_norm": 0.6417868766704653, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 44871 + }, + { + "epoch": 0.44872, + "grad_norm": 0.6655943558221773, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 44872 + }, + { + "epoch": 0.44873, + "grad_norm": 0.6900401030072422, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 44873 + }, + { + "epoch": 0.44874, + "grad_norm": 0.8172488323524848, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 44874 + }, + { + "epoch": 0.44875, + "grad_norm": 0.8439962368223305, + "learning_rate": 0.003, + "loss": 3.9444, + "step": 44875 + }, + { + "epoch": 0.44876, + "grad_norm": 0.8260745930531789, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 44876 + }, + { + "epoch": 0.44877, + "grad_norm": 0.8370109087998194, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 44877 + }, + { + "epoch": 0.44878, + "grad_norm": 0.7305482443367034, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 44878 + }, + { + "epoch": 0.44879, + "grad_norm": 0.715121232935516, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 44879 + }, + { + "epoch": 0.4488, + "grad_norm": 0.7317264635507803, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 44880 + }, + { + "epoch": 0.44881, + "grad_norm": 0.7019015312041867, + "learning_rate": 0.003, + "loss": 3.999, + "step": 44881 + }, + { + "epoch": 0.44882, + "grad_norm": 0.7195342560521861, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 44882 + }, + { + "epoch": 0.44883, + "grad_norm": 0.8928083100478361, + "learning_rate": 0.003, + "loss": 3.9449, + "step": 44883 + }, + { + "epoch": 0.44884, + "grad_norm": 1.0898053070605629, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 44884 + }, + { + "epoch": 0.44885, + "grad_norm": 0.9372364653096461, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 44885 + }, + { + "epoch": 0.44886, + "grad_norm": 0.8693471407486012, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 44886 + }, + { + "epoch": 0.44887, + "grad_norm": 0.8628417623525472, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 44887 + }, + { + "epoch": 0.44888, + "grad_norm": 0.6956947146351796, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 44888 + }, + { + "epoch": 0.44889, + "grad_norm": 0.6902251160518267, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 44889 + }, + { + "epoch": 0.4489, + "grad_norm": 0.7041615080548945, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 44890 + }, + { + "epoch": 0.44891, + "grad_norm": 0.7106489008037673, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 44891 + }, + { + "epoch": 0.44892, + "grad_norm": 0.6788853317535873, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 44892 + }, + { + "epoch": 0.44893, + "grad_norm": 0.6157558766607435, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 44893 + }, + { + "epoch": 0.44894, + "grad_norm": 0.6262751344876626, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 44894 + }, + { + "epoch": 0.44895, + "grad_norm": 0.7705708912047928, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 44895 + }, + { + "epoch": 0.44896, + "grad_norm": 0.8550435884360175, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 44896 + }, + { + "epoch": 0.44897, + "grad_norm": 0.8386021292302223, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 44897 + }, + { + "epoch": 0.44898, + "grad_norm": 0.8689896506839441, + "learning_rate": 0.003, + "loss": 3.97, + "step": 44898 + }, + { + "epoch": 0.44899, + "grad_norm": 0.8372487430057667, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 44899 + }, + { + "epoch": 0.449, + "grad_norm": 1.0385707535833968, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 44900 + }, + { + "epoch": 0.44901, + "grad_norm": 1.2187230605273074, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 44901 + }, + { + "epoch": 0.44902, + "grad_norm": 0.5937890818096078, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 44902 + }, + { + "epoch": 0.44903, + "grad_norm": 0.6739510929609569, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 44903 + }, + { + "epoch": 0.44904, + "grad_norm": 0.9413523934348774, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 44904 + }, + { + "epoch": 0.44905, + "grad_norm": 1.218496287389657, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 44905 + }, + { + "epoch": 0.44906, + "grad_norm": 0.9596759148113043, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 44906 + }, + { + "epoch": 0.44907, + "grad_norm": 0.9884902075807431, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 44907 + }, + { + "epoch": 0.44908, + "grad_norm": 0.959118181411373, + "learning_rate": 0.003, + "loss": 3.999, + "step": 44908 + }, + { + "epoch": 0.44909, + "grad_norm": 0.7934045012087764, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 44909 + }, + { + "epoch": 0.4491, + "grad_norm": 0.8329496415360008, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 44910 + }, + { + "epoch": 0.44911, + "grad_norm": 0.8376941903901546, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 44911 + }, + { + "epoch": 0.44912, + "grad_norm": 0.7540973532637311, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 44912 + }, + { + "epoch": 0.44913, + "grad_norm": 0.7328655242078264, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 44913 + }, + { + "epoch": 0.44914, + "grad_norm": 0.7930611221166639, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 44914 + }, + { + "epoch": 0.44915, + "grad_norm": 0.753019813615504, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 44915 + }, + { + "epoch": 0.44916, + "grad_norm": 0.7409586832482415, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 44916 + }, + { + "epoch": 0.44917, + "grad_norm": 0.5865177794928943, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 44917 + }, + { + "epoch": 0.44918, + "grad_norm": 0.5984947694656442, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 44918 + }, + { + "epoch": 0.44919, + "grad_norm": 0.5633918942984907, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 44919 + }, + { + "epoch": 0.4492, + "grad_norm": 0.6057960096326921, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 44920 + }, + { + "epoch": 0.44921, + "grad_norm": 0.5891497357308993, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 44921 + }, + { + "epoch": 0.44922, + "grad_norm": 0.5424486439111248, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 44922 + }, + { + "epoch": 0.44923, + "grad_norm": 0.5327241422913672, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 44923 + }, + { + "epoch": 0.44924, + "grad_norm": 0.5216294114921726, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 44924 + }, + { + "epoch": 0.44925, + "grad_norm": 0.5004260941765873, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 44925 + }, + { + "epoch": 0.44926, + "grad_norm": 0.5075347268264596, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 44926 + }, + { + "epoch": 0.44927, + "grad_norm": 0.5096284990377679, + "learning_rate": 0.003, + "loss": 3.9432, + "step": 44927 + }, + { + "epoch": 0.44928, + "grad_norm": 0.5654856182343981, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 44928 + }, + { + "epoch": 0.44929, + "grad_norm": 0.5635671138408438, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 44929 + }, + { + "epoch": 0.4493, + "grad_norm": 0.5439941813506035, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 44930 + }, + { + "epoch": 0.44931, + "grad_norm": 0.6185721224219822, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 44931 + }, + { + "epoch": 0.44932, + "grad_norm": 0.7679588879427953, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 44932 + }, + { + "epoch": 0.44933, + "grad_norm": 0.822946416088607, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 44933 + }, + { + "epoch": 0.44934, + "grad_norm": 0.7575706450443965, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 44934 + }, + { + "epoch": 0.44935, + "grad_norm": 0.9084229106983815, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 44935 + }, + { + "epoch": 0.44936, + "grad_norm": 1.2386070327763767, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 44936 + }, + { + "epoch": 0.44937, + "grad_norm": 0.9504473660792394, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 44937 + }, + { + "epoch": 0.44938, + "grad_norm": 0.8758226493158018, + "learning_rate": 0.003, + "loss": 3.983, + "step": 44938 + }, + { + "epoch": 0.44939, + "grad_norm": 0.8476741234681979, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 44939 + }, + { + "epoch": 0.4494, + "grad_norm": 0.817838509372428, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 44940 + }, + { + "epoch": 0.44941, + "grad_norm": 0.9011248323208022, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 44941 + }, + { + "epoch": 0.44942, + "grad_norm": 1.0008025143196178, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 44942 + }, + { + "epoch": 0.44943, + "grad_norm": 1.104190930932033, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 44943 + }, + { + "epoch": 0.44944, + "grad_norm": 1.129067044493158, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 44944 + }, + { + "epoch": 0.44945, + "grad_norm": 1.0890430186402407, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 44945 + }, + { + "epoch": 0.44946, + "grad_norm": 0.9130578073243808, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 44946 + }, + { + "epoch": 0.44947, + "grad_norm": 0.7269671913770522, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 44947 + }, + { + "epoch": 0.44948, + "grad_norm": 0.6745076017108538, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 44948 + }, + { + "epoch": 0.44949, + "grad_norm": 0.7658152741003185, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 44949 + }, + { + "epoch": 0.4495, + "grad_norm": 0.7690333568543959, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 44950 + }, + { + "epoch": 0.44951, + "grad_norm": 0.7713418054233311, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 44951 + }, + { + "epoch": 0.44952, + "grad_norm": 0.7857543230525745, + "learning_rate": 0.003, + "loss": 3.985, + "step": 44952 + }, + { + "epoch": 0.44953, + "grad_norm": 0.7307775077916251, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 44953 + }, + { + "epoch": 0.44954, + "grad_norm": 0.6663906829214604, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 44954 + }, + { + "epoch": 0.44955, + "grad_norm": 0.649805278544578, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 44955 + }, + { + "epoch": 0.44956, + "grad_norm": 0.6416724472124663, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 44956 + }, + { + "epoch": 0.44957, + "grad_norm": 0.6347013668326181, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 44957 + }, + { + "epoch": 0.44958, + "grad_norm": 0.6834998560677319, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 44958 + }, + { + "epoch": 0.44959, + "grad_norm": 0.6431264230762189, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 44959 + }, + { + "epoch": 0.4496, + "grad_norm": 0.6941831283120358, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 44960 + }, + { + "epoch": 0.44961, + "grad_norm": 0.6710203169078052, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 44961 + }, + { + "epoch": 0.44962, + "grad_norm": 0.7125359303448605, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 44962 + }, + { + "epoch": 0.44963, + "grad_norm": 0.630943932125791, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 44963 + }, + { + "epoch": 0.44964, + "grad_norm": 0.5925127902820339, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 44964 + }, + { + "epoch": 0.44965, + "grad_norm": 0.7139929553941594, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 44965 + }, + { + "epoch": 0.44966, + "grad_norm": 0.8090956238752497, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 44966 + }, + { + "epoch": 0.44967, + "grad_norm": 0.9366400071432621, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 44967 + }, + { + "epoch": 0.44968, + "grad_norm": 0.9736952749878663, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 44968 + }, + { + "epoch": 0.44969, + "grad_norm": 0.9226370947688397, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 44969 + }, + { + "epoch": 0.4497, + "grad_norm": 0.8506364250342818, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 44970 + }, + { + "epoch": 0.44971, + "grad_norm": 0.7596078040809056, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 44971 + }, + { + "epoch": 0.44972, + "grad_norm": 0.736656525786227, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 44972 + }, + { + "epoch": 0.44973, + "grad_norm": 0.7334950040713587, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 44973 + }, + { + "epoch": 0.44974, + "grad_norm": 0.7190621732621273, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 44974 + }, + { + "epoch": 0.44975, + "grad_norm": 0.743833820818463, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 44975 + }, + { + "epoch": 0.44976, + "grad_norm": 0.6944107036489275, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 44976 + }, + { + "epoch": 0.44977, + "grad_norm": 0.6705822120072672, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 44977 + }, + { + "epoch": 0.44978, + "grad_norm": 0.6856701215925577, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 44978 + }, + { + "epoch": 0.44979, + "grad_norm": 0.7440629656958015, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 44979 + }, + { + "epoch": 0.4498, + "grad_norm": 0.8687245367613129, + "learning_rate": 0.003, + "loss": 3.9369, + "step": 44980 + }, + { + "epoch": 0.44981, + "grad_norm": 0.8994485242249206, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 44981 + }, + { + "epoch": 0.44982, + "grad_norm": 0.9814494075608894, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 44982 + }, + { + "epoch": 0.44983, + "grad_norm": 0.9620579315674701, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 44983 + }, + { + "epoch": 0.44984, + "grad_norm": 0.947671375151728, + "learning_rate": 0.003, + "loss": 4.009, + "step": 44984 + }, + { + "epoch": 0.44985, + "grad_norm": 0.9923147775978702, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 44985 + }, + { + "epoch": 0.44986, + "grad_norm": 1.0510552132246158, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 44986 + }, + { + "epoch": 0.44987, + "grad_norm": 1.028999414252556, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 44987 + }, + { + "epoch": 0.44988, + "grad_norm": 0.9315557324312764, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 44988 + }, + { + "epoch": 0.44989, + "grad_norm": 0.8390007512550849, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 44989 + }, + { + "epoch": 0.4499, + "grad_norm": 0.8834636929915393, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 44990 + }, + { + "epoch": 0.44991, + "grad_norm": 0.9768923485504487, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 44991 + }, + { + "epoch": 0.44992, + "grad_norm": 1.0936645657846449, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 44992 + }, + { + "epoch": 0.44993, + "grad_norm": 0.9353181131328192, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 44993 + }, + { + "epoch": 0.44994, + "grad_norm": 0.8559739196140097, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 44994 + }, + { + "epoch": 0.44995, + "grad_norm": 0.904641101961563, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 44995 + }, + { + "epoch": 0.44996, + "grad_norm": 0.9507012689326771, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 44996 + }, + { + "epoch": 0.44997, + "grad_norm": 1.0054353491051427, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 44997 + }, + { + "epoch": 0.44998, + "grad_norm": 0.8619636463637582, + "learning_rate": 0.003, + "loss": 4.0154, + "step": 44998 + }, + { + "epoch": 0.44999, + "grad_norm": 0.8959100434803337, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 44999 + }, + { + "epoch": 0.45, + "grad_norm": 0.8819704007701576, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 45000 + }, + { + "epoch": 0.45001, + "grad_norm": 0.7264228065649394, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 45001 + }, + { + "epoch": 0.45002, + "grad_norm": 0.7384946259402132, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 45002 + }, + { + "epoch": 0.45003, + "grad_norm": 0.8040648815454287, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 45003 + }, + { + "epoch": 0.45004, + "grad_norm": 0.7783421440277923, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 45004 + }, + { + "epoch": 0.45005, + "grad_norm": 0.7327488390980456, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 45005 + }, + { + "epoch": 0.45006, + "grad_norm": 0.7980026169583074, + "learning_rate": 0.003, + "loss": 3.9285, + "step": 45006 + }, + { + "epoch": 0.45007, + "grad_norm": 0.7785562451551346, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 45007 + }, + { + "epoch": 0.45008, + "grad_norm": 0.6718145740867841, + "learning_rate": 0.003, + "loss": 3.986, + "step": 45008 + }, + { + "epoch": 0.45009, + "grad_norm": 0.6404567476824338, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 45009 + }, + { + "epoch": 0.4501, + "grad_norm": 0.6131652004065644, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 45010 + }, + { + "epoch": 0.45011, + "grad_norm": 0.6498407200833823, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 45011 + }, + { + "epoch": 0.45012, + "grad_norm": 0.5848487108120227, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 45012 + }, + { + "epoch": 0.45013, + "grad_norm": 0.5942764426617564, + "learning_rate": 0.003, + "loss": 3.963, + "step": 45013 + }, + { + "epoch": 0.45014, + "grad_norm": 0.6354447518232861, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 45014 + }, + { + "epoch": 0.45015, + "grad_norm": 0.662717732508906, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 45015 + }, + { + "epoch": 0.45016, + "grad_norm": 0.6370640484014818, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 45016 + }, + { + "epoch": 0.45017, + "grad_norm": 0.6091455690914913, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 45017 + }, + { + "epoch": 0.45018, + "grad_norm": 0.679030846436583, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 45018 + }, + { + "epoch": 0.45019, + "grad_norm": 0.7288141121294094, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 45019 + }, + { + "epoch": 0.4502, + "grad_norm": 0.7844059266490877, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 45020 + }, + { + "epoch": 0.45021, + "grad_norm": 0.8555496343975995, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 45021 + }, + { + "epoch": 0.45022, + "grad_norm": 0.7872966415703593, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 45022 + }, + { + "epoch": 0.45023, + "grad_norm": 0.6588163887725943, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 45023 + }, + { + "epoch": 0.45024, + "grad_norm": 0.6280388079695371, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 45024 + }, + { + "epoch": 0.45025, + "grad_norm": 0.5936615899395555, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 45025 + }, + { + "epoch": 0.45026, + "grad_norm": 0.6173497563976472, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 45026 + }, + { + "epoch": 0.45027, + "grad_norm": 0.6658885174675073, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 45027 + }, + { + "epoch": 0.45028, + "grad_norm": 0.6712220726996848, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 45028 + }, + { + "epoch": 0.45029, + "grad_norm": 0.7566923555574339, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 45029 + }, + { + "epoch": 0.4503, + "grad_norm": 0.7653148471682575, + "learning_rate": 0.003, + "loss": 3.972, + "step": 45030 + }, + { + "epoch": 0.45031, + "grad_norm": 0.6758601447872407, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 45031 + }, + { + "epoch": 0.45032, + "grad_norm": 0.7400013280472038, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 45032 + }, + { + "epoch": 0.45033, + "grad_norm": 0.8172617853607174, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 45033 + }, + { + "epoch": 0.45034, + "grad_norm": 1.0357369592783456, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 45034 + }, + { + "epoch": 0.45035, + "grad_norm": 1.1879268836323473, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 45035 + }, + { + "epoch": 0.45036, + "grad_norm": 0.7686661549305419, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 45036 + }, + { + "epoch": 0.45037, + "grad_norm": 0.7989267821258204, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 45037 + }, + { + "epoch": 0.45038, + "grad_norm": 0.8913139870407444, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 45038 + }, + { + "epoch": 0.45039, + "grad_norm": 1.0301957147088128, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 45039 + }, + { + "epoch": 0.4504, + "grad_norm": 0.8762903808774347, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 45040 + }, + { + "epoch": 0.45041, + "grad_norm": 0.7693994933102961, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 45041 + }, + { + "epoch": 0.45042, + "grad_norm": 0.6964033543236873, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 45042 + }, + { + "epoch": 0.45043, + "grad_norm": 0.7220647812057049, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 45043 + }, + { + "epoch": 0.45044, + "grad_norm": 0.6858044940318466, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 45044 + }, + { + "epoch": 0.45045, + "grad_norm": 0.7160175395860481, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 45045 + }, + { + "epoch": 0.45046, + "grad_norm": 0.7672056881323315, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 45046 + }, + { + "epoch": 0.45047, + "grad_norm": 0.8450993984035985, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 45047 + }, + { + "epoch": 0.45048, + "grad_norm": 0.7962824618400679, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 45048 + }, + { + "epoch": 0.45049, + "grad_norm": 0.7819827685881438, + "learning_rate": 0.003, + "loss": 3.981, + "step": 45049 + }, + { + "epoch": 0.4505, + "grad_norm": 0.8785890216255355, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 45050 + }, + { + "epoch": 0.45051, + "grad_norm": 0.9428329781563483, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 45051 + }, + { + "epoch": 0.45052, + "grad_norm": 0.8844839443583842, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 45052 + }, + { + "epoch": 0.45053, + "grad_norm": 0.9321214039240969, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 45053 + }, + { + "epoch": 0.45054, + "grad_norm": 0.9635247818344389, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 45054 + }, + { + "epoch": 0.45055, + "grad_norm": 0.8555687956942737, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 45055 + }, + { + "epoch": 0.45056, + "grad_norm": 0.7017587769465868, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 45056 + }, + { + "epoch": 0.45057, + "grad_norm": 0.6090692471901317, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 45057 + }, + { + "epoch": 0.45058, + "grad_norm": 0.6706613470157102, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 45058 + }, + { + "epoch": 0.45059, + "grad_norm": 0.6794096644624387, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 45059 + }, + { + "epoch": 0.4506, + "grad_norm": 0.7260036811866986, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 45060 + }, + { + "epoch": 0.45061, + "grad_norm": 0.75909538867594, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 45061 + }, + { + "epoch": 0.45062, + "grad_norm": 0.7172376083326747, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 45062 + }, + { + "epoch": 0.45063, + "grad_norm": 0.6123851652243252, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 45063 + }, + { + "epoch": 0.45064, + "grad_norm": 0.6001553535736835, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 45064 + }, + { + "epoch": 0.45065, + "grad_norm": 0.5708038039589499, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 45065 + }, + { + "epoch": 0.45066, + "grad_norm": 0.5629100293764104, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 45066 + }, + { + "epoch": 0.45067, + "grad_norm": 0.5474589750501018, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 45067 + }, + { + "epoch": 0.45068, + "grad_norm": 0.5488663273134836, + "learning_rate": 0.003, + "loss": 3.99, + "step": 45068 + }, + { + "epoch": 0.45069, + "grad_norm": 0.6951464299720393, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 45069 + }, + { + "epoch": 0.4507, + "grad_norm": 0.8794020012658604, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 45070 + }, + { + "epoch": 0.45071, + "grad_norm": 1.0962543326831413, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 45071 + }, + { + "epoch": 0.45072, + "grad_norm": 0.9312727632860068, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 45072 + }, + { + "epoch": 0.45073, + "grad_norm": 1.0006222823488906, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 45073 + }, + { + "epoch": 0.45074, + "grad_norm": 0.9045519642028359, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 45074 + }, + { + "epoch": 0.45075, + "grad_norm": 0.7423528283262444, + "learning_rate": 0.003, + "loss": 3.977, + "step": 45075 + }, + { + "epoch": 0.45076, + "grad_norm": 0.7753103718187838, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 45076 + }, + { + "epoch": 0.45077, + "grad_norm": 0.8303011600241376, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 45077 + }, + { + "epoch": 0.45078, + "grad_norm": 0.8402898091281452, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 45078 + }, + { + "epoch": 0.45079, + "grad_norm": 0.9333127909704836, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 45079 + }, + { + "epoch": 0.4508, + "grad_norm": 0.8340892475056632, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 45080 + }, + { + "epoch": 0.45081, + "grad_norm": 0.7956400844080108, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 45081 + }, + { + "epoch": 0.45082, + "grad_norm": 0.9797204750425048, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 45082 + }, + { + "epoch": 0.45083, + "grad_norm": 1.1047900349403335, + "learning_rate": 0.003, + "loss": 3.983, + "step": 45083 + }, + { + "epoch": 0.45084, + "grad_norm": 0.916824632802459, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 45084 + }, + { + "epoch": 0.45085, + "grad_norm": 0.7692804863452868, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 45085 + }, + { + "epoch": 0.45086, + "grad_norm": 0.85664751945575, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 45086 + }, + { + "epoch": 0.45087, + "grad_norm": 0.8743405508599777, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 45087 + }, + { + "epoch": 0.45088, + "grad_norm": 0.7741273129963251, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 45088 + }, + { + "epoch": 0.45089, + "grad_norm": 0.7888964400899915, + "learning_rate": 0.003, + "loss": 3.998, + "step": 45089 + }, + { + "epoch": 0.4509, + "grad_norm": 0.781597574175159, + "learning_rate": 0.003, + "loss": 4.0269, + "step": 45090 + }, + { + "epoch": 0.45091, + "grad_norm": 0.7242664496139888, + "learning_rate": 0.003, + "loss": 3.975, + "step": 45091 + }, + { + "epoch": 0.45092, + "grad_norm": 0.7772779239680977, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 45092 + }, + { + "epoch": 0.45093, + "grad_norm": 0.8091224405981323, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 45093 + }, + { + "epoch": 0.45094, + "grad_norm": 0.7813276098237115, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 45094 + }, + { + "epoch": 0.45095, + "grad_norm": 0.8057410217674493, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 45095 + }, + { + "epoch": 0.45096, + "grad_norm": 0.8925225068042929, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 45096 + }, + { + "epoch": 0.45097, + "grad_norm": 0.8560028590488776, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 45097 + }, + { + "epoch": 0.45098, + "grad_norm": 0.7443021606599535, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 45098 + }, + { + "epoch": 0.45099, + "grad_norm": 0.6318862974378813, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 45099 + }, + { + "epoch": 0.451, + "grad_norm": 0.6206730542927089, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 45100 + }, + { + "epoch": 0.45101, + "grad_norm": 0.599365174259168, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 45101 + }, + { + "epoch": 0.45102, + "grad_norm": 0.6514532314778362, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45102 + }, + { + "epoch": 0.45103, + "grad_norm": 0.6908591966663157, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 45103 + }, + { + "epoch": 0.45104, + "grad_norm": 0.6947579029348503, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 45104 + }, + { + "epoch": 0.45105, + "grad_norm": 0.7184991933286682, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 45105 + }, + { + "epoch": 0.45106, + "grad_norm": 0.7212558244688857, + "learning_rate": 0.003, + "loss": 3.973, + "step": 45106 + }, + { + "epoch": 0.45107, + "grad_norm": 0.7295006143032212, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 45107 + }, + { + "epoch": 0.45108, + "grad_norm": 0.710625005695475, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 45108 + }, + { + "epoch": 0.45109, + "grad_norm": 0.8271386656593125, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 45109 + }, + { + "epoch": 0.4511, + "grad_norm": 1.0810330823830447, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 45110 + }, + { + "epoch": 0.45111, + "grad_norm": 1.1321205564981656, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 45111 + }, + { + "epoch": 0.45112, + "grad_norm": 0.881742542848341, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 45112 + }, + { + "epoch": 0.45113, + "grad_norm": 0.8772121379856903, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 45113 + }, + { + "epoch": 0.45114, + "grad_norm": 0.8658765992882373, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 45114 + }, + { + "epoch": 0.45115, + "grad_norm": 0.9141465121264153, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 45115 + }, + { + "epoch": 0.45116, + "grad_norm": 1.0240670056706622, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 45116 + }, + { + "epoch": 0.45117, + "grad_norm": 0.924139227550852, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 45117 + }, + { + "epoch": 0.45118, + "grad_norm": 0.8009387245004979, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 45118 + }, + { + "epoch": 0.45119, + "grad_norm": 0.7648450092540902, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 45119 + }, + { + "epoch": 0.4512, + "grad_norm": 0.808929726019306, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 45120 + }, + { + "epoch": 0.45121, + "grad_norm": 0.8349950610405767, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 45121 + }, + { + "epoch": 0.45122, + "grad_norm": 0.8028690643953206, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 45122 + }, + { + "epoch": 0.45123, + "grad_norm": 0.8270045509960386, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 45123 + }, + { + "epoch": 0.45124, + "grad_norm": 0.8621975356917722, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 45124 + }, + { + "epoch": 0.45125, + "grad_norm": 0.7973024287689251, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 45125 + }, + { + "epoch": 0.45126, + "grad_norm": 0.8081050963551419, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 45126 + }, + { + "epoch": 0.45127, + "grad_norm": 0.8591389600689414, + "learning_rate": 0.003, + "loss": 4.011, + "step": 45127 + }, + { + "epoch": 0.45128, + "grad_norm": 0.9348696724513067, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 45128 + }, + { + "epoch": 0.45129, + "grad_norm": 0.9479235949765845, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 45129 + }, + { + "epoch": 0.4513, + "grad_norm": 1.0219925010284558, + "learning_rate": 0.003, + "loss": 3.985, + "step": 45130 + }, + { + "epoch": 0.45131, + "grad_norm": 0.8405167953492616, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 45131 + }, + { + "epoch": 0.45132, + "grad_norm": 0.6752681298338077, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 45132 + }, + { + "epoch": 0.45133, + "grad_norm": 0.6586196038277864, + "learning_rate": 0.003, + "loss": 3.944, + "step": 45133 + }, + { + "epoch": 0.45134, + "grad_norm": 0.6966765655648429, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 45134 + }, + { + "epoch": 0.45135, + "grad_norm": 0.6676423378850137, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 45135 + }, + { + "epoch": 0.45136, + "grad_norm": 0.5740464912407609, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 45136 + }, + { + "epoch": 0.45137, + "grad_norm": 0.5418281354543315, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 45137 + }, + { + "epoch": 0.45138, + "grad_norm": 0.5697869580456553, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 45138 + }, + { + "epoch": 0.45139, + "grad_norm": 0.562074462600689, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 45139 + }, + { + "epoch": 0.4514, + "grad_norm": 0.6580986183381122, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 45140 + }, + { + "epoch": 0.45141, + "grad_norm": 0.6520261209418561, + "learning_rate": 0.003, + "loss": 3.9398, + "step": 45141 + }, + { + "epoch": 0.45142, + "grad_norm": 0.6169091973899907, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 45142 + }, + { + "epoch": 0.45143, + "grad_norm": 0.5832958006311852, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 45143 + }, + { + "epoch": 0.45144, + "grad_norm": 0.6388806283765136, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 45144 + }, + { + "epoch": 0.45145, + "grad_norm": 0.6258294479090474, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 45145 + }, + { + "epoch": 0.45146, + "grad_norm": 0.6295051104134491, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 45146 + }, + { + "epoch": 0.45147, + "grad_norm": 0.6655489569286575, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 45147 + }, + { + "epoch": 0.45148, + "grad_norm": 0.7414952537589227, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 45148 + }, + { + "epoch": 0.45149, + "grad_norm": 0.7879427220808684, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 45149 + }, + { + "epoch": 0.4515, + "grad_norm": 0.8435903828727238, + "learning_rate": 0.003, + "loss": 3.97, + "step": 45150 + }, + { + "epoch": 0.45151, + "grad_norm": 0.9264195175398358, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 45151 + }, + { + "epoch": 0.45152, + "grad_norm": 0.9653685051142915, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 45152 + }, + { + "epoch": 0.45153, + "grad_norm": 0.8614280125151258, + "learning_rate": 0.003, + "loss": 4.0298, + "step": 45153 + }, + { + "epoch": 0.45154, + "grad_norm": 0.906672709141107, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 45154 + }, + { + "epoch": 0.45155, + "grad_norm": 0.819378191361562, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 45155 + }, + { + "epoch": 0.45156, + "grad_norm": 0.7591460602242927, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 45156 + }, + { + "epoch": 0.45157, + "grad_norm": 0.7100562458827522, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45157 + }, + { + "epoch": 0.45158, + "grad_norm": 0.822204061749465, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 45158 + }, + { + "epoch": 0.45159, + "grad_norm": 1.10229700106646, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 45159 + }, + { + "epoch": 0.4516, + "grad_norm": 1.1292006958378218, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 45160 + }, + { + "epoch": 0.45161, + "grad_norm": 0.8650606021207712, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 45161 + }, + { + "epoch": 0.45162, + "grad_norm": 0.710655630166072, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 45162 + }, + { + "epoch": 0.45163, + "grad_norm": 0.7883792050744353, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 45163 + }, + { + "epoch": 0.45164, + "grad_norm": 0.7828626293082551, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 45164 + }, + { + "epoch": 0.45165, + "grad_norm": 0.8740995981925036, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 45165 + }, + { + "epoch": 0.45166, + "grad_norm": 0.9170809336981851, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 45166 + }, + { + "epoch": 0.45167, + "grad_norm": 0.9221152751719878, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 45167 + }, + { + "epoch": 0.45168, + "grad_norm": 0.8812918338360375, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 45168 + }, + { + "epoch": 0.45169, + "grad_norm": 0.9246100610308987, + "learning_rate": 0.003, + "loss": 3.983, + "step": 45169 + }, + { + "epoch": 0.4517, + "grad_norm": 1.0224333194961286, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 45170 + }, + { + "epoch": 0.45171, + "grad_norm": 0.992806558424103, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 45171 + }, + { + "epoch": 0.45172, + "grad_norm": 0.9890811550411731, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 45172 + }, + { + "epoch": 0.45173, + "grad_norm": 0.9267167964229666, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 45173 + }, + { + "epoch": 0.45174, + "grad_norm": 0.840051684833458, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 45174 + }, + { + "epoch": 0.45175, + "grad_norm": 0.8506088573481809, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 45175 + }, + { + "epoch": 0.45176, + "grad_norm": 0.7866404665900004, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 45176 + }, + { + "epoch": 0.45177, + "grad_norm": 0.7245653273708909, + "learning_rate": 0.003, + "loss": 4.001, + "step": 45177 + }, + { + "epoch": 0.45178, + "grad_norm": 0.7325755333527538, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 45178 + }, + { + "epoch": 0.45179, + "grad_norm": 0.7695036192328675, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 45179 + }, + { + "epoch": 0.4518, + "grad_norm": 0.8033442631586895, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 45180 + }, + { + "epoch": 0.45181, + "grad_norm": 0.732383399038429, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 45181 + }, + { + "epoch": 0.45182, + "grad_norm": 0.7597571951128708, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 45182 + }, + { + "epoch": 0.45183, + "grad_norm": 0.854956445712504, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 45183 + }, + { + "epoch": 0.45184, + "grad_norm": 0.7477501393565286, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 45184 + }, + { + "epoch": 0.45185, + "grad_norm": 0.6634354018718079, + "learning_rate": 0.003, + "loss": 3.9546, + "step": 45185 + }, + { + "epoch": 0.45186, + "grad_norm": 0.6556383108440265, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 45186 + }, + { + "epoch": 0.45187, + "grad_norm": 0.6793623559968293, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 45187 + }, + { + "epoch": 0.45188, + "grad_norm": 0.570361917471078, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 45188 + }, + { + "epoch": 0.45189, + "grad_norm": 0.5783961152678179, + "learning_rate": 0.003, + "loss": 3.979, + "step": 45189 + }, + { + "epoch": 0.4519, + "grad_norm": 0.6247129600022477, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 45190 + }, + { + "epoch": 0.45191, + "grad_norm": 0.6822300128772073, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 45191 + }, + { + "epoch": 0.45192, + "grad_norm": 0.7288663943825778, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 45192 + }, + { + "epoch": 0.45193, + "grad_norm": 0.6918217116394885, + "learning_rate": 0.003, + "loss": 4.009, + "step": 45193 + }, + { + "epoch": 0.45194, + "grad_norm": 0.6274905410120919, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 45194 + }, + { + "epoch": 0.45195, + "grad_norm": 0.698192555484593, + "learning_rate": 0.003, + "loss": 3.977, + "step": 45195 + }, + { + "epoch": 0.45196, + "grad_norm": 0.9115732968910162, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 45196 + }, + { + "epoch": 0.45197, + "grad_norm": 1.110837943674115, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 45197 + }, + { + "epoch": 0.45198, + "grad_norm": 0.8593035361709188, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 45198 + }, + { + "epoch": 0.45199, + "grad_norm": 0.7869872767105385, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 45199 + }, + { + "epoch": 0.452, + "grad_norm": 0.6428662703204711, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 45200 + }, + { + "epoch": 0.45201, + "grad_norm": 0.6259910283284236, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 45201 + }, + { + "epoch": 0.45202, + "grad_norm": 0.7329680665369563, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 45202 + }, + { + "epoch": 0.45203, + "grad_norm": 0.7029481800189611, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 45203 + }, + { + "epoch": 0.45204, + "grad_norm": 0.7287804191724134, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 45204 + }, + { + "epoch": 0.45205, + "grad_norm": 0.7938178643422896, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 45205 + }, + { + "epoch": 0.45206, + "grad_norm": 0.7983546760426392, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 45206 + }, + { + "epoch": 0.45207, + "grad_norm": 0.7291199144712439, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 45207 + }, + { + "epoch": 0.45208, + "grad_norm": 0.8026624055655669, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 45208 + }, + { + "epoch": 0.45209, + "grad_norm": 0.8756950379080274, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 45209 + }, + { + "epoch": 0.4521, + "grad_norm": 0.9447977344850691, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 45210 + }, + { + "epoch": 0.45211, + "grad_norm": 0.9749487646118469, + "learning_rate": 0.003, + "loss": 3.999, + "step": 45211 + }, + { + "epoch": 0.45212, + "grad_norm": 1.0035645009977232, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 45212 + }, + { + "epoch": 0.45213, + "grad_norm": 1.137219920222207, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 45213 + }, + { + "epoch": 0.45214, + "grad_norm": 0.8274249570192742, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 45214 + }, + { + "epoch": 0.45215, + "grad_norm": 0.7329943326421435, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 45215 + }, + { + "epoch": 0.45216, + "grad_norm": 0.825016035009505, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 45216 + }, + { + "epoch": 0.45217, + "grad_norm": 0.956895691903889, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 45217 + }, + { + "epoch": 0.45218, + "grad_norm": 1.05200521855725, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 45218 + }, + { + "epoch": 0.45219, + "grad_norm": 0.8537255655618091, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 45219 + }, + { + "epoch": 0.4522, + "grad_norm": 0.7565638174448209, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 45220 + }, + { + "epoch": 0.45221, + "grad_norm": 0.7493860799476457, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 45221 + }, + { + "epoch": 0.45222, + "grad_norm": 0.7273174568509051, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 45222 + }, + { + "epoch": 0.45223, + "grad_norm": 0.6821672819769934, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 45223 + }, + { + "epoch": 0.45224, + "grad_norm": 0.693353757414468, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 45224 + }, + { + "epoch": 0.45225, + "grad_norm": 0.6258221697859929, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 45225 + }, + { + "epoch": 0.45226, + "grad_norm": 0.5332933907461458, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 45226 + }, + { + "epoch": 0.45227, + "grad_norm": 0.6133737090222838, + "learning_rate": 0.003, + "loss": 3.998, + "step": 45227 + }, + { + "epoch": 0.45228, + "grad_norm": 0.7279148395913378, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 45228 + }, + { + "epoch": 0.45229, + "grad_norm": 0.7982051822871864, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 45229 + }, + { + "epoch": 0.4523, + "grad_norm": 0.7392857397840056, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 45230 + }, + { + "epoch": 0.45231, + "grad_norm": 0.6899837547580432, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 45231 + }, + { + "epoch": 0.45232, + "grad_norm": 0.7364677536180866, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 45232 + }, + { + "epoch": 0.45233, + "grad_norm": 0.6402987171774006, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 45233 + }, + { + "epoch": 0.45234, + "grad_norm": 0.7758270042229868, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 45234 + }, + { + "epoch": 0.45235, + "grad_norm": 1.0149946575933226, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 45235 + }, + { + "epoch": 0.45236, + "grad_norm": 1.0472663032028418, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 45236 + }, + { + "epoch": 0.45237, + "grad_norm": 0.7618566581655363, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 45237 + }, + { + "epoch": 0.45238, + "grad_norm": 0.70391489701402, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 45238 + }, + { + "epoch": 0.45239, + "grad_norm": 0.7876269832076599, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 45239 + }, + { + "epoch": 0.4524, + "grad_norm": 0.8348169802802172, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 45240 + }, + { + "epoch": 0.45241, + "grad_norm": 0.7758806202334577, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 45241 + }, + { + "epoch": 0.45242, + "grad_norm": 0.832281612391265, + "learning_rate": 0.003, + "loss": 3.967, + "step": 45242 + }, + { + "epoch": 0.45243, + "grad_norm": 0.8048107789990141, + "learning_rate": 0.003, + "loss": 3.972, + "step": 45243 + }, + { + "epoch": 0.45244, + "grad_norm": 0.7757557272993799, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 45244 + }, + { + "epoch": 0.45245, + "grad_norm": 0.7477836548376766, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 45245 + }, + { + "epoch": 0.45246, + "grad_norm": 0.728696707043395, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 45246 + }, + { + "epoch": 0.45247, + "grad_norm": 0.880156364022067, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 45247 + }, + { + "epoch": 0.45248, + "grad_norm": 0.9326797066297934, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 45248 + }, + { + "epoch": 0.45249, + "grad_norm": 0.8843444591365085, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 45249 + }, + { + "epoch": 0.4525, + "grad_norm": 0.8054171796553921, + "learning_rate": 0.003, + "loss": 4.007, + "step": 45250 + }, + { + "epoch": 0.45251, + "grad_norm": 0.776712180167148, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 45251 + }, + { + "epoch": 0.45252, + "grad_norm": 0.7918196735156233, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 45252 + }, + { + "epoch": 0.45253, + "grad_norm": 0.7141000554084498, + "learning_rate": 0.003, + "loss": 4.008, + "step": 45253 + }, + { + "epoch": 0.45254, + "grad_norm": 0.8357720871758445, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 45254 + }, + { + "epoch": 0.45255, + "grad_norm": 0.7831909328560575, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 45255 + }, + { + "epoch": 0.45256, + "grad_norm": 0.6650146464725919, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 45256 + }, + { + "epoch": 0.45257, + "grad_norm": 0.6438946945363706, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 45257 + }, + { + "epoch": 0.45258, + "grad_norm": 0.6228854778086922, + "learning_rate": 0.003, + "loss": 4.008, + "step": 45258 + }, + { + "epoch": 0.45259, + "grad_norm": 0.6771663863399296, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 45259 + }, + { + "epoch": 0.4526, + "grad_norm": 0.7367204413333821, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 45260 + }, + { + "epoch": 0.45261, + "grad_norm": 0.9047452209163767, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 45261 + }, + { + "epoch": 0.45262, + "grad_norm": 0.9592694030726563, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 45262 + }, + { + "epoch": 0.45263, + "grad_norm": 1.0257032306353375, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 45263 + }, + { + "epoch": 0.45264, + "grad_norm": 0.9313624657296223, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 45264 + }, + { + "epoch": 0.45265, + "grad_norm": 0.7940394213522042, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 45265 + }, + { + "epoch": 0.45266, + "grad_norm": 0.7015139257059423, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 45266 + }, + { + "epoch": 0.45267, + "grad_norm": 0.701290710947746, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 45267 + }, + { + "epoch": 0.45268, + "grad_norm": 0.8699781579599847, + "learning_rate": 0.003, + "loss": 3.978, + "step": 45268 + }, + { + "epoch": 0.45269, + "grad_norm": 1.0597479365820142, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 45269 + }, + { + "epoch": 0.4527, + "grad_norm": 0.881572102683861, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 45270 + }, + { + "epoch": 0.45271, + "grad_norm": 0.8586195885898978, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 45271 + }, + { + "epoch": 0.45272, + "grad_norm": 0.8961660108789269, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 45272 + }, + { + "epoch": 0.45273, + "grad_norm": 0.8958313390789074, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 45273 + }, + { + "epoch": 0.45274, + "grad_norm": 0.8053568722796046, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45274 + }, + { + "epoch": 0.45275, + "grad_norm": 0.7330829785527649, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 45275 + }, + { + "epoch": 0.45276, + "grad_norm": 0.6720507788403192, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 45276 + }, + { + "epoch": 0.45277, + "grad_norm": 0.6956773219204111, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 45277 + }, + { + "epoch": 0.45278, + "grad_norm": 0.6451204239293237, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 45278 + }, + { + "epoch": 0.45279, + "grad_norm": 0.7210858044947319, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 45279 + }, + { + "epoch": 0.4528, + "grad_norm": 0.7659019556320177, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 45280 + }, + { + "epoch": 0.45281, + "grad_norm": 0.693990043790657, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 45281 + }, + { + "epoch": 0.45282, + "grad_norm": 0.7529480177930983, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 45282 + }, + { + "epoch": 0.45283, + "grad_norm": 0.7922050164956234, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 45283 + }, + { + "epoch": 0.45284, + "grad_norm": 0.7236582882784142, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 45284 + }, + { + "epoch": 0.45285, + "grad_norm": 0.7056942499950617, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 45285 + }, + { + "epoch": 0.45286, + "grad_norm": 0.7173210008807863, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 45286 + }, + { + "epoch": 0.45287, + "grad_norm": 0.76362369996366, + "learning_rate": 0.003, + "loss": 4.0185, + "step": 45287 + }, + { + "epoch": 0.45288, + "grad_norm": 0.8895392506428386, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 45288 + }, + { + "epoch": 0.45289, + "grad_norm": 1.0005357395908436, + "learning_rate": 0.003, + "loss": 3.943, + "step": 45289 + }, + { + "epoch": 0.4529, + "grad_norm": 1.1763065786054843, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 45290 + }, + { + "epoch": 0.45291, + "grad_norm": 0.6054105622964204, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 45291 + }, + { + "epoch": 0.45292, + "grad_norm": 0.6734536714377931, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 45292 + }, + { + "epoch": 0.45293, + "grad_norm": 0.7543221443104233, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 45293 + }, + { + "epoch": 0.45294, + "grad_norm": 0.7613500373224276, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 45294 + }, + { + "epoch": 0.45295, + "grad_norm": 0.6930013771069629, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 45295 + }, + { + "epoch": 0.45296, + "grad_norm": 0.7924375222105171, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 45296 + }, + { + "epoch": 0.45297, + "grad_norm": 0.7610766736174783, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 45297 + }, + { + "epoch": 0.45298, + "grad_norm": 0.7009365853113341, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 45298 + }, + { + "epoch": 0.45299, + "grad_norm": 0.5954704550895125, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 45299 + }, + { + "epoch": 0.453, + "grad_norm": 0.735094952890486, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 45300 + }, + { + "epoch": 0.45301, + "grad_norm": 0.8930149031547798, + "learning_rate": 0.003, + "loss": 4.007, + "step": 45301 + }, + { + "epoch": 0.45302, + "grad_norm": 0.92874620640782, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 45302 + }, + { + "epoch": 0.45303, + "grad_norm": 0.8610911422856351, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 45303 + }, + { + "epoch": 0.45304, + "grad_norm": 0.8251780822509832, + "learning_rate": 0.003, + "loss": 3.981, + "step": 45304 + }, + { + "epoch": 0.45305, + "grad_norm": 0.9236122818893808, + "learning_rate": 0.003, + "loss": 3.997, + "step": 45305 + }, + { + "epoch": 0.45306, + "grad_norm": 0.9268364655409754, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 45306 + }, + { + "epoch": 0.45307, + "grad_norm": 0.9237686920961468, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 45307 + }, + { + "epoch": 0.45308, + "grad_norm": 0.9060263340920449, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 45308 + }, + { + "epoch": 0.45309, + "grad_norm": 0.7721658625024301, + "learning_rate": 0.003, + "loss": 4.013, + "step": 45309 + }, + { + "epoch": 0.4531, + "grad_norm": 0.6645541814636751, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 45310 + }, + { + "epoch": 0.45311, + "grad_norm": 0.5976295081987903, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 45311 + }, + { + "epoch": 0.45312, + "grad_norm": 0.6400612873626031, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 45312 + }, + { + "epoch": 0.45313, + "grad_norm": 0.6021399580617541, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 45313 + }, + { + "epoch": 0.45314, + "grad_norm": 0.7770062269059809, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 45314 + }, + { + "epoch": 0.45315, + "grad_norm": 0.8852396645526359, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 45315 + }, + { + "epoch": 0.45316, + "grad_norm": 0.9613155278937966, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 45316 + }, + { + "epoch": 0.45317, + "grad_norm": 0.919648559394291, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 45317 + }, + { + "epoch": 0.45318, + "grad_norm": 0.8163708490443745, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 45318 + }, + { + "epoch": 0.45319, + "grad_norm": 0.6717826069166934, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 45319 + }, + { + "epoch": 0.4532, + "grad_norm": 0.6022791305996721, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 45320 + }, + { + "epoch": 0.45321, + "grad_norm": 0.545510489453022, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 45321 + }, + { + "epoch": 0.45322, + "grad_norm": 0.5174542295096327, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 45322 + }, + { + "epoch": 0.45323, + "grad_norm": 0.5760999723828201, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 45323 + }, + { + "epoch": 0.45324, + "grad_norm": 0.5996864896030648, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 45324 + }, + { + "epoch": 0.45325, + "grad_norm": 0.6501552912888694, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 45325 + }, + { + "epoch": 0.45326, + "grad_norm": 0.6968577027969679, + "learning_rate": 0.003, + "loss": 3.9405, + "step": 45326 + }, + { + "epoch": 0.45327, + "grad_norm": 0.8781116541221309, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 45327 + }, + { + "epoch": 0.45328, + "grad_norm": 1.09227964828566, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 45328 + }, + { + "epoch": 0.45329, + "grad_norm": 0.866344614238971, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 45329 + }, + { + "epoch": 0.4533, + "grad_norm": 0.7372064960971974, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 45330 + }, + { + "epoch": 0.45331, + "grad_norm": 0.7622464514729851, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 45331 + }, + { + "epoch": 0.45332, + "grad_norm": 0.8496779305536505, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45332 + }, + { + "epoch": 0.45333, + "grad_norm": 0.8056528293097679, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 45333 + }, + { + "epoch": 0.45334, + "grad_norm": 0.8271579215373916, + "learning_rate": 0.003, + "loss": 3.9373, + "step": 45334 + }, + { + "epoch": 0.45335, + "grad_norm": 0.7915962082528251, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 45335 + }, + { + "epoch": 0.45336, + "grad_norm": 0.7661746818332765, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 45336 + }, + { + "epoch": 0.45337, + "grad_norm": 0.7634486809892018, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 45337 + }, + { + "epoch": 0.45338, + "grad_norm": 0.8196440687510488, + "learning_rate": 0.003, + "loss": 3.975, + "step": 45338 + }, + { + "epoch": 0.45339, + "grad_norm": 0.8870466955593728, + "learning_rate": 0.003, + "loss": 3.969, + "step": 45339 + }, + { + "epoch": 0.4534, + "grad_norm": 0.8901220017534228, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 45340 + }, + { + "epoch": 0.45341, + "grad_norm": 0.8020732417883176, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 45341 + }, + { + "epoch": 0.45342, + "grad_norm": 0.8272826937827907, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 45342 + }, + { + "epoch": 0.45343, + "grad_norm": 0.8602952503057342, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 45343 + }, + { + "epoch": 0.45344, + "grad_norm": 0.9714855849471719, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 45344 + }, + { + "epoch": 0.45345, + "grad_norm": 0.8714652191453082, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 45345 + }, + { + "epoch": 0.45346, + "grad_norm": 0.8063273818042924, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 45346 + }, + { + "epoch": 0.45347, + "grad_norm": 0.8715869290113583, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 45347 + }, + { + "epoch": 0.45348, + "grad_norm": 0.7590011923259175, + "learning_rate": 0.003, + "loss": 3.9447, + "step": 45348 + }, + { + "epoch": 0.45349, + "grad_norm": 0.7395195172183704, + "learning_rate": 0.003, + "loss": 3.988, + "step": 45349 + }, + { + "epoch": 0.4535, + "grad_norm": 0.864538819310165, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 45350 + }, + { + "epoch": 0.45351, + "grad_norm": 1.0067511558804556, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 45351 + }, + { + "epoch": 0.45352, + "grad_norm": 0.9555662962620162, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 45352 + }, + { + "epoch": 0.45353, + "grad_norm": 0.8643108730554392, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 45353 + }, + { + "epoch": 0.45354, + "grad_norm": 0.7417476583231002, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 45354 + }, + { + "epoch": 0.45355, + "grad_norm": 0.6196897365363273, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 45355 + }, + { + "epoch": 0.45356, + "grad_norm": 0.6810748923910754, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 45356 + }, + { + "epoch": 0.45357, + "grad_norm": 0.6967290557250767, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 45357 + }, + { + "epoch": 0.45358, + "grad_norm": 0.7562392850694643, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 45358 + }, + { + "epoch": 0.45359, + "grad_norm": 0.6650439605571975, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 45359 + }, + { + "epoch": 0.4536, + "grad_norm": 0.5894674908931687, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 45360 + }, + { + "epoch": 0.45361, + "grad_norm": 0.5192758428715629, + "learning_rate": 0.003, + "loss": 3.977, + "step": 45361 + }, + { + "epoch": 0.45362, + "grad_norm": 0.5661634476649664, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 45362 + }, + { + "epoch": 0.45363, + "grad_norm": 0.6174845917215905, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 45363 + }, + { + "epoch": 0.45364, + "grad_norm": 0.6865344207223165, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 45364 + }, + { + "epoch": 0.45365, + "grad_norm": 0.8889944517158422, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 45365 + }, + { + "epoch": 0.45366, + "grad_norm": 1.1294056739581588, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 45366 + }, + { + "epoch": 0.45367, + "grad_norm": 0.8510477398092497, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 45367 + }, + { + "epoch": 0.45368, + "grad_norm": 0.8074659152141824, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 45368 + }, + { + "epoch": 0.45369, + "grad_norm": 0.7663884581744733, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 45369 + }, + { + "epoch": 0.4537, + "grad_norm": 0.836099009119525, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 45370 + }, + { + "epoch": 0.45371, + "grad_norm": 0.9538827722793011, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 45371 + }, + { + "epoch": 0.45372, + "grad_norm": 1.1581144628589135, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 45372 + }, + { + "epoch": 0.45373, + "grad_norm": 0.9286613671256464, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 45373 + }, + { + "epoch": 0.45374, + "grad_norm": 0.7037829972816109, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 45374 + }, + { + "epoch": 0.45375, + "grad_norm": 0.7712579370499831, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 45375 + }, + { + "epoch": 0.45376, + "grad_norm": 0.6666373475358606, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 45376 + }, + { + "epoch": 0.45377, + "grad_norm": 0.5825575744598614, + "learning_rate": 0.003, + "loss": 3.983, + "step": 45377 + }, + { + "epoch": 0.45378, + "grad_norm": 0.6147493251663287, + "learning_rate": 0.003, + "loss": 3.986, + "step": 45378 + }, + { + "epoch": 0.45379, + "grad_norm": 0.7094564936142354, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 45379 + }, + { + "epoch": 0.4538, + "grad_norm": 0.8096867783462184, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 45380 + }, + { + "epoch": 0.45381, + "grad_norm": 1.047871383842373, + "learning_rate": 0.003, + "loss": 3.9416, + "step": 45381 + }, + { + "epoch": 0.45382, + "grad_norm": 1.320864841227125, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 45382 + }, + { + "epoch": 0.45383, + "grad_norm": 0.8196649145338029, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 45383 + }, + { + "epoch": 0.45384, + "grad_norm": 1.2633158471842951, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 45384 + }, + { + "epoch": 0.45385, + "grad_norm": 1.3948914345859718, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 45385 + }, + { + "epoch": 0.45386, + "grad_norm": 0.8010219633237441, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 45386 + }, + { + "epoch": 0.45387, + "grad_norm": 0.8388795071523178, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 45387 + }, + { + "epoch": 0.45388, + "grad_norm": 0.8319161205155211, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 45388 + }, + { + "epoch": 0.45389, + "grad_norm": 0.82543856857801, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 45389 + }, + { + "epoch": 0.4539, + "grad_norm": 0.8572461486366272, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 45390 + }, + { + "epoch": 0.45391, + "grad_norm": 1.003853534194317, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 45391 + }, + { + "epoch": 0.45392, + "grad_norm": 1.000133807715818, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 45392 + }, + { + "epoch": 0.45393, + "grad_norm": 0.7931903568361903, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 45393 + }, + { + "epoch": 0.45394, + "grad_norm": 0.7017141908957154, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 45394 + }, + { + "epoch": 0.45395, + "grad_norm": 0.7391816570519869, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 45395 + }, + { + "epoch": 0.45396, + "grad_norm": 0.749054147626167, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 45396 + }, + { + "epoch": 0.45397, + "grad_norm": 0.8778219140443172, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 45397 + }, + { + "epoch": 0.45398, + "grad_norm": 0.8873543163784685, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 45398 + }, + { + "epoch": 0.45399, + "grad_norm": 0.8740669267118016, + "learning_rate": 0.003, + "loss": 4.013, + "step": 45399 + }, + { + "epoch": 0.454, + "grad_norm": 0.9532886562981311, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 45400 + }, + { + "epoch": 0.45401, + "grad_norm": 0.8078127323658699, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 45401 + }, + { + "epoch": 0.45402, + "grad_norm": 0.7924828262055066, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 45402 + }, + { + "epoch": 0.45403, + "grad_norm": 0.8681065907666419, + "learning_rate": 0.003, + "loss": 4.016, + "step": 45403 + }, + { + "epoch": 0.45404, + "grad_norm": 0.8643231719531678, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 45404 + }, + { + "epoch": 0.45405, + "grad_norm": 0.8298735941920354, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 45405 + }, + { + "epoch": 0.45406, + "grad_norm": 0.8354908884925731, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 45406 + }, + { + "epoch": 0.45407, + "grad_norm": 0.9222127651872238, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 45407 + }, + { + "epoch": 0.45408, + "grad_norm": 1.0205620200212497, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 45408 + }, + { + "epoch": 0.45409, + "grad_norm": 1.0186483710564178, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 45409 + }, + { + "epoch": 0.4541, + "grad_norm": 0.8121957604816648, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 45410 + }, + { + "epoch": 0.45411, + "grad_norm": 0.8160086496934339, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 45411 + }, + { + "epoch": 0.45412, + "grad_norm": 0.9111105361585168, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 45412 + }, + { + "epoch": 0.45413, + "grad_norm": 0.9153082825023009, + "learning_rate": 0.003, + "loss": 3.999, + "step": 45413 + }, + { + "epoch": 0.45414, + "grad_norm": 0.8988603171457507, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 45414 + }, + { + "epoch": 0.45415, + "grad_norm": 0.8460336028357321, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 45415 + }, + { + "epoch": 0.45416, + "grad_norm": 0.7218775518159363, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 45416 + }, + { + "epoch": 0.45417, + "grad_norm": 0.668337484729157, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 45417 + }, + { + "epoch": 0.45418, + "grad_norm": 0.6473353037230845, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 45418 + }, + { + "epoch": 0.45419, + "grad_norm": 0.6387524629134926, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 45419 + }, + { + "epoch": 0.4542, + "grad_norm": 0.680572844324892, + "learning_rate": 0.003, + "loss": 3.981, + "step": 45420 + }, + { + "epoch": 0.45421, + "grad_norm": 0.7004814705658409, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 45421 + }, + { + "epoch": 0.45422, + "grad_norm": 0.7043561356163953, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 45422 + }, + { + "epoch": 0.45423, + "grad_norm": 0.7631933402722428, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 45423 + }, + { + "epoch": 0.45424, + "grad_norm": 0.6995826910115548, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 45424 + }, + { + "epoch": 0.45425, + "grad_norm": 0.6028592839163733, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 45425 + }, + { + "epoch": 0.45426, + "grad_norm": 0.5665044369236976, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 45426 + }, + { + "epoch": 0.45427, + "grad_norm": 0.6800545971798166, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 45427 + }, + { + "epoch": 0.45428, + "grad_norm": 0.8038082135401067, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 45428 + }, + { + "epoch": 0.45429, + "grad_norm": 0.8700831547455755, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 45429 + }, + { + "epoch": 0.4543, + "grad_norm": 0.8129351007174656, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 45430 + }, + { + "epoch": 0.45431, + "grad_norm": 0.9602173907377167, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 45431 + }, + { + "epoch": 0.45432, + "grad_norm": 1.0657380773346181, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 45432 + }, + { + "epoch": 0.45433, + "grad_norm": 0.8509973994927459, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 45433 + }, + { + "epoch": 0.45434, + "grad_norm": 0.7901959127257974, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 45434 + }, + { + "epoch": 0.45435, + "grad_norm": 0.7876628404559417, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 45435 + }, + { + "epoch": 0.45436, + "grad_norm": 0.7919648055818004, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 45436 + }, + { + "epoch": 0.45437, + "grad_norm": 0.8285649203289401, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 45437 + }, + { + "epoch": 0.45438, + "grad_norm": 0.7842027804560376, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 45438 + }, + { + "epoch": 0.45439, + "grad_norm": 0.8010663807305365, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 45439 + }, + { + "epoch": 0.4544, + "grad_norm": 0.899449243437036, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 45440 + }, + { + "epoch": 0.45441, + "grad_norm": 0.9203859014723935, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 45441 + }, + { + "epoch": 0.45442, + "grad_norm": 1.0053236773889223, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 45442 + }, + { + "epoch": 0.45443, + "grad_norm": 1.0735379466413413, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 45443 + }, + { + "epoch": 0.45444, + "grad_norm": 0.77708828303217, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 45444 + }, + { + "epoch": 0.45445, + "grad_norm": 0.6717674902885453, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 45445 + }, + { + "epoch": 0.45446, + "grad_norm": 0.6585475810040012, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 45446 + }, + { + "epoch": 0.45447, + "grad_norm": 0.6359519486260687, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 45447 + }, + { + "epoch": 0.45448, + "grad_norm": 0.6604038077142177, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 45448 + }, + { + "epoch": 0.45449, + "grad_norm": 0.5895144099544211, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 45449 + }, + { + "epoch": 0.4545, + "grad_norm": 0.6389917165172833, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 45450 + }, + { + "epoch": 0.45451, + "grad_norm": 0.5617679446883083, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 45451 + }, + { + "epoch": 0.45452, + "grad_norm": 0.5364294132557039, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 45452 + }, + { + "epoch": 0.45453, + "grad_norm": 0.5061947255385736, + "learning_rate": 0.003, + "loss": 3.988, + "step": 45453 + }, + { + "epoch": 0.45454, + "grad_norm": 0.46715459866520254, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 45454 + }, + { + "epoch": 0.45455, + "grad_norm": 0.4460978858895178, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 45455 + }, + { + "epoch": 0.45456, + "grad_norm": 0.4323789737209894, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 45456 + }, + { + "epoch": 0.45457, + "grad_norm": 0.5343970353158671, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 45457 + }, + { + "epoch": 0.45458, + "grad_norm": 0.724686176862932, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 45458 + }, + { + "epoch": 0.45459, + "grad_norm": 0.9452668728162629, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 45459 + }, + { + "epoch": 0.4546, + "grad_norm": 1.1481793468123584, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 45460 + }, + { + "epoch": 0.45461, + "grad_norm": 0.904263617971631, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 45461 + }, + { + "epoch": 0.45462, + "grad_norm": 0.8175211269308519, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 45462 + }, + { + "epoch": 0.45463, + "grad_norm": 0.8142191297925827, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 45463 + }, + { + "epoch": 0.45464, + "grad_norm": 0.9405682303469054, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45464 + }, + { + "epoch": 0.45465, + "grad_norm": 0.8649611933816987, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 45465 + }, + { + "epoch": 0.45466, + "grad_norm": 0.7195349581099277, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45466 + }, + { + "epoch": 0.45467, + "grad_norm": 0.6973252718334111, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 45467 + }, + { + "epoch": 0.45468, + "grad_norm": 0.8872261838258909, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 45468 + }, + { + "epoch": 0.45469, + "grad_norm": 0.9295851913205073, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 45469 + }, + { + "epoch": 0.4547, + "grad_norm": 0.896092251455999, + "learning_rate": 0.003, + "loss": 3.959, + "step": 45470 + }, + { + "epoch": 0.45471, + "grad_norm": 0.9362438937452818, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 45471 + }, + { + "epoch": 0.45472, + "grad_norm": 0.9019348266265392, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 45472 + }, + { + "epoch": 0.45473, + "grad_norm": 0.946161623366752, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 45473 + }, + { + "epoch": 0.45474, + "grad_norm": 0.8641567860527934, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 45474 + }, + { + "epoch": 0.45475, + "grad_norm": 0.8750848367544719, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 45475 + }, + { + "epoch": 0.45476, + "grad_norm": 1.054938614591537, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 45476 + }, + { + "epoch": 0.45477, + "grad_norm": 0.9007211292198729, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 45477 + }, + { + "epoch": 0.45478, + "grad_norm": 0.8918209510680783, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 45478 + }, + { + "epoch": 0.45479, + "grad_norm": 0.8648978919355045, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 45479 + }, + { + "epoch": 0.4548, + "grad_norm": 0.7287917846128954, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 45480 + }, + { + "epoch": 0.45481, + "grad_norm": 0.7457429714593048, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 45481 + }, + { + "epoch": 0.45482, + "grad_norm": 0.8131895715003703, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 45482 + }, + { + "epoch": 0.45483, + "grad_norm": 0.8701482170707605, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 45483 + }, + { + "epoch": 0.45484, + "grad_norm": 0.9154965007176787, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 45484 + }, + { + "epoch": 0.45485, + "grad_norm": 0.9581077886551714, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 45485 + }, + { + "epoch": 0.45486, + "grad_norm": 1.017073394269003, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 45486 + }, + { + "epoch": 0.45487, + "grad_norm": 0.8102218377914956, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 45487 + }, + { + "epoch": 0.45488, + "grad_norm": 0.6928879444922476, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 45488 + }, + { + "epoch": 0.45489, + "grad_norm": 0.65642109056485, + "learning_rate": 0.003, + "loss": 3.985, + "step": 45489 + }, + { + "epoch": 0.4549, + "grad_norm": 0.6725302148385858, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 45490 + }, + { + "epoch": 0.45491, + "grad_norm": 0.6194774227886575, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 45491 + }, + { + "epoch": 0.45492, + "grad_norm": 0.5955374506466906, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 45492 + }, + { + "epoch": 0.45493, + "grad_norm": 0.5573964917786265, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 45493 + }, + { + "epoch": 0.45494, + "grad_norm": 0.5161834199582703, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 45494 + }, + { + "epoch": 0.45495, + "grad_norm": 0.503412602000699, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 45495 + }, + { + "epoch": 0.45496, + "grad_norm": 0.4406179137335162, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 45496 + }, + { + "epoch": 0.45497, + "grad_norm": 0.44113013891496755, + "learning_rate": 0.003, + "loss": 3.987, + "step": 45497 + }, + { + "epoch": 0.45498, + "grad_norm": 0.4667816117391842, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 45498 + }, + { + "epoch": 0.45499, + "grad_norm": 0.5156305148068051, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 45499 + }, + { + "epoch": 0.455, + "grad_norm": 0.5559882272476145, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 45500 + }, + { + "epoch": 0.45501, + "grad_norm": 0.6631865780703511, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 45501 + }, + { + "epoch": 0.45502, + "grad_norm": 0.7877902490308232, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 45502 + }, + { + "epoch": 0.45503, + "grad_norm": 0.9132095310258281, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 45503 + }, + { + "epoch": 0.45504, + "grad_norm": 0.9000113929398172, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 45504 + }, + { + "epoch": 0.45505, + "grad_norm": 0.8627094406422318, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 45505 + }, + { + "epoch": 0.45506, + "grad_norm": 0.7803131706582389, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 45506 + }, + { + "epoch": 0.45507, + "grad_norm": 0.7735950266057972, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 45507 + }, + { + "epoch": 0.45508, + "grad_norm": 0.845811466303726, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 45508 + }, + { + "epoch": 0.45509, + "grad_norm": 0.8762325948074823, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 45509 + }, + { + "epoch": 0.4551, + "grad_norm": 0.8658944883506343, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 45510 + }, + { + "epoch": 0.45511, + "grad_norm": 0.8403570561348318, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 45511 + }, + { + "epoch": 0.45512, + "grad_norm": 0.8335970201162434, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 45512 + }, + { + "epoch": 0.45513, + "grad_norm": 0.7839156402402789, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 45513 + }, + { + "epoch": 0.45514, + "grad_norm": 0.8602204340828224, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 45514 + }, + { + "epoch": 0.45515, + "grad_norm": 0.7977965144981469, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 45515 + }, + { + "epoch": 0.45516, + "grad_norm": 0.8069330106618048, + "learning_rate": 0.003, + "loss": 3.999, + "step": 45516 + }, + { + "epoch": 0.45517, + "grad_norm": 0.8972767495765298, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 45517 + }, + { + "epoch": 0.45518, + "grad_norm": 0.8998107911580336, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 45518 + }, + { + "epoch": 0.45519, + "grad_norm": 0.9454198846750205, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 45519 + }, + { + "epoch": 0.4552, + "grad_norm": 0.9384109559845203, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 45520 + }, + { + "epoch": 0.45521, + "grad_norm": 0.8214698279770398, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 45521 + }, + { + "epoch": 0.45522, + "grad_norm": 0.7579362077362953, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 45522 + }, + { + "epoch": 0.45523, + "grad_norm": 0.7486464583016912, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 45523 + }, + { + "epoch": 0.45524, + "grad_norm": 0.7429071142654132, + "learning_rate": 0.003, + "loss": 3.987, + "step": 45524 + }, + { + "epoch": 0.45525, + "grad_norm": 0.6918024373785049, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 45525 + }, + { + "epoch": 0.45526, + "grad_norm": 0.7648090118977261, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 45526 + }, + { + "epoch": 0.45527, + "grad_norm": 0.8037602036909873, + "learning_rate": 0.003, + "loss": 3.987, + "step": 45527 + }, + { + "epoch": 0.45528, + "grad_norm": 0.8149907684449018, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 45528 + }, + { + "epoch": 0.45529, + "grad_norm": 0.7593805055074554, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 45529 + }, + { + "epoch": 0.4553, + "grad_norm": 0.8941566735927677, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 45530 + }, + { + "epoch": 0.45531, + "grad_norm": 0.9048114949970116, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 45531 + }, + { + "epoch": 0.45532, + "grad_norm": 0.9643573660276864, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 45532 + }, + { + "epoch": 0.45533, + "grad_norm": 1.0331079221229902, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 45533 + }, + { + "epoch": 0.45534, + "grad_norm": 0.9451608526741121, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 45534 + }, + { + "epoch": 0.45535, + "grad_norm": 0.9794347710872855, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 45535 + }, + { + "epoch": 0.45536, + "grad_norm": 0.9446044800647164, + "learning_rate": 0.003, + "loss": 4.009, + "step": 45536 + }, + { + "epoch": 0.45537, + "grad_norm": 0.7893431029072405, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 45537 + }, + { + "epoch": 0.45538, + "grad_norm": 0.78189411985945, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 45538 + }, + { + "epoch": 0.45539, + "grad_norm": 0.6846854388424217, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 45539 + }, + { + "epoch": 0.4554, + "grad_norm": 0.7178795601693921, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 45540 + }, + { + "epoch": 0.45541, + "grad_norm": 0.675985038073533, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 45541 + }, + { + "epoch": 0.45542, + "grad_norm": 0.5686668300271944, + "learning_rate": 0.003, + "loss": 3.996, + "step": 45542 + }, + { + "epoch": 0.45543, + "grad_norm": 0.5525916613180375, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 45543 + }, + { + "epoch": 0.45544, + "grad_norm": 0.6618999963171046, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 45544 + }, + { + "epoch": 0.45545, + "grad_norm": 0.7340412319236179, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 45545 + }, + { + "epoch": 0.45546, + "grad_norm": 0.8427681097801174, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 45546 + }, + { + "epoch": 0.45547, + "grad_norm": 0.8770776259385593, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 45547 + }, + { + "epoch": 0.45548, + "grad_norm": 0.9068221043200931, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 45548 + }, + { + "epoch": 0.45549, + "grad_norm": 0.8304015763517681, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 45549 + }, + { + "epoch": 0.4555, + "grad_norm": 0.8057906856530065, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 45550 + }, + { + "epoch": 0.45551, + "grad_norm": 0.758218945877282, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 45551 + }, + { + "epoch": 0.45552, + "grad_norm": 0.709240875602472, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 45552 + }, + { + "epoch": 0.45553, + "grad_norm": 0.707872848147174, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 45553 + }, + { + "epoch": 0.45554, + "grad_norm": 0.7144878177811759, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 45554 + }, + { + "epoch": 0.45555, + "grad_norm": 0.6836335094433283, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 45555 + }, + { + "epoch": 0.45556, + "grad_norm": 0.7011657820782939, + "learning_rate": 0.003, + "loss": 3.983, + "step": 45556 + }, + { + "epoch": 0.45557, + "grad_norm": 0.6148885067459354, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 45557 + }, + { + "epoch": 0.45558, + "grad_norm": 0.6734059629693977, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 45558 + }, + { + "epoch": 0.45559, + "grad_norm": 0.682904000526697, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 45559 + }, + { + "epoch": 0.4556, + "grad_norm": 0.7136668924296632, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 45560 + }, + { + "epoch": 0.45561, + "grad_norm": 0.7925307106666153, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45561 + }, + { + "epoch": 0.45562, + "grad_norm": 0.793966236884669, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 45562 + }, + { + "epoch": 0.45563, + "grad_norm": 0.8038709910313695, + "learning_rate": 0.003, + "loss": 3.983, + "step": 45563 + }, + { + "epoch": 0.45564, + "grad_norm": 0.7058947260003671, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 45564 + }, + { + "epoch": 0.45565, + "grad_norm": 0.6801206576165771, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 45565 + }, + { + "epoch": 0.45566, + "grad_norm": 0.7139838497867845, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 45566 + }, + { + "epoch": 0.45567, + "grad_norm": 0.6488689357933746, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 45567 + }, + { + "epoch": 0.45568, + "grad_norm": 0.5846234825814421, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 45568 + }, + { + "epoch": 0.45569, + "grad_norm": 0.6966262067444033, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 45569 + }, + { + "epoch": 0.4557, + "grad_norm": 0.7092226456873107, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 45570 + }, + { + "epoch": 0.45571, + "grad_norm": 0.675252898553252, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 45571 + }, + { + "epoch": 0.45572, + "grad_norm": 0.6136675138860493, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 45572 + }, + { + "epoch": 0.45573, + "grad_norm": 0.6512096400242554, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 45573 + }, + { + "epoch": 0.45574, + "grad_norm": 0.868182455308154, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 45574 + }, + { + "epoch": 0.45575, + "grad_norm": 0.9920599053095914, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 45575 + }, + { + "epoch": 0.45576, + "grad_norm": 1.1256710330974338, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 45576 + }, + { + "epoch": 0.45577, + "grad_norm": 0.712935417169233, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 45577 + }, + { + "epoch": 0.45578, + "grad_norm": 0.6550194356266597, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 45578 + }, + { + "epoch": 0.45579, + "grad_norm": 0.7455116068661438, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 45579 + }, + { + "epoch": 0.4558, + "grad_norm": 0.8103274814045874, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 45580 + }, + { + "epoch": 0.45581, + "grad_norm": 0.9599058983458455, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 45581 + }, + { + "epoch": 0.45582, + "grad_norm": 1.0448414682594465, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 45582 + }, + { + "epoch": 0.45583, + "grad_norm": 0.8681894224426633, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 45583 + }, + { + "epoch": 0.45584, + "grad_norm": 0.7271135897317385, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 45584 + }, + { + "epoch": 0.45585, + "grad_norm": 0.6840466241024591, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 45585 + }, + { + "epoch": 0.45586, + "grad_norm": 0.7412076203809608, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 45586 + }, + { + "epoch": 0.45587, + "grad_norm": 0.8734467340969576, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 45587 + }, + { + "epoch": 0.45588, + "grad_norm": 1.0021840317813968, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 45588 + }, + { + "epoch": 0.45589, + "grad_norm": 1.016290371508774, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 45589 + }, + { + "epoch": 0.4559, + "grad_norm": 0.9331283268855902, + "learning_rate": 0.003, + "loss": 3.956, + "step": 45590 + }, + { + "epoch": 0.45591, + "grad_norm": 1.0287679195575496, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 45591 + }, + { + "epoch": 0.45592, + "grad_norm": 1.0517225626067073, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 45592 + }, + { + "epoch": 0.45593, + "grad_norm": 0.9714182864295821, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 45593 + }, + { + "epoch": 0.45594, + "grad_norm": 1.0533223124266387, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 45594 + }, + { + "epoch": 0.45595, + "grad_norm": 0.9942148368532145, + "learning_rate": 0.003, + "loss": 4.0294, + "step": 45595 + }, + { + "epoch": 0.45596, + "grad_norm": 1.1345649871092147, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 45596 + }, + { + "epoch": 0.45597, + "grad_norm": 0.8106234499518922, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 45597 + }, + { + "epoch": 0.45598, + "grad_norm": 0.8298775576248788, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 45598 + }, + { + "epoch": 0.45599, + "grad_norm": 0.8494442753828984, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 45599 + }, + { + "epoch": 0.456, + "grad_norm": 0.9137966343677753, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 45600 + }, + { + "epoch": 0.45601, + "grad_norm": 0.9974772300196978, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 45601 + }, + { + "epoch": 0.45602, + "grad_norm": 0.7688277544774433, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 45602 + }, + { + "epoch": 0.45603, + "grad_norm": 0.6992408273216382, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 45603 + }, + { + "epoch": 0.45604, + "grad_norm": 0.6392823025117437, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 45604 + }, + { + "epoch": 0.45605, + "grad_norm": 0.673334200649027, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 45605 + }, + { + "epoch": 0.45606, + "grad_norm": 0.6004831768185338, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 45606 + }, + { + "epoch": 0.45607, + "grad_norm": 0.6328219962104127, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 45607 + }, + { + "epoch": 0.45608, + "grad_norm": 0.6307102632716421, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 45608 + }, + { + "epoch": 0.45609, + "grad_norm": 0.5995214621886394, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 45609 + }, + { + "epoch": 0.4561, + "grad_norm": 0.6062074860898554, + "learning_rate": 0.003, + "loss": 3.9426, + "step": 45610 + }, + { + "epoch": 0.45611, + "grad_norm": 0.5495739827423586, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 45611 + }, + { + "epoch": 0.45612, + "grad_norm": 0.6156699649021932, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 45612 + }, + { + "epoch": 0.45613, + "grad_norm": 0.6162397678170497, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 45613 + }, + { + "epoch": 0.45614, + "grad_norm": 0.5300159541390091, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 45614 + }, + { + "epoch": 0.45615, + "grad_norm": 0.6069958285407404, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 45615 + }, + { + "epoch": 0.45616, + "grad_norm": 0.7569876382752376, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 45616 + }, + { + "epoch": 0.45617, + "grad_norm": 0.9325969115047812, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 45617 + }, + { + "epoch": 0.45618, + "grad_norm": 1.1334499012786017, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 45618 + }, + { + "epoch": 0.45619, + "grad_norm": 0.8450207967193522, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 45619 + }, + { + "epoch": 0.4562, + "grad_norm": 0.7727178416706865, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 45620 + }, + { + "epoch": 0.45621, + "grad_norm": 0.743771316241705, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 45621 + }, + { + "epoch": 0.45622, + "grad_norm": 0.7466046259112569, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 45622 + }, + { + "epoch": 0.45623, + "grad_norm": 0.8104093684916878, + "learning_rate": 0.003, + "loss": 3.9386, + "step": 45623 + }, + { + "epoch": 0.45624, + "grad_norm": 0.8580212000192161, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 45624 + }, + { + "epoch": 0.45625, + "grad_norm": 0.8985817542488218, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 45625 + }, + { + "epoch": 0.45626, + "grad_norm": 0.9551366161295227, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 45626 + }, + { + "epoch": 0.45627, + "grad_norm": 0.8503354410371803, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 45627 + }, + { + "epoch": 0.45628, + "grad_norm": 0.6198292334067557, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 45628 + }, + { + "epoch": 0.45629, + "grad_norm": 0.6263111579650588, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 45629 + }, + { + "epoch": 0.4563, + "grad_norm": 0.6852287920663076, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 45630 + }, + { + "epoch": 0.45631, + "grad_norm": 0.6272925130353005, + "learning_rate": 0.003, + "loss": 3.998, + "step": 45631 + }, + { + "epoch": 0.45632, + "grad_norm": 0.6437444279629397, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 45632 + }, + { + "epoch": 0.45633, + "grad_norm": 0.6077922476949752, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 45633 + }, + { + "epoch": 0.45634, + "grad_norm": 0.5430034061998632, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 45634 + }, + { + "epoch": 0.45635, + "grad_norm": 0.5219076278959098, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 45635 + }, + { + "epoch": 0.45636, + "grad_norm": 0.5504242679681124, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 45636 + }, + { + "epoch": 0.45637, + "grad_norm": 0.5567071435515073, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 45637 + }, + { + "epoch": 0.45638, + "grad_norm": 0.6062342112513596, + "learning_rate": 0.003, + "loss": 3.961, + "step": 45638 + }, + { + "epoch": 0.45639, + "grad_norm": 0.6739435681972836, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 45639 + }, + { + "epoch": 0.4564, + "grad_norm": 0.696790919026797, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 45640 + }, + { + "epoch": 0.45641, + "grad_norm": 0.8183554182292611, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 45641 + }, + { + "epoch": 0.45642, + "grad_norm": 0.7666502830772114, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 45642 + }, + { + "epoch": 0.45643, + "grad_norm": 0.8528100763599464, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 45643 + }, + { + "epoch": 0.45644, + "grad_norm": 0.8100241916457832, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 45644 + }, + { + "epoch": 0.45645, + "grad_norm": 0.9072376998740908, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 45645 + }, + { + "epoch": 0.45646, + "grad_norm": 0.8861350303322013, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 45646 + }, + { + "epoch": 0.45647, + "grad_norm": 1.0437999771439423, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 45647 + }, + { + "epoch": 0.45648, + "grad_norm": 0.9826015745630009, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 45648 + }, + { + "epoch": 0.45649, + "grad_norm": 0.9381670521960334, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 45649 + }, + { + "epoch": 0.4565, + "grad_norm": 0.9554472319224372, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 45650 + }, + { + "epoch": 0.45651, + "grad_norm": 1.056286713457018, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 45651 + }, + { + "epoch": 0.45652, + "grad_norm": 0.999106419760802, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 45652 + }, + { + "epoch": 0.45653, + "grad_norm": 1.1032980016645528, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 45653 + }, + { + "epoch": 0.45654, + "grad_norm": 0.9786126754891192, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 45654 + }, + { + "epoch": 0.45655, + "grad_norm": 1.1040261504444704, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 45655 + }, + { + "epoch": 0.45656, + "grad_norm": 0.8317369983659528, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 45656 + }, + { + "epoch": 0.45657, + "grad_norm": 0.7101916289998461, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 45657 + }, + { + "epoch": 0.45658, + "grad_norm": 0.7116014903252393, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 45658 + }, + { + "epoch": 0.45659, + "grad_norm": 0.6504174255289084, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 45659 + }, + { + "epoch": 0.4566, + "grad_norm": 0.5723642159689043, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 45660 + }, + { + "epoch": 0.45661, + "grad_norm": 0.5229643536062787, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 45661 + }, + { + "epoch": 0.45662, + "grad_norm": 0.48086158338889207, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 45662 + }, + { + "epoch": 0.45663, + "grad_norm": 0.47363342506713907, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 45663 + }, + { + "epoch": 0.45664, + "grad_norm": 0.5272739524575845, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 45664 + }, + { + "epoch": 0.45665, + "grad_norm": 0.6147533453216922, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 45665 + }, + { + "epoch": 0.45666, + "grad_norm": 0.7158299995160774, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 45666 + }, + { + "epoch": 0.45667, + "grad_norm": 0.8979686390763398, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 45667 + }, + { + "epoch": 0.45668, + "grad_norm": 0.8469949210904583, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 45668 + }, + { + "epoch": 0.45669, + "grad_norm": 0.8074693185686732, + "learning_rate": 0.003, + "loss": 3.985, + "step": 45669 + }, + { + "epoch": 0.4567, + "grad_norm": 0.8091543898553957, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 45670 + }, + { + "epoch": 0.45671, + "grad_norm": 0.7744186848618614, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 45671 + }, + { + "epoch": 0.45672, + "grad_norm": 0.814209805898811, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 45672 + }, + { + "epoch": 0.45673, + "grad_norm": 0.8628120168254901, + "learning_rate": 0.003, + "loss": 3.964, + "step": 45673 + }, + { + "epoch": 0.45674, + "grad_norm": 0.9528181214734293, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 45674 + }, + { + "epoch": 0.45675, + "grad_norm": 0.9665878012783699, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 45675 + }, + { + "epoch": 0.45676, + "grad_norm": 1.1143220718010147, + "learning_rate": 0.003, + "loss": 3.995, + "step": 45676 + }, + { + "epoch": 0.45677, + "grad_norm": 0.9071239880323297, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 45677 + }, + { + "epoch": 0.45678, + "grad_norm": 0.8863761327611743, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 45678 + }, + { + "epoch": 0.45679, + "grad_norm": 0.8022022862620004, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 45679 + }, + { + "epoch": 0.4568, + "grad_norm": 0.7783605931688978, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 45680 + }, + { + "epoch": 0.45681, + "grad_norm": 0.6473350021835713, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 45681 + }, + { + "epoch": 0.45682, + "grad_norm": 0.7016262644739695, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 45682 + }, + { + "epoch": 0.45683, + "grad_norm": 0.6857151413044349, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 45683 + }, + { + "epoch": 0.45684, + "grad_norm": 0.8263308374618812, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 45684 + }, + { + "epoch": 0.45685, + "grad_norm": 1.041840509921614, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 45685 + }, + { + "epoch": 0.45686, + "grad_norm": 0.934283356695278, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 45686 + }, + { + "epoch": 0.45687, + "grad_norm": 0.970559637094736, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 45687 + }, + { + "epoch": 0.45688, + "grad_norm": 1.0521291878898988, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 45688 + }, + { + "epoch": 0.45689, + "grad_norm": 0.9451524880023253, + "learning_rate": 0.003, + "loss": 4.011, + "step": 45689 + }, + { + "epoch": 0.4569, + "grad_norm": 0.8197598696989136, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 45690 + }, + { + "epoch": 0.45691, + "grad_norm": 0.715358545187743, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 45691 + }, + { + "epoch": 0.45692, + "grad_norm": 0.7028258162137898, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 45692 + }, + { + "epoch": 0.45693, + "grad_norm": 0.7891318275142851, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 45693 + }, + { + "epoch": 0.45694, + "grad_norm": 0.6777110689404189, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 45694 + }, + { + "epoch": 0.45695, + "grad_norm": 0.5237442776878143, + "learning_rate": 0.003, + "loss": 3.985, + "step": 45695 + }, + { + "epoch": 0.45696, + "grad_norm": 0.5121859168135737, + "learning_rate": 0.003, + "loss": 3.952, + "step": 45696 + }, + { + "epoch": 0.45697, + "grad_norm": 0.5017265838821792, + "learning_rate": 0.003, + "loss": 3.9465, + "step": 45697 + }, + { + "epoch": 0.45698, + "grad_norm": 0.5364682203411074, + "learning_rate": 0.003, + "loss": 3.9208, + "step": 45698 + }, + { + "epoch": 0.45699, + "grad_norm": 0.5253534988735356, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 45699 + }, + { + "epoch": 0.457, + "grad_norm": 0.5565120784202015, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 45700 + }, + { + "epoch": 0.45701, + "grad_norm": 0.5451934845598218, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 45701 + }, + { + "epoch": 0.45702, + "grad_norm": 0.5963280020801823, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 45702 + }, + { + "epoch": 0.45703, + "grad_norm": 0.7340920555787113, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 45703 + }, + { + "epoch": 0.45704, + "grad_norm": 0.7988103260756452, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 45704 + }, + { + "epoch": 0.45705, + "grad_norm": 0.885226422670474, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 45705 + }, + { + "epoch": 0.45706, + "grad_norm": 0.9252239283301738, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 45706 + }, + { + "epoch": 0.45707, + "grad_norm": 0.8183972182823731, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 45707 + }, + { + "epoch": 0.45708, + "grad_norm": 0.806023749577273, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 45708 + }, + { + "epoch": 0.45709, + "grad_norm": 0.8446415341918426, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 45709 + }, + { + "epoch": 0.4571, + "grad_norm": 0.8851454182874815, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 45710 + }, + { + "epoch": 0.45711, + "grad_norm": 0.9748625992276146, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 45711 + }, + { + "epoch": 0.45712, + "grad_norm": 0.9692772295534702, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 45712 + }, + { + "epoch": 0.45713, + "grad_norm": 0.999659291451751, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 45713 + }, + { + "epoch": 0.45714, + "grad_norm": 1.153860171024719, + "learning_rate": 0.003, + "loss": 3.986, + "step": 45714 + }, + { + "epoch": 0.45715, + "grad_norm": 0.8400015116238666, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 45715 + }, + { + "epoch": 0.45716, + "grad_norm": 0.8062840314024727, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 45716 + }, + { + "epoch": 0.45717, + "grad_norm": 0.7705627704172701, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 45717 + }, + { + "epoch": 0.45718, + "grad_norm": 0.6913247530583061, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 45718 + }, + { + "epoch": 0.45719, + "grad_norm": 0.6401571242826288, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 45719 + }, + { + "epoch": 0.4572, + "grad_norm": 0.68207146300672, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 45720 + }, + { + "epoch": 0.45721, + "grad_norm": 0.8105779179333442, + "learning_rate": 0.003, + "loss": 3.965, + "step": 45721 + }, + { + "epoch": 0.45722, + "grad_norm": 0.93728587072804, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 45722 + }, + { + "epoch": 0.45723, + "grad_norm": 1.0421306603791882, + "learning_rate": 0.003, + "loss": 4.0237, + "step": 45723 + }, + { + "epoch": 0.45724, + "grad_norm": 0.8764198259817266, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 45724 + }, + { + "epoch": 0.45725, + "grad_norm": 0.713109292419786, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 45725 + }, + { + "epoch": 0.45726, + "grad_norm": 0.5804562834231327, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 45726 + }, + { + "epoch": 0.45727, + "grad_norm": 0.5965442018428443, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 45727 + }, + { + "epoch": 0.45728, + "grad_norm": 0.6132554631170459, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 45728 + }, + { + "epoch": 0.45729, + "grad_norm": 0.6346741283971237, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 45729 + }, + { + "epoch": 0.4573, + "grad_norm": 0.6700254466264967, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 45730 + }, + { + "epoch": 0.45731, + "grad_norm": 0.6505071158369381, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 45731 + }, + { + "epoch": 0.45732, + "grad_norm": 0.6151065304725196, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 45732 + }, + { + "epoch": 0.45733, + "grad_norm": 0.5512409242001275, + "learning_rate": 0.003, + "loss": 3.985, + "step": 45733 + }, + { + "epoch": 0.45734, + "grad_norm": 0.5729012715946286, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 45734 + }, + { + "epoch": 0.45735, + "grad_norm": 0.6735645765539628, + "learning_rate": 0.003, + "loss": 4.005, + "step": 45735 + }, + { + "epoch": 0.45736, + "grad_norm": 0.8123138634066195, + "learning_rate": 0.003, + "loss": 3.972, + "step": 45736 + }, + { + "epoch": 0.45737, + "grad_norm": 0.9202870740319876, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 45737 + }, + { + "epoch": 0.45738, + "grad_norm": 1.0329322477972767, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 45738 + }, + { + "epoch": 0.45739, + "grad_norm": 1.051365767030947, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 45739 + }, + { + "epoch": 0.4574, + "grad_norm": 0.9372561894500323, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 45740 + }, + { + "epoch": 0.45741, + "grad_norm": 1.0015601904069023, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 45741 + }, + { + "epoch": 0.45742, + "grad_norm": 1.063675107382115, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 45742 + }, + { + "epoch": 0.45743, + "grad_norm": 0.8436096424838971, + "learning_rate": 0.003, + "loss": 3.975, + "step": 45743 + }, + { + "epoch": 0.45744, + "grad_norm": 0.7623817134870421, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 45744 + }, + { + "epoch": 0.45745, + "grad_norm": 0.8594438148582373, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 45745 + }, + { + "epoch": 0.45746, + "grad_norm": 1.0309298988115176, + "learning_rate": 0.003, + "loss": 3.985, + "step": 45746 + }, + { + "epoch": 0.45747, + "grad_norm": 0.9280470154898603, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 45747 + }, + { + "epoch": 0.45748, + "grad_norm": 0.7937257182716413, + "learning_rate": 0.003, + "loss": 3.994, + "step": 45748 + }, + { + "epoch": 0.45749, + "grad_norm": 0.7331558458942304, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 45749 + }, + { + "epoch": 0.4575, + "grad_norm": 0.7033171975734619, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 45750 + }, + { + "epoch": 0.45751, + "grad_norm": 0.6926444869936295, + "learning_rate": 0.003, + "loss": 3.998, + "step": 45751 + }, + { + "epoch": 0.45752, + "grad_norm": 0.6413156676557258, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 45752 + }, + { + "epoch": 0.45753, + "grad_norm": 0.6168099775106948, + "learning_rate": 0.003, + "loss": 3.98, + "step": 45753 + }, + { + "epoch": 0.45754, + "grad_norm": 0.5832025174647519, + "learning_rate": 0.003, + "loss": 3.966, + "step": 45754 + }, + { + "epoch": 0.45755, + "grad_norm": 0.6395737796900409, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 45755 + }, + { + "epoch": 0.45756, + "grad_norm": 0.7037332319953713, + "learning_rate": 0.003, + "loss": 3.989, + "step": 45756 + }, + { + "epoch": 0.45757, + "grad_norm": 0.6921068818640714, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 45757 + }, + { + "epoch": 0.45758, + "grad_norm": 0.5545783552571333, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 45758 + }, + { + "epoch": 0.45759, + "grad_norm": 0.6033971546913773, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 45759 + }, + { + "epoch": 0.4576, + "grad_norm": 0.5994678832639725, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 45760 + }, + { + "epoch": 0.45761, + "grad_norm": 0.571903229788691, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 45761 + }, + { + "epoch": 0.45762, + "grad_norm": 0.6074384730059643, + "learning_rate": 0.003, + "loss": 3.958, + "step": 45762 + }, + { + "epoch": 0.45763, + "grad_norm": 0.7175069825263919, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 45763 + }, + { + "epoch": 0.45764, + "grad_norm": 0.7670980707882391, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 45764 + }, + { + "epoch": 0.45765, + "grad_norm": 0.7463628951767689, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 45765 + }, + { + "epoch": 0.45766, + "grad_norm": 0.6541104527240139, + "learning_rate": 0.003, + "loss": 3.971, + "step": 45766 + }, + { + "epoch": 0.45767, + "grad_norm": 0.6872331203084575, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 45767 + }, + { + "epoch": 0.45768, + "grad_norm": 0.6269640407834098, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 45768 + }, + { + "epoch": 0.45769, + "grad_norm": 0.5954490731200482, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 45769 + }, + { + "epoch": 0.4577, + "grad_norm": 0.774441930893577, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 45770 + }, + { + "epoch": 0.45771, + "grad_norm": 0.8891169885288305, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 45771 + }, + { + "epoch": 0.45772, + "grad_norm": 0.9057662304370047, + "learning_rate": 0.003, + "loss": 3.994, + "step": 45772 + }, + { + "epoch": 0.45773, + "grad_norm": 1.043548239782826, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 45773 + }, + { + "epoch": 0.45774, + "grad_norm": 1.1819937016822248, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 45774 + }, + { + "epoch": 0.45775, + "grad_norm": 0.8825144663382547, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 45775 + }, + { + "epoch": 0.45776, + "grad_norm": 0.8450264619257709, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 45776 + }, + { + "epoch": 0.45777, + "grad_norm": 0.7705734230983637, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 45777 + }, + { + "epoch": 0.45778, + "grad_norm": 0.6926199615014754, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 45778 + }, + { + "epoch": 0.45779, + "grad_norm": 0.6931960341192454, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 45779 + }, + { + "epoch": 0.4578, + "grad_norm": 0.77030849180695, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 45780 + }, + { + "epoch": 0.45781, + "grad_norm": 0.8974213115071699, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 45781 + }, + { + "epoch": 0.45782, + "grad_norm": 1.0228036182010936, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 45782 + }, + { + "epoch": 0.45783, + "grad_norm": 0.9488728889735519, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 45783 + }, + { + "epoch": 0.45784, + "grad_norm": 0.824934385493851, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 45784 + }, + { + "epoch": 0.45785, + "grad_norm": 0.7768751329130706, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 45785 + }, + { + "epoch": 0.45786, + "grad_norm": 0.9144839401304478, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 45786 + }, + { + "epoch": 0.45787, + "grad_norm": 0.9212796816598507, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 45787 + }, + { + "epoch": 0.45788, + "grad_norm": 0.9118153802738785, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 45788 + }, + { + "epoch": 0.45789, + "grad_norm": 0.9166764875261599, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 45789 + }, + { + "epoch": 0.4579, + "grad_norm": 0.8314688632180776, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 45790 + }, + { + "epoch": 0.45791, + "grad_norm": 0.9237786430173612, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 45791 + }, + { + "epoch": 0.45792, + "grad_norm": 0.8249324836603644, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 45792 + }, + { + "epoch": 0.45793, + "grad_norm": 0.8258205332529419, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 45793 + }, + { + "epoch": 0.45794, + "grad_norm": 0.7297577516002661, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 45794 + }, + { + "epoch": 0.45795, + "grad_norm": 0.7018579325643166, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 45795 + }, + { + "epoch": 0.45796, + "grad_norm": 0.6438970923942916, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 45796 + }, + { + "epoch": 0.45797, + "grad_norm": 0.5969100334695672, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 45797 + }, + { + "epoch": 0.45798, + "grad_norm": 0.6569747900761541, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 45798 + }, + { + "epoch": 0.45799, + "grad_norm": 0.7619237821827067, + "learning_rate": 0.003, + "loss": 3.987, + "step": 45799 + }, + { + "epoch": 0.458, + "grad_norm": 0.9176996012267716, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 45800 + }, + { + "epoch": 0.45801, + "grad_norm": 0.8697620249700254, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 45801 + }, + { + "epoch": 0.45802, + "grad_norm": 0.7471087461317211, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 45802 + }, + { + "epoch": 0.45803, + "grad_norm": 0.7146216982426684, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 45803 + }, + { + "epoch": 0.45804, + "grad_norm": 0.7042577621975549, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 45804 + }, + { + "epoch": 0.45805, + "grad_norm": 0.7080291804919574, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 45805 + }, + { + "epoch": 0.45806, + "grad_norm": 0.7102484934130313, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 45806 + }, + { + "epoch": 0.45807, + "grad_norm": 0.8183132158301788, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 45807 + }, + { + "epoch": 0.45808, + "grad_norm": 1.002275149585218, + "learning_rate": 0.003, + "loss": 3.994, + "step": 45808 + }, + { + "epoch": 0.45809, + "grad_norm": 1.0194910499574221, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 45809 + }, + { + "epoch": 0.4581, + "grad_norm": 0.9047332857732249, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 45810 + }, + { + "epoch": 0.45811, + "grad_norm": 0.8501567817376345, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 45811 + }, + { + "epoch": 0.45812, + "grad_norm": 0.8052245585708121, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 45812 + }, + { + "epoch": 0.45813, + "grad_norm": 0.8685629649235613, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 45813 + }, + { + "epoch": 0.45814, + "grad_norm": 0.8517836058666015, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 45814 + }, + { + "epoch": 0.45815, + "grad_norm": 0.7086433332677913, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 45815 + }, + { + "epoch": 0.45816, + "grad_norm": 0.6195434837045773, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 45816 + }, + { + "epoch": 0.45817, + "grad_norm": 0.6622252857079914, + "learning_rate": 0.003, + "loss": 4.02, + "step": 45817 + }, + { + "epoch": 0.45818, + "grad_norm": 0.7084746195375453, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 45818 + }, + { + "epoch": 0.45819, + "grad_norm": 0.6264848239885143, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 45819 + }, + { + "epoch": 0.4582, + "grad_norm": 0.5904309995756575, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 45820 + }, + { + "epoch": 0.45821, + "grad_norm": 0.5838172332815187, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 45821 + }, + { + "epoch": 0.45822, + "grad_norm": 0.6453520475432133, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 45822 + }, + { + "epoch": 0.45823, + "grad_norm": 0.682078847308575, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 45823 + }, + { + "epoch": 0.45824, + "grad_norm": 0.674881282413307, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 45824 + }, + { + "epoch": 0.45825, + "grad_norm": 0.6712638407839026, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 45825 + }, + { + "epoch": 0.45826, + "grad_norm": 0.6507943860158157, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 45826 + }, + { + "epoch": 0.45827, + "grad_norm": 0.614023403316421, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 45827 + }, + { + "epoch": 0.45828, + "grad_norm": 0.5894527466537016, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45828 + }, + { + "epoch": 0.45829, + "grad_norm": 0.654837637980983, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 45829 + }, + { + "epoch": 0.4583, + "grad_norm": 0.6555215020704841, + "learning_rate": 0.003, + "loss": 3.9413, + "step": 45830 + }, + { + "epoch": 0.45831, + "grad_norm": 0.6126411024479983, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 45831 + }, + { + "epoch": 0.45832, + "grad_norm": 0.7249262426793532, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 45832 + }, + { + "epoch": 0.45833, + "grad_norm": 0.8021022280303687, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 45833 + }, + { + "epoch": 0.45834, + "grad_norm": 0.8156407822692996, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 45834 + }, + { + "epoch": 0.45835, + "grad_norm": 0.9670780121678768, + "learning_rate": 0.003, + "loss": 3.9432, + "step": 45835 + }, + { + "epoch": 0.45836, + "grad_norm": 1.163920056723688, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 45836 + }, + { + "epoch": 0.45837, + "grad_norm": 1.0483297317483808, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 45837 + }, + { + "epoch": 0.45838, + "grad_norm": 0.881931762446981, + "learning_rate": 0.003, + "loss": 4.001, + "step": 45838 + }, + { + "epoch": 0.45839, + "grad_norm": 0.8273865010637333, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 45839 + }, + { + "epoch": 0.4584, + "grad_norm": 0.955231472515739, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 45840 + }, + { + "epoch": 0.45841, + "grad_norm": 0.9764321251446725, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 45841 + }, + { + "epoch": 0.45842, + "grad_norm": 1.0423964258787002, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 45842 + }, + { + "epoch": 0.45843, + "grad_norm": 1.1009450724955847, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 45843 + }, + { + "epoch": 0.45844, + "grad_norm": 0.9240391241184909, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 45844 + }, + { + "epoch": 0.45845, + "grad_norm": 0.9281252751281811, + "learning_rate": 0.003, + "loss": 4.0321, + "step": 45845 + }, + { + "epoch": 0.45846, + "grad_norm": 0.9539279992759069, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 45846 + }, + { + "epoch": 0.45847, + "grad_norm": 0.8663734423811819, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 45847 + }, + { + "epoch": 0.45848, + "grad_norm": 0.721926670987644, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 45848 + }, + { + "epoch": 0.45849, + "grad_norm": 0.7319750993179213, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 45849 + }, + { + "epoch": 0.4585, + "grad_norm": 0.7579821447195934, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 45850 + }, + { + "epoch": 0.45851, + "grad_norm": 0.6986833066492012, + "learning_rate": 0.003, + "loss": 3.976, + "step": 45851 + }, + { + "epoch": 0.45852, + "grad_norm": 0.6315811768845188, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 45852 + }, + { + "epoch": 0.45853, + "grad_norm": 0.523507121905459, + "learning_rate": 0.003, + "loss": 3.9409, + "step": 45853 + }, + { + "epoch": 0.45854, + "grad_norm": 0.5079046627127866, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 45854 + }, + { + "epoch": 0.45855, + "grad_norm": 0.5538115025493991, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 45855 + }, + { + "epoch": 0.45856, + "grad_norm": 0.6211510255424588, + "learning_rate": 0.003, + "loss": 3.99, + "step": 45856 + }, + { + "epoch": 0.45857, + "grad_norm": 0.7173991277350621, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 45857 + }, + { + "epoch": 0.45858, + "grad_norm": 0.8183309879737046, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 45858 + }, + { + "epoch": 0.45859, + "grad_norm": 0.8575832143818269, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 45859 + }, + { + "epoch": 0.4586, + "grad_norm": 0.8655088655650656, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 45860 + }, + { + "epoch": 0.45861, + "grad_norm": 0.8526661731154948, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 45861 + }, + { + "epoch": 0.45862, + "grad_norm": 0.8528345133399087, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 45862 + }, + { + "epoch": 0.45863, + "grad_norm": 0.8009735741049023, + "learning_rate": 0.003, + "loss": 3.994, + "step": 45863 + }, + { + "epoch": 0.45864, + "grad_norm": 0.7937037498078957, + "learning_rate": 0.003, + "loss": 3.981, + "step": 45864 + }, + { + "epoch": 0.45865, + "grad_norm": 0.8287110753582085, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 45865 + }, + { + "epoch": 0.45866, + "grad_norm": 0.9522767108726267, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 45866 + }, + { + "epoch": 0.45867, + "grad_norm": 0.872912110271035, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 45867 + }, + { + "epoch": 0.45868, + "grad_norm": 0.844443152876898, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 45868 + }, + { + "epoch": 0.45869, + "grad_norm": 0.8082232666202104, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 45869 + }, + { + "epoch": 0.4587, + "grad_norm": 0.7084271927193077, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 45870 + }, + { + "epoch": 0.45871, + "grad_norm": 0.7199901078082452, + "learning_rate": 0.003, + "loss": 3.962, + "step": 45871 + }, + { + "epoch": 0.45872, + "grad_norm": 0.7331127983475109, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 45872 + }, + { + "epoch": 0.45873, + "grad_norm": 0.7276185960730007, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 45873 + }, + { + "epoch": 0.45874, + "grad_norm": 0.7130035453416493, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 45874 + }, + { + "epoch": 0.45875, + "grad_norm": 0.6867207715652746, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 45875 + }, + { + "epoch": 0.45876, + "grad_norm": 0.7703738918400634, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 45876 + }, + { + "epoch": 0.45877, + "grad_norm": 0.9128114994476844, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 45877 + }, + { + "epoch": 0.45878, + "grad_norm": 0.9151183590144825, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 45878 + }, + { + "epoch": 0.45879, + "grad_norm": 0.9342049382866018, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 45879 + }, + { + "epoch": 0.4588, + "grad_norm": 0.9910920516138355, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 45880 + }, + { + "epoch": 0.45881, + "grad_norm": 1.0243001194381793, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 45881 + }, + { + "epoch": 0.45882, + "grad_norm": 1.0684505009183658, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 45882 + }, + { + "epoch": 0.45883, + "grad_norm": 0.9698655695662842, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 45883 + }, + { + "epoch": 0.45884, + "grad_norm": 0.9802154396157289, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 45884 + }, + { + "epoch": 0.45885, + "grad_norm": 1.086509080613651, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 45885 + }, + { + "epoch": 0.45886, + "grad_norm": 0.8825920925570511, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 45886 + }, + { + "epoch": 0.45887, + "grad_norm": 0.9636488821308378, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 45887 + }, + { + "epoch": 0.45888, + "grad_norm": 1.0545093424352876, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 45888 + }, + { + "epoch": 0.45889, + "grad_norm": 0.8955683539900625, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 45889 + }, + { + "epoch": 0.4589, + "grad_norm": 0.9748349859362654, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 45890 + }, + { + "epoch": 0.45891, + "grad_norm": 1.0329553316423228, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 45891 + }, + { + "epoch": 0.45892, + "grad_norm": 0.837940708183608, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 45892 + }, + { + "epoch": 0.45893, + "grad_norm": 0.7127481658618146, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 45893 + }, + { + "epoch": 0.45894, + "grad_norm": 0.6237095771616661, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 45894 + }, + { + "epoch": 0.45895, + "grad_norm": 0.5865867263956949, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 45895 + }, + { + "epoch": 0.45896, + "grad_norm": 0.5951859079881158, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 45896 + }, + { + "epoch": 0.45897, + "grad_norm": 0.5173683846715972, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 45897 + }, + { + "epoch": 0.45898, + "grad_norm": 0.5023468866660654, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 45898 + }, + { + "epoch": 0.45899, + "grad_norm": 0.4846041407213096, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 45899 + }, + { + "epoch": 0.459, + "grad_norm": 0.4927437885904906, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 45900 + }, + { + "epoch": 0.45901, + "grad_norm": 0.5145781219417481, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 45901 + }, + { + "epoch": 0.45902, + "grad_norm": 0.6064354538436008, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 45902 + }, + { + "epoch": 0.45903, + "grad_norm": 0.7519799134833279, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 45903 + }, + { + "epoch": 0.45904, + "grad_norm": 0.7730065981437327, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 45904 + }, + { + "epoch": 0.45905, + "grad_norm": 0.7005027388216094, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 45905 + }, + { + "epoch": 0.45906, + "grad_norm": 0.5518657163396068, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 45906 + }, + { + "epoch": 0.45907, + "grad_norm": 0.5445593733978049, + "learning_rate": 0.003, + "loss": 3.974, + "step": 45907 + }, + { + "epoch": 0.45908, + "grad_norm": 0.5623465638830137, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 45908 + }, + { + "epoch": 0.45909, + "grad_norm": 0.5983517618710721, + "learning_rate": 0.003, + "loss": 3.9195, + "step": 45909 + }, + { + "epoch": 0.4591, + "grad_norm": 0.639111256984333, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 45910 + }, + { + "epoch": 0.45911, + "grad_norm": 0.6641897519795208, + "learning_rate": 0.003, + "loss": 3.9412, + "step": 45911 + }, + { + "epoch": 0.45912, + "grad_norm": 0.6841897734418015, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 45912 + }, + { + "epoch": 0.45913, + "grad_norm": 0.6145335855331135, + "learning_rate": 0.003, + "loss": 3.9336, + "step": 45913 + }, + { + "epoch": 0.45914, + "grad_norm": 0.6222841107175213, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 45914 + }, + { + "epoch": 0.45915, + "grad_norm": 0.7319003342406967, + "learning_rate": 0.003, + "loss": 3.9427, + "step": 45915 + }, + { + "epoch": 0.45916, + "grad_norm": 0.8115756972002274, + "learning_rate": 0.003, + "loss": 3.972, + "step": 45916 + }, + { + "epoch": 0.45917, + "grad_norm": 0.8529820387514025, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 45917 + }, + { + "epoch": 0.45918, + "grad_norm": 0.9141062470529091, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 45918 + }, + { + "epoch": 0.45919, + "grad_norm": 0.9098808704667959, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 45919 + }, + { + "epoch": 0.4592, + "grad_norm": 0.8862199775994878, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 45920 + }, + { + "epoch": 0.45921, + "grad_norm": 0.7358173381594926, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 45921 + }, + { + "epoch": 0.45922, + "grad_norm": 0.67864261831632, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 45922 + }, + { + "epoch": 0.45923, + "grad_norm": 0.7032330310480808, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 45923 + }, + { + "epoch": 0.45924, + "grad_norm": 0.7326312295779683, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 45924 + }, + { + "epoch": 0.45925, + "grad_norm": 0.815134695620021, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 45925 + }, + { + "epoch": 0.45926, + "grad_norm": 0.8480978214300064, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 45926 + }, + { + "epoch": 0.45927, + "grad_norm": 0.8681794735682964, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 45927 + }, + { + "epoch": 0.45928, + "grad_norm": 0.9989692998042118, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 45928 + }, + { + "epoch": 0.45929, + "grad_norm": 0.9852089085928891, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 45929 + }, + { + "epoch": 0.4593, + "grad_norm": 1.0540625428019892, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 45930 + }, + { + "epoch": 0.45931, + "grad_norm": 0.9821868582117134, + "learning_rate": 0.003, + "loss": 4.014, + "step": 45931 + }, + { + "epoch": 0.45932, + "grad_norm": 0.9316066459295945, + "learning_rate": 0.003, + "loss": 3.988, + "step": 45932 + }, + { + "epoch": 0.45933, + "grad_norm": 0.7894300592585765, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 45933 + }, + { + "epoch": 0.45934, + "grad_norm": 0.9256868122109813, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 45934 + }, + { + "epoch": 0.45935, + "grad_norm": 0.9827669544632331, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 45935 + }, + { + "epoch": 0.45936, + "grad_norm": 0.9722550703558717, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 45936 + }, + { + "epoch": 0.45937, + "grad_norm": 0.7533034392084588, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 45937 + }, + { + "epoch": 0.45938, + "grad_norm": 0.6840067079859504, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 45938 + }, + { + "epoch": 0.45939, + "grad_norm": 0.6239530033221399, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 45939 + }, + { + "epoch": 0.4594, + "grad_norm": 0.6431481195171072, + "learning_rate": 0.003, + "loss": 3.971, + "step": 45940 + }, + { + "epoch": 0.45941, + "grad_norm": 0.6718078999726267, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 45941 + }, + { + "epoch": 0.45942, + "grad_norm": 0.6671374079963648, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 45942 + }, + { + "epoch": 0.45943, + "grad_norm": 0.7268664553316169, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 45943 + }, + { + "epoch": 0.45944, + "grad_norm": 0.7859427955243546, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 45944 + }, + { + "epoch": 0.45945, + "grad_norm": 0.825997432307094, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 45945 + }, + { + "epoch": 0.45946, + "grad_norm": 0.7952239918208981, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 45946 + }, + { + "epoch": 0.45947, + "grad_norm": 0.8282561460114867, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 45947 + }, + { + "epoch": 0.45948, + "grad_norm": 0.8377546959133135, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 45948 + }, + { + "epoch": 0.45949, + "grad_norm": 0.8131627182902379, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 45949 + }, + { + "epoch": 0.4595, + "grad_norm": 0.6975175112669217, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 45950 + }, + { + "epoch": 0.45951, + "grad_norm": 0.7354058965807864, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 45951 + }, + { + "epoch": 0.45952, + "grad_norm": 0.7992446075535626, + "learning_rate": 0.003, + "loss": 3.978, + "step": 45952 + }, + { + "epoch": 0.45953, + "grad_norm": 0.8441747534331915, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 45953 + }, + { + "epoch": 0.45954, + "grad_norm": 0.9367397799301774, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 45954 + }, + { + "epoch": 0.45955, + "grad_norm": 0.8482503864920434, + "learning_rate": 0.003, + "loss": 3.982, + "step": 45955 + }, + { + "epoch": 0.45956, + "grad_norm": 0.7795030290407405, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 45956 + }, + { + "epoch": 0.45957, + "grad_norm": 0.6437184834308076, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 45957 + }, + { + "epoch": 0.45958, + "grad_norm": 0.6819080715499801, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 45958 + }, + { + "epoch": 0.45959, + "grad_norm": 0.673166042445817, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 45959 + }, + { + "epoch": 0.4596, + "grad_norm": 0.7374983780159573, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 45960 + }, + { + "epoch": 0.45961, + "grad_norm": 0.7643474383267415, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 45961 + }, + { + "epoch": 0.45962, + "grad_norm": 0.7728762209316329, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 45962 + }, + { + "epoch": 0.45963, + "grad_norm": 0.8154230998192461, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 45963 + }, + { + "epoch": 0.45964, + "grad_norm": 0.8699756011477405, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 45964 + }, + { + "epoch": 0.45965, + "grad_norm": 0.8108081279058741, + "learning_rate": 0.003, + "loss": 3.984, + "step": 45965 + }, + { + "epoch": 0.45966, + "grad_norm": 0.7301025276506278, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 45966 + }, + { + "epoch": 0.45967, + "grad_norm": 0.7168774882943625, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 45967 + }, + { + "epoch": 0.45968, + "grad_norm": 0.7637641031480135, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 45968 + }, + { + "epoch": 0.45969, + "grad_norm": 0.8747385000536887, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 45969 + }, + { + "epoch": 0.4597, + "grad_norm": 0.8877741166708035, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 45970 + }, + { + "epoch": 0.45971, + "grad_norm": 0.906806688039604, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 45971 + }, + { + "epoch": 0.45972, + "grad_norm": 0.8788181707888396, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 45972 + }, + { + "epoch": 0.45973, + "grad_norm": 0.8692234918616335, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 45973 + }, + { + "epoch": 0.45974, + "grad_norm": 0.9673536797111825, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 45974 + }, + { + "epoch": 0.45975, + "grad_norm": 0.9280511417241692, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 45975 + }, + { + "epoch": 0.45976, + "grad_norm": 0.7005914073348503, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 45976 + }, + { + "epoch": 0.45977, + "grad_norm": 0.670359225736178, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 45977 + }, + { + "epoch": 0.45978, + "grad_norm": 0.652281533385934, + "learning_rate": 0.003, + "loss": 3.976, + "step": 45978 + }, + { + "epoch": 0.45979, + "grad_norm": 0.6590844988557982, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 45979 + }, + { + "epoch": 0.4598, + "grad_norm": 0.7398137745718162, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 45980 + }, + { + "epoch": 0.45981, + "grad_norm": 0.7170254174198463, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 45981 + }, + { + "epoch": 0.45982, + "grad_norm": 0.6991159999637168, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 45982 + }, + { + "epoch": 0.45983, + "grad_norm": 0.7026325597110185, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 45983 + }, + { + "epoch": 0.45984, + "grad_norm": 0.7632630882456943, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 45984 + }, + { + "epoch": 0.45985, + "grad_norm": 0.7592339946185284, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 45985 + }, + { + "epoch": 0.45986, + "grad_norm": 0.6387128313831643, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 45986 + }, + { + "epoch": 0.45987, + "grad_norm": 0.6537458757077941, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 45987 + }, + { + "epoch": 0.45988, + "grad_norm": 0.7434480489739486, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 45988 + }, + { + "epoch": 0.45989, + "grad_norm": 0.7263686180577892, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 45989 + }, + { + "epoch": 0.4599, + "grad_norm": 0.7126769109182134, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 45990 + }, + { + "epoch": 0.45991, + "grad_norm": 0.7080296995399536, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 45991 + }, + { + "epoch": 0.45992, + "grad_norm": 0.6873240274017708, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 45992 + }, + { + "epoch": 0.45993, + "grad_norm": 0.7217096703952773, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 45993 + }, + { + "epoch": 0.45994, + "grad_norm": 0.8600780137170366, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 45994 + }, + { + "epoch": 0.45995, + "grad_norm": 0.9682439400963625, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 45995 + }, + { + "epoch": 0.45996, + "grad_norm": 0.9545583572241373, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 45996 + }, + { + "epoch": 0.45997, + "grad_norm": 0.9372598130450152, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 45997 + }, + { + "epoch": 0.45998, + "grad_norm": 0.9939639316846165, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 45998 + }, + { + "epoch": 0.45999, + "grad_norm": 0.9171461181695216, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 45999 + }, + { + "epoch": 0.46, + "grad_norm": 0.9214990674662328, + "learning_rate": 0.003, + "loss": 3.997, + "step": 46000 + }, + { + "epoch": 0.46001, + "grad_norm": 0.7945693660709936, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46001 + }, + { + "epoch": 0.46002, + "grad_norm": 0.8464697791342256, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 46002 + }, + { + "epoch": 0.46003, + "grad_norm": 0.8835518847883034, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 46003 + }, + { + "epoch": 0.46004, + "grad_norm": 0.9368008740715354, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 46004 + }, + { + "epoch": 0.46005, + "grad_norm": 0.9376566702210118, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 46005 + }, + { + "epoch": 0.46006, + "grad_norm": 1.0022362466926815, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 46006 + }, + { + "epoch": 0.46007, + "grad_norm": 0.9252477484925699, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 46007 + }, + { + "epoch": 0.46008, + "grad_norm": 0.8593056034617549, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 46008 + }, + { + "epoch": 0.46009, + "grad_norm": 0.8483348637891964, + "learning_rate": 0.003, + "loss": 4.0205, + "step": 46009 + }, + { + "epoch": 0.4601, + "grad_norm": 0.9488629418472476, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 46010 + }, + { + "epoch": 0.46011, + "grad_norm": 0.9325805302306269, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 46011 + }, + { + "epoch": 0.46012, + "grad_norm": 1.026237007452606, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 46012 + }, + { + "epoch": 0.46013, + "grad_norm": 0.9842001652287132, + "learning_rate": 0.003, + "loss": 3.983, + "step": 46013 + }, + { + "epoch": 0.46014, + "grad_norm": 0.8175636738321571, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 46014 + }, + { + "epoch": 0.46015, + "grad_norm": 0.719027827451472, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 46015 + }, + { + "epoch": 0.46016, + "grad_norm": 0.5884856492161473, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 46016 + }, + { + "epoch": 0.46017, + "grad_norm": 0.6197381274390735, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 46017 + }, + { + "epoch": 0.46018, + "grad_norm": 0.5024633177335326, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 46018 + }, + { + "epoch": 0.46019, + "grad_norm": 0.5132819970587487, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 46019 + }, + { + "epoch": 0.4602, + "grad_norm": 0.46350687918201816, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 46020 + }, + { + "epoch": 0.46021, + "grad_norm": 0.5323562881145368, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 46021 + }, + { + "epoch": 0.46022, + "grad_norm": 0.5704943850557213, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 46022 + }, + { + "epoch": 0.46023, + "grad_norm": 0.5984404056323166, + "learning_rate": 0.003, + "loss": 3.9346, + "step": 46023 + }, + { + "epoch": 0.46024, + "grad_norm": 0.6301061775961129, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 46024 + }, + { + "epoch": 0.46025, + "grad_norm": 0.6268581856788747, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 46025 + }, + { + "epoch": 0.46026, + "grad_norm": 0.5874639167426927, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 46026 + }, + { + "epoch": 0.46027, + "grad_norm": 0.5411681831829592, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 46027 + }, + { + "epoch": 0.46028, + "grad_norm": 0.5389658179692829, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 46028 + }, + { + "epoch": 0.46029, + "grad_norm": 0.5000080696859871, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 46029 + }, + { + "epoch": 0.4603, + "grad_norm": 0.49780956071390897, + "learning_rate": 0.003, + "loss": 3.947, + "step": 46030 + }, + { + "epoch": 0.46031, + "grad_norm": 0.5064195598186843, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 46031 + }, + { + "epoch": 0.46032, + "grad_norm": 0.6619026886949401, + "learning_rate": 0.003, + "loss": 3.9395, + "step": 46032 + }, + { + "epoch": 0.46033, + "grad_norm": 0.8421161079294226, + "learning_rate": 0.003, + "loss": 3.9311, + "step": 46033 + }, + { + "epoch": 0.46034, + "grad_norm": 1.0285685430519242, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 46034 + }, + { + "epoch": 0.46035, + "grad_norm": 1.0339086612582982, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 46035 + }, + { + "epoch": 0.46036, + "grad_norm": 0.861391934316792, + "learning_rate": 0.003, + "loss": 3.959, + "step": 46036 + }, + { + "epoch": 0.46037, + "grad_norm": 0.8187930628006375, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 46037 + }, + { + "epoch": 0.46038, + "grad_norm": 0.781729035146361, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 46038 + }, + { + "epoch": 0.46039, + "grad_norm": 0.796730975927588, + "learning_rate": 0.003, + "loss": 3.988, + "step": 46039 + }, + { + "epoch": 0.4604, + "grad_norm": 0.6618665858328932, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 46040 + }, + { + "epoch": 0.46041, + "grad_norm": 0.6204910715703043, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 46041 + }, + { + "epoch": 0.46042, + "grad_norm": 0.6559821631847058, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 46042 + }, + { + "epoch": 0.46043, + "grad_norm": 0.6615109021986246, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 46043 + }, + { + "epoch": 0.46044, + "grad_norm": 0.7156353345401439, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 46044 + }, + { + "epoch": 0.46045, + "grad_norm": 0.7804907856122351, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 46045 + }, + { + "epoch": 0.46046, + "grad_norm": 0.8434066137733828, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 46046 + }, + { + "epoch": 0.46047, + "grad_norm": 0.9637577507811026, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 46047 + }, + { + "epoch": 0.46048, + "grad_norm": 1.193753437460881, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 46048 + }, + { + "epoch": 0.46049, + "grad_norm": 1.015525954019732, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 46049 + }, + { + "epoch": 0.4605, + "grad_norm": 1.0329972134130412, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 46050 + }, + { + "epoch": 0.46051, + "grad_norm": 1.1602118680657967, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 46051 + }, + { + "epoch": 0.46052, + "grad_norm": 0.8162655696176662, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 46052 + }, + { + "epoch": 0.46053, + "grad_norm": 0.7632198779119507, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 46053 + }, + { + "epoch": 0.46054, + "grad_norm": 0.7355472220883392, + "learning_rate": 0.003, + "loss": 3.982, + "step": 46054 + }, + { + "epoch": 0.46055, + "grad_norm": 0.7450308085616708, + "learning_rate": 0.003, + "loss": 3.966, + "step": 46055 + }, + { + "epoch": 0.46056, + "grad_norm": 0.8474237873465453, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 46056 + }, + { + "epoch": 0.46057, + "grad_norm": 0.8664266357919659, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 46057 + }, + { + "epoch": 0.46058, + "grad_norm": 0.7671729937779257, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 46058 + }, + { + "epoch": 0.46059, + "grad_norm": 0.731086701698021, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 46059 + }, + { + "epoch": 0.4606, + "grad_norm": 0.7045042022081376, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 46060 + }, + { + "epoch": 0.46061, + "grad_norm": 0.794936363449463, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 46061 + }, + { + "epoch": 0.46062, + "grad_norm": 0.7806836687771108, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 46062 + }, + { + "epoch": 0.46063, + "grad_norm": 0.7439140729860022, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 46063 + }, + { + "epoch": 0.46064, + "grad_norm": 0.6409719599662987, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 46064 + }, + { + "epoch": 0.46065, + "grad_norm": 0.6249656890881152, + "learning_rate": 0.003, + "loss": 4.004, + "step": 46065 + }, + { + "epoch": 0.46066, + "grad_norm": 0.6478988803104786, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 46066 + }, + { + "epoch": 0.46067, + "grad_norm": 0.7167363681195962, + "learning_rate": 0.003, + "loss": 3.995, + "step": 46067 + }, + { + "epoch": 0.46068, + "grad_norm": 0.852538734062243, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 46068 + }, + { + "epoch": 0.46069, + "grad_norm": 0.9243938508958368, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 46069 + }, + { + "epoch": 0.4607, + "grad_norm": 1.0346103518647547, + "learning_rate": 0.003, + "loss": 3.9408, + "step": 46070 + }, + { + "epoch": 0.46071, + "grad_norm": 1.0268264181092464, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 46071 + }, + { + "epoch": 0.46072, + "grad_norm": 0.8386338804819159, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 46072 + }, + { + "epoch": 0.46073, + "grad_norm": 0.8441400971699398, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 46073 + }, + { + "epoch": 0.46074, + "grad_norm": 0.7539030020451548, + "learning_rate": 0.003, + "loss": 4.0181, + "step": 46074 + }, + { + "epoch": 0.46075, + "grad_norm": 0.7015553679550032, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 46075 + }, + { + "epoch": 0.46076, + "grad_norm": 0.7220729169194872, + "learning_rate": 0.003, + "loss": 3.966, + "step": 46076 + }, + { + "epoch": 0.46077, + "grad_norm": 0.9075621885403641, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 46077 + }, + { + "epoch": 0.46078, + "grad_norm": 0.8923137914102691, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 46078 + }, + { + "epoch": 0.46079, + "grad_norm": 0.8699352712527414, + "learning_rate": 0.003, + "loss": 4.03, + "step": 46079 + }, + { + "epoch": 0.4608, + "grad_norm": 0.7843795732275294, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 46080 + }, + { + "epoch": 0.46081, + "grad_norm": 0.7667373421079948, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 46081 + }, + { + "epoch": 0.46082, + "grad_norm": 0.6931905299081694, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 46082 + }, + { + "epoch": 0.46083, + "grad_norm": 0.5474838749058973, + "learning_rate": 0.003, + "loss": 3.978, + "step": 46083 + }, + { + "epoch": 0.46084, + "grad_norm": 0.5459488659059257, + "learning_rate": 0.003, + "loss": 3.967, + "step": 46084 + }, + { + "epoch": 0.46085, + "grad_norm": 0.5945738148122804, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 46085 + }, + { + "epoch": 0.46086, + "grad_norm": 0.7040884456925613, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 46086 + }, + { + "epoch": 0.46087, + "grad_norm": 0.7488013887671129, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 46087 + }, + { + "epoch": 0.46088, + "grad_norm": 0.6626752032781599, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 46088 + }, + { + "epoch": 0.46089, + "grad_norm": 0.6483425559655348, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 46089 + }, + { + "epoch": 0.4609, + "grad_norm": 0.674508799670295, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 46090 + }, + { + "epoch": 0.46091, + "grad_norm": 0.732170902994375, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 46091 + }, + { + "epoch": 0.46092, + "grad_norm": 0.7640636641006977, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 46092 + }, + { + "epoch": 0.46093, + "grad_norm": 0.853814012178034, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 46093 + }, + { + "epoch": 0.46094, + "grad_norm": 1.0062926451473615, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 46094 + }, + { + "epoch": 0.46095, + "grad_norm": 1.0819083783361492, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 46095 + }, + { + "epoch": 0.46096, + "grad_norm": 0.8403748790661945, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 46096 + }, + { + "epoch": 0.46097, + "grad_norm": 0.8955068315328364, + "learning_rate": 0.003, + "loss": 4.01, + "step": 46097 + }, + { + "epoch": 0.46098, + "grad_norm": 0.8560859697849938, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 46098 + }, + { + "epoch": 0.46099, + "grad_norm": 0.8423360857786364, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 46099 + }, + { + "epoch": 0.461, + "grad_norm": 0.7983836190122336, + "learning_rate": 0.003, + "loss": 3.969, + "step": 46100 + }, + { + "epoch": 0.46101, + "grad_norm": 0.7705520567330981, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 46101 + }, + { + "epoch": 0.46102, + "grad_norm": 0.7956174141964488, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 46102 + }, + { + "epoch": 0.46103, + "grad_norm": 0.7636728669858597, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 46103 + }, + { + "epoch": 0.46104, + "grad_norm": 0.8247677136187849, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 46104 + }, + { + "epoch": 0.46105, + "grad_norm": 0.9283479175286721, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 46105 + }, + { + "epoch": 0.46106, + "grad_norm": 0.9545633424223205, + "learning_rate": 0.003, + "loss": 3.988, + "step": 46106 + }, + { + "epoch": 0.46107, + "grad_norm": 0.9606806757175189, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 46107 + }, + { + "epoch": 0.46108, + "grad_norm": 0.934198652510119, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 46108 + }, + { + "epoch": 0.46109, + "grad_norm": 0.9272538443288518, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 46109 + }, + { + "epoch": 0.4611, + "grad_norm": 0.9185493545229426, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 46110 + }, + { + "epoch": 0.46111, + "grad_norm": 0.8694535552130924, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 46111 + }, + { + "epoch": 0.46112, + "grad_norm": 0.9761741266282735, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 46112 + }, + { + "epoch": 0.46113, + "grad_norm": 1.2405985783842788, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 46113 + }, + { + "epoch": 0.46114, + "grad_norm": 0.9001046446634914, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 46114 + }, + { + "epoch": 0.46115, + "grad_norm": 0.8121474195584287, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 46115 + }, + { + "epoch": 0.46116, + "grad_norm": 0.774263864853858, + "learning_rate": 0.003, + "loss": 3.993, + "step": 46116 + }, + { + "epoch": 0.46117, + "grad_norm": 0.7141660636900601, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 46117 + }, + { + "epoch": 0.46118, + "grad_norm": 0.7969617286685999, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 46118 + }, + { + "epoch": 0.46119, + "grad_norm": 0.7961524746307774, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 46119 + }, + { + "epoch": 0.4612, + "grad_norm": 0.6731190223968893, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 46120 + }, + { + "epoch": 0.46121, + "grad_norm": 0.6401503393846736, + "learning_rate": 0.003, + "loss": 4.0236, + "step": 46121 + }, + { + "epoch": 0.46122, + "grad_norm": 0.6448155027402772, + "learning_rate": 0.003, + "loss": 4.004, + "step": 46122 + }, + { + "epoch": 0.46123, + "grad_norm": 0.6673836124568621, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 46123 + }, + { + "epoch": 0.46124, + "grad_norm": 0.7266344184003495, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 46124 + }, + { + "epoch": 0.46125, + "grad_norm": 0.6941791625682858, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 46125 + }, + { + "epoch": 0.46126, + "grad_norm": 0.6551257282034285, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 46126 + }, + { + "epoch": 0.46127, + "grad_norm": 0.683078841384787, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 46127 + }, + { + "epoch": 0.46128, + "grad_norm": 0.7361976345526366, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 46128 + }, + { + "epoch": 0.46129, + "grad_norm": 0.8356025457599914, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 46129 + }, + { + "epoch": 0.4613, + "grad_norm": 0.869298550456343, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 46130 + }, + { + "epoch": 0.46131, + "grad_norm": 0.7565007716719779, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 46131 + }, + { + "epoch": 0.46132, + "grad_norm": 0.7262560064308989, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 46132 + }, + { + "epoch": 0.46133, + "grad_norm": 0.7564745979145113, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 46133 + }, + { + "epoch": 0.46134, + "grad_norm": 0.8091674296707665, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 46134 + }, + { + "epoch": 0.46135, + "grad_norm": 0.7733282009734743, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 46135 + }, + { + "epoch": 0.46136, + "grad_norm": 0.710790171211667, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 46136 + }, + { + "epoch": 0.46137, + "grad_norm": 0.7519632456751361, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 46137 + }, + { + "epoch": 0.46138, + "grad_norm": 0.7780379341657814, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 46138 + }, + { + "epoch": 0.46139, + "grad_norm": 0.7235979917729689, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 46139 + }, + { + "epoch": 0.4614, + "grad_norm": 0.7593030399000641, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 46140 + }, + { + "epoch": 0.46141, + "grad_norm": 0.8498714167369619, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 46141 + }, + { + "epoch": 0.46142, + "grad_norm": 0.8633024791421033, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 46142 + }, + { + "epoch": 0.46143, + "grad_norm": 0.9303701236548307, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 46143 + }, + { + "epoch": 0.46144, + "grad_norm": 0.8445521480526309, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 46144 + }, + { + "epoch": 0.46145, + "grad_norm": 0.7561366521387777, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 46145 + }, + { + "epoch": 0.46146, + "grad_norm": 0.6717462763929125, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 46146 + }, + { + "epoch": 0.46147, + "grad_norm": 0.7296823683657341, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 46147 + }, + { + "epoch": 0.46148, + "grad_norm": 0.8125849415899478, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 46148 + }, + { + "epoch": 0.46149, + "grad_norm": 0.829336773357504, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 46149 + }, + { + "epoch": 0.4615, + "grad_norm": 0.8245396899722678, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 46150 + }, + { + "epoch": 0.46151, + "grad_norm": 0.7687590327407453, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 46151 + }, + { + "epoch": 0.46152, + "grad_norm": 0.6392530753970026, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 46152 + }, + { + "epoch": 0.46153, + "grad_norm": 0.534543120007805, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 46153 + }, + { + "epoch": 0.46154, + "grad_norm": 0.5196071819820476, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 46154 + }, + { + "epoch": 0.46155, + "grad_norm": 0.49664568520665325, + "learning_rate": 0.003, + "loss": 3.981, + "step": 46155 + }, + { + "epoch": 0.46156, + "grad_norm": 0.5497125593889314, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 46156 + }, + { + "epoch": 0.46157, + "grad_norm": 0.6703604605774172, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 46157 + }, + { + "epoch": 0.46158, + "grad_norm": 0.7878285423098444, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 46158 + }, + { + "epoch": 0.46159, + "grad_norm": 0.8679938172953491, + "learning_rate": 0.003, + "loss": 3.9435, + "step": 46159 + }, + { + "epoch": 0.4616, + "grad_norm": 0.8535259179522828, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 46160 + }, + { + "epoch": 0.46161, + "grad_norm": 0.8163142340692621, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 46161 + }, + { + "epoch": 0.46162, + "grad_norm": 0.8060294366568058, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 46162 + }, + { + "epoch": 0.46163, + "grad_norm": 0.8323422863794439, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 46163 + }, + { + "epoch": 0.46164, + "grad_norm": 0.8436092453493211, + "learning_rate": 0.003, + "loss": 4.0188, + "step": 46164 + }, + { + "epoch": 0.46165, + "grad_norm": 0.832219871588543, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 46165 + }, + { + "epoch": 0.46166, + "grad_norm": 0.86471897967413, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 46166 + }, + { + "epoch": 0.46167, + "grad_norm": 0.8770047136667483, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 46167 + }, + { + "epoch": 0.46168, + "grad_norm": 0.9408458827235473, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 46168 + }, + { + "epoch": 0.46169, + "grad_norm": 1.0012761277584115, + "learning_rate": 0.003, + "loss": 3.997, + "step": 46169 + }, + { + "epoch": 0.4617, + "grad_norm": 0.9546975895108062, + "learning_rate": 0.003, + "loss": 4.0, + "step": 46170 + }, + { + "epoch": 0.46171, + "grad_norm": 0.9048528286601103, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 46171 + }, + { + "epoch": 0.46172, + "grad_norm": 0.8837453315474972, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 46172 + }, + { + "epoch": 0.46173, + "grad_norm": 0.9000116698647799, + "learning_rate": 0.003, + "loss": 3.995, + "step": 46173 + }, + { + "epoch": 0.46174, + "grad_norm": 0.8011555457760788, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 46174 + }, + { + "epoch": 0.46175, + "grad_norm": 0.9851516956787489, + "learning_rate": 0.003, + "loss": 3.9449, + "step": 46175 + }, + { + "epoch": 0.46176, + "grad_norm": 1.1030772002138476, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 46176 + }, + { + "epoch": 0.46177, + "grad_norm": 0.7914077714376997, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 46177 + }, + { + "epoch": 0.46178, + "grad_norm": 0.7460688137753876, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 46178 + }, + { + "epoch": 0.46179, + "grad_norm": 0.7743927460519637, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 46179 + }, + { + "epoch": 0.4618, + "grad_norm": 0.7752277541356939, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 46180 + }, + { + "epoch": 0.46181, + "grad_norm": 0.8318341627791387, + "learning_rate": 0.003, + "loss": 3.992, + "step": 46181 + }, + { + "epoch": 0.46182, + "grad_norm": 0.9978629329380362, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 46182 + }, + { + "epoch": 0.46183, + "grad_norm": 1.1539690537116398, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 46183 + }, + { + "epoch": 0.46184, + "grad_norm": 0.7843764773900718, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 46184 + }, + { + "epoch": 0.46185, + "grad_norm": 0.742935876899624, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 46185 + }, + { + "epoch": 0.46186, + "grad_norm": 0.7800364029862336, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 46186 + }, + { + "epoch": 0.46187, + "grad_norm": 0.7988240521523147, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 46187 + }, + { + "epoch": 0.46188, + "grad_norm": 0.936711281905032, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 46188 + }, + { + "epoch": 0.46189, + "grad_norm": 0.951868519294804, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 46189 + }, + { + "epoch": 0.4619, + "grad_norm": 0.9178832048991714, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 46190 + }, + { + "epoch": 0.46191, + "grad_norm": 0.8118299065819827, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 46191 + }, + { + "epoch": 0.46192, + "grad_norm": 0.72081709506864, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 46192 + }, + { + "epoch": 0.46193, + "grad_norm": 0.7167607230025047, + "learning_rate": 0.003, + "loss": 3.983, + "step": 46193 + }, + { + "epoch": 0.46194, + "grad_norm": 0.7774853246558611, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 46194 + }, + { + "epoch": 0.46195, + "grad_norm": 0.7109103826908839, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 46195 + }, + { + "epoch": 0.46196, + "grad_norm": 0.6635986372991618, + "learning_rate": 0.003, + "loss": 3.96, + "step": 46196 + }, + { + "epoch": 0.46197, + "grad_norm": 0.7328105567685727, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 46197 + }, + { + "epoch": 0.46198, + "grad_norm": 0.8457936989376855, + "learning_rate": 0.003, + "loss": 4.0267, + "step": 46198 + }, + { + "epoch": 0.46199, + "grad_norm": 0.9556305641457998, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 46199 + }, + { + "epoch": 0.462, + "grad_norm": 0.8506267635450259, + "learning_rate": 0.003, + "loss": 3.968, + "step": 46200 + }, + { + "epoch": 0.46201, + "grad_norm": 0.697616838495345, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 46201 + }, + { + "epoch": 0.46202, + "grad_norm": 0.598365154174751, + "learning_rate": 0.003, + "loss": 3.988, + "step": 46202 + }, + { + "epoch": 0.46203, + "grad_norm": 0.6491319628507128, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 46203 + }, + { + "epoch": 0.46204, + "grad_norm": 0.7234093842313428, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 46204 + }, + { + "epoch": 0.46205, + "grad_norm": 0.670213054292606, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 46205 + }, + { + "epoch": 0.46206, + "grad_norm": 0.6799429866481086, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 46206 + }, + { + "epoch": 0.46207, + "grad_norm": 0.7117253278472223, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 46207 + }, + { + "epoch": 0.46208, + "grad_norm": 0.6257628739034952, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 46208 + }, + { + "epoch": 0.46209, + "grad_norm": 0.6325995003665369, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 46209 + }, + { + "epoch": 0.4621, + "grad_norm": 0.6449915142295872, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 46210 + }, + { + "epoch": 0.46211, + "grad_norm": 0.7222609815147819, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 46211 + }, + { + "epoch": 0.46212, + "grad_norm": 0.7246301752949615, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 46212 + }, + { + "epoch": 0.46213, + "grad_norm": 0.7472387920654417, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 46213 + }, + { + "epoch": 0.46214, + "grad_norm": 0.7432552802411749, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 46214 + }, + { + "epoch": 0.46215, + "grad_norm": 0.6426093500521299, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 46215 + }, + { + "epoch": 0.46216, + "grad_norm": 0.6759047791618987, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 46216 + }, + { + "epoch": 0.46217, + "grad_norm": 0.7459662700185898, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 46217 + }, + { + "epoch": 0.46218, + "grad_norm": 0.7515146455865624, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 46218 + }, + { + "epoch": 0.46219, + "grad_norm": 0.8425824850013869, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 46219 + }, + { + "epoch": 0.4622, + "grad_norm": 1.0325399437493, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 46220 + }, + { + "epoch": 0.46221, + "grad_norm": 1.194597200686082, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 46221 + }, + { + "epoch": 0.46222, + "grad_norm": 0.7477463266491715, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 46222 + }, + { + "epoch": 0.46223, + "grad_norm": 0.6777338940089018, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 46223 + }, + { + "epoch": 0.46224, + "grad_norm": 0.6624070277262407, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 46224 + }, + { + "epoch": 0.46225, + "grad_norm": 0.7305060669700033, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 46225 + }, + { + "epoch": 0.46226, + "grad_norm": 0.7951338466811549, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 46226 + }, + { + "epoch": 0.46227, + "grad_norm": 0.7891517508220959, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46227 + }, + { + "epoch": 0.46228, + "grad_norm": 0.7005503322879093, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 46228 + }, + { + "epoch": 0.46229, + "grad_norm": 0.7003324074964248, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 46229 + }, + { + "epoch": 0.4623, + "grad_norm": 0.6922280014665422, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 46230 + }, + { + "epoch": 0.46231, + "grad_norm": 0.7173128251376121, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 46231 + }, + { + "epoch": 0.46232, + "grad_norm": 0.79498622252163, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 46232 + }, + { + "epoch": 0.46233, + "grad_norm": 0.8519465076517279, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 46233 + }, + { + "epoch": 0.46234, + "grad_norm": 1.0029961872902402, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 46234 + }, + { + "epoch": 0.46235, + "grad_norm": 1.3280935406085954, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 46235 + }, + { + "epoch": 0.46236, + "grad_norm": 0.7603010005279883, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 46236 + }, + { + "epoch": 0.46237, + "grad_norm": 0.7585955555855337, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 46237 + }, + { + "epoch": 0.46238, + "grad_norm": 0.739921999650996, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 46238 + }, + { + "epoch": 0.46239, + "grad_norm": 0.7529842904689223, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 46239 + }, + { + "epoch": 0.4624, + "grad_norm": 0.7759813165778462, + "learning_rate": 0.003, + "loss": 3.981, + "step": 46240 + }, + { + "epoch": 0.46241, + "grad_norm": 0.845785230640751, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 46241 + }, + { + "epoch": 0.46242, + "grad_norm": 0.7421522665619219, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46242 + }, + { + "epoch": 0.46243, + "grad_norm": 0.653916669204992, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 46243 + }, + { + "epoch": 0.46244, + "grad_norm": 0.6732274063530488, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 46244 + }, + { + "epoch": 0.46245, + "grad_norm": 0.7083630644669818, + "learning_rate": 0.003, + "loss": 3.9348, + "step": 46245 + }, + { + "epoch": 0.46246, + "grad_norm": 0.6395465448169065, + "learning_rate": 0.003, + "loss": 3.978, + "step": 46246 + }, + { + "epoch": 0.46247, + "grad_norm": 0.6938289041478275, + "learning_rate": 0.003, + "loss": 3.9275, + "step": 46247 + }, + { + "epoch": 0.46248, + "grad_norm": 0.6949319374265978, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 46248 + }, + { + "epoch": 0.46249, + "grad_norm": 0.7757403075927536, + "learning_rate": 0.003, + "loss": 3.981, + "step": 46249 + }, + { + "epoch": 0.4625, + "grad_norm": 0.706791006426814, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 46250 + }, + { + "epoch": 0.46251, + "grad_norm": 0.7112388303162562, + "learning_rate": 0.003, + "loss": 3.965, + "step": 46251 + }, + { + "epoch": 0.46252, + "grad_norm": 0.705760113409512, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 46252 + }, + { + "epoch": 0.46253, + "grad_norm": 0.9217922102899526, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 46253 + }, + { + "epoch": 0.46254, + "grad_norm": 1.1381803476561427, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 46254 + }, + { + "epoch": 0.46255, + "grad_norm": 0.8332563505496833, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 46255 + }, + { + "epoch": 0.46256, + "grad_norm": 0.6668494330696547, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 46256 + }, + { + "epoch": 0.46257, + "grad_norm": 0.6305570621570368, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 46257 + }, + { + "epoch": 0.46258, + "grad_norm": 0.5975674984083665, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 46258 + }, + { + "epoch": 0.46259, + "grad_norm": 0.7292826431853713, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 46259 + }, + { + "epoch": 0.4626, + "grad_norm": 0.7028282751489373, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 46260 + }, + { + "epoch": 0.46261, + "grad_norm": 0.7858564001756767, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 46261 + }, + { + "epoch": 0.46262, + "grad_norm": 0.7734033178762917, + "learning_rate": 0.003, + "loss": 3.981, + "step": 46262 + }, + { + "epoch": 0.46263, + "grad_norm": 0.7401515965705364, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 46263 + }, + { + "epoch": 0.46264, + "grad_norm": 0.6588713485541382, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46264 + }, + { + "epoch": 0.46265, + "grad_norm": 0.6281714513822332, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 46265 + }, + { + "epoch": 0.46266, + "grad_norm": 0.5990275687850257, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 46266 + }, + { + "epoch": 0.46267, + "grad_norm": 0.7160168495523183, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 46267 + }, + { + "epoch": 0.46268, + "grad_norm": 0.7232423587681381, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 46268 + }, + { + "epoch": 0.46269, + "grad_norm": 0.7481375178370783, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 46269 + }, + { + "epoch": 0.4627, + "grad_norm": 0.728308578042047, + "learning_rate": 0.003, + "loss": 3.989, + "step": 46270 + }, + { + "epoch": 0.46271, + "grad_norm": 0.9105590241482242, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 46271 + }, + { + "epoch": 0.46272, + "grad_norm": 1.083303870157831, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 46272 + }, + { + "epoch": 0.46273, + "grad_norm": 1.00691695688253, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 46273 + }, + { + "epoch": 0.46274, + "grad_norm": 1.0227439597594339, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 46274 + }, + { + "epoch": 0.46275, + "grad_norm": 0.9069179728775554, + "learning_rate": 0.003, + "loss": 3.979, + "step": 46275 + }, + { + "epoch": 0.46276, + "grad_norm": 0.8173769097072481, + "learning_rate": 0.003, + "loss": 4.0295, + "step": 46276 + }, + { + "epoch": 0.46277, + "grad_norm": 0.822541394102342, + "learning_rate": 0.003, + "loss": 3.998, + "step": 46277 + }, + { + "epoch": 0.46278, + "grad_norm": 0.8223709492013216, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 46278 + }, + { + "epoch": 0.46279, + "grad_norm": 0.995496608179802, + "learning_rate": 0.003, + "loss": 3.985, + "step": 46279 + }, + { + "epoch": 0.4628, + "grad_norm": 1.170548091340935, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 46280 + }, + { + "epoch": 0.46281, + "grad_norm": 0.9642143650365168, + "learning_rate": 0.003, + "loss": 3.976, + "step": 46281 + }, + { + "epoch": 0.46282, + "grad_norm": 1.0676746266016834, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 46282 + }, + { + "epoch": 0.46283, + "grad_norm": 0.9583635915672908, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 46283 + }, + { + "epoch": 0.46284, + "grad_norm": 0.7966846266549803, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 46284 + }, + { + "epoch": 0.46285, + "grad_norm": 0.84666475072917, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 46285 + }, + { + "epoch": 0.46286, + "grad_norm": 0.8855661946499901, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 46286 + }, + { + "epoch": 0.46287, + "grad_norm": 0.7836439913074331, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 46287 + }, + { + "epoch": 0.46288, + "grad_norm": 0.6658467344665019, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 46288 + }, + { + "epoch": 0.46289, + "grad_norm": 0.6619983746433484, + "learning_rate": 0.003, + "loss": 3.969, + "step": 46289 + }, + { + "epoch": 0.4629, + "grad_norm": 0.5706791421682932, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46290 + }, + { + "epoch": 0.46291, + "grad_norm": 0.5402878092150121, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 46291 + }, + { + "epoch": 0.46292, + "grad_norm": 0.5676775655509568, + "learning_rate": 0.003, + "loss": 3.967, + "step": 46292 + }, + { + "epoch": 0.46293, + "grad_norm": 0.5852553879625921, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 46293 + }, + { + "epoch": 0.46294, + "grad_norm": 0.5613697920364583, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 46294 + }, + { + "epoch": 0.46295, + "grad_norm": 0.5775348965868908, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 46295 + }, + { + "epoch": 0.46296, + "grad_norm": 0.5739665052728381, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 46296 + }, + { + "epoch": 0.46297, + "grad_norm": 0.5689807426652945, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 46297 + }, + { + "epoch": 0.46298, + "grad_norm": 0.7079538682744421, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 46298 + }, + { + "epoch": 0.46299, + "grad_norm": 1.018571414380638, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 46299 + }, + { + "epoch": 0.463, + "grad_norm": 1.2502870635724925, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 46300 + }, + { + "epoch": 0.46301, + "grad_norm": 0.7601223181256347, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 46301 + }, + { + "epoch": 0.46302, + "grad_norm": 0.7386904543614131, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 46302 + }, + { + "epoch": 0.46303, + "grad_norm": 0.662648183273251, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46303 + }, + { + "epoch": 0.46304, + "grad_norm": 0.625517667986332, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 46304 + }, + { + "epoch": 0.46305, + "grad_norm": 0.5855129028779539, + "learning_rate": 0.003, + "loss": 3.9395, + "step": 46305 + }, + { + "epoch": 0.46306, + "grad_norm": 0.5804698950918122, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 46306 + }, + { + "epoch": 0.46307, + "grad_norm": 0.6097536635654853, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 46307 + }, + { + "epoch": 0.46308, + "grad_norm": 0.6540870460279794, + "learning_rate": 0.003, + "loss": 4.006, + "step": 46308 + }, + { + "epoch": 0.46309, + "grad_norm": 0.7118001610795277, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 46309 + }, + { + "epoch": 0.4631, + "grad_norm": 0.7409367072440245, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 46310 + }, + { + "epoch": 0.46311, + "grad_norm": 0.652956337365434, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 46311 + }, + { + "epoch": 0.46312, + "grad_norm": 0.6538322254194736, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 46312 + }, + { + "epoch": 0.46313, + "grad_norm": 0.7576465501919833, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 46313 + }, + { + "epoch": 0.46314, + "grad_norm": 0.841642079341252, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 46314 + }, + { + "epoch": 0.46315, + "grad_norm": 1.0016132254770462, + "learning_rate": 0.003, + "loss": 3.9424, + "step": 46315 + }, + { + "epoch": 0.46316, + "grad_norm": 1.0057039180449807, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 46316 + }, + { + "epoch": 0.46317, + "grad_norm": 0.960152701652127, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 46317 + }, + { + "epoch": 0.46318, + "grad_norm": 0.7912956155624002, + "learning_rate": 0.003, + "loss": 3.963, + "step": 46318 + }, + { + "epoch": 0.46319, + "grad_norm": 0.7702370161302693, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 46319 + }, + { + "epoch": 0.4632, + "grad_norm": 0.8892871444248774, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 46320 + }, + { + "epoch": 0.46321, + "grad_norm": 0.8880682587908104, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 46321 + }, + { + "epoch": 0.46322, + "grad_norm": 0.935237042436257, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 46322 + }, + { + "epoch": 0.46323, + "grad_norm": 0.9937418284043154, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 46323 + }, + { + "epoch": 0.46324, + "grad_norm": 0.9958749968691105, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 46324 + }, + { + "epoch": 0.46325, + "grad_norm": 1.1075384592766666, + "learning_rate": 0.003, + "loss": 3.993, + "step": 46325 + }, + { + "epoch": 0.46326, + "grad_norm": 0.7568576255475864, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 46326 + }, + { + "epoch": 0.46327, + "grad_norm": 0.6085528734008655, + "learning_rate": 0.003, + "loss": 3.9543, + "step": 46327 + }, + { + "epoch": 0.46328, + "grad_norm": 0.5843247294244058, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 46328 + }, + { + "epoch": 0.46329, + "grad_norm": 0.6090964186560367, + "learning_rate": 0.003, + "loss": 3.9414, + "step": 46329 + }, + { + "epoch": 0.4633, + "grad_norm": 0.7078076900035776, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 46330 + }, + { + "epoch": 0.46331, + "grad_norm": 0.8658878609964398, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 46331 + }, + { + "epoch": 0.46332, + "grad_norm": 1.0366436439249729, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 46332 + }, + { + "epoch": 0.46333, + "grad_norm": 0.9616709047439519, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 46333 + }, + { + "epoch": 0.46334, + "grad_norm": 1.0008652786003338, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 46334 + }, + { + "epoch": 0.46335, + "grad_norm": 1.036945627047381, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 46335 + }, + { + "epoch": 0.46336, + "grad_norm": 0.8115642746426299, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 46336 + }, + { + "epoch": 0.46337, + "grad_norm": 0.7652721635760016, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 46337 + }, + { + "epoch": 0.46338, + "grad_norm": 0.8263645247477285, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 46338 + }, + { + "epoch": 0.46339, + "grad_norm": 0.8165499309809056, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 46339 + }, + { + "epoch": 0.4634, + "grad_norm": 0.8719951259207123, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 46340 + }, + { + "epoch": 0.46341, + "grad_norm": 0.887818131024306, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 46341 + }, + { + "epoch": 0.46342, + "grad_norm": 0.7547870973948934, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 46342 + }, + { + "epoch": 0.46343, + "grad_norm": 0.8709034838428417, + "learning_rate": 0.003, + "loss": 3.994, + "step": 46343 + }, + { + "epoch": 0.46344, + "grad_norm": 1.030198167119345, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 46344 + }, + { + "epoch": 0.46345, + "grad_norm": 0.9741762658941431, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 46345 + }, + { + "epoch": 0.46346, + "grad_norm": 0.7701556592711276, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 46346 + }, + { + "epoch": 0.46347, + "grad_norm": 0.7538509920889235, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 46347 + }, + { + "epoch": 0.46348, + "grad_norm": 0.6732549484301388, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 46348 + }, + { + "epoch": 0.46349, + "grad_norm": 0.717636928327057, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 46349 + }, + { + "epoch": 0.4635, + "grad_norm": 0.6719017424571267, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 46350 + }, + { + "epoch": 0.46351, + "grad_norm": 0.6829045637921138, + "learning_rate": 0.003, + "loss": 3.9396, + "step": 46351 + }, + { + "epoch": 0.46352, + "grad_norm": 0.6965234785178215, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 46352 + }, + { + "epoch": 0.46353, + "grad_norm": 0.7656951551056823, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 46353 + }, + { + "epoch": 0.46354, + "grad_norm": 0.9695692944890874, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 46354 + }, + { + "epoch": 0.46355, + "grad_norm": 1.2088559695792913, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 46355 + }, + { + "epoch": 0.46356, + "grad_norm": 0.7774197790639472, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 46356 + }, + { + "epoch": 0.46357, + "grad_norm": 0.659641682517417, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 46357 + }, + { + "epoch": 0.46358, + "grad_norm": 0.6686533928336333, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 46358 + }, + { + "epoch": 0.46359, + "grad_norm": 0.6773976891019196, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 46359 + }, + { + "epoch": 0.4636, + "grad_norm": 0.6960039793463295, + "learning_rate": 0.003, + "loss": 3.982, + "step": 46360 + }, + { + "epoch": 0.46361, + "grad_norm": 0.6942125496490018, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 46361 + }, + { + "epoch": 0.46362, + "grad_norm": 0.6080716905526087, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 46362 + }, + { + "epoch": 0.46363, + "grad_norm": 0.5507695510695167, + "learning_rate": 0.003, + "loss": 3.993, + "step": 46363 + }, + { + "epoch": 0.46364, + "grad_norm": 0.562527106911726, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 46364 + }, + { + "epoch": 0.46365, + "grad_norm": 0.6461638631998633, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46365 + }, + { + "epoch": 0.46366, + "grad_norm": 0.7253262881817144, + "learning_rate": 0.003, + "loss": 3.974, + "step": 46366 + }, + { + "epoch": 0.46367, + "grad_norm": 0.754558245265746, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 46367 + }, + { + "epoch": 0.46368, + "grad_norm": 0.6991236742813753, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 46368 + }, + { + "epoch": 0.46369, + "grad_norm": 0.7135271596743125, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 46369 + }, + { + "epoch": 0.4637, + "grad_norm": 0.6883863976525191, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 46370 + }, + { + "epoch": 0.46371, + "grad_norm": 0.7398381400892182, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 46371 + }, + { + "epoch": 0.46372, + "grad_norm": 0.8030695676255072, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 46372 + }, + { + "epoch": 0.46373, + "grad_norm": 0.8693023991527065, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 46373 + }, + { + "epoch": 0.46374, + "grad_norm": 0.866842835361332, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 46374 + }, + { + "epoch": 0.46375, + "grad_norm": 0.9368358516547253, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 46375 + }, + { + "epoch": 0.46376, + "grad_norm": 0.9257829050642995, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 46376 + }, + { + "epoch": 0.46377, + "grad_norm": 0.8995883336494885, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 46377 + }, + { + "epoch": 0.46378, + "grad_norm": 0.80119793044248, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 46378 + }, + { + "epoch": 0.46379, + "grad_norm": 0.6667930936982303, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 46379 + }, + { + "epoch": 0.4638, + "grad_norm": 0.6743618553007573, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 46380 + }, + { + "epoch": 0.46381, + "grad_norm": 0.6834996812285783, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 46381 + }, + { + "epoch": 0.46382, + "grad_norm": 0.8084831922778744, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 46382 + }, + { + "epoch": 0.46383, + "grad_norm": 0.8463119400567184, + "learning_rate": 0.003, + "loss": 3.9407, + "step": 46383 + }, + { + "epoch": 0.46384, + "grad_norm": 0.7486560676852836, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 46384 + }, + { + "epoch": 0.46385, + "grad_norm": 0.7343887262628164, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 46385 + }, + { + "epoch": 0.46386, + "grad_norm": 0.682768884267967, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 46386 + }, + { + "epoch": 0.46387, + "grad_norm": 0.6673850284425459, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 46387 + }, + { + "epoch": 0.46388, + "grad_norm": 0.6591345108057289, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 46388 + }, + { + "epoch": 0.46389, + "grad_norm": 0.7294590740548467, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 46389 + }, + { + "epoch": 0.4639, + "grad_norm": 0.8901979561958527, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 46390 + }, + { + "epoch": 0.46391, + "grad_norm": 1.015205764588102, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 46391 + }, + { + "epoch": 0.46392, + "grad_norm": 0.9646208731440735, + "learning_rate": 0.003, + "loss": 3.984, + "step": 46392 + }, + { + "epoch": 0.46393, + "grad_norm": 0.8969744893889869, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 46393 + }, + { + "epoch": 0.46394, + "grad_norm": 0.7675761077268521, + "learning_rate": 0.003, + "loss": 3.9432, + "step": 46394 + }, + { + "epoch": 0.46395, + "grad_norm": 0.7221674704635824, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 46395 + }, + { + "epoch": 0.46396, + "grad_norm": 0.760415921462264, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 46396 + }, + { + "epoch": 0.46397, + "grad_norm": 0.7619400578340789, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 46397 + }, + { + "epoch": 0.46398, + "grad_norm": 0.8985073718636332, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 46398 + }, + { + "epoch": 0.46399, + "grad_norm": 0.8801897811267838, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 46399 + }, + { + "epoch": 0.464, + "grad_norm": 0.990352268191268, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 46400 + }, + { + "epoch": 0.46401, + "grad_norm": 1.1100163378953962, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 46401 + }, + { + "epoch": 0.46402, + "grad_norm": 0.883177035145137, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 46402 + }, + { + "epoch": 0.46403, + "grad_norm": 0.803794299611985, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 46403 + }, + { + "epoch": 0.46404, + "grad_norm": 0.9071342560771948, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 46404 + }, + { + "epoch": 0.46405, + "grad_norm": 0.9729573132646535, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 46405 + }, + { + "epoch": 0.46406, + "grad_norm": 0.9304040402915039, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 46406 + }, + { + "epoch": 0.46407, + "grad_norm": 0.8244410373249064, + "learning_rate": 0.003, + "loss": 4.003, + "step": 46407 + }, + { + "epoch": 0.46408, + "grad_norm": 0.8301022491627844, + "learning_rate": 0.003, + "loss": 4.016, + "step": 46408 + }, + { + "epoch": 0.46409, + "grad_norm": 0.878018685877541, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 46409 + }, + { + "epoch": 0.4641, + "grad_norm": 0.8749104574857131, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 46410 + }, + { + "epoch": 0.46411, + "grad_norm": 0.7894494829874069, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 46411 + }, + { + "epoch": 0.46412, + "grad_norm": 0.7104058346101614, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 46412 + }, + { + "epoch": 0.46413, + "grad_norm": 0.7280539406994576, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 46413 + }, + { + "epoch": 0.46414, + "grad_norm": 0.8772536059990104, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 46414 + }, + { + "epoch": 0.46415, + "grad_norm": 0.9515576552763895, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 46415 + }, + { + "epoch": 0.46416, + "grad_norm": 1.02526139955615, + "learning_rate": 0.003, + "loss": 4.012, + "step": 46416 + }, + { + "epoch": 0.46417, + "grad_norm": 0.9776488496630199, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 46417 + }, + { + "epoch": 0.46418, + "grad_norm": 0.893275143821093, + "learning_rate": 0.003, + "loss": 4.0213, + "step": 46418 + }, + { + "epoch": 0.46419, + "grad_norm": 0.8313554981757233, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 46419 + }, + { + "epoch": 0.4642, + "grad_norm": 0.867746606030377, + "learning_rate": 0.003, + "loss": 4.024, + "step": 46420 + }, + { + "epoch": 0.46421, + "grad_norm": 0.8086933354833992, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 46421 + }, + { + "epoch": 0.46422, + "grad_norm": 0.8869680706350013, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46422 + }, + { + "epoch": 0.46423, + "grad_norm": 0.8667220227119891, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 46423 + }, + { + "epoch": 0.46424, + "grad_norm": 0.8043535207656605, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 46424 + }, + { + "epoch": 0.46425, + "grad_norm": 0.7551482581900543, + "learning_rate": 0.003, + "loss": 3.9447, + "step": 46425 + }, + { + "epoch": 0.46426, + "grad_norm": 0.7516698586701372, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 46426 + }, + { + "epoch": 0.46427, + "grad_norm": 0.6671879514757904, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 46427 + }, + { + "epoch": 0.46428, + "grad_norm": 0.6430522327751562, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 46428 + }, + { + "epoch": 0.46429, + "grad_norm": 0.6300518962071661, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 46429 + }, + { + "epoch": 0.4643, + "grad_norm": 0.6084271460414201, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 46430 + }, + { + "epoch": 0.46431, + "grad_norm": 0.6520694814502295, + "learning_rate": 0.003, + "loss": 3.9438, + "step": 46431 + }, + { + "epoch": 0.46432, + "grad_norm": 0.5702770090735999, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 46432 + }, + { + "epoch": 0.46433, + "grad_norm": 0.5473870972749146, + "learning_rate": 0.003, + "loss": 3.971, + "step": 46433 + }, + { + "epoch": 0.46434, + "grad_norm": 0.5238486645015744, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 46434 + }, + { + "epoch": 0.46435, + "grad_norm": 0.476485275760457, + "learning_rate": 0.003, + "loss": 3.946, + "step": 46435 + }, + { + "epoch": 0.46436, + "grad_norm": 0.5142204441810805, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 46436 + }, + { + "epoch": 0.46437, + "grad_norm": 0.6018710252489368, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 46437 + }, + { + "epoch": 0.46438, + "grad_norm": 0.7244211496180328, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 46438 + }, + { + "epoch": 0.46439, + "grad_norm": 0.8536841805202073, + "learning_rate": 0.003, + "loss": 4.006, + "step": 46439 + }, + { + "epoch": 0.4644, + "grad_norm": 0.8755578066413254, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 46440 + }, + { + "epoch": 0.46441, + "grad_norm": 0.8231804267653136, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 46441 + }, + { + "epoch": 0.46442, + "grad_norm": 0.8031031295265496, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 46442 + }, + { + "epoch": 0.46443, + "grad_norm": 0.786739038973472, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 46443 + }, + { + "epoch": 0.46444, + "grad_norm": 0.7620920450098863, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 46444 + }, + { + "epoch": 0.46445, + "grad_norm": 0.8052558699751219, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 46445 + }, + { + "epoch": 0.46446, + "grad_norm": 0.7016825176096508, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 46446 + }, + { + "epoch": 0.46447, + "grad_norm": 0.8111246280303054, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 46447 + }, + { + "epoch": 0.46448, + "grad_norm": 0.9926628134516027, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 46448 + }, + { + "epoch": 0.46449, + "grad_norm": 1.203237101552385, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 46449 + }, + { + "epoch": 0.4645, + "grad_norm": 0.8532252890870962, + "learning_rate": 0.003, + "loss": 3.998, + "step": 46450 + }, + { + "epoch": 0.46451, + "grad_norm": 0.7456594020727968, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 46451 + }, + { + "epoch": 0.46452, + "grad_norm": 0.7164628742708826, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 46452 + }, + { + "epoch": 0.46453, + "grad_norm": 0.7058195836768859, + "learning_rate": 0.003, + "loss": 3.984, + "step": 46453 + }, + { + "epoch": 0.46454, + "grad_norm": 0.7594132166662131, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 46454 + }, + { + "epoch": 0.46455, + "grad_norm": 0.6963953065717493, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 46455 + }, + { + "epoch": 0.46456, + "grad_norm": 0.6732334454400104, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 46456 + }, + { + "epoch": 0.46457, + "grad_norm": 0.725895075415639, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 46457 + }, + { + "epoch": 0.46458, + "grad_norm": 0.7843670721943763, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 46458 + }, + { + "epoch": 0.46459, + "grad_norm": 0.8413155964103024, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 46459 + }, + { + "epoch": 0.4646, + "grad_norm": 0.8778009793651151, + "learning_rate": 0.003, + "loss": 3.983, + "step": 46460 + }, + { + "epoch": 0.46461, + "grad_norm": 1.0361809677162799, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 46461 + }, + { + "epoch": 0.46462, + "grad_norm": 0.9747566746300849, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 46462 + }, + { + "epoch": 0.46463, + "grad_norm": 0.8811768911275394, + "learning_rate": 0.003, + "loss": 3.982, + "step": 46463 + }, + { + "epoch": 0.46464, + "grad_norm": 0.7305207233591509, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 46464 + }, + { + "epoch": 0.46465, + "grad_norm": 0.6904009611280963, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 46465 + }, + { + "epoch": 0.46466, + "grad_norm": 0.7553858920011703, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 46466 + }, + { + "epoch": 0.46467, + "grad_norm": 0.7888568327164343, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 46467 + }, + { + "epoch": 0.46468, + "grad_norm": 0.6787983129180369, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 46468 + }, + { + "epoch": 0.46469, + "grad_norm": 0.6919923141095066, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 46469 + }, + { + "epoch": 0.4647, + "grad_norm": 0.81996605192093, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 46470 + }, + { + "epoch": 0.46471, + "grad_norm": 0.8685069173291512, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 46471 + }, + { + "epoch": 0.46472, + "grad_norm": 0.9420845876044346, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 46472 + }, + { + "epoch": 0.46473, + "grad_norm": 0.815110620308834, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 46473 + }, + { + "epoch": 0.46474, + "grad_norm": 0.7516110778039417, + "learning_rate": 0.003, + "loss": 3.987, + "step": 46474 + }, + { + "epoch": 0.46475, + "grad_norm": 0.8873650567869436, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 46475 + }, + { + "epoch": 0.46476, + "grad_norm": 0.851505750216038, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 46476 + }, + { + "epoch": 0.46477, + "grad_norm": 0.7129192754374976, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 46477 + }, + { + "epoch": 0.46478, + "grad_norm": 0.6629890618221737, + "learning_rate": 0.003, + "loss": 3.991, + "step": 46478 + }, + { + "epoch": 0.46479, + "grad_norm": 0.5720286399686928, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 46479 + }, + { + "epoch": 0.4648, + "grad_norm": 0.6785450010244412, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 46480 + }, + { + "epoch": 0.46481, + "grad_norm": 0.9015076257262047, + "learning_rate": 0.003, + "loss": 3.999, + "step": 46481 + }, + { + "epoch": 0.46482, + "grad_norm": 0.9763804023235685, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 46482 + }, + { + "epoch": 0.46483, + "grad_norm": 0.8164606455470603, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 46483 + }, + { + "epoch": 0.46484, + "grad_norm": 0.6722938527119537, + "learning_rate": 0.003, + "loss": 3.974, + "step": 46484 + }, + { + "epoch": 0.46485, + "grad_norm": 0.6080462336423563, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 46485 + }, + { + "epoch": 0.46486, + "grad_norm": 0.6195243903582683, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 46486 + }, + { + "epoch": 0.46487, + "grad_norm": 0.7072186210417766, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 46487 + }, + { + "epoch": 0.46488, + "grad_norm": 0.6633504280053322, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 46488 + }, + { + "epoch": 0.46489, + "grad_norm": 0.614018308571964, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 46489 + }, + { + "epoch": 0.4649, + "grad_norm": 0.6284612537241508, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 46490 + }, + { + "epoch": 0.46491, + "grad_norm": 0.6912478514821406, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 46491 + }, + { + "epoch": 0.46492, + "grad_norm": 0.6915631519268347, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 46492 + }, + { + "epoch": 0.46493, + "grad_norm": 0.6482641456860452, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 46493 + }, + { + "epoch": 0.46494, + "grad_norm": 0.7624619187474425, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 46494 + }, + { + "epoch": 0.46495, + "grad_norm": 0.7806573558883897, + "learning_rate": 0.003, + "loss": 3.9486, + "step": 46495 + }, + { + "epoch": 0.46496, + "grad_norm": 0.8053637875219066, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 46496 + }, + { + "epoch": 0.46497, + "grad_norm": 0.8531683367120544, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 46497 + }, + { + "epoch": 0.46498, + "grad_norm": 0.8739702816577511, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 46498 + }, + { + "epoch": 0.46499, + "grad_norm": 0.7211871431440285, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 46499 + }, + { + "epoch": 0.465, + "grad_norm": 0.662669138030735, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 46500 + }, + { + "epoch": 0.46501, + "grad_norm": 0.6389182693968946, + "learning_rate": 0.003, + "loss": 3.9399, + "step": 46501 + }, + { + "epoch": 0.46502, + "grad_norm": 0.7735774649734808, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 46502 + }, + { + "epoch": 0.46503, + "grad_norm": 0.7307938342714599, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 46503 + }, + { + "epoch": 0.46504, + "grad_norm": 0.6688834590834551, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 46504 + }, + { + "epoch": 0.46505, + "grad_norm": 0.7114141908164743, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 46505 + }, + { + "epoch": 0.46506, + "grad_norm": 0.8002558859981722, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 46506 + }, + { + "epoch": 0.46507, + "grad_norm": 0.768580481119052, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 46507 + }, + { + "epoch": 0.46508, + "grad_norm": 0.7954219899644831, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 46508 + }, + { + "epoch": 0.46509, + "grad_norm": 0.8764101869685761, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 46509 + }, + { + "epoch": 0.4651, + "grad_norm": 1.0002910315204874, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 46510 + }, + { + "epoch": 0.46511, + "grad_norm": 1.2446340300519456, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 46511 + }, + { + "epoch": 0.46512, + "grad_norm": 1.0131515414789942, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 46512 + }, + { + "epoch": 0.46513, + "grad_norm": 0.9430084535933839, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 46513 + }, + { + "epoch": 0.46514, + "grad_norm": 0.9500522411337291, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 46514 + }, + { + "epoch": 0.46515, + "grad_norm": 0.9784719879829479, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 46515 + }, + { + "epoch": 0.46516, + "grad_norm": 0.8288044707676038, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 46516 + }, + { + "epoch": 0.46517, + "grad_norm": 0.7444819741329266, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 46517 + }, + { + "epoch": 0.46518, + "grad_norm": 0.7852861128919576, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 46518 + }, + { + "epoch": 0.46519, + "grad_norm": 0.9375902743667008, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 46519 + }, + { + "epoch": 0.4652, + "grad_norm": 1.0758793160974998, + "learning_rate": 0.003, + "loss": 4.019, + "step": 46520 + }, + { + "epoch": 0.46521, + "grad_norm": 0.8805644813229665, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46521 + }, + { + "epoch": 0.46522, + "grad_norm": 0.7833232110584601, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 46522 + }, + { + "epoch": 0.46523, + "grad_norm": 0.6940326699459892, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 46523 + }, + { + "epoch": 0.46524, + "grad_norm": 0.7169429683690883, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 46524 + }, + { + "epoch": 0.46525, + "grad_norm": 0.7308922203079051, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 46525 + }, + { + "epoch": 0.46526, + "grad_norm": 0.7545592947595391, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 46526 + }, + { + "epoch": 0.46527, + "grad_norm": 0.6910751005854648, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 46527 + }, + { + "epoch": 0.46528, + "grad_norm": 0.621705787064052, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 46528 + }, + { + "epoch": 0.46529, + "grad_norm": 0.5279482411044917, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 46529 + }, + { + "epoch": 0.4653, + "grad_norm": 0.5737454110880386, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 46530 + }, + { + "epoch": 0.46531, + "grad_norm": 0.6091514826985286, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 46531 + }, + { + "epoch": 0.46532, + "grad_norm": 0.7345038559281951, + "learning_rate": 0.003, + "loss": 3.975, + "step": 46532 + }, + { + "epoch": 0.46533, + "grad_norm": 0.8056056097433238, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 46533 + }, + { + "epoch": 0.46534, + "grad_norm": 0.7858723634700038, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 46534 + }, + { + "epoch": 0.46535, + "grad_norm": 0.8818152361156659, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 46535 + }, + { + "epoch": 0.46536, + "grad_norm": 0.9708590003200621, + "learning_rate": 0.003, + "loss": 4.0281, + "step": 46536 + }, + { + "epoch": 0.46537, + "grad_norm": 0.9762375834655882, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 46537 + }, + { + "epoch": 0.46538, + "grad_norm": 0.9257653149415933, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 46538 + }, + { + "epoch": 0.46539, + "grad_norm": 0.8181554898991495, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 46539 + }, + { + "epoch": 0.4654, + "grad_norm": 0.8083695838039902, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 46540 + }, + { + "epoch": 0.46541, + "grad_norm": 0.8325801001773332, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 46541 + }, + { + "epoch": 0.46542, + "grad_norm": 0.9148680933230161, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 46542 + }, + { + "epoch": 0.46543, + "grad_norm": 0.8479519601297116, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 46543 + }, + { + "epoch": 0.46544, + "grad_norm": 0.9492826692360017, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 46544 + }, + { + "epoch": 0.46545, + "grad_norm": 0.9540422718223937, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 46545 + }, + { + "epoch": 0.46546, + "grad_norm": 0.8034742963134194, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 46546 + }, + { + "epoch": 0.46547, + "grad_norm": 0.8688221032284031, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 46547 + }, + { + "epoch": 0.46548, + "grad_norm": 0.8912800320272162, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 46548 + }, + { + "epoch": 0.46549, + "grad_norm": 0.9102593499529318, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 46549 + }, + { + "epoch": 0.4655, + "grad_norm": 1.0528239163762383, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 46550 + }, + { + "epoch": 0.46551, + "grad_norm": 1.0122853000492409, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 46551 + }, + { + "epoch": 0.46552, + "grad_norm": 1.0493706683694155, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 46552 + }, + { + "epoch": 0.46553, + "grad_norm": 0.8367330336552604, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 46553 + }, + { + "epoch": 0.46554, + "grad_norm": 0.8193650133707449, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 46554 + }, + { + "epoch": 0.46555, + "grad_norm": 0.8672125483186375, + "learning_rate": 0.003, + "loss": 4.009, + "step": 46555 + }, + { + "epoch": 0.46556, + "grad_norm": 0.7722292523183327, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 46556 + }, + { + "epoch": 0.46557, + "grad_norm": 0.7090677261811024, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 46557 + }, + { + "epoch": 0.46558, + "grad_norm": 0.718211458641562, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 46558 + }, + { + "epoch": 0.46559, + "grad_norm": 0.6708061909909895, + "learning_rate": 0.003, + "loss": 3.98, + "step": 46559 + }, + { + "epoch": 0.4656, + "grad_norm": 0.6499605717018847, + "learning_rate": 0.003, + "loss": 3.985, + "step": 46560 + }, + { + "epoch": 0.46561, + "grad_norm": 0.7014978691713903, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 46561 + }, + { + "epoch": 0.46562, + "grad_norm": 0.6916293674346055, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 46562 + }, + { + "epoch": 0.46563, + "grad_norm": 0.7480185214196544, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 46563 + }, + { + "epoch": 0.46564, + "grad_norm": 0.7339068956345262, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 46564 + }, + { + "epoch": 0.46565, + "grad_norm": 0.7102357454737236, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 46565 + }, + { + "epoch": 0.46566, + "grad_norm": 0.6027149897519993, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 46566 + }, + { + "epoch": 0.46567, + "grad_norm": 0.5447758893888874, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 46567 + }, + { + "epoch": 0.46568, + "grad_norm": 0.602136526415101, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 46568 + }, + { + "epoch": 0.46569, + "grad_norm": 0.5374944144951567, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 46569 + }, + { + "epoch": 0.4657, + "grad_norm": 0.5005561984147123, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 46570 + }, + { + "epoch": 0.46571, + "grad_norm": 0.5400812852432911, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 46571 + }, + { + "epoch": 0.46572, + "grad_norm": 0.6557173734004227, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 46572 + }, + { + "epoch": 0.46573, + "grad_norm": 0.6744887571374, + "learning_rate": 0.003, + "loss": 3.9369, + "step": 46573 + }, + { + "epoch": 0.46574, + "grad_norm": 0.647319073432573, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 46574 + }, + { + "epoch": 0.46575, + "grad_norm": 0.7622923486104772, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 46575 + }, + { + "epoch": 0.46576, + "grad_norm": 0.9124721446459043, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 46576 + }, + { + "epoch": 0.46577, + "grad_norm": 0.9200844578685616, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 46577 + }, + { + "epoch": 0.46578, + "grad_norm": 0.9108353057125842, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 46578 + }, + { + "epoch": 0.46579, + "grad_norm": 0.9949335370929286, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 46579 + }, + { + "epoch": 0.4658, + "grad_norm": 1.0565975891990167, + "learning_rate": 0.003, + "loss": 3.971, + "step": 46580 + }, + { + "epoch": 0.46581, + "grad_norm": 0.7949565649007602, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 46581 + }, + { + "epoch": 0.46582, + "grad_norm": 0.7907341693692125, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 46582 + }, + { + "epoch": 0.46583, + "grad_norm": 0.7551237799712205, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 46583 + }, + { + "epoch": 0.46584, + "grad_norm": 0.8495654302248551, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 46584 + }, + { + "epoch": 0.46585, + "grad_norm": 0.9077162114255208, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 46585 + }, + { + "epoch": 0.46586, + "grad_norm": 0.8482594445458578, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 46586 + }, + { + "epoch": 0.46587, + "grad_norm": 0.8053458712332797, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 46587 + }, + { + "epoch": 0.46588, + "grad_norm": 0.6971700601937412, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 46588 + }, + { + "epoch": 0.46589, + "grad_norm": 0.6800578958401152, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 46589 + }, + { + "epoch": 0.4659, + "grad_norm": 0.6234365177032627, + "learning_rate": 0.003, + "loss": 3.977, + "step": 46590 + }, + { + "epoch": 0.46591, + "grad_norm": 0.6870069733833603, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 46591 + }, + { + "epoch": 0.46592, + "grad_norm": 0.7791831525397065, + "learning_rate": 0.003, + "loss": 3.952, + "step": 46592 + }, + { + "epoch": 0.46593, + "grad_norm": 0.794495350669247, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 46593 + }, + { + "epoch": 0.46594, + "grad_norm": 0.8357790496245853, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 46594 + }, + { + "epoch": 0.46595, + "grad_norm": 0.7609346573822696, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 46595 + }, + { + "epoch": 0.46596, + "grad_norm": 0.7817708218421041, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 46596 + }, + { + "epoch": 0.46597, + "grad_norm": 0.8785028741020913, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 46597 + }, + { + "epoch": 0.46598, + "grad_norm": 0.9255496810374985, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 46598 + }, + { + "epoch": 0.46599, + "grad_norm": 1.101983883174918, + "learning_rate": 0.003, + "loss": 3.965, + "step": 46599 + }, + { + "epoch": 0.466, + "grad_norm": 0.8165799378483372, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 46600 + }, + { + "epoch": 0.46601, + "grad_norm": 0.7549449775581419, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 46601 + }, + { + "epoch": 0.46602, + "grad_norm": 0.7308591798678318, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 46602 + }, + { + "epoch": 0.46603, + "grad_norm": 0.7488843196774412, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 46603 + }, + { + "epoch": 0.46604, + "grad_norm": 0.757140417159377, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 46604 + }, + { + "epoch": 0.46605, + "grad_norm": 0.7487711991278502, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 46605 + }, + { + "epoch": 0.46606, + "grad_norm": 0.8157033294358406, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 46606 + }, + { + "epoch": 0.46607, + "grad_norm": 0.8133243660720515, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 46607 + }, + { + "epoch": 0.46608, + "grad_norm": 0.7749737654714256, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 46608 + }, + { + "epoch": 0.46609, + "grad_norm": 0.7186452305624756, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 46609 + }, + { + "epoch": 0.4661, + "grad_norm": 0.6983386563072892, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 46610 + }, + { + "epoch": 0.46611, + "grad_norm": 0.7587231402740232, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46611 + }, + { + "epoch": 0.46612, + "grad_norm": 0.7999690243265566, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 46612 + }, + { + "epoch": 0.46613, + "grad_norm": 0.894491704490435, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 46613 + }, + { + "epoch": 0.46614, + "grad_norm": 1.024372517731427, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 46614 + }, + { + "epoch": 0.46615, + "grad_norm": 0.8851346224265155, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 46615 + }, + { + "epoch": 0.46616, + "grad_norm": 0.8267289613552107, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 46616 + }, + { + "epoch": 0.46617, + "grad_norm": 0.8051084739519568, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 46617 + }, + { + "epoch": 0.46618, + "grad_norm": 0.8198397417628511, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 46618 + }, + { + "epoch": 0.46619, + "grad_norm": 0.8243073832409293, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 46619 + }, + { + "epoch": 0.4662, + "grad_norm": 0.8156436047370904, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 46620 + }, + { + "epoch": 0.46621, + "grad_norm": 0.8567215221496167, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 46621 + }, + { + "epoch": 0.46622, + "grad_norm": 0.8781070422427459, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 46622 + }, + { + "epoch": 0.46623, + "grad_norm": 0.8887464049434299, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 46623 + }, + { + "epoch": 0.46624, + "grad_norm": 0.8701206214068196, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 46624 + }, + { + "epoch": 0.46625, + "grad_norm": 0.9119787279690154, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 46625 + }, + { + "epoch": 0.46626, + "grad_norm": 0.9224153974056739, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 46626 + }, + { + "epoch": 0.46627, + "grad_norm": 0.9024439682263719, + "learning_rate": 0.003, + "loss": 4.002, + "step": 46627 + }, + { + "epoch": 0.46628, + "grad_norm": 0.8489112937407901, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 46628 + }, + { + "epoch": 0.46629, + "grad_norm": 0.8113720830281596, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 46629 + }, + { + "epoch": 0.4663, + "grad_norm": 0.7751362000247697, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 46630 + }, + { + "epoch": 0.46631, + "grad_norm": 0.8036215072970075, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 46631 + }, + { + "epoch": 0.46632, + "grad_norm": 0.7233192407966437, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 46632 + }, + { + "epoch": 0.46633, + "grad_norm": 0.658157726214627, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 46633 + }, + { + "epoch": 0.46634, + "grad_norm": 0.6202875242965311, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 46634 + }, + { + "epoch": 0.46635, + "grad_norm": 0.6100520771208217, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 46635 + }, + { + "epoch": 0.46636, + "grad_norm": 0.5962725374975687, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 46636 + }, + { + "epoch": 0.46637, + "grad_norm": 0.5718933224701519, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 46637 + }, + { + "epoch": 0.46638, + "grad_norm": 0.6460715921956364, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 46638 + }, + { + "epoch": 0.46639, + "grad_norm": 0.7760145472630556, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 46639 + }, + { + "epoch": 0.4664, + "grad_norm": 1.0388310459991208, + "learning_rate": 0.003, + "loss": 4.012, + "step": 46640 + }, + { + "epoch": 0.46641, + "grad_norm": 1.0508557277154005, + "learning_rate": 0.003, + "loss": 3.974, + "step": 46641 + }, + { + "epoch": 0.46642, + "grad_norm": 0.807016445142203, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 46642 + }, + { + "epoch": 0.46643, + "grad_norm": 0.5816840444834572, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 46643 + }, + { + "epoch": 0.46644, + "grad_norm": 0.7039528455342585, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 46644 + }, + { + "epoch": 0.46645, + "grad_norm": 0.7988020768228793, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 46645 + }, + { + "epoch": 0.46646, + "grad_norm": 0.8390886780175661, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 46646 + }, + { + "epoch": 0.46647, + "grad_norm": 0.7349208187319899, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 46647 + }, + { + "epoch": 0.46648, + "grad_norm": 0.6901076198106698, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 46648 + }, + { + "epoch": 0.46649, + "grad_norm": 0.648901508058625, + "learning_rate": 0.003, + "loss": 3.987, + "step": 46649 + }, + { + "epoch": 0.4665, + "grad_norm": 0.6781968537839146, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 46650 + }, + { + "epoch": 0.46651, + "grad_norm": 0.6819521515068233, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 46651 + }, + { + "epoch": 0.46652, + "grad_norm": 0.7727747533640723, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 46652 + }, + { + "epoch": 0.46653, + "grad_norm": 0.8429977804555789, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 46653 + }, + { + "epoch": 0.46654, + "grad_norm": 0.8434748842791501, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 46654 + }, + { + "epoch": 0.46655, + "grad_norm": 0.8434947676331576, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 46655 + }, + { + "epoch": 0.46656, + "grad_norm": 0.8597580940631795, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 46656 + }, + { + "epoch": 0.46657, + "grad_norm": 0.814131301509419, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 46657 + }, + { + "epoch": 0.46658, + "grad_norm": 0.7584671137279733, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 46658 + }, + { + "epoch": 0.46659, + "grad_norm": 0.7885180920514531, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 46659 + }, + { + "epoch": 0.4666, + "grad_norm": 0.7398463693388758, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 46660 + }, + { + "epoch": 0.46661, + "grad_norm": 0.7792432748543168, + "learning_rate": 0.003, + "loss": 3.995, + "step": 46661 + }, + { + "epoch": 0.46662, + "grad_norm": 0.9010074101249669, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 46662 + }, + { + "epoch": 0.46663, + "grad_norm": 1.0586517457287887, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 46663 + }, + { + "epoch": 0.46664, + "grad_norm": 0.9408822140525552, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 46664 + }, + { + "epoch": 0.46665, + "grad_norm": 0.9681992740850253, + "learning_rate": 0.003, + "loss": 4.012, + "step": 46665 + }, + { + "epoch": 0.46666, + "grad_norm": 0.9600936498352444, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 46666 + }, + { + "epoch": 0.46667, + "grad_norm": 0.8953255317281751, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 46667 + }, + { + "epoch": 0.46668, + "grad_norm": 0.9221335068119045, + "learning_rate": 0.003, + "loss": 4.017, + "step": 46668 + }, + { + "epoch": 0.46669, + "grad_norm": 0.9485963041064666, + "learning_rate": 0.003, + "loss": 3.971, + "step": 46669 + }, + { + "epoch": 0.4667, + "grad_norm": 0.8913240123158085, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 46670 + }, + { + "epoch": 0.46671, + "grad_norm": 0.9476151771836543, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 46671 + }, + { + "epoch": 0.46672, + "grad_norm": 0.895384688654582, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 46672 + }, + { + "epoch": 0.46673, + "grad_norm": 0.7220838838077761, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 46673 + }, + { + "epoch": 0.46674, + "grad_norm": 0.6330499644969994, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 46674 + }, + { + "epoch": 0.46675, + "grad_norm": 0.7980642627243162, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 46675 + }, + { + "epoch": 0.46676, + "grad_norm": 1.0037036512615742, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 46676 + }, + { + "epoch": 0.46677, + "grad_norm": 1.0059401991373196, + "learning_rate": 0.003, + "loss": 3.966, + "step": 46677 + }, + { + "epoch": 0.46678, + "grad_norm": 0.9074952995850377, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 46678 + }, + { + "epoch": 0.46679, + "grad_norm": 0.7780261732267572, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 46679 + }, + { + "epoch": 0.4668, + "grad_norm": 0.7275300052180578, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 46680 + }, + { + "epoch": 0.46681, + "grad_norm": 0.6686863165165987, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 46681 + }, + { + "epoch": 0.46682, + "grad_norm": 0.5300696506161525, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 46682 + }, + { + "epoch": 0.46683, + "grad_norm": 0.48997046414212836, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 46683 + }, + { + "epoch": 0.46684, + "grad_norm": 0.45743552132303067, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 46684 + }, + { + "epoch": 0.46685, + "grad_norm": 0.4889811139962784, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 46685 + }, + { + "epoch": 0.46686, + "grad_norm": 0.4680025121254483, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 46686 + }, + { + "epoch": 0.46687, + "grad_norm": 0.4266939230649972, + "learning_rate": 0.003, + "loss": 3.9416, + "step": 46687 + }, + { + "epoch": 0.46688, + "grad_norm": 0.4256100031519534, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 46688 + }, + { + "epoch": 0.46689, + "grad_norm": 0.4619331061408679, + "learning_rate": 0.003, + "loss": 3.966, + "step": 46689 + }, + { + "epoch": 0.4669, + "grad_norm": 0.47204047801163496, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 46690 + }, + { + "epoch": 0.46691, + "grad_norm": 0.5219208865824942, + "learning_rate": 0.003, + "loss": 3.972, + "step": 46691 + }, + { + "epoch": 0.46692, + "grad_norm": 0.6478039332544222, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 46692 + }, + { + "epoch": 0.46693, + "grad_norm": 0.7228882546548213, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 46693 + }, + { + "epoch": 0.46694, + "grad_norm": 0.7403111169450444, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 46694 + }, + { + "epoch": 0.46695, + "grad_norm": 0.872102971595808, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 46695 + }, + { + "epoch": 0.46696, + "grad_norm": 1.1291925162378365, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 46696 + }, + { + "epoch": 0.46697, + "grad_norm": 0.7995952843530204, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 46697 + }, + { + "epoch": 0.46698, + "grad_norm": 0.699137429719369, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 46698 + }, + { + "epoch": 0.46699, + "grad_norm": 0.7874435592083444, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 46699 + }, + { + "epoch": 0.467, + "grad_norm": 0.8960190660337644, + "learning_rate": 0.003, + "loss": 3.961, + "step": 46700 + }, + { + "epoch": 0.46701, + "grad_norm": 0.8502487104154317, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 46701 + }, + { + "epoch": 0.46702, + "grad_norm": 0.8834150842030594, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 46702 + }, + { + "epoch": 0.46703, + "grad_norm": 0.9150602801405903, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 46703 + }, + { + "epoch": 0.46704, + "grad_norm": 0.9777041761940443, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 46704 + }, + { + "epoch": 0.46705, + "grad_norm": 1.1335445964043591, + "learning_rate": 0.003, + "loss": 4.005, + "step": 46705 + }, + { + "epoch": 0.46706, + "grad_norm": 0.9522345780262294, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 46706 + }, + { + "epoch": 0.46707, + "grad_norm": 0.8008617955290828, + "learning_rate": 0.003, + "loss": 3.9361, + "step": 46707 + }, + { + "epoch": 0.46708, + "grad_norm": 0.8358837320789477, + "learning_rate": 0.003, + "loss": 3.968, + "step": 46708 + }, + { + "epoch": 0.46709, + "grad_norm": 0.8559713096611573, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 46709 + }, + { + "epoch": 0.4671, + "grad_norm": 0.9818321714095293, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 46710 + }, + { + "epoch": 0.46711, + "grad_norm": 1.02276075841639, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 46711 + }, + { + "epoch": 0.46712, + "grad_norm": 0.8210384634715571, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 46712 + }, + { + "epoch": 0.46713, + "grad_norm": 0.8076201878829108, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46713 + }, + { + "epoch": 0.46714, + "grad_norm": 0.8435632610526991, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 46714 + }, + { + "epoch": 0.46715, + "grad_norm": 0.754846214822983, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 46715 + }, + { + "epoch": 0.46716, + "grad_norm": 0.7866159270859169, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 46716 + }, + { + "epoch": 0.46717, + "grad_norm": 0.8206597470089844, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 46717 + }, + { + "epoch": 0.46718, + "grad_norm": 0.8990824093893356, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 46718 + }, + { + "epoch": 0.46719, + "grad_norm": 0.9610948519708007, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 46719 + }, + { + "epoch": 0.4672, + "grad_norm": 1.0530854151253266, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 46720 + }, + { + "epoch": 0.46721, + "grad_norm": 0.9311107512801475, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 46721 + }, + { + "epoch": 0.46722, + "grad_norm": 0.8562944466764287, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 46722 + }, + { + "epoch": 0.46723, + "grad_norm": 0.8398388187620583, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 46723 + }, + { + "epoch": 0.46724, + "grad_norm": 0.8266463390032696, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46724 + }, + { + "epoch": 0.46725, + "grad_norm": 0.7707350070242224, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 46725 + }, + { + "epoch": 0.46726, + "grad_norm": 0.7504714705828304, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 46726 + }, + { + "epoch": 0.46727, + "grad_norm": 0.8383135880031857, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 46727 + }, + { + "epoch": 0.46728, + "grad_norm": 0.82875956701226, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 46728 + }, + { + "epoch": 0.46729, + "grad_norm": 0.7251482361917533, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 46729 + }, + { + "epoch": 0.4673, + "grad_norm": 0.5987507216582056, + "learning_rate": 0.003, + "loss": 3.987, + "step": 46730 + }, + { + "epoch": 0.46731, + "grad_norm": 0.5199442642174402, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 46731 + }, + { + "epoch": 0.46732, + "grad_norm": 0.5905462619895276, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46732 + }, + { + "epoch": 0.46733, + "grad_norm": 0.675199635985549, + "learning_rate": 0.003, + "loss": 3.977, + "step": 46733 + }, + { + "epoch": 0.46734, + "grad_norm": 0.7003329311008752, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 46734 + }, + { + "epoch": 0.46735, + "grad_norm": 0.6797032433521768, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 46735 + }, + { + "epoch": 0.46736, + "grad_norm": 0.5465897724436884, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 46736 + }, + { + "epoch": 0.46737, + "grad_norm": 0.42855599336342476, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 46737 + }, + { + "epoch": 0.46738, + "grad_norm": 0.5484203277918834, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 46738 + }, + { + "epoch": 0.46739, + "grad_norm": 0.5967586786764493, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 46739 + }, + { + "epoch": 0.4674, + "grad_norm": 0.6525485111205472, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 46740 + }, + { + "epoch": 0.46741, + "grad_norm": 0.696645959020276, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 46741 + }, + { + "epoch": 0.46742, + "grad_norm": 0.6786980038022649, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 46742 + }, + { + "epoch": 0.46743, + "grad_norm": 0.534295897592977, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 46743 + }, + { + "epoch": 0.46744, + "grad_norm": 0.5372235545830454, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 46744 + }, + { + "epoch": 0.46745, + "grad_norm": 0.5622174363697647, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 46745 + }, + { + "epoch": 0.46746, + "grad_norm": 0.646492028510629, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 46746 + }, + { + "epoch": 0.46747, + "grad_norm": 0.7272710191728204, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 46747 + }, + { + "epoch": 0.46748, + "grad_norm": 0.8190308488470613, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 46748 + }, + { + "epoch": 0.46749, + "grad_norm": 0.9583450309322215, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 46749 + }, + { + "epoch": 0.4675, + "grad_norm": 1.136133136621964, + "learning_rate": 0.003, + "loss": 3.969, + "step": 46750 + }, + { + "epoch": 0.46751, + "grad_norm": 0.8126240670722678, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 46751 + }, + { + "epoch": 0.46752, + "grad_norm": 0.6494493675326036, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 46752 + }, + { + "epoch": 0.46753, + "grad_norm": 0.6522639367605269, + "learning_rate": 0.003, + "loss": 3.959, + "step": 46753 + }, + { + "epoch": 0.46754, + "grad_norm": 0.6947406839844467, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 46754 + }, + { + "epoch": 0.46755, + "grad_norm": 0.744482160377522, + "learning_rate": 0.003, + "loss": 3.953, + "step": 46755 + }, + { + "epoch": 0.46756, + "grad_norm": 0.9527089450860022, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 46756 + }, + { + "epoch": 0.46757, + "grad_norm": 0.9977540781096232, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 46757 + }, + { + "epoch": 0.46758, + "grad_norm": 0.8737726971370813, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 46758 + }, + { + "epoch": 0.46759, + "grad_norm": 0.8081561573403189, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 46759 + }, + { + "epoch": 0.4676, + "grad_norm": 0.8480576563364727, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 46760 + }, + { + "epoch": 0.46761, + "grad_norm": 0.8272620687596196, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 46761 + }, + { + "epoch": 0.46762, + "grad_norm": 0.7927468791813252, + "learning_rate": 0.003, + "loss": 3.994, + "step": 46762 + }, + { + "epoch": 0.46763, + "grad_norm": 0.8054786745773272, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 46763 + }, + { + "epoch": 0.46764, + "grad_norm": 0.765105455018216, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 46764 + }, + { + "epoch": 0.46765, + "grad_norm": 0.7192264513058666, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 46765 + }, + { + "epoch": 0.46766, + "grad_norm": 0.7092195844222775, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 46766 + }, + { + "epoch": 0.46767, + "grad_norm": 0.836548823323464, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 46767 + }, + { + "epoch": 0.46768, + "grad_norm": 0.8333613055868141, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 46768 + }, + { + "epoch": 0.46769, + "grad_norm": 0.9720356074516051, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 46769 + }, + { + "epoch": 0.4677, + "grad_norm": 0.9294552231633649, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 46770 + }, + { + "epoch": 0.46771, + "grad_norm": 0.949362518583993, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 46771 + }, + { + "epoch": 0.46772, + "grad_norm": 1.0650492062539987, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 46772 + }, + { + "epoch": 0.46773, + "grad_norm": 0.9331620452061177, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 46773 + }, + { + "epoch": 0.46774, + "grad_norm": 0.9370115393826328, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 46774 + }, + { + "epoch": 0.46775, + "grad_norm": 0.9051234778601319, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 46775 + }, + { + "epoch": 0.46776, + "grad_norm": 0.9827772739997279, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 46776 + }, + { + "epoch": 0.46777, + "grad_norm": 1.2066934983041515, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 46777 + }, + { + "epoch": 0.46778, + "grad_norm": 0.6996962097504326, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 46778 + }, + { + "epoch": 0.46779, + "grad_norm": 0.6036977326287093, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 46779 + }, + { + "epoch": 0.4678, + "grad_norm": 0.6089728835689733, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 46780 + }, + { + "epoch": 0.46781, + "grad_norm": 0.7170817087345145, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 46781 + }, + { + "epoch": 0.46782, + "grad_norm": 0.7269736496580339, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 46782 + }, + { + "epoch": 0.46783, + "grad_norm": 0.7964941955764615, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 46783 + }, + { + "epoch": 0.46784, + "grad_norm": 0.8047968366197146, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 46784 + }, + { + "epoch": 0.46785, + "grad_norm": 0.8943158022118403, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 46785 + }, + { + "epoch": 0.46786, + "grad_norm": 1.0285626536657368, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 46786 + }, + { + "epoch": 0.46787, + "grad_norm": 1.0472740053908678, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 46787 + }, + { + "epoch": 0.46788, + "grad_norm": 0.9205616084224406, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 46788 + }, + { + "epoch": 0.46789, + "grad_norm": 0.8361664200197041, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 46789 + }, + { + "epoch": 0.4679, + "grad_norm": 0.8007008342715514, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 46790 + }, + { + "epoch": 0.46791, + "grad_norm": 0.7747265959939471, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 46791 + }, + { + "epoch": 0.46792, + "grad_norm": 0.7299269507721875, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 46792 + }, + { + "epoch": 0.46793, + "grad_norm": 0.6610684617954945, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 46793 + }, + { + "epoch": 0.46794, + "grad_norm": 0.7174588762277474, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 46794 + }, + { + "epoch": 0.46795, + "grad_norm": 0.7319761393409604, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 46795 + }, + { + "epoch": 0.46796, + "grad_norm": 0.6912631867325196, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 46796 + }, + { + "epoch": 0.46797, + "grad_norm": 0.6572372576068534, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 46797 + }, + { + "epoch": 0.46798, + "grad_norm": 0.5702054817096465, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 46798 + }, + { + "epoch": 0.46799, + "grad_norm": 0.6026675417844238, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 46799 + }, + { + "epoch": 0.468, + "grad_norm": 0.5892325576771484, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 46800 + }, + { + "epoch": 0.46801, + "grad_norm": 0.5649989144220903, + "learning_rate": 0.003, + "loss": 3.981, + "step": 46801 + }, + { + "epoch": 0.46802, + "grad_norm": 0.5590159736299799, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 46802 + }, + { + "epoch": 0.46803, + "grad_norm": 0.5575845380010044, + "learning_rate": 0.003, + "loss": 3.967, + "step": 46803 + }, + { + "epoch": 0.46804, + "grad_norm": 0.5246798704284585, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 46804 + }, + { + "epoch": 0.46805, + "grad_norm": 0.5127546216438476, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 46805 + }, + { + "epoch": 0.46806, + "grad_norm": 0.5766740203836743, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 46806 + }, + { + "epoch": 0.46807, + "grad_norm": 0.5944535415805664, + "learning_rate": 0.003, + "loss": 3.9474, + "step": 46807 + }, + { + "epoch": 0.46808, + "grad_norm": 0.5996820288161155, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 46808 + }, + { + "epoch": 0.46809, + "grad_norm": 0.6135298303259986, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 46809 + }, + { + "epoch": 0.4681, + "grad_norm": 0.7472909605594333, + "learning_rate": 0.003, + "loss": 3.949, + "step": 46810 + }, + { + "epoch": 0.46811, + "grad_norm": 0.8563074875748485, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 46811 + }, + { + "epoch": 0.46812, + "grad_norm": 0.9436360770620852, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 46812 + }, + { + "epoch": 0.46813, + "grad_norm": 1.0282097163086525, + "learning_rate": 0.003, + "loss": 3.986, + "step": 46813 + }, + { + "epoch": 0.46814, + "grad_norm": 1.0821864971604245, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 46814 + }, + { + "epoch": 0.46815, + "grad_norm": 0.7770056769045659, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 46815 + }, + { + "epoch": 0.46816, + "grad_norm": 0.8598038245871467, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 46816 + }, + { + "epoch": 0.46817, + "grad_norm": 0.8750181575353908, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 46817 + }, + { + "epoch": 0.46818, + "grad_norm": 0.8441792120273808, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 46818 + }, + { + "epoch": 0.46819, + "grad_norm": 0.9298358885111744, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 46819 + }, + { + "epoch": 0.4682, + "grad_norm": 1.0860756523540502, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 46820 + }, + { + "epoch": 0.46821, + "grad_norm": 1.0041455155019101, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 46821 + }, + { + "epoch": 0.46822, + "grad_norm": 0.8168816815901089, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 46822 + }, + { + "epoch": 0.46823, + "grad_norm": 0.8198496168928685, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 46823 + }, + { + "epoch": 0.46824, + "grad_norm": 0.802124592337894, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 46824 + }, + { + "epoch": 0.46825, + "grad_norm": 0.8010692867769369, + "learning_rate": 0.003, + "loss": 4.004, + "step": 46825 + }, + { + "epoch": 0.46826, + "grad_norm": 0.8690917492831143, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 46826 + }, + { + "epoch": 0.46827, + "grad_norm": 0.9147756605684808, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 46827 + }, + { + "epoch": 0.46828, + "grad_norm": 0.9371810645642488, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 46828 + }, + { + "epoch": 0.46829, + "grad_norm": 0.9428277975582163, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 46829 + }, + { + "epoch": 0.4683, + "grad_norm": 0.8665662611639066, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 46830 + }, + { + "epoch": 0.46831, + "grad_norm": 0.794975058242521, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 46831 + }, + { + "epoch": 0.46832, + "grad_norm": 0.7722044629536422, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 46832 + }, + { + "epoch": 0.46833, + "grad_norm": 0.8085277299192872, + "learning_rate": 0.003, + "loss": 4.0148, + "step": 46833 + }, + { + "epoch": 0.46834, + "grad_norm": 0.9155482439816348, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 46834 + }, + { + "epoch": 0.46835, + "grad_norm": 0.8778298460629649, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 46835 + }, + { + "epoch": 0.46836, + "grad_norm": 0.8968013017827069, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 46836 + }, + { + "epoch": 0.46837, + "grad_norm": 0.8489564926675333, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 46837 + }, + { + "epoch": 0.46838, + "grad_norm": 0.7897121097359009, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 46838 + }, + { + "epoch": 0.46839, + "grad_norm": 0.8097944754510563, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 46839 + }, + { + "epoch": 0.4684, + "grad_norm": 0.8277557114846429, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 46840 + }, + { + "epoch": 0.46841, + "grad_norm": 0.9576223787821765, + "learning_rate": 0.003, + "loss": 3.976, + "step": 46841 + }, + { + "epoch": 0.46842, + "grad_norm": 1.038229254768518, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 46842 + }, + { + "epoch": 0.46843, + "grad_norm": 0.8601091978295222, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 46843 + }, + { + "epoch": 0.46844, + "grad_norm": 0.7095096763432996, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 46844 + }, + { + "epoch": 0.46845, + "grad_norm": 0.6240410453446861, + "learning_rate": 0.003, + "loss": 3.99, + "step": 46845 + }, + { + "epoch": 0.46846, + "grad_norm": 0.7308900711219597, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 46846 + }, + { + "epoch": 0.46847, + "grad_norm": 0.8930871900220579, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 46847 + }, + { + "epoch": 0.46848, + "grad_norm": 0.8760755191867275, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 46848 + }, + { + "epoch": 0.46849, + "grad_norm": 0.8170533852547909, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 46849 + }, + { + "epoch": 0.4685, + "grad_norm": 0.6736578641020697, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 46850 + }, + { + "epoch": 0.46851, + "grad_norm": 0.6036632130869116, + "learning_rate": 0.003, + "loss": 4.016, + "step": 46851 + }, + { + "epoch": 0.46852, + "grad_norm": 0.6172621229272017, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 46852 + }, + { + "epoch": 0.46853, + "grad_norm": 0.6648760462953841, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 46853 + }, + { + "epoch": 0.46854, + "grad_norm": 0.6621437623824559, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 46854 + }, + { + "epoch": 0.46855, + "grad_norm": 0.6839294327707499, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 46855 + }, + { + "epoch": 0.46856, + "grad_norm": 0.694187900796603, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 46856 + }, + { + "epoch": 0.46857, + "grad_norm": 0.6005774867292824, + "learning_rate": 0.003, + "loss": 3.999, + "step": 46857 + }, + { + "epoch": 0.46858, + "grad_norm": 0.5830288449251663, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 46858 + }, + { + "epoch": 0.46859, + "grad_norm": 0.5953359781039975, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 46859 + }, + { + "epoch": 0.4686, + "grad_norm": 0.636286985624577, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 46860 + }, + { + "epoch": 0.46861, + "grad_norm": 0.8860694373196301, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 46861 + }, + { + "epoch": 0.46862, + "grad_norm": 1.0879476572401892, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 46862 + }, + { + "epoch": 0.46863, + "grad_norm": 0.9542194804888936, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 46863 + }, + { + "epoch": 0.46864, + "grad_norm": 0.8650580971122754, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 46864 + }, + { + "epoch": 0.46865, + "grad_norm": 0.822756577472074, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 46865 + }, + { + "epoch": 0.46866, + "grad_norm": 0.6847637521969462, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 46866 + }, + { + "epoch": 0.46867, + "grad_norm": 0.6488069361593791, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 46867 + }, + { + "epoch": 0.46868, + "grad_norm": 0.6063441827247688, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 46868 + }, + { + "epoch": 0.46869, + "grad_norm": 0.611563139090343, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 46869 + }, + { + "epoch": 0.4687, + "grad_norm": 0.7097131587536799, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 46870 + }, + { + "epoch": 0.46871, + "grad_norm": 0.8011316985288159, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 46871 + }, + { + "epoch": 0.46872, + "grad_norm": 0.9107390185366074, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 46872 + }, + { + "epoch": 0.46873, + "grad_norm": 0.834037134136384, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 46873 + }, + { + "epoch": 0.46874, + "grad_norm": 0.7449452381493714, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 46874 + }, + { + "epoch": 0.46875, + "grad_norm": 0.737202996086516, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 46875 + }, + { + "epoch": 0.46876, + "grad_norm": 0.7996261228952046, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 46876 + }, + { + "epoch": 0.46877, + "grad_norm": 0.8593662344551095, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 46877 + }, + { + "epoch": 0.46878, + "grad_norm": 0.8287914156152979, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 46878 + }, + { + "epoch": 0.46879, + "grad_norm": 0.7960220339526677, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 46879 + }, + { + "epoch": 0.4688, + "grad_norm": 0.9155911908649867, + "learning_rate": 0.003, + "loss": 3.977, + "step": 46880 + }, + { + "epoch": 0.46881, + "grad_norm": 1.0065041641077153, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 46881 + }, + { + "epoch": 0.46882, + "grad_norm": 0.9464315303202064, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 46882 + }, + { + "epoch": 0.46883, + "grad_norm": 0.9092083530658578, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 46883 + }, + { + "epoch": 0.46884, + "grad_norm": 1.071783456956112, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 46884 + }, + { + "epoch": 0.46885, + "grad_norm": 1.0732278671312212, + "learning_rate": 0.003, + "loss": 4.0342, + "step": 46885 + }, + { + "epoch": 0.46886, + "grad_norm": 0.8775191639557538, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 46886 + }, + { + "epoch": 0.46887, + "grad_norm": 0.7693380086909187, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 46887 + }, + { + "epoch": 0.46888, + "grad_norm": 0.7184657985635258, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 46888 + }, + { + "epoch": 0.46889, + "grad_norm": 0.6435790567575349, + "learning_rate": 0.003, + "loss": 4.0227, + "step": 46889 + }, + { + "epoch": 0.4689, + "grad_norm": 0.7397505347023045, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 46890 + }, + { + "epoch": 0.46891, + "grad_norm": 0.8297521960080787, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 46891 + }, + { + "epoch": 0.46892, + "grad_norm": 0.9083789738172482, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 46892 + }, + { + "epoch": 0.46893, + "grad_norm": 0.888365590872734, + "learning_rate": 0.003, + "loss": 3.985, + "step": 46893 + }, + { + "epoch": 0.46894, + "grad_norm": 0.9302804375874723, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 46894 + }, + { + "epoch": 0.46895, + "grad_norm": 0.9394106232367749, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 46895 + }, + { + "epoch": 0.46896, + "grad_norm": 0.8903335643046912, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 46896 + }, + { + "epoch": 0.46897, + "grad_norm": 0.8561052285329004, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 46897 + }, + { + "epoch": 0.46898, + "grad_norm": 0.7602884854214625, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 46898 + }, + { + "epoch": 0.46899, + "grad_norm": 0.8096251235032806, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 46899 + }, + { + "epoch": 0.469, + "grad_norm": 0.9063654999401527, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 46900 + }, + { + "epoch": 0.46901, + "grad_norm": 1.0029771759005035, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 46901 + }, + { + "epoch": 0.46902, + "grad_norm": 1.0566790413406546, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 46902 + }, + { + "epoch": 0.46903, + "grad_norm": 0.830879472063994, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 46903 + }, + { + "epoch": 0.46904, + "grad_norm": 0.6436077425847695, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 46904 + }, + { + "epoch": 0.46905, + "grad_norm": 0.6570983002088067, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 46905 + }, + { + "epoch": 0.46906, + "grad_norm": 0.598747062441732, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 46906 + }, + { + "epoch": 0.46907, + "grad_norm": 0.5475571266402862, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 46907 + }, + { + "epoch": 0.46908, + "grad_norm": 0.4750542632311205, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 46908 + }, + { + "epoch": 0.46909, + "grad_norm": 0.5231527930274049, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 46909 + }, + { + "epoch": 0.4691, + "grad_norm": 0.5614371278002933, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 46910 + }, + { + "epoch": 0.46911, + "grad_norm": 0.624550921393328, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 46911 + }, + { + "epoch": 0.46912, + "grad_norm": 0.6678993328059725, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 46912 + }, + { + "epoch": 0.46913, + "grad_norm": 0.6330434200766246, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 46913 + }, + { + "epoch": 0.46914, + "grad_norm": 0.592371742183363, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 46914 + }, + { + "epoch": 0.46915, + "grad_norm": 0.6001910408393349, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 46915 + }, + { + "epoch": 0.46916, + "grad_norm": 0.5498405433449401, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 46916 + }, + { + "epoch": 0.46917, + "grad_norm": 0.5562345446037505, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 46917 + }, + { + "epoch": 0.46918, + "grad_norm": 0.5862455091365936, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 46918 + }, + { + "epoch": 0.46919, + "grad_norm": 0.6066189735815005, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 46919 + }, + { + "epoch": 0.4692, + "grad_norm": 0.5513609141096899, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 46920 + }, + { + "epoch": 0.46921, + "grad_norm": 0.515532272839453, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 46921 + }, + { + "epoch": 0.46922, + "grad_norm": 0.5862155357290344, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 46922 + }, + { + "epoch": 0.46923, + "grad_norm": 0.6223765568003554, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 46923 + }, + { + "epoch": 0.46924, + "grad_norm": 0.6741591351419732, + "learning_rate": 0.003, + "loss": 3.955, + "step": 46924 + }, + { + "epoch": 0.46925, + "grad_norm": 0.7507908488226924, + "learning_rate": 0.003, + "loss": 3.9369, + "step": 46925 + }, + { + "epoch": 0.46926, + "grad_norm": 0.6483901524295642, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 46926 + }, + { + "epoch": 0.46927, + "grad_norm": 0.5884062063540709, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 46927 + }, + { + "epoch": 0.46928, + "grad_norm": 0.7401741466400001, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 46928 + }, + { + "epoch": 0.46929, + "grad_norm": 0.9958897712321229, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 46929 + }, + { + "epoch": 0.4693, + "grad_norm": 1.17121897119667, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 46930 + }, + { + "epoch": 0.46931, + "grad_norm": 0.8679494978390588, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 46931 + }, + { + "epoch": 0.46932, + "grad_norm": 0.7639615742193918, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 46932 + }, + { + "epoch": 0.46933, + "grad_norm": 0.8804862491246708, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 46933 + }, + { + "epoch": 0.46934, + "grad_norm": 0.9170589555949085, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 46934 + }, + { + "epoch": 0.46935, + "grad_norm": 1.078336088324518, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 46935 + }, + { + "epoch": 0.46936, + "grad_norm": 1.0158853250087905, + "learning_rate": 0.003, + "loss": 3.984, + "step": 46936 + }, + { + "epoch": 0.46937, + "grad_norm": 0.9645771163799186, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 46937 + }, + { + "epoch": 0.46938, + "grad_norm": 0.9544572775846123, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 46938 + }, + { + "epoch": 0.46939, + "grad_norm": 1.0292730789580564, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 46939 + }, + { + "epoch": 0.4694, + "grad_norm": 1.0760851215341796, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 46940 + }, + { + "epoch": 0.46941, + "grad_norm": 0.7484445016820592, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 46941 + }, + { + "epoch": 0.46942, + "grad_norm": 0.8434071372396238, + "learning_rate": 0.003, + "loss": 4.0182, + "step": 46942 + }, + { + "epoch": 0.46943, + "grad_norm": 0.8685072610621557, + "learning_rate": 0.003, + "loss": 4.006, + "step": 46943 + }, + { + "epoch": 0.46944, + "grad_norm": 1.078444362103051, + "learning_rate": 0.003, + "loss": 4.027, + "step": 46944 + }, + { + "epoch": 0.46945, + "grad_norm": 1.0056608205076447, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 46945 + }, + { + "epoch": 0.46946, + "grad_norm": 0.9850802604735257, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 46946 + }, + { + "epoch": 0.46947, + "grad_norm": 1.1233450258196722, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 46947 + }, + { + "epoch": 0.46948, + "grad_norm": 0.87984109060922, + "learning_rate": 0.003, + "loss": 4.0268, + "step": 46948 + }, + { + "epoch": 0.46949, + "grad_norm": 0.9628239195682803, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 46949 + }, + { + "epoch": 0.4695, + "grad_norm": 1.0103209055161346, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 46950 + }, + { + "epoch": 0.46951, + "grad_norm": 0.9441785432784094, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 46951 + }, + { + "epoch": 0.46952, + "grad_norm": 0.9785153192668672, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 46952 + }, + { + "epoch": 0.46953, + "grad_norm": 0.8885284712082806, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 46953 + }, + { + "epoch": 0.46954, + "grad_norm": 0.9274858937351282, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 46954 + }, + { + "epoch": 0.46955, + "grad_norm": 0.9233374325063993, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 46955 + }, + { + "epoch": 0.46956, + "grad_norm": 0.8941288540346164, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 46956 + }, + { + "epoch": 0.46957, + "grad_norm": 0.8516752366553556, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 46957 + }, + { + "epoch": 0.46958, + "grad_norm": 0.813901777703451, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 46958 + }, + { + "epoch": 0.46959, + "grad_norm": 0.6973803773467224, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 46959 + }, + { + "epoch": 0.4696, + "grad_norm": 0.6784898068567361, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 46960 + }, + { + "epoch": 0.46961, + "grad_norm": 0.6606948514969297, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 46961 + }, + { + "epoch": 0.46962, + "grad_norm": 0.7481949428627195, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 46962 + }, + { + "epoch": 0.46963, + "grad_norm": 0.8909775782174565, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 46963 + }, + { + "epoch": 0.46964, + "grad_norm": 0.9211278475709174, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 46964 + }, + { + "epoch": 0.46965, + "grad_norm": 0.8052782919578275, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 46965 + }, + { + "epoch": 0.46966, + "grad_norm": 0.6641772018698358, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 46966 + }, + { + "epoch": 0.46967, + "grad_norm": 0.5468537423327078, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 46967 + }, + { + "epoch": 0.46968, + "grad_norm": 0.5241570029311021, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 46968 + }, + { + "epoch": 0.46969, + "grad_norm": 0.4446284708012736, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 46969 + }, + { + "epoch": 0.4697, + "grad_norm": 0.4778565528138039, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 46970 + }, + { + "epoch": 0.46971, + "grad_norm": 0.5362580571946972, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 46971 + }, + { + "epoch": 0.46972, + "grad_norm": 0.5995775139081446, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 46972 + }, + { + "epoch": 0.46973, + "grad_norm": 0.6177748451439617, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 46973 + }, + { + "epoch": 0.46974, + "grad_norm": 0.5016466941364403, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 46974 + }, + { + "epoch": 0.46975, + "grad_norm": 0.465021995760637, + "learning_rate": 0.003, + "loss": 3.9115, + "step": 46975 + }, + { + "epoch": 0.46976, + "grad_norm": 0.44781174771073096, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 46976 + }, + { + "epoch": 0.46977, + "grad_norm": 0.4439795227982857, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 46977 + }, + { + "epoch": 0.46978, + "grad_norm": 0.5168156173500201, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 46978 + }, + { + "epoch": 0.46979, + "grad_norm": 0.6065046656761204, + "learning_rate": 0.003, + "loss": 3.9404, + "step": 46979 + }, + { + "epoch": 0.4698, + "grad_norm": 0.7306877574063667, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 46980 + }, + { + "epoch": 0.46981, + "grad_norm": 0.8376574386927432, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 46981 + }, + { + "epoch": 0.46982, + "grad_norm": 0.8991338478372287, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 46982 + }, + { + "epoch": 0.46983, + "grad_norm": 0.8906465206571146, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 46983 + }, + { + "epoch": 0.46984, + "grad_norm": 0.8852297833574749, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 46984 + }, + { + "epoch": 0.46985, + "grad_norm": 0.834493709124405, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 46985 + }, + { + "epoch": 0.46986, + "grad_norm": 0.8987471548836309, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 46986 + }, + { + "epoch": 0.46987, + "grad_norm": 0.9837313727780626, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 46987 + }, + { + "epoch": 0.46988, + "grad_norm": 0.8582621728633725, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 46988 + }, + { + "epoch": 0.46989, + "grad_norm": 0.7789274721151219, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 46989 + }, + { + "epoch": 0.4699, + "grad_norm": 0.8614184927724415, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 46990 + }, + { + "epoch": 0.46991, + "grad_norm": 0.7769900283644453, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 46991 + }, + { + "epoch": 0.46992, + "grad_norm": 0.7383564950848069, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 46992 + }, + { + "epoch": 0.46993, + "grad_norm": 0.7743377532145285, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 46993 + }, + { + "epoch": 0.46994, + "grad_norm": 0.8869243699108973, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 46994 + }, + { + "epoch": 0.46995, + "grad_norm": 0.9908902043853998, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 46995 + }, + { + "epoch": 0.46996, + "grad_norm": 1.0177972190571978, + "learning_rate": 0.003, + "loss": 3.975, + "step": 46996 + }, + { + "epoch": 0.46997, + "grad_norm": 0.9198909371185565, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 46997 + }, + { + "epoch": 0.46998, + "grad_norm": 0.8932008278655216, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 46998 + }, + { + "epoch": 0.46999, + "grad_norm": 0.9054694021379021, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 46999 + }, + { + "epoch": 0.47, + "grad_norm": 0.8213141869974503, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 47000 + }, + { + "epoch": 0.47001, + "grad_norm": 0.747875715015047, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 47001 + }, + { + "epoch": 0.47002, + "grad_norm": 0.9114278960564977, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 47002 + }, + { + "epoch": 0.47003, + "grad_norm": 0.9883592328990211, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 47003 + }, + { + "epoch": 0.47004, + "grad_norm": 0.9292807910984607, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 47004 + }, + { + "epoch": 0.47005, + "grad_norm": 0.8530058585759173, + "learning_rate": 0.003, + "loss": 3.967, + "step": 47005 + }, + { + "epoch": 0.47006, + "grad_norm": 0.8024261749107751, + "learning_rate": 0.003, + "loss": 3.957, + "step": 47006 + }, + { + "epoch": 0.47007, + "grad_norm": 0.7334039726798804, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 47007 + }, + { + "epoch": 0.47008, + "grad_norm": 0.65964611925669, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 47008 + }, + { + "epoch": 0.47009, + "grad_norm": 0.6271417886729114, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 47009 + }, + { + "epoch": 0.4701, + "grad_norm": 0.6769489143219881, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 47010 + }, + { + "epoch": 0.47011, + "grad_norm": 0.8208594316544231, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 47011 + }, + { + "epoch": 0.47012, + "grad_norm": 1.077193657931963, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 47012 + }, + { + "epoch": 0.47013, + "grad_norm": 0.92764727986197, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 47013 + }, + { + "epoch": 0.47014, + "grad_norm": 0.7866197988790653, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 47014 + }, + { + "epoch": 0.47015, + "grad_norm": 0.70826708963553, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 47015 + }, + { + "epoch": 0.47016, + "grad_norm": 0.7503211987042646, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 47016 + }, + { + "epoch": 0.47017, + "grad_norm": 0.629402978316606, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 47017 + }, + { + "epoch": 0.47018, + "grad_norm": 0.6636199793328755, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 47018 + }, + { + "epoch": 0.47019, + "grad_norm": 0.8181477903407466, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 47019 + }, + { + "epoch": 0.4702, + "grad_norm": 0.8117006626281956, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 47020 + }, + { + "epoch": 0.47021, + "grad_norm": 0.6856454275810056, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 47021 + }, + { + "epoch": 0.47022, + "grad_norm": 0.6532082636093172, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 47022 + }, + { + "epoch": 0.47023, + "grad_norm": 0.6860980840768574, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 47023 + }, + { + "epoch": 0.47024, + "grad_norm": 0.6830253400900101, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 47024 + }, + { + "epoch": 0.47025, + "grad_norm": 0.7810174195923986, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 47025 + }, + { + "epoch": 0.47026, + "grad_norm": 0.9287240283198575, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 47026 + }, + { + "epoch": 0.47027, + "grad_norm": 0.9078818017092718, + "learning_rate": 0.003, + "loss": 4.0122, + "step": 47027 + }, + { + "epoch": 0.47028, + "grad_norm": 0.8073694939076342, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 47028 + }, + { + "epoch": 0.47029, + "grad_norm": 0.6155373701678561, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 47029 + }, + { + "epoch": 0.4703, + "grad_norm": 0.592933033482125, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 47030 + }, + { + "epoch": 0.47031, + "grad_norm": 0.6102979056190457, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 47031 + }, + { + "epoch": 0.47032, + "grad_norm": 0.6310823831405185, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 47032 + }, + { + "epoch": 0.47033, + "grad_norm": 0.5965683482779892, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 47033 + }, + { + "epoch": 0.47034, + "grad_norm": 0.5965972304411118, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 47034 + }, + { + "epoch": 0.47035, + "grad_norm": 0.6542498281735639, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 47035 + }, + { + "epoch": 0.47036, + "grad_norm": 0.6850461138273444, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 47036 + }, + { + "epoch": 0.47037, + "grad_norm": 0.6158037796035429, + "learning_rate": 0.003, + "loss": 3.996, + "step": 47037 + }, + { + "epoch": 0.47038, + "grad_norm": 0.60196405355758, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 47038 + }, + { + "epoch": 0.47039, + "grad_norm": 0.6503223773070101, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 47039 + }, + { + "epoch": 0.4704, + "grad_norm": 0.7031824274986416, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 47040 + }, + { + "epoch": 0.47041, + "grad_norm": 0.7800634365272356, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 47041 + }, + { + "epoch": 0.47042, + "grad_norm": 0.8370703623609113, + "learning_rate": 0.003, + "loss": 3.956, + "step": 47042 + }, + { + "epoch": 0.47043, + "grad_norm": 0.7934680017897011, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 47043 + }, + { + "epoch": 0.47044, + "grad_norm": 0.7268540297409626, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 47044 + }, + { + "epoch": 0.47045, + "grad_norm": 0.7323190920721484, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 47045 + }, + { + "epoch": 0.47046, + "grad_norm": 0.7597558709190158, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 47046 + }, + { + "epoch": 0.47047, + "grad_norm": 0.9920615637970766, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 47047 + }, + { + "epoch": 0.47048, + "grad_norm": 1.2921031177082096, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 47048 + }, + { + "epoch": 0.47049, + "grad_norm": 0.6573976205951793, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 47049 + }, + { + "epoch": 0.4705, + "grad_norm": 0.7580420759535045, + "learning_rate": 0.003, + "loss": 3.9456, + "step": 47050 + }, + { + "epoch": 0.47051, + "grad_norm": 0.9444359105570252, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 47051 + }, + { + "epoch": 0.47052, + "grad_norm": 1.097198675899508, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 47052 + }, + { + "epoch": 0.47053, + "grad_norm": 1.034505280802518, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 47053 + }, + { + "epoch": 0.47054, + "grad_norm": 0.8743715835117473, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 47054 + }, + { + "epoch": 0.47055, + "grad_norm": 0.8702196184026795, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 47055 + }, + { + "epoch": 0.47056, + "grad_norm": 0.8468815043079084, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 47056 + }, + { + "epoch": 0.47057, + "grad_norm": 0.8391573610022082, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 47057 + }, + { + "epoch": 0.47058, + "grad_norm": 0.8139686821439538, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 47058 + }, + { + "epoch": 0.47059, + "grad_norm": 0.8078793693291818, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 47059 + }, + { + "epoch": 0.4706, + "grad_norm": 0.8003767748595868, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 47060 + }, + { + "epoch": 0.47061, + "grad_norm": 0.7382789569510737, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 47061 + }, + { + "epoch": 0.47062, + "grad_norm": 0.6919760453851579, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 47062 + }, + { + "epoch": 0.47063, + "grad_norm": 0.6471659215083624, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 47063 + }, + { + "epoch": 0.47064, + "grad_norm": 0.5978914013003054, + "learning_rate": 0.003, + "loss": 3.9437, + "step": 47064 + }, + { + "epoch": 0.47065, + "grad_norm": 0.5979260981694648, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 47065 + }, + { + "epoch": 0.47066, + "grad_norm": 0.6969351317615258, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 47066 + }, + { + "epoch": 0.47067, + "grad_norm": 0.8105617261354031, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 47067 + }, + { + "epoch": 0.47068, + "grad_norm": 0.8194389873947074, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 47068 + }, + { + "epoch": 0.47069, + "grad_norm": 0.772692520508916, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 47069 + }, + { + "epoch": 0.4707, + "grad_norm": 0.8531848214456538, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 47070 + }, + { + "epoch": 0.47071, + "grad_norm": 1.0056287410691755, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 47071 + }, + { + "epoch": 0.47072, + "grad_norm": 0.8684437824680936, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 47072 + }, + { + "epoch": 0.47073, + "grad_norm": 0.7585706903129267, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 47073 + }, + { + "epoch": 0.47074, + "grad_norm": 0.7721876771004637, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 47074 + }, + { + "epoch": 0.47075, + "grad_norm": 0.8050947602865342, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 47075 + }, + { + "epoch": 0.47076, + "grad_norm": 0.8650746907799904, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 47076 + }, + { + "epoch": 0.47077, + "grad_norm": 0.8483885164441232, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 47077 + }, + { + "epoch": 0.47078, + "grad_norm": 0.9330765953440657, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 47078 + }, + { + "epoch": 0.47079, + "grad_norm": 0.9087268147000899, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 47079 + }, + { + "epoch": 0.4708, + "grad_norm": 0.8957176317594341, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 47080 + }, + { + "epoch": 0.47081, + "grad_norm": 0.812671913441309, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 47081 + }, + { + "epoch": 0.47082, + "grad_norm": 0.7047371668620679, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 47082 + }, + { + "epoch": 0.47083, + "grad_norm": 0.6686854512834434, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 47083 + }, + { + "epoch": 0.47084, + "grad_norm": 0.7702734275977965, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 47084 + }, + { + "epoch": 0.47085, + "grad_norm": 0.8075200327606356, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 47085 + }, + { + "epoch": 0.47086, + "grad_norm": 0.7576292901150377, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 47086 + }, + { + "epoch": 0.47087, + "grad_norm": 0.6724060727902528, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 47087 + }, + { + "epoch": 0.47088, + "grad_norm": 0.6262204202538405, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 47088 + }, + { + "epoch": 0.47089, + "grad_norm": 0.7384345758256587, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 47089 + }, + { + "epoch": 0.4709, + "grad_norm": 0.7672172975837668, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 47090 + }, + { + "epoch": 0.47091, + "grad_norm": 0.8073693022767934, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 47091 + }, + { + "epoch": 0.47092, + "grad_norm": 0.8370989087992187, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 47092 + }, + { + "epoch": 0.47093, + "grad_norm": 0.8227206921441732, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 47093 + }, + { + "epoch": 0.47094, + "grad_norm": 0.7985266387717094, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 47094 + }, + { + "epoch": 0.47095, + "grad_norm": 0.7813131215745424, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 47095 + }, + { + "epoch": 0.47096, + "grad_norm": 0.7066562574553549, + "learning_rate": 0.003, + "loss": 3.99, + "step": 47096 + }, + { + "epoch": 0.47097, + "grad_norm": 0.7553266881947762, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 47097 + }, + { + "epoch": 0.47098, + "grad_norm": 0.7239556252196815, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 47098 + }, + { + "epoch": 0.47099, + "grad_norm": 0.6143505479509489, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 47099 + }, + { + "epoch": 0.471, + "grad_norm": 0.5714621028086425, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 47100 + }, + { + "epoch": 0.47101, + "grad_norm": 0.6274267518824764, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 47101 + }, + { + "epoch": 0.47102, + "grad_norm": 0.7052502718050365, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 47102 + }, + { + "epoch": 0.47103, + "grad_norm": 0.8856634658016668, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 47103 + }, + { + "epoch": 0.47104, + "grad_norm": 1.0801913061242086, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 47104 + }, + { + "epoch": 0.47105, + "grad_norm": 1.0859930438505703, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 47105 + }, + { + "epoch": 0.47106, + "grad_norm": 0.7883188386847574, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 47106 + }, + { + "epoch": 0.47107, + "grad_norm": 0.7905995240327387, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 47107 + }, + { + "epoch": 0.47108, + "grad_norm": 1.0550748598300368, + "learning_rate": 0.003, + "loss": 3.979, + "step": 47108 + }, + { + "epoch": 0.47109, + "grad_norm": 1.1930697517548188, + "learning_rate": 0.003, + "loss": 4.005, + "step": 47109 + }, + { + "epoch": 0.4711, + "grad_norm": 0.8125902393983145, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 47110 + }, + { + "epoch": 0.47111, + "grad_norm": 0.7454627792462746, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 47111 + }, + { + "epoch": 0.47112, + "grad_norm": 0.7298973071837287, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 47112 + }, + { + "epoch": 0.47113, + "grad_norm": 0.7553586690268628, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 47113 + }, + { + "epoch": 0.47114, + "grad_norm": 0.7867508164223014, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 47114 + }, + { + "epoch": 0.47115, + "grad_norm": 0.7950076238252513, + "learning_rate": 0.003, + "loss": 4.003, + "step": 47115 + }, + { + "epoch": 0.47116, + "grad_norm": 0.7446837671476165, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 47116 + }, + { + "epoch": 0.47117, + "grad_norm": 0.6705736216355046, + "learning_rate": 0.003, + "loss": 4.005, + "step": 47117 + }, + { + "epoch": 0.47118, + "grad_norm": 0.744500485026805, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 47118 + }, + { + "epoch": 0.47119, + "grad_norm": 0.7549975766189815, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 47119 + }, + { + "epoch": 0.4712, + "grad_norm": 0.9040922477330234, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 47120 + }, + { + "epoch": 0.47121, + "grad_norm": 1.0004555898698066, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 47121 + }, + { + "epoch": 0.47122, + "grad_norm": 0.9060098985788146, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 47122 + }, + { + "epoch": 0.47123, + "grad_norm": 0.7725669375278873, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 47123 + }, + { + "epoch": 0.47124, + "grad_norm": 0.6106982621114778, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 47124 + }, + { + "epoch": 0.47125, + "grad_norm": 0.5446732595978525, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 47125 + }, + { + "epoch": 0.47126, + "grad_norm": 0.49226384983515764, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 47126 + }, + { + "epoch": 0.47127, + "grad_norm": 0.46903338235543574, + "learning_rate": 0.003, + "loss": 3.997, + "step": 47127 + }, + { + "epoch": 0.47128, + "grad_norm": 0.4588700251347558, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 47128 + }, + { + "epoch": 0.47129, + "grad_norm": 0.4919151485191485, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 47129 + }, + { + "epoch": 0.4713, + "grad_norm": 0.5342771346406966, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 47130 + }, + { + "epoch": 0.47131, + "grad_norm": 0.5900484017429928, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 47131 + }, + { + "epoch": 0.47132, + "grad_norm": 0.7857062851106144, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 47132 + }, + { + "epoch": 0.47133, + "grad_norm": 0.9782382747673265, + "learning_rate": 0.003, + "loss": 3.992, + "step": 47133 + }, + { + "epoch": 0.47134, + "grad_norm": 0.9209771794050431, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 47134 + }, + { + "epoch": 0.47135, + "grad_norm": 0.7883648263430499, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 47135 + }, + { + "epoch": 0.47136, + "grad_norm": 0.6728544919425054, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 47136 + }, + { + "epoch": 0.47137, + "grad_norm": 0.7703535686661174, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 47137 + }, + { + "epoch": 0.47138, + "grad_norm": 0.8314674306866573, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 47138 + }, + { + "epoch": 0.47139, + "grad_norm": 0.8892050938976641, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 47139 + }, + { + "epoch": 0.4714, + "grad_norm": 0.9428666535904777, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 47140 + }, + { + "epoch": 0.47141, + "grad_norm": 0.9333628396558781, + "learning_rate": 0.003, + "loss": 3.979, + "step": 47141 + }, + { + "epoch": 0.47142, + "grad_norm": 0.8443200489037515, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 47142 + }, + { + "epoch": 0.47143, + "grad_norm": 0.9112299919852556, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 47143 + }, + { + "epoch": 0.47144, + "grad_norm": 0.8774771979893352, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 47144 + }, + { + "epoch": 0.47145, + "grad_norm": 0.9465217671985884, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 47145 + }, + { + "epoch": 0.47146, + "grad_norm": 1.0164843295842785, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 47146 + }, + { + "epoch": 0.47147, + "grad_norm": 0.9703548302080093, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 47147 + }, + { + "epoch": 0.47148, + "grad_norm": 1.0518750682554967, + "learning_rate": 0.003, + "loss": 3.963, + "step": 47148 + }, + { + "epoch": 0.47149, + "grad_norm": 0.8052121422234338, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 47149 + }, + { + "epoch": 0.4715, + "grad_norm": 0.7735703067222534, + "learning_rate": 0.003, + "loss": 4.0104, + "step": 47150 + }, + { + "epoch": 0.47151, + "grad_norm": 0.7711858951551293, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 47151 + }, + { + "epoch": 0.47152, + "grad_norm": 0.7068780725887387, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 47152 + }, + { + "epoch": 0.47153, + "grad_norm": 0.66433722551341, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 47153 + }, + { + "epoch": 0.47154, + "grad_norm": 0.6168509495878155, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 47154 + }, + { + "epoch": 0.47155, + "grad_norm": 0.6541207788033752, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 47155 + }, + { + "epoch": 0.47156, + "grad_norm": 0.679636015547371, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 47156 + }, + { + "epoch": 0.47157, + "grad_norm": 0.6116461500018188, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 47157 + }, + { + "epoch": 0.47158, + "grad_norm": 0.5950968697075256, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 47158 + }, + { + "epoch": 0.47159, + "grad_norm": 0.6028701975673867, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 47159 + }, + { + "epoch": 0.4716, + "grad_norm": 0.6030296434697265, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 47160 + }, + { + "epoch": 0.47161, + "grad_norm": 0.55732881088136, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 47161 + }, + { + "epoch": 0.47162, + "grad_norm": 0.5488437489010789, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 47162 + }, + { + "epoch": 0.47163, + "grad_norm": 0.5886666566249754, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 47163 + }, + { + "epoch": 0.47164, + "grad_norm": 0.6418088374207214, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 47164 + }, + { + "epoch": 0.47165, + "grad_norm": 0.6406097668830932, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 47165 + }, + { + "epoch": 0.47166, + "grad_norm": 0.6669040683254972, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 47166 + }, + { + "epoch": 0.47167, + "grad_norm": 0.7438926174391618, + "learning_rate": 0.003, + "loss": 3.986, + "step": 47167 + }, + { + "epoch": 0.47168, + "grad_norm": 0.784135854674901, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 47168 + }, + { + "epoch": 0.47169, + "grad_norm": 0.8308884599541639, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 47169 + }, + { + "epoch": 0.4717, + "grad_norm": 1.0035759502502763, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 47170 + }, + { + "epoch": 0.47171, + "grad_norm": 1.142665990801784, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 47171 + }, + { + "epoch": 0.47172, + "grad_norm": 0.8518383738888258, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 47172 + }, + { + "epoch": 0.47173, + "grad_norm": 0.8183797422499349, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 47173 + }, + { + "epoch": 0.47174, + "grad_norm": 0.8039312477557363, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 47174 + }, + { + "epoch": 0.47175, + "grad_norm": 0.7838576611349541, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 47175 + }, + { + "epoch": 0.47176, + "grad_norm": 0.7391653193581814, + "learning_rate": 0.003, + "loss": 3.993, + "step": 47176 + }, + { + "epoch": 0.47177, + "grad_norm": 0.760192249631394, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 47177 + }, + { + "epoch": 0.47178, + "grad_norm": 0.8453333311195346, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 47178 + }, + { + "epoch": 0.47179, + "grad_norm": 1.0132692244816057, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 47179 + }, + { + "epoch": 0.4718, + "grad_norm": 1.1100689016903795, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 47180 + }, + { + "epoch": 0.47181, + "grad_norm": 0.8977904403023571, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 47181 + }, + { + "epoch": 0.47182, + "grad_norm": 0.9450637705392514, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 47182 + }, + { + "epoch": 0.47183, + "grad_norm": 1.0327039619515352, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 47183 + }, + { + "epoch": 0.47184, + "grad_norm": 0.9097079052909173, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 47184 + }, + { + "epoch": 0.47185, + "grad_norm": 0.8292747636502508, + "learning_rate": 0.003, + "loss": 4.017, + "step": 47185 + }, + { + "epoch": 0.47186, + "grad_norm": 0.9035848151270831, + "learning_rate": 0.003, + "loss": 4.001, + "step": 47186 + }, + { + "epoch": 0.47187, + "grad_norm": 1.0420268882999615, + "learning_rate": 0.003, + "loss": 4.0302, + "step": 47187 + }, + { + "epoch": 0.47188, + "grad_norm": 1.006903734979543, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 47188 + }, + { + "epoch": 0.47189, + "grad_norm": 0.8822500157370391, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 47189 + }, + { + "epoch": 0.4719, + "grad_norm": 0.8197314871244366, + "learning_rate": 0.003, + "loss": 4.0341, + "step": 47190 + }, + { + "epoch": 0.47191, + "grad_norm": 0.8226612776763579, + "learning_rate": 0.003, + "loss": 3.996, + "step": 47191 + }, + { + "epoch": 0.47192, + "grad_norm": 0.8719453564118039, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 47192 + }, + { + "epoch": 0.47193, + "grad_norm": 0.8463076330148865, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 47193 + }, + { + "epoch": 0.47194, + "grad_norm": 0.8209693555810298, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 47194 + }, + { + "epoch": 0.47195, + "grad_norm": 0.7786001415300662, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 47195 + }, + { + "epoch": 0.47196, + "grad_norm": 0.7153040362572491, + "learning_rate": 0.003, + "loss": 4.0191, + "step": 47196 + }, + { + "epoch": 0.47197, + "grad_norm": 0.7765831461823127, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 47197 + }, + { + "epoch": 0.47198, + "grad_norm": 0.9155403681413742, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 47198 + }, + { + "epoch": 0.47199, + "grad_norm": 0.9790773091858803, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 47199 + }, + { + "epoch": 0.472, + "grad_norm": 0.9664601662808848, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 47200 + }, + { + "epoch": 0.47201, + "grad_norm": 0.80862814038624, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 47201 + }, + { + "epoch": 0.47202, + "grad_norm": 0.7127191525590779, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 47202 + }, + { + "epoch": 0.47203, + "grad_norm": 0.6577098411165299, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 47203 + }, + { + "epoch": 0.47204, + "grad_norm": 0.6030271464554936, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 47204 + }, + { + "epoch": 0.47205, + "grad_norm": 0.5447000738520629, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 47205 + }, + { + "epoch": 0.47206, + "grad_norm": 0.5086567527759724, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 47206 + }, + { + "epoch": 0.47207, + "grad_norm": 0.5186791541496254, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 47207 + }, + { + "epoch": 0.47208, + "grad_norm": 0.5441399941426959, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 47208 + }, + { + "epoch": 0.47209, + "grad_norm": 0.5965083810900886, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 47209 + }, + { + "epoch": 0.4721, + "grad_norm": 0.631699184828222, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 47210 + }, + { + "epoch": 0.47211, + "grad_norm": 0.6672677327226865, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 47211 + }, + { + "epoch": 0.47212, + "grad_norm": 0.7518078551026693, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 47212 + }, + { + "epoch": 0.47213, + "grad_norm": 0.6888660488200726, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 47213 + }, + { + "epoch": 0.47214, + "grad_norm": 0.5684201794982218, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 47214 + }, + { + "epoch": 0.47215, + "grad_norm": 0.6061812000505798, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 47215 + }, + { + "epoch": 0.47216, + "grad_norm": 0.5885765768105649, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 47216 + }, + { + "epoch": 0.47217, + "grad_norm": 0.6548104781298878, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 47217 + }, + { + "epoch": 0.47218, + "grad_norm": 0.7677379649692656, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 47218 + }, + { + "epoch": 0.47219, + "grad_norm": 0.7381726823553868, + "learning_rate": 0.003, + "loss": 3.9392, + "step": 47219 + }, + { + "epoch": 0.4722, + "grad_norm": 0.5983634749254773, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 47220 + }, + { + "epoch": 0.47221, + "grad_norm": 0.5058152078439926, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 47221 + }, + { + "epoch": 0.47222, + "grad_norm": 0.49557451308158235, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 47222 + }, + { + "epoch": 0.47223, + "grad_norm": 0.6489412481511926, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 47223 + }, + { + "epoch": 0.47224, + "grad_norm": 0.9074152172364284, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 47224 + }, + { + "epoch": 0.47225, + "grad_norm": 1.08275147392886, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 47225 + }, + { + "epoch": 0.47226, + "grad_norm": 0.9383000828015052, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 47226 + }, + { + "epoch": 0.47227, + "grad_norm": 0.7371165250645286, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 47227 + }, + { + "epoch": 0.47228, + "grad_norm": 0.7438105266593671, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 47228 + }, + { + "epoch": 0.47229, + "grad_norm": 0.7986019584920355, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 47229 + }, + { + "epoch": 0.4723, + "grad_norm": 0.8265922043879949, + "learning_rate": 0.003, + "loss": 3.993, + "step": 47230 + }, + { + "epoch": 0.47231, + "grad_norm": 1.0519663126318368, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 47231 + }, + { + "epoch": 0.47232, + "grad_norm": 1.1350485217020694, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 47232 + }, + { + "epoch": 0.47233, + "grad_norm": 0.8292999792158746, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 47233 + }, + { + "epoch": 0.47234, + "grad_norm": 0.8075452552219657, + "learning_rate": 0.003, + "loss": 4.0262, + "step": 47234 + }, + { + "epoch": 0.47235, + "grad_norm": 0.8311100177111589, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 47235 + }, + { + "epoch": 0.47236, + "grad_norm": 0.8766040232409847, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 47236 + }, + { + "epoch": 0.47237, + "grad_norm": 0.7968243809023619, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 47237 + }, + { + "epoch": 0.47238, + "grad_norm": 0.7990795318885994, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 47238 + }, + { + "epoch": 0.47239, + "grad_norm": 0.7423281726640862, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 47239 + }, + { + "epoch": 0.4724, + "grad_norm": 0.672570469291204, + "learning_rate": 0.003, + "loss": 3.9514, + "step": 47240 + }, + { + "epoch": 0.47241, + "grad_norm": 0.7230376126199621, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 47241 + }, + { + "epoch": 0.47242, + "grad_norm": 0.815063886575986, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 47242 + }, + { + "epoch": 0.47243, + "grad_norm": 0.8328160698035616, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 47243 + }, + { + "epoch": 0.47244, + "grad_norm": 0.9531498261461293, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 47244 + }, + { + "epoch": 0.47245, + "grad_norm": 0.861418534565271, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 47245 + }, + { + "epoch": 0.47246, + "grad_norm": 0.7649044614028815, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 47246 + }, + { + "epoch": 0.47247, + "grad_norm": 0.904774382690852, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 47247 + }, + { + "epoch": 0.47248, + "grad_norm": 1.096331175582275, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 47248 + }, + { + "epoch": 0.47249, + "grad_norm": 0.9761027805053227, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 47249 + }, + { + "epoch": 0.4725, + "grad_norm": 1.0491887347764868, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 47250 + }, + { + "epoch": 0.47251, + "grad_norm": 0.8305428267534896, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 47251 + }, + { + "epoch": 0.47252, + "grad_norm": 0.7161097468503851, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 47252 + }, + { + "epoch": 0.47253, + "grad_norm": 0.6078224978944969, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 47253 + }, + { + "epoch": 0.47254, + "grad_norm": 0.5918251224549805, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 47254 + }, + { + "epoch": 0.47255, + "grad_norm": 0.6059866991563845, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 47255 + }, + { + "epoch": 0.47256, + "grad_norm": 0.709073002680074, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 47256 + }, + { + "epoch": 0.47257, + "grad_norm": 0.7256640913365824, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 47257 + }, + { + "epoch": 0.47258, + "grad_norm": 0.8077264841893045, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 47258 + }, + { + "epoch": 0.47259, + "grad_norm": 0.8860362633335284, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 47259 + }, + { + "epoch": 0.4726, + "grad_norm": 0.9308760485175135, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 47260 + }, + { + "epoch": 0.47261, + "grad_norm": 0.8773892643695159, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 47261 + }, + { + "epoch": 0.47262, + "grad_norm": 0.8266910442963192, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 47262 + }, + { + "epoch": 0.47263, + "grad_norm": 0.8134067161051061, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 47263 + }, + { + "epoch": 0.47264, + "grad_norm": 0.7791740598737378, + "learning_rate": 0.003, + "loss": 4.0289, + "step": 47264 + }, + { + "epoch": 0.47265, + "grad_norm": 0.8054417663136418, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 47265 + }, + { + "epoch": 0.47266, + "grad_norm": 0.8167856101148026, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 47266 + }, + { + "epoch": 0.47267, + "grad_norm": 0.8525132657607216, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 47267 + }, + { + "epoch": 0.47268, + "grad_norm": 0.9412019316235396, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 47268 + }, + { + "epoch": 0.47269, + "grad_norm": 0.8669492146918771, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 47269 + }, + { + "epoch": 0.4727, + "grad_norm": 0.9011796253781312, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 47270 + }, + { + "epoch": 0.47271, + "grad_norm": 0.869389582332447, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 47271 + }, + { + "epoch": 0.47272, + "grad_norm": 0.7659733847373822, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 47272 + }, + { + "epoch": 0.47273, + "grad_norm": 0.6759031937194008, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 47273 + }, + { + "epoch": 0.47274, + "grad_norm": 0.5986101321379166, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 47274 + }, + { + "epoch": 0.47275, + "grad_norm": 0.5895249457780657, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 47275 + }, + { + "epoch": 0.47276, + "grad_norm": 0.5710912791720425, + "learning_rate": 0.003, + "loss": 3.9319, + "step": 47276 + }, + { + "epoch": 0.47277, + "grad_norm": 0.6178788042436376, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 47277 + }, + { + "epoch": 0.47278, + "grad_norm": 0.6060401437438775, + "learning_rate": 0.003, + "loss": 3.962, + "step": 47278 + }, + { + "epoch": 0.47279, + "grad_norm": 0.6430692460678639, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 47279 + }, + { + "epoch": 0.4728, + "grad_norm": 0.689835799095437, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 47280 + }, + { + "epoch": 0.47281, + "grad_norm": 0.7069218308051661, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 47281 + }, + { + "epoch": 0.47282, + "grad_norm": 0.8285040396877086, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 47282 + }, + { + "epoch": 0.47283, + "grad_norm": 0.9189999162159705, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 47283 + }, + { + "epoch": 0.47284, + "grad_norm": 1.0232055873421058, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 47284 + }, + { + "epoch": 0.47285, + "grad_norm": 1.2468147062079165, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 47285 + }, + { + "epoch": 0.47286, + "grad_norm": 0.7322385924513303, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 47286 + }, + { + "epoch": 0.47287, + "grad_norm": 0.7311193002284246, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 47287 + }, + { + "epoch": 0.47288, + "grad_norm": 0.957885744226408, + "learning_rate": 0.003, + "loss": 3.991, + "step": 47288 + }, + { + "epoch": 0.47289, + "grad_norm": 1.0736638896112598, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 47289 + }, + { + "epoch": 0.4729, + "grad_norm": 0.977322660968102, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 47290 + }, + { + "epoch": 0.47291, + "grad_norm": 0.8903159939242309, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 47291 + }, + { + "epoch": 0.47292, + "grad_norm": 0.7498513728194165, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 47292 + }, + { + "epoch": 0.47293, + "grad_norm": 0.6158787219137739, + "learning_rate": 0.003, + "loss": 3.955, + "step": 47293 + }, + { + "epoch": 0.47294, + "grad_norm": 0.58102775092391, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 47294 + }, + { + "epoch": 0.47295, + "grad_norm": 0.5531199303921409, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 47295 + }, + { + "epoch": 0.47296, + "grad_norm": 0.6535530041755843, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 47296 + }, + { + "epoch": 0.47297, + "grad_norm": 0.706598780637852, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 47297 + }, + { + "epoch": 0.47298, + "grad_norm": 0.7405166803908585, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 47298 + }, + { + "epoch": 0.47299, + "grad_norm": 0.7483132598260792, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 47299 + }, + { + "epoch": 0.473, + "grad_norm": 0.7023589093314154, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 47300 + }, + { + "epoch": 0.47301, + "grad_norm": 0.6440529392880929, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 47301 + }, + { + "epoch": 0.47302, + "grad_norm": 0.6138622740013635, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 47302 + }, + { + "epoch": 0.47303, + "grad_norm": 0.6216318755338568, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 47303 + }, + { + "epoch": 0.47304, + "grad_norm": 0.5309145226274209, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 47304 + }, + { + "epoch": 0.47305, + "grad_norm": 0.6060201941699151, + "learning_rate": 0.003, + "loss": 3.978, + "step": 47305 + }, + { + "epoch": 0.47306, + "grad_norm": 0.633036595688522, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 47306 + }, + { + "epoch": 0.47307, + "grad_norm": 0.639296435524978, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 47307 + }, + { + "epoch": 0.47308, + "grad_norm": 0.6847288942733117, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 47308 + }, + { + "epoch": 0.47309, + "grad_norm": 0.7370639358590636, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 47309 + }, + { + "epoch": 0.4731, + "grad_norm": 0.7344303774877302, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 47310 + }, + { + "epoch": 0.47311, + "grad_norm": 0.7520522925610932, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 47311 + }, + { + "epoch": 0.47312, + "grad_norm": 0.8579413685807654, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 47312 + }, + { + "epoch": 0.47313, + "grad_norm": 0.9149031034515098, + "learning_rate": 0.003, + "loss": 3.991, + "step": 47313 + }, + { + "epoch": 0.47314, + "grad_norm": 0.8296964248185134, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 47314 + }, + { + "epoch": 0.47315, + "grad_norm": 0.6944922412094247, + "learning_rate": 0.003, + "loss": 4.004, + "step": 47315 + }, + { + "epoch": 0.47316, + "grad_norm": 0.6071298961167346, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 47316 + }, + { + "epoch": 0.47317, + "grad_norm": 0.6855763791596355, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 47317 + }, + { + "epoch": 0.47318, + "grad_norm": 0.745823422271628, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 47318 + }, + { + "epoch": 0.47319, + "grad_norm": 0.812933714266878, + "learning_rate": 0.003, + "loss": 4.007, + "step": 47319 + }, + { + "epoch": 0.4732, + "grad_norm": 0.9038914861666985, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 47320 + }, + { + "epoch": 0.47321, + "grad_norm": 1.0093892163470006, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 47321 + }, + { + "epoch": 0.47322, + "grad_norm": 0.9633284867672932, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 47322 + }, + { + "epoch": 0.47323, + "grad_norm": 0.834121306630602, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 47323 + }, + { + "epoch": 0.47324, + "grad_norm": 0.9043047269680909, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 47324 + }, + { + "epoch": 0.47325, + "grad_norm": 0.9346326632072867, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 47325 + }, + { + "epoch": 0.47326, + "grad_norm": 0.9348295408297788, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 47326 + }, + { + "epoch": 0.47327, + "grad_norm": 1.0746628091372747, + "learning_rate": 0.003, + "loss": 4.0278, + "step": 47327 + }, + { + "epoch": 0.47328, + "grad_norm": 1.0722667389561849, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 47328 + }, + { + "epoch": 0.47329, + "grad_norm": 0.9333100451862552, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 47329 + }, + { + "epoch": 0.4733, + "grad_norm": 0.869173509500971, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 47330 + }, + { + "epoch": 0.47331, + "grad_norm": 0.8742645156280802, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 47331 + }, + { + "epoch": 0.47332, + "grad_norm": 0.8760119915077598, + "learning_rate": 0.003, + "loss": 3.994, + "step": 47332 + }, + { + "epoch": 0.47333, + "grad_norm": 0.8642339407696991, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 47333 + }, + { + "epoch": 0.47334, + "grad_norm": 0.8741660744216334, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 47334 + }, + { + "epoch": 0.47335, + "grad_norm": 0.8932817927411121, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 47335 + }, + { + "epoch": 0.47336, + "grad_norm": 0.9994775737818842, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 47336 + }, + { + "epoch": 0.47337, + "grad_norm": 1.151391562744872, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 47337 + }, + { + "epoch": 0.47338, + "grad_norm": 0.7928911273558771, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 47338 + }, + { + "epoch": 0.47339, + "grad_norm": 0.7316254990022991, + "learning_rate": 0.003, + "loss": 4.001, + "step": 47339 + }, + { + "epoch": 0.4734, + "grad_norm": 0.745390795417945, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 47340 + }, + { + "epoch": 0.47341, + "grad_norm": 0.6680246950212585, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 47341 + }, + { + "epoch": 0.47342, + "grad_norm": 0.6721689288263185, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 47342 + }, + { + "epoch": 0.47343, + "grad_norm": 0.6479358532135986, + "learning_rate": 0.003, + "loss": 4.0219, + "step": 47343 + }, + { + "epoch": 0.47344, + "grad_norm": 0.6308377430361357, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 47344 + }, + { + "epoch": 0.47345, + "grad_norm": 0.7310895363385221, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 47345 + }, + { + "epoch": 0.47346, + "grad_norm": 0.8064163970901601, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 47346 + }, + { + "epoch": 0.47347, + "grad_norm": 0.9270962331591112, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 47347 + }, + { + "epoch": 0.47348, + "grad_norm": 1.021881490184415, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 47348 + }, + { + "epoch": 0.47349, + "grad_norm": 0.9532928810598897, + "learning_rate": 0.003, + "loss": 3.996, + "step": 47349 + }, + { + "epoch": 0.4735, + "grad_norm": 0.8918700690415361, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 47350 + }, + { + "epoch": 0.47351, + "grad_norm": 0.7804485818297342, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 47351 + }, + { + "epoch": 0.47352, + "grad_norm": 0.6851365922838767, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 47352 + }, + { + "epoch": 0.47353, + "grad_norm": 0.620322278618122, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 47353 + }, + { + "epoch": 0.47354, + "grad_norm": 0.651325258957151, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 47354 + }, + { + "epoch": 0.47355, + "grad_norm": 0.5676835407311095, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 47355 + }, + { + "epoch": 0.47356, + "grad_norm": 0.6076635626557941, + "learning_rate": 0.003, + "loss": 4.0384, + "step": 47356 + }, + { + "epoch": 0.47357, + "grad_norm": 0.6361154507030721, + "learning_rate": 0.003, + "loss": 3.928, + "step": 47357 + }, + { + "epoch": 0.47358, + "grad_norm": 0.6874253920713265, + "learning_rate": 0.003, + "loss": 3.974, + "step": 47358 + }, + { + "epoch": 0.47359, + "grad_norm": 0.7457277300623645, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 47359 + }, + { + "epoch": 0.4736, + "grad_norm": 0.9431394097972252, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 47360 + }, + { + "epoch": 0.47361, + "grad_norm": 1.1635119081707475, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 47361 + }, + { + "epoch": 0.47362, + "grad_norm": 0.8915546065274247, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 47362 + }, + { + "epoch": 0.47363, + "grad_norm": 0.7602514081621212, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 47363 + }, + { + "epoch": 0.47364, + "grad_norm": 0.6467050792166074, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 47364 + }, + { + "epoch": 0.47365, + "grad_norm": 0.6577273724314089, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 47365 + }, + { + "epoch": 0.47366, + "grad_norm": 0.6814910467574117, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 47366 + }, + { + "epoch": 0.47367, + "grad_norm": 0.6359573770114713, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 47367 + }, + { + "epoch": 0.47368, + "grad_norm": 0.5943313619572039, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 47368 + }, + { + "epoch": 0.47369, + "grad_norm": 0.5890227613123179, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 47369 + }, + { + "epoch": 0.4737, + "grad_norm": 0.6444533004820572, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 47370 + }, + { + "epoch": 0.47371, + "grad_norm": 0.7464633960956373, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 47371 + }, + { + "epoch": 0.47372, + "grad_norm": 0.8475097601213554, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 47372 + }, + { + "epoch": 0.47373, + "grad_norm": 0.948468923874196, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 47373 + }, + { + "epoch": 0.47374, + "grad_norm": 0.9071247673340175, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 47374 + }, + { + "epoch": 0.47375, + "grad_norm": 0.8284435727134256, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 47375 + }, + { + "epoch": 0.47376, + "grad_norm": 0.7888940566065273, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 47376 + }, + { + "epoch": 0.47377, + "grad_norm": 0.7973966080689325, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 47377 + }, + { + "epoch": 0.47378, + "grad_norm": 0.8507972825720594, + "learning_rate": 0.003, + "loss": 3.998, + "step": 47378 + }, + { + "epoch": 0.47379, + "grad_norm": 0.8619564921078594, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 47379 + }, + { + "epoch": 0.4738, + "grad_norm": 0.7861515813858652, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 47380 + }, + { + "epoch": 0.47381, + "grad_norm": 0.791197434331603, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 47381 + }, + { + "epoch": 0.47382, + "grad_norm": 0.8232903042336931, + "learning_rate": 0.003, + "loss": 4.012, + "step": 47382 + }, + { + "epoch": 0.47383, + "grad_norm": 0.8135899081107109, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 47383 + }, + { + "epoch": 0.47384, + "grad_norm": 0.8257468892998187, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 47384 + }, + { + "epoch": 0.47385, + "grad_norm": 0.9346256561600931, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 47385 + }, + { + "epoch": 0.47386, + "grad_norm": 1.0085369790930336, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 47386 + }, + { + "epoch": 0.47387, + "grad_norm": 0.9757929827596898, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 47387 + }, + { + "epoch": 0.47388, + "grad_norm": 0.8879195619737555, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 47388 + }, + { + "epoch": 0.47389, + "grad_norm": 0.8212858747359074, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 47389 + }, + { + "epoch": 0.4739, + "grad_norm": 0.9107288974844356, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 47390 + }, + { + "epoch": 0.47391, + "grad_norm": 0.8881229261572338, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 47391 + }, + { + "epoch": 0.47392, + "grad_norm": 0.8990576468660846, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 47392 + }, + { + "epoch": 0.47393, + "grad_norm": 0.8436260148557401, + "learning_rate": 0.003, + "loss": 4.009, + "step": 47393 + }, + { + "epoch": 0.47394, + "grad_norm": 0.8141723744962505, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 47394 + }, + { + "epoch": 0.47395, + "grad_norm": 0.7739600329842414, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 47395 + }, + { + "epoch": 0.47396, + "grad_norm": 0.8338118097981341, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 47396 + }, + { + "epoch": 0.47397, + "grad_norm": 0.8685356708592453, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 47397 + }, + { + "epoch": 0.47398, + "grad_norm": 0.7357428051941676, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 47398 + }, + { + "epoch": 0.47399, + "grad_norm": 0.6771712402860267, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 47399 + }, + { + "epoch": 0.474, + "grad_norm": 0.6498734418656238, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 47400 + }, + { + "epoch": 0.47401, + "grad_norm": 0.5785288127519128, + "learning_rate": 0.003, + "loss": 3.977, + "step": 47401 + }, + { + "epoch": 0.47402, + "grad_norm": 0.5146137600134751, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 47402 + }, + { + "epoch": 0.47403, + "grad_norm": 0.4537484703252305, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 47403 + }, + { + "epoch": 0.47404, + "grad_norm": 0.48989387192893186, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 47404 + }, + { + "epoch": 0.47405, + "grad_norm": 0.5018145592393474, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 47405 + }, + { + "epoch": 0.47406, + "grad_norm": 0.6631386647249996, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 47406 + }, + { + "epoch": 0.47407, + "grad_norm": 0.8462275106245543, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 47407 + }, + { + "epoch": 0.47408, + "grad_norm": 0.9149168563722088, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 47408 + }, + { + "epoch": 0.47409, + "grad_norm": 0.8647862211740337, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 47409 + }, + { + "epoch": 0.4741, + "grad_norm": 0.7878761914853947, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 47410 + }, + { + "epoch": 0.47411, + "grad_norm": 0.7384827126651806, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 47411 + }, + { + "epoch": 0.47412, + "grad_norm": 0.7755688444852831, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 47412 + }, + { + "epoch": 0.47413, + "grad_norm": 0.7576678481327865, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 47413 + }, + { + "epoch": 0.47414, + "grad_norm": 0.7289184930537588, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 47414 + }, + { + "epoch": 0.47415, + "grad_norm": 0.8579433173878523, + "learning_rate": 0.003, + "loss": 3.9427, + "step": 47415 + }, + { + "epoch": 0.47416, + "grad_norm": 0.8984391065989059, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 47416 + }, + { + "epoch": 0.47417, + "grad_norm": 0.9828227925588676, + "learning_rate": 0.003, + "loss": 3.99, + "step": 47417 + }, + { + "epoch": 0.47418, + "grad_norm": 0.870328170040356, + "learning_rate": 0.003, + "loss": 3.975, + "step": 47418 + }, + { + "epoch": 0.47419, + "grad_norm": 0.8021682859662856, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 47419 + }, + { + "epoch": 0.4742, + "grad_norm": 0.8033265514009371, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 47420 + }, + { + "epoch": 0.47421, + "grad_norm": 0.8280257839463617, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 47421 + }, + { + "epoch": 0.47422, + "grad_norm": 0.8168321346430537, + "learning_rate": 0.003, + "loss": 3.967, + "step": 47422 + }, + { + "epoch": 0.47423, + "grad_norm": 0.8652974219831482, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 47423 + }, + { + "epoch": 0.47424, + "grad_norm": 0.9160271734010871, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 47424 + }, + { + "epoch": 0.47425, + "grad_norm": 0.9238674101865078, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 47425 + }, + { + "epoch": 0.47426, + "grad_norm": 0.7633078892303563, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 47426 + }, + { + "epoch": 0.47427, + "grad_norm": 0.7752625356758428, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 47427 + }, + { + "epoch": 0.47428, + "grad_norm": 0.979668496401578, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 47428 + }, + { + "epoch": 0.47429, + "grad_norm": 1.1504026228861892, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 47429 + }, + { + "epoch": 0.4743, + "grad_norm": 0.7808073694338103, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 47430 + }, + { + "epoch": 0.47431, + "grad_norm": 0.7881192022198753, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 47431 + }, + { + "epoch": 0.47432, + "grad_norm": 0.787141704791941, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 47432 + }, + { + "epoch": 0.47433, + "grad_norm": 0.6927778065831123, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 47433 + }, + { + "epoch": 0.47434, + "grad_norm": 0.7541630866211593, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 47434 + }, + { + "epoch": 0.47435, + "grad_norm": 0.7227851084388526, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 47435 + }, + { + "epoch": 0.47436, + "grad_norm": 0.7398144035110042, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 47436 + }, + { + "epoch": 0.47437, + "grad_norm": 0.7097396837343076, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 47437 + }, + { + "epoch": 0.47438, + "grad_norm": 0.7043537555948215, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 47438 + }, + { + "epoch": 0.47439, + "grad_norm": 0.7097603314648028, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 47439 + }, + { + "epoch": 0.4744, + "grad_norm": 0.7316099173944577, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 47440 + }, + { + "epoch": 0.47441, + "grad_norm": 0.7415032422246725, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 47441 + }, + { + "epoch": 0.47442, + "grad_norm": 0.6431000114578188, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 47442 + }, + { + "epoch": 0.47443, + "grad_norm": 0.5603183032985929, + "learning_rate": 0.003, + "loss": 3.972, + "step": 47443 + }, + { + "epoch": 0.47444, + "grad_norm": 0.5425514001524504, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 47444 + }, + { + "epoch": 0.47445, + "grad_norm": 0.5752699207156472, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 47445 + }, + { + "epoch": 0.47446, + "grad_norm": 0.5381476014234244, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 47446 + }, + { + "epoch": 0.47447, + "grad_norm": 0.5474558812537941, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 47447 + }, + { + "epoch": 0.47448, + "grad_norm": 0.5879505911525446, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 47448 + }, + { + "epoch": 0.47449, + "grad_norm": 0.7385466945216433, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 47449 + }, + { + "epoch": 0.4745, + "grad_norm": 1.0469155474232636, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 47450 + }, + { + "epoch": 0.47451, + "grad_norm": 1.142086246347804, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 47451 + }, + { + "epoch": 0.47452, + "grad_norm": 0.6479111881498791, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 47452 + }, + { + "epoch": 0.47453, + "grad_norm": 0.7405937461792969, + "learning_rate": 0.003, + "loss": 3.962, + "step": 47453 + }, + { + "epoch": 0.47454, + "grad_norm": 0.8471126537275429, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 47454 + }, + { + "epoch": 0.47455, + "grad_norm": 0.9637560773473249, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 47455 + }, + { + "epoch": 0.47456, + "grad_norm": 0.9707214434727646, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 47456 + }, + { + "epoch": 0.47457, + "grad_norm": 0.9152242290361646, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 47457 + }, + { + "epoch": 0.47458, + "grad_norm": 0.889600904908398, + "learning_rate": 0.003, + "loss": 3.974, + "step": 47458 + }, + { + "epoch": 0.47459, + "grad_norm": 0.9115027075433291, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 47459 + }, + { + "epoch": 0.4746, + "grad_norm": 0.8307464112705704, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 47460 + }, + { + "epoch": 0.47461, + "grad_norm": 0.7883137231719285, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 47461 + }, + { + "epoch": 0.47462, + "grad_norm": 0.8604496879689819, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 47462 + }, + { + "epoch": 0.47463, + "grad_norm": 1.0030416470302599, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 47463 + }, + { + "epoch": 0.47464, + "grad_norm": 1.00749206669894, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 47464 + }, + { + "epoch": 0.47465, + "grad_norm": 1.0355817309891768, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 47465 + }, + { + "epoch": 0.47466, + "grad_norm": 0.93037910623268, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 47466 + }, + { + "epoch": 0.47467, + "grad_norm": 0.7873803776357692, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 47467 + }, + { + "epoch": 0.47468, + "grad_norm": 0.7135561316643982, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 47468 + }, + { + "epoch": 0.47469, + "grad_norm": 0.6868524303745709, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 47469 + }, + { + "epoch": 0.4747, + "grad_norm": 0.6333489587282518, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 47470 + }, + { + "epoch": 0.47471, + "grad_norm": 0.6009350361345522, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 47471 + }, + { + "epoch": 0.47472, + "grad_norm": 0.6208058920113657, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 47472 + }, + { + "epoch": 0.47473, + "grad_norm": 0.6106378795710525, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 47473 + }, + { + "epoch": 0.47474, + "grad_norm": 0.6938516488507234, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 47474 + }, + { + "epoch": 0.47475, + "grad_norm": 0.8045190297410927, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 47475 + }, + { + "epoch": 0.47476, + "grad_norm": 0.8096566815730457, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 47476 + }, + { + "epoch": 0.47477, + "grad_norm": 0.7934828475196696, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 47477 + }, + { + "epoch": 0.47478, + "grad_norm": 0.7446222210108197, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 47478 + }, + { + "epoch": 0.47479, + "grad_norm": 0.8545218768823379, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 47479 + }, + { + "epoch": 0.4748, + "grad_norm": 0.8195505511510027, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 47480 + }, + { + "epoch": 0.47481, + "grad_norm": 0.7707598219584848, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 47481 + }, + { + "epoch": 0.47482, + "grad_norm": 0.7259072466438894, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 47482 + }, + { + "epoch": 0.47483, + "grad_norm": 0.7577273856348736, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 47483 + }, + { + "epoch": 0.47484, + "grad_norm": 0.6876102182613274, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 47484 + }, + { + "epoch": 0.47485, + "grad_norm": 0.5700616197590991, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 47485 + }, + { + "epoch": 0.47486, + "grad_norm": 0.5303009379455769, + "learning_rate": 0.003, + "loss": 3.98, + "step": 47486 + }, + { + "epoch": 0.47487, + "grad_norm": 0.5332441490596497, + "learning_rate": 0.003, + "loss": 3.945, + "step": 47487 + }, + { + "epoch": 0.47488, + "grad_norm": 0.6289127777953477, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 47488 + }, + { + "epoch": 0.47489, + "grad_norm": 0.8002960982802394, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 47489 + }, + { + "epoch": 0.4749, + "grad_norm": 1.0045831562631433, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 47490 + }, + { + "epoch": 0.47491, + "grad_norm": 1.1347642025890168, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 47491 + }, + { + "epoch": 0.47492, + "grad_norm": 0.7834067376343348, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 47492 + }, + { + "epoch": 0.47493, + "grad_norm": 0.6362102650247178, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 47493 + }, + { + "epoch": 0.47494, + "grad_norm": 0.7365016342980737, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 47494 + }, + { + "epoch": 0.47495, + "grad_norm": 0.8440878692826157, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 47495 + }, + { + "epoch": 0.47496, + "grad_norm": 0.8805136273689144, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 47496 + }, + { + "epoch": 0.47497, + "grad_norm": 0.8031938624971061, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 47497 + }, + { + "epoch": 0.47498, + "grad_norm": 0.653439007749304, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 47498 + }, + { + "epoch": 0.47499, + "grad_norm": 0.7008018342986431, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 47499 + }, + { + "epoch": 0.475, + "grad_norm": 0.8248185477494873, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 47500 + }, + { + "epoch": 0.47501, + "grad_norm": 0.9553507484092393, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 47501 + }, + { + "epoch": 0.47502, + "grad_norm": 0.8047907233321658, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 47502 + }, + { + "epoch": 0.47503, + "grad_norm": 0.8377214042027429, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 47503 + }, + { + "epoch": 0.47504, + "grad_norm": 0.8990987391306018, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 47504 + }, + { + "epoch": 0.47505, + "grad_norm": 0.9597970447778333, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 47505 + }, + { + "epoch": 0.47506, + "grad_norm": 1.157494825185325, + "learning_rate": 0.003, + "loss": 4.0512, + "step": 47506 + }, + { + "epoch": 0.47507, + "grad_norm": 0.973212445062747, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 47507 + }, + { + "epoch": 0.47508, + "grad_norm": 0.8390280091011729, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 47508 + }, + { + "epoch": 0.47509, + "grad_norm": 0.8209586609153676, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 47509 + }, + { + "epoch": 0.4751, + "grad_norm": 0.8478332616806138, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 47510 + }, + { + "epoch": 0.47511, + "grad_norm": 0.9165551261978242, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 47511 + }, + { + "epoch": 0.47512, + "grad_norm": 1.0310636381923803, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 47512 + }, + { + "epoch": 0.47513, + "grad_norm": 0.9631408156423246, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 47513 + }, + { + "epoch": 0.47514, + "grad_norm": 1.1340739484469207, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 47514 + }, + { + "epoch": 0.47515, + "grad_norm": 0.8353407291228397, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 47515 + }, + { + "epoch": 0.47516, + "grad_norm": 0.6855016532766716, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 47516 + }, + { + "epoch": 0.47517, + "grad_norm": 0.6327178567387534, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 47517 + }, + { + "epoch": 0.47518, + "grad_norm": 0.5879372328437759, + "learning_rate": 0.003, + "loss": 4.0092, + "step": 47518 + }, + { + "epoch": 0.47519, + "grad_norm": 0.5926724130045354, + "learning_rate": 0.003, + "loss": 3.984, + "step": 47519 + }, + { + "epoch": 0.4752, + "grad_norm": 0.6877070168386127, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 47520 + }, + { + "epoch": 0.47521, + "grad_norm": 0.7197288823239641, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 47521 + }, + { + "epoch": 0.47522, + "grad_norm": 0.761282852821606, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 47522 + }, + { + "epoch": 0.47523, + "grad_norm": 0.8093207965871082, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 47523 + }, + { + "epoch": 0.47524, + "grad_norm": 0.7791719966169707, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 47524 + }, + { + "epoch": 0.47525, + "grad_norm": 0.8162005942347034, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 47525 + }, + { + "epoch": 0.47526, + "grad_norm": 0.8315405280171928, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 47526 + }, + { + "epoch": 0.47527, + "grad_norm": 0.9202595046014133, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 47527 + }, + { + "epoch": 0.47528, + "grad_norm": 0.8442729497622418, + "learning_rate": 0.003, + "loss": 3.996, + "step": 47528 + }, + { + "epoch": 0.47529, + "grad_norm": 0.8643161585052942, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 47529 + }, + { + "epoch": 0.4753, + "grad_norm": 0.6858079404922124, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 47530 + }, + { + "epoch": 0.47531, + "grad_norm": 0.6521453067574374, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 47531 + }, + { + "epoch": 0.47532, + "grad_norm": 0.6007821115296551, + "learning_rate": 0.003, + "loss": 3.9325, + "step": 47532 + }, + { + "epoch": 0.47533, + "grad_norm": 0.618911579562635, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 47533 + }, + { + "epoch": 0.47534, + "grad_norm": 0.5356435649202794, + "learning_rate": 0.003, + "loss": 3.977, + "step": 47534 + }, + { + "epoch": 0.47535, + "grad_norm": 0.5320850462126658, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 47535 + }, + { + "epoch": 0.47536, + "grad_norm": 0.46416782025138653, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 47536 + }, + { + "epoch": 0.47537, + "grad_norm": 0.4382139639890125, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 47537 + }, + { + "epoch": 0.47538, + "grad_norm": 0.5858071494965267, + "learning_rate": 0.003, + "loss": 3.942, + "step": 47538 + }, + { + "epoch": 0.47539, + "grad_norm": 0.6187936501184047, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 47539 + }, + { + "epoch": 0.4754, + "grad_norm": 0.6900205706278274, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 47540 + }, + { + "epoch": 0.47541, + "grad_norm": 0.8724409268066005, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 47541 + }, + { + "epoch": 0.47542, + "grad_norm": 1.069479290859462, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 47542 + }, + { + "epoch": 0.47543, + "grad_norm": 1.0928217297796388, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 47543 + }, + { + "epoch": 0.47544, + "grad_norm": 0.7982572530791529, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 47544 + }, + { + "epoch": 0.47545, + "grad_norm": 0.8008160756856753, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 47545 + }, + { + "epoch": 0.47546, + "grad_norm": 0.855960489653465, + "learning_rate": 0.003, + "loss": 3.984, + "step": 47546 + }, + { + "epoch": 0.47547, + "grad_norm": 0.9819851501078978, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 47547 + }, + { + "epoch": 0.47548, + "grad_norm": 1.0472766343741273, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 47548 + }, + { + "epoch": 0.47549, + "grad_norm": 0.7822785478279989, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 47549 + }, + { + "epoch": 0.4755, + "grad_norm": 0.9335950243906049, + "learning_rate": 0.003, + "loss": 4.0283, + "step": 47550 + }, + { + "epoch": 0.47551, + "grad_norm": 0.9589732576704345, + "learning_rate": 0.003, + "loss": 3.997, + "step": 47551 + }, + { + "epoch": 0.47552, + "grad_norm": 0.8653472171321698, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 47552 + }, + { + "epoch": 0.47553, + "grad_norm": 0.8168151860079704, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 47553 + }, + { + "epoch": 0.47554, + "grad_norm": 0.7453634077898619, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 47554 + }, + { + "epoch": 0.47555, + "grad_norm": 0.7464006758515439, + "learning_rate": 0.003, + "loss": 3.998, + "step": 47555 + }, + { + "epoch": 0.47556, + "grad_norm": 0.7537913781122394, + "learning_rate": 0.003, + "loss": 4.013, + "step": 47556 + }, + { + "epoch": 0.47557, + "grad_norm": 0.7546337385636708, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 47557 + }, + { + "epoch": 0.47558, + "grad_norm": 0.7317768188864334, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 47558 + }, + { + "epoch": 0.47559, + "grad_norm": 0.7529984008658, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 47559 + }, + { + "epoch": 0.4756, + "grad_norm": 0.8624618744117852, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 47560 + }, + { + "epoch": 0.47561, + "grad_norm": 0.9480551648074212, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 47561 + }, + { + "epoch": 0.47562, + "grad_norm": 0.9745821796995222, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 47562 + }, + { + "epoch": 0.47563, + "grad_norm": 0.9055866068198548, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 47563 + }, + { + "epoch": 0.47564, + "grad_norm": 0.7903422171892982, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 47564 + }, + { + "epoch": 0.47565, + "grad_norm": 0.6244759023520979, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 47565 + }, + { + "epoch": 0.47566, + "grad_norm": 0.6299202127376073, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 47566 + }, + { + "epoch": 0.47567, + "grad_norm": 0.5384173039036099, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 47567 + }, + { + "epoch": 0.47568, + "grad_norm": 0.5165169516851611, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 47568 + }, + { + "epoch": 0.47569, + "grad_norm": 0.5167891999736908, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 47569 + }, + { + "epoch": 0.4757, + "grad_norm": 0.5388757882929676, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 47570 + }, + { + "epoch": 0.47571, + "grad_norm": 0.6088376644803185, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 47571 + }, + { + "epoch": 0.47572, + "grad_norm": 0.691380342477591, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 47572 + }, + { + "epoch": 0.47573, + "grad_norm": 0.65685434957146, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 47573 + }, + { + "epoch": 0.47574, + "grad_norm": 0.7024549231130366, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 47574 + }, + { + "epoch": 0.47575, + "grad_norm": 0.7566219178582384, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 47575 + }, + { + "epoch": 0.47576, + "grad_norm": 0.8668135407899642, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 47576 + }, + { + "epoch": 0.47577, + "grad_norm": 0.9721045159283079, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 47577 + }, + { + "epoch": 0.47578, + "grad_norm": 0.9064367316444788, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 47578 + }, + { + "epoch": 0.47579, + "grad_norm": 0.6954263639131038, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 47579 + }, + { + "epoch": 0.4758, + "grad_norm": 0.6312917343453837, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 47580 + }, + { + "epoch": 0.47581, + "grad_norm": 0.6291058547790587, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 47581 + }, + { + "epoch": 0.47582, + "grad_norm": 0.6468191722859197, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 47582 + }, + { + "epoch": 0.47583, + "grad_norm": 0.6278062518180093, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 47583 + }, + { + "epoch": 0.47584, + "grad_norm": 0.6867564494808335, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 47584 + }, + { + "epoch": 0.47585, + "grad_norm": 0.7843884886494147, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 47585 + }, + { + "epoch": 0.47586, + "grad_norm": 0.9729215018507501, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 47586 + }, + { + "epoch": 0.47587, + "grad_norm": 1.1254180785930823, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 47587 + }, + { + "epoch": 0.47588, + "grad_norm": 0.8187368221969433, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 47588 + }, + { + "epoch": 0.47589, + "grad_norm": 0.8409234259126228, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 47589 + }, + { + "epoch": 0.4759, + "grad_norm": 0.8662759234250236, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 47590 + }, + { + "epoch": 0.47591, + "grad_norm": 0.8463317505410337, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 47591 + }, + { + "epoch": 0.47592, + "grad_norm": 0.9551146542378713, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 47592 + }, + { + "epoch": 0.47593, + "grad_norm": 1.1512294250669488, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 47593 + }, + { + "epoch": 0.47594, + "grad_norm": 0.7567175921467661, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 47594 + }, + { + "epoch": 0.47595, + "grad_norm": 0.7304758654570488, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 47595 + }, + { + "epoch": 0.47596, + "grad_norm": 0.658253633738951, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 47596 + }, + { + "epoch": 0.47597, + "grad_norm": 0.6365427533628638, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 47597 + }, + { + "epoch": 0.47598, + "grad_norm": 0.6612219249810363, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 47598 + }, + { + "epoch": 0.47599, + "grad_norm": 0.717902716937759, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 47599 + }, + { + "epoch": 0.476, + "grad_norm": 0.7770061359052305, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 47600 + }, + { + "epoch": 0.47601, + "grad_norm": 0.8656633469963074, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 47601 + }, + { + "epoch": 0.47602, + "grad_norm": 0.9237058023605236, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 47602 + }, + { + "epoch": 0.47603, + "grad_norm": 0.9050612249972814, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 47603 + }, + { + "epoch": 0.47604, + "grad_norm": 0.9575004881065737, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 47604 + }, + { + "epoch": 0.47605, + "grad_norm": 1.0162168965789233, + "learning_rate": 0.003, + "loss": 4.013, + "step": 47605 + }, + { + "epoch": 0.47606, + "grad_norm": 0.9335992707018607, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 47606 + }, + { + "epoch": 0.47607, + "grad_norm": 0.8269743986386406, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 47607 + }, + { + "epoch": 0.47608, + "grad_norm": 0.7878688607179206, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 47608 + }, + { + "epoch": 0.47609, + "grad_norm": 0.7891559301493458, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 47609 + }, + { + "epoch": 0.4761, + "grad_norm": 0.8916173371455612, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 47610 + }, + { + "epoch": 0.47611, + "grad_norm": 0.9829518134643661, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 47611 + }, + { + "epoch": 0.47612, + "grad_norm": 0.8989906967945624, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 47612 + }, + { + "epoch": 0.47613, + "grad_norm": 0.870288657103143, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 47613 + }, + { + "epoch": 0.47614, + "grad_norm": 0.8542473865212752, + "learning_rate": 0.003, + "loss": 3.984, + "step": 47614 + }, + { + "epoch": 0.47615, + "grad_norm": 0.7970575737503044, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 47615 + }, + { + "epoch": 0.47616, + "grad_norm": 0.673571866219935, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 47616 + }, + { + "epoch": 0.47617, + "grad_norm": 0.6612343852044388, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 47617 + }, + { + "epoch": 0.47618, + "grad_norm": 0.6724221567667473, + "learning_rate": 0.003, + "loss": 3.973, + "step": 47618 + }, + { + "epoch": 0.47619, + "grad_norm": 0.6542949953094124, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 47619 + }, + { + "epoch": 0.4762, + "grad_norm": 0.6626878587657076, + "learning_rate": 0.003, + "loss": 3.975, + "step": 47620 + }, + { + "epoch": 0.47621, + "grad_norm": 0.6485432676001988, + "learning_rate": 0.003, + "loss": 3.99, + "step": 47621 + }, + { + "epoch": 0.47622, + "grad_norm": 0.6296133774355623, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 47622 + }, + { + "epoch": 0.47623, + "grad_norm": 0.6498619914810236, + "learning_rate": 0.003, + "loss": 3.974, + "step": 47623 + }, + { + "epoch": 0.47624, + "grad_norm": 0.6144668231091809, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 47624 + }, + { + "epoch": 0.47625, + "grad_norm": 0.7688202714521777, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 47625 + }, + { + "epoch": 0.47626, + "grad_norm": 0.8546166850136624, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 47626 + }, + { + "epoch": 0.47627, + "grad_norm": 0.7777715620975542, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 47627 + }, + { + "epoch": 0.47628, + "grad_norm": 0.7026406690019137, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 47628 + }, + { + "epoch": 0.47629, + "grad_norm": 0.7084082316578608, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 47629 + }, + { + "epoch": 0.4763, + "grad_norm": 0.7510941794346014, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 47630 + }, + { + "epoch": 0.47631, + "grad_norm": 0.8228850296587118, + "learning_rate": 0.003, + "loss": 4.004, + "step": 47631 + }, + { + "epoch": 0.47632, + "grad_norm": 0.8683153173576683, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 47632 + }, + { + "epoch": 0.47633, + "grad_norm": 0.8026020715088721, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 47633 + }, + { + "epoch": 0.47634, + "grad_norm": 0.727421035112356, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 47634 + }, + { + "epoch": 0.47635, + "grad_norm": 0.7585047903986433, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 47635 + }, + { + "epoch": 0.47636, + "grad_norm": 0.8026893100820564, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 47636 + }, + { + "epoch": 0.47637, + "grad_norm": 0.7437662335192315, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 47637 + }, + { + "epoch": 0.47638, + "grad_norm": 0.7289277147947969, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 47638 + }, + { + "epoch": 0.47639, + "grad_norm": 0.7817025703302178, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 47639 + }, + { + "epoch": 0.4764, + "grad_norm": 0.7758525610813205, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 47640 + }, + { + "epoch": 0.47641, + "grad_norm": 0.7011771284739615, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 47641 + }, + { + "epoch": 0.47642, + "grad_norm": 0.7359569875838591, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 47642 + }, + { + "epoch": 0.47643, + "grad_norm": 0.8195706920968207, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 47643 + }, + { + "epoch": 0.47644, + "grad_norm": 0.8878186510046845, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 47644 + }, + { + "epoch": 0.47645, + "grad_norm": 0.9062580293430279, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 47645 + }, + { + "epoch": 0.47646, + "grad_norm": 0.92372996801626, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 47646 + }, + { + "epoch": 0.47647, + "grad_norm": 1.0795527487471863, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 47647 + }, + { + "epoch": 0.47648, + "grad_norm": 0.8673890200566345, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 47648 + }, + { + "epoch": 0.47649, + "grad_norm": 0.7947039490193445, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 47649 + }, + { + "epoch": 0.4765, + "grad_norm": 0.7523241482217755, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 47650 + }, + { + "epoch": 0.47651, + "grad_norm": 0.7540693693087575, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 47651 + }, + { + "epoch": 0.47652, + "grad_norm": 0.7706081473601534, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 47652 + }, + { + "epoch": 0.47653, + "grad_norm": 0.8238972689056997, + "learning_rate": 0.003, + "loss": 4.004, + "step": 47653 + }, + { + "epoch": 0.47654, + "grad_norm": 0.840137536601707, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 47654 + }, + { + "epoch": 0.47655, + "grad_norm": 0.9007138938140736, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 47655 + }, + { + "epoch": 0.47656, + "grad_norm": 0.901847284552641, + "learning_rate": 0.003, + "loss": 3.962, + "step": 47656 + }, + { + "epoch": 0.47657, + "grad_norm": 0.6983475143241749, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 47657 + }, + { + "epoch": 0.47658, + "grad_norm": 0.6629535297658214, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 47658 + }, + { + "epoch": 0.47659, + "grad_norm": 0.7826319758381384, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 47659 + }, + { + "epoch": 0.4766, + "grad_norm": 0.7331975250585286, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 47660 + }, + { + "epoch": 0.47661, + "grad_norm": 0.7451522509814117, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 47661 + }, + { + "epoch": 0.47662, + "grad_norm": 0.79034427606473, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 47662 + }, + { + "epoch": 0.47663, + "grad_norm": 0.812417561584997, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 47663 + }, + { + "epoch": 0.47664, + "grad_norm": 0.7988824968194251, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 47664 + }, + { + "epoch": 0.47665, + "grad_norm": 0.7647487075983482, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 47665 + }, + { + "epoch": 0.47666, + "grad_norm": 0.7046244286879608, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 47666 + }, + { + "epoch": 0.47667, + "grad_norm": 0.7962654368913972, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 47667 + }, + { + "epoch": 0.47668, + "grad_norm": 0.7928280864576004, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 47668 + }, + { + "epoch": 0.47669, + "grad_norm": 0.7268360617154023, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 47669 + }, + { + "epoch": 0.4767, + "grad_norm": 0.7024104393765291, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 47670 + }, + { + "epoch": 0.47671, + "grad_norm": 0.5958798142184806, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 47671 + }, + { + "epoch": 0.47672, + "grad_norm": 0.5073181120669975, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 47672 + }, + { + "epoch": 0.47673, + "grad_norm": 0.6364266116001905, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 47673 + }, + { + "epoch": 0.47674, + "grad_norm": 0.7458534480362994, + "learning_rate": 0.003, + "loss": 4.001, + "step": 47674 + }, + { + "epoch": 0.47675, + "grad_norm": 0.8528406862403523, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 47675 + }, + { + "epoch": 0.47676, + "grad_norm": 0.9431118934954136, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 47676 + }, + { + "epoch": 0.47677, + "grad_norm": 1.0200475672098288, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 47677 + }, + { + "epoch": 0.47678, + "grad_norm": 1.0757741196988186, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 47678 + }, + { + "epoch": 0.47679, + "grad_norm": 0.9495132177867237, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 47679 + }, + { + "epoch": 0.4768, + "grad_norm": 0.8276927143910425, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 47680 + }, + { + "epoch": 0.47681, + "grad_norm": 0.8097162219461526, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 47681 + }, + { + "epoch": 0.47682, + "grad_norm": 0.7463035682168934, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 47682 + }, + { + "epoch": 0.47683, + "grad_norm": 0.7909813835063716, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 47683 + }, + { + "epoch": 0.47684, + "grad_norm": 0.7713998457314368, + "learning_rate": 0.003, + "loss": 3.975, + "step": 47684 + }, + { + "epoch": 0.47685, + "grad_norm": 0.6954379759080078, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 47685 + }, + { + "epoch": 0.47686, + "grad_norm": 0.7215877894338508, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 47686 + }, + { + "epoch": 0.47687, + "grad_norm": 0.7662742834547378, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 47687 + }, + { + "epoch": 0.47688, + "grad_norm": 0.9271752489799445, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 47688 + }, + { + "epoch": 0.47689, + "grad_norm": 0.9751294406888266, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 47689 + }, + { + "epoch": 0.4769, + "grad_norm": 0.8994576687893435, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 47690 + }, + { + "epoch": 0.47691, + "grad_norm": 0.8730103835941124, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 47691 + }, + { + "epoch": 0.47692, + "grad_norm": 0.8086551112353585, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 47692 + }, + { + "epoch": 0.47693, + "grad_norm": 0.562669833205488, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 47693 + }, + { + "epoch": 0.47694, + "grad_norm": 0.6090223858311209, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 47694 + }, + { + "epoch": 0.47695, + "grad_norm": 0.5961722518367865, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 47695 + }, + { + "epoch": 0.47696, + "grad_norm": 0.6102876332034336, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 47696 + }, + { + "epoch": 0.47697, + "grad_norm": 0.7312281684648155, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 47697 + }, + { + "epoch": 0.47698, + "grad_norm": 0.8194058835913122, + "learning_rate": 0.003, + "loss": 3.989, + "step": 47698 + }, + { + "epoch": 0.47699, + "grad_norm": 0.7608188257909749, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 47699 + }, + { + "epoch": 0.477, + "grad_norm": 0.6941637400794692, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 47700 + }, + { + "epoch": 0.47701, + "grad_norm": 0.8350547010115815, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 47701 + }, + { + "epoch": 0.47702, + "grad_norm": 0.7268854771239796, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 47702 + }, + { + "epoch": 0.47703, + "grad_norm": 0.7102010429310209, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 47703 + }, + { + "epoch": 0.47704, + "grad_norm": 0.8115231619362215, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 47704 + }, + { + "epoch": 0.47705, + "grad_norm": 0.8492215226728994, + "learning_rate": 0.003, + "loss": 3.985, + "step": 47705 + }, + { + "epoch": 0.47706, + "grad_norm": 0.9692695842057709, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 47706 + }, + { + "epoch": 0.47707, + "grad_norm": 1.0501032317481407, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 47707 + }, + { + "epoch": 0.47708, + "grad_norm": 1.0016592931021964, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 47708 + }, + { + "epoch": 0.47709, + "grad_norm": 0.9200403011079212, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 47709 + }, + { + "epoch": 0.4771, + "grad_norm": 0.7168631906655512, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 47710 + }, + { + "epoch": 0.47711, + "grad_norm": 0.6942459309818011, + "learning_rate": 0.003, + "loss": 3.996, + "step": 47711 + }, + { + "epoch": 0.47712, + "grad_norm": 0.8386955530989085, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 47712 + }, + { + "epoch": 0.47713, + "grad_norm": 0.883473260898607, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 47713 + }, + { + "epoch": 0.47714, + "grad_norm": 0.8564794927051947, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 47714 + }, + { + "epoch": 0.47715, + "grad_norm": 0.7762384961500278, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 47715 + }, + { + "epoch": 0.47716, + "grad_norm": 0.8110315516638277, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 47716 + }, + { + "epoch": 0.47717, + "grad_norm": 0.8231309294800653, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 47717 + }, + { + "epoch": 0.47718, + "grad_norm": 0.6863678246896507, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 47718 + }, + { + "epoch": 0.47719, + "grad_norm": 0.6996653106373792, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 47719 + }, + { + "epoch": 0.4772, + "grad_norm": 0.7501305607176235, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 47720 + }, + { + "epoch": 0.47721, + "grad_norm": 0.8673359470969905, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 47721 + }, + { + "epoch": 0.47722, + "grad_norm": 0.8489341257106228, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 47722 + }, + { + "epoch": 0.47723, + "grad_norm": 0.7388800506335641, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 47723 + }, + { + "epoch": 0.47724, + "grad_norm": 0.6950009737566392, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 47724 + }, + { + "epoch": 0.47725, + "grad_norm": 0.6929068852671212, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 47725 + }, + { + "epoch": 0.47726, + "grad_norm": 0.7736944753643312, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 47726 + }, + { + "epoch": 0.47727, + "grad_norm": 0.871750713233798, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 47727 + }, + { + "epoch": 0.47728, + "grad_norm": 0.9385676532263043, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 47728 + }, + { + "epoch": 0.47729, + "grad_norm": 1.002867498284725, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 47729 + }, + { + "epoch": 0.4773, + "grad_norm": 0.9217244705717861, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 47730 + }, + { + "epoch": 0.47731, + "grad_norm": 0.9875859877145615, + "learning_rate": 0.003, + "loss": 3.994, + "step": 47731 + }, + { + "epoch": 0.47732, + "grad_norm": 0.947594313705036, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 47732 + }, + { + "epoch": 0.47733, + "grad_norm": 0.9832268494288808, + "learning_rate": 0.003, + "loss": 4.006, + "step": 47733 + }, + { + "epoch": 0.47734, + "grad_norm": 0.9683253190074271, + "learning_rate": 0.003, + "loss": 4.0393, + "step": 47734 + }, + { + "epoch": 0.47735, + "grad_norm": 0.9420344472666172, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 47735 + }, + { + "epoch": 0.47736, + "grad_norm": 0.9187997794928204, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 47736 + }, + { + "epoch": 0.47737, + "grad_norm": 0.9027781277432755, + "learning_rate": 0.003, + "loss": 4.0204, + "step": 47737 + }, + { + "epoch": 0.47738, + "grad_norm": 0.9085173128339313, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 47738 + }, + { + "epoch": 0.47739, + "grad_norm": 0.8597613881811277, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 47739 + }, + { + "epoch": 0.4774, + "grad_norm": 0.6585305352682941, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 47740 + }, + { + "epoch": 0.47741, + "grad_norm": 0.5744638865624365, + "learning_rate": 0.003, + "loss": 4.0039, + "step": 47741 + }, + { + "epoch": 0.47742, + "grad_norm": 0.5791158966512956, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 47742 + }, + { + "epoch": 0.47743, + "grad_norm": 0.594010127430501, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 47743 + }, + { + "epoch": 0.47744, + "grad_norm": 0.7219453432564873, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 47744 + }, + { + "epoch": 0.47745, + "grad_norm": 0.7324373748654369, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 47745 + }, + { + "epoch": 0.47746, + "grad_norm": 0.5891514353470979, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 47746 + }, + { + "epoch": 0.47747, + "grad_norm": 0.6179056958960585, + "learning_rate": 0.003, + "loss": 3.961, + "step": 47747 + }, + { + "epoch": 0.47748, + "grad_norm": 0.6833111373066117, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 47748 + }, + { + "epoch": 0.47749, + "grad_norm": 0.8040055480790904, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 47749 + }, + { + "epoch": 0.4775, + "grad_norm": 0.8683424853404167, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 47750 + }, + { + "epoch": 0.47751, + "grad_norm": 0.8643367350885004, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 47751 + }, + { + "epoch": 0.47752, + "grad_norm": 0.8402787844841716, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 47752 + }, + { + "epoch": 0.47753, + "grad_norm": 0.8055620135636062, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 47753 + }, + { + "epoch": 0.47754, + "grad_norm": 0.8062249299912202, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 47754 + }, + { + "epoch": 0.47755, + "grad_norm": 0.7269049031070335, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 47755 + }, + { + "epoch": 0.47756, + "grad_norm": 0.7028730863042688, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 47756 + }, + { + "epoch": 0.47757, + "grad_norm": 0.6139925408757865, + "learning_rate": 0.003, + "loss": 3.965, + "step": 47757 + }, + { + "epoch": 0.47758, + "grad_norm": 0.6379371181269605, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 47758 + }, + { + "epoch": 0.47759, + "grad_norm": 0.5869365164244115, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 47759 + }, + { + "epoch": 0.4776, + "grad_norm": 0.5564452515374105, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 47760 + }, + { + "epoch": 0.47761, + "grad_norm": 0.5171165934615999, + "learning_rate": 0.003, + "loss": 4.002, + "step": 47761 + }, + { + "epoch": 0.47762, + "grad_norm": 0.4833329231316688, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 47762 + }, + { + "epoch": 0.47763, + "grad_norm": 0.5285209437603542, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 47763 + }, + { + "epoch": 0.47764, + "grad_norm": 0.5821692579488225, + "learning_rate": 0.003, + "loss": 3.942, + "step": 47764 + }, + { + "epoch": 0.47765, + "grad_norm": 0.6890122854893809, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 47765 + }, + { + "epoch": 0.47766, + "grad_norm": 0.8645770093824408, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 47766 + }, + { + "epoch": 0.47767, + "grad_norm": 1.0362514084822876, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 47767 + }, + { + "epoch": 0.47768, + "grad_norm": 0.8036540266179978, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 47768 + }, + { + "epoch": 0.47769, + "grad_norm": 0.828953229893694, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 47769 + }, + { + "epoch": 0.4777, + "grad_norm": 0.9990733639881566, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 47770 + }, + { + "epoch": 0.47771, + "grad_norm": 1.1333770132934118, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 47771 + }, + { + "epoch": 0.47772, + "grad_norm": 0.71293988358807, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 47772 + }, + { + "epoch": 0.47773, + "grad_norm": 0.718548282417372, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 47773 + }, + { + "epoch": 0.47774, + "grad_norm": 0.7610862867915986, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 47774 + }, + { + "epoch": 0.47775, + "grad_norm": 0.7024222072249467, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 47775 + }, + { + "epoch": 0.47776, + "grad_norm": 0.6016258172881644, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 47776 + }, + { + "epoch": 0.47777, + "grad_norm": 0.5911522362334819, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 47777 + }, + { + "epoch": 0.47778, + "grad_norm": 0.6660240391743508, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 47778 + }, + { + "epoch": 0.47779, + "grad_norm": 0.7350513456904105, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 47779 + }, + { + "epoch": 0.4778, + "grad_norm": 0.8137336449371746, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 47780 + }, + { + "epoch": 0.47781, + "grad_norm": 0.722904640856469, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 47781 + }, + { + "epoch": 0.47782, + "grad_norm": 0.6983196599934531, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 47782 + }, + { + "epoch": 0.47783, + "grad_norm": 0.7561278877439501, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 47783 + }, + { + "epoch": 0.47784, + "grad_norm": 0.9539019313673942, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 47784 + }, + { + "epoch": 0.47785, + "grad_norm": 1.0511181918423258, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 47785 + }, + { + "epoch": 0.47786, + "grad_norm": 1.0299573271404914, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 47786 + }, + { + "epoch": 0.47787, + "grad_norm": 0.9909336203155836, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 47787 + }, + { + "epoch": 0.47788, + "grad_norm": 0.8816487742513328, + "learning_rate": 0.003, + "loss": 3.985, + "step": 47788 + }, + { + "epoch": 0.47789, + "grad_norm": 0.8399429610862007, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 47789 + }, + { + "epoch": 0.4779, + "grad_norm": 0.8562012546661159, + "learning_rate": 0.003, + "loss": 4.017, + "step": 47790 + }, + { + "epoch": 0.47791, + "grad_norm": 0.7420684119648524, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 47791 + }, + { + "epoch": 0.47792, + "grad_norm": 0.7314985196097037, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 47792 + }, + { + "epoch": 0.47793, + "grad_norm": 0.8593359224633232, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 47793 + }, + { + "epoch": 0.47794, + "grad_norm": 1.048374247144701, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 47794 + }, + { + "epoch": 0.47795, + "grad_norm": 1.0743976550188812, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 47795 + }, + { + "epoch": 0.47796, + "grad_norm": 0.9220442083435549, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 47796 + }, + { + "epoch": 0.47797, + "grad_norm": 0.861520116908853, + "learning_rate": 0.003, + "loss": 3.996, + "step": 47797 + }, + { + "epoch": 0.47798, + "grad_norm": 0.8435621921005869, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 47798 + }, + { + "epoch": 0.47799, + "grad_norm": 0.8122647389719769, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 47799 + }, + { + "epoch": 0.478, + "grad_norm": 0.8742683245617667, + "learning_rate": 0.003, + "loss": 3.993, + "step": 47800 + }, + { + "epoch": 0.47801, + "grad_norm": 0.9114975003305117, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 47801 + }, + { + "epoch": 0.47802, + "grad_norm": 0.9822933031034322, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 47802 + }, + { + "epoch": 0.47803, + "grad_norm": 0.9969470743470826, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 47803 + }, + { + "epoch": 0.47804, + "grad_norm": 0.908461500332866, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 47804 + }, + { + "epoch": 0.47805, + "grad_norm": 0.858948576186745, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 47805 + }, + { + "epoch": 0.47806, + "grad_norm": 0.8194395786424477, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 47806 + }, + { + "epoch": 0.47807, + "grad_norm": 0.6910558287353098, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 47807 + }, + { + "epoch": 0.47808, + "grad_norm": 0.6560735735527073, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 47808 + }, + { + "epoch": 0.47809, + "grad_norm": 0.5984381200322784, + "learning_rate": 0.003, + "loss": 3.985, + "step": 47809 + }, + { + "epoch": 0.4781, + "grad_norm": 0.606965083897499, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 47810 + }, + { + "epoch": 0.47811, + "grad_norm": 0.6056252502650438, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 47811 + }, + { + "epoch": 0.47812, + "grad_norm": 0.5402849095040014, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 47812 + }, + { + "epoch": 0.47813, + "grad_norm": 0.5860899012103203, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 47813 + }, + { + "epoch": 0.47814, + "grad_norm": 0.6090656548084942, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 47814 + }, + { + "epoch": 0.47815, + "grad_norm": 0.5866523262137021, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 47815 + }, + { + "epoch": 0.47816, + "grad_norm": 0.604700321464007, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 47816 + }, + { + "epoch": 0.47817, + "grad_norm": 0.565158616834833, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 47817 + }, + { + "epoch": 0.47818, + "grad_norm": 0.5930552135950996, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 47818 + }, + { + "epoch": 0.47819, + "grad_norm": 0.681518462287789, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 47819 + }, + { + "epoch": 0.4782, + "grad_norm": 0.7387590124998612, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 47820 + }, + { + "epoch": 0.47821, + "grad_norm": 0.8196129983987969, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 47821 + }, + { + "epoch": 0.47822, + "grad_norm": 0.8147438029978511, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 47822 + }, + { + "epoch": 0.47823, + "grad_norm": 0.9070517922901147, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 47823 + }, + { + "epoch": 0.47824, + "grad_norm": 1.0004253162523173, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 47824 + }, + { + "epoch": 0.47825, + "grad_norm": 1.0907792503382943, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 47825 + }, + { + "epoch": 0.47826, + "grad_norm": 0.97923931783567, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 47826 + }, + { + "epoch": 0.47827, + "grad_norm": 0.9024553278959124, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 47827 + }, + { + "epoch": 0.47828, + "grad_norm": 0.9058484143483199, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 47828 + }, + { + "epoch": 0.47829, + "grad_norm": 0.9404207234079212, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 47829 + }, + { + "epoch": 0.4783, + "grad_norm": 0.8548018285380775, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 47830 + }, + { + "epoch": 0.47831, + "grad_norm": 0.8246311005733216, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 47831 + }, + { + "epoch": 0.47832, + "grad_norm": 0.7488745609189696, + "learning_rate": 0.003, + "loss": 3.94, + "step": 47832 + }, + { + "epoch": 0.47833, + "grad_norm": 0.6588385436775067, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 47833 + }, + { + "epoch": 0.47834, + "grad_norm": 0.6315308409372294, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 47834 + }, + { + "epoch": 0.47835, + "grad_norm": 0.6056451427895266, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 47835 + }, + { + "epoch": 0.47836, + "grad_norm": 0.8022838446088001, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 47836 + }, + { + "epoch": 0.47837, + "grad_norm": 1.006931233284138, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 47837 + }, + { + "epoch": 0.47838, + "grad_norm": 1.012703135985482, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 47838 + }, + { + "epoch": 0.47839, + "grad_norm": 0.8463202752510618, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 47839 + }, + { + "epoch": 0.4784, + "grad_norm": 0.6852317902682046, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 47840 + }, + { + "epoch": 0.47841, + "grad_norm": 0.67717332183975, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 47841 + }, + { + "epoch": 0.47842, + "grad_norm": 0.715960379307335, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 47842 + }, + { + "epoch": 0.47843, + "grad_norm": 0.7319900452732347, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 47843 + }, + { + "epoch": 0.47844, + "grad_norm": 0.7573333141147893, + "learning_rate": 0.003, + "loss": 4.001, + "step": 47844 + }, + { + "epoch": 0.47845, + "grad_norm": 0.8709466289924057, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 47845 + }, + { + "epoch": 0.47846, + "grad_norm": 0.9330636567994143, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 47846 + }, + { + "epoch": 0.47847, + "grad_norm": 1.0619271545099769, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 47847 + }, + { + "epoch": 0.47848, + "grad_norm": 0.9284344703288095, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 47848 + }, + { + "epoch": 0.47849, + "grad_norm": 0.8127895750853147, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 47849 + }, + { + "epoch": 0.4785, + "grad_norm": 0.7586012310828116, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 47850 + }, + { + "epoch": 0.47851, + "grad_norm": 0.6904467218396902, + "learning_rate": 0.003, + "loss": 3.971, + "step": 47851 + }, + { + "epoch": 0.47852, + "grad_norm": 0.654722969337331, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 47852 + }, + { + "epoch": 0.47853, + "grad_norm": 0.6719118300707259, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 47853 + }, + { + "epoch": 0.47854, + "grad_norm": 0.749547264041494, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 47854 + }, + { + "epoch": 0.47855, + "grad_norm": 0.8138873824005519, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 47855 + }, + { + "epoch": 0.47856, + "grad_norm": 0.8253211568612036, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 47856 + }, + { + "epoch": 0.47857, + "grad_norm": 0.8568751344575204, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 47857 + }, + { + "epoch": 0.47858, + "grad_norm": 0.9472434042210819, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 47858 + }, + { + "epoch": 0.47859, + "grad_norm": 0.8384333508004288, + "learning_rate": 0.003, + "loss": 3.9252, + "step": 47859 + }, + { + "epoch": 0.4786, + "grad_norm": 0.7105932060192394, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 47860 + }, + { + "epoch": 0.47861, + "grad_norm": 0.6255278049594869, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 47861 + }, + { + "epoch": 0.47862, + "grad_norm": 0.6959562938274282, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 47862 + }, + { + "epoch": 0.47863, + "grad_norm": 0.7467258980816571, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 47863 + }, + { + "epoch": 0.47864, + "grad_norm": 0.7349297932114462, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 47864 + }, + { + "epoch": 0.47865, + "grad_norm": 0.7030767743456536, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 47865 + }, + { + "epoch": 0.47866, + "grad_norm": 0.6817377594774302, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 47866 + }, + { + "epoch": 0.47867, + "grad_norm": 0.6639275581578923, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 47867 + }, + { + "epoch": 0.47868, + "grad_norm": 0.6761591337672216, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 47868 + }, + { + "epoch": 0.47869, + "grad_norm": 0.7868747289555323, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 47869 + }, + { + "epoch": 0.4787, + "grad_norm": 0.7502443838735883, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 47870 + }, + { + "epoch": 0.47871, + "grad_norm": 0.7392118891867828, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 47871 + }, + { + "epoch": 0.47872, + "grad_norm": 0.7555924176497165, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 47872 + }, + { + "epoch": 0.47873, + "grad_norm": 0.7094657994427362, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 47873 + }, + { + "epoch": 0.47874, + "grad_norm": 0.7418263495448576, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 47874 + }, + { + "epoch": 0.47875, + "grad_norm": 0.9467482366347004, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 47875 + }, + { + "epoch": 0.47876, + "grad_norm": 0.9839770675896341, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 47876 + }, + { + "epoch": 0.47877, + "grad_norm": 0.8874771279523064, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 47877 + }, + { + "epoch": 0.47878, + "grad_norm": 0.8006776311484799, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 47878 + }, + { + "epoch": 0.47879, + "grad_norm": 0.9166674754259656, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 47879 + }, + { + "epoch": 0.4788, + "grad_norm": 1.019717268196481, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 47880 + }, + { + "epoch": 0.47881, + "grad_norm": 0.8293863676096486, + "learning_rate": 0.003, + "loss": 4.0, + "step": 47881 + }, + { + "epoch": 0.47882, + "grad_norm": 0.7772478915313992, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 47882 + }, + { + "epoch": 0.47883, + "grad_norm": 0.8698981577555857, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 47883 + }, + { + "epoch": 0.47884, + "grad_norm": 0.9379820494101071, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 47884 + }, + { + "epoch": 0.47885, + "grad_norm": 0.9724978865207167, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 47885 + }, + { + "epoch": 0.47886, + "grad_norm": 0.9991425305777722, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 47886 + }, + { + "epoch": 0.47887, + "grad_norm": 0.8122575964606684, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 47887 + }, + { + "epoch": 0.47888, + "grad_norm": 0.8275500306584024, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 47888 + }, + { + "epoch": 0.47889, + "grad_norm": 0.8115639677246481, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 47889 + }, + { + "epoch": 0.4789, + "grad_norm": 0.7595870608213149, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 47890 + }, + { + "epoch": 0.47891, + "grad_norm": 0.7883914643489442, + "learning_rate": 0.003, + "loss": 3.96, + "step": 47891 + }, + { + "epoch": 0.47892, + "grad_norm": 1.0122286702524559, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 47892 + }, + { + "epoch": 0.47893, + "grad_norm": 0.9398517198725216, + "learning_rate": 0.003, + "loss": 4.019, + "step": 47893 + }, + { + "epoch": 0.47894, + "grad_norm": 0.7924866510234423, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 47894 + }, + { + "epoch": 0.47895, + "grad_norm": 0.7043106057776343, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 47895 + }, + { + "epoch": 0.47896, + "grad_norm": 0.6161369690760247, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 47896 + }, + { + "epoch": 0.47897, + "grad_norm": 0.6172491591383096, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 47897 + }, + { + "epoch": 0.47898, + "grad_norm": 0.6351984238836058, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 47898 + }, + { + "epoch": 0.47899, + "grad_norm": 0.6754295124272242, + "learning_rate": 0.003, + "loss": 3.976, + "step": 47899 + }, + { + "epoch": 0.479, + "grad_norm": 0.6198933390410192, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 47900 + }, + { + "epoch": 0.47901, + "grad_norm": 0.6530373297198904, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 47901 + }, + { + "epoch": 0.47902, + "grad_norm": 0.6107915466240024, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 47902 + }, + { + "epoch": 0.47903, + "grad_norm": 0.6480618262156835, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 47903 + }, + { + "epoch": 0.47904, + "grad_norm": 0.6514014196214424, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 47904 + }, + { + "epoch": 0.47905, + "grad_norm": 0.6123557597807223, + "learning_rate": 0.003, + "loss": 3.956, + "step": 47905 + }, + { + "epoch": 0.47906, + "grad_norm": 0.5670994546247652, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 47906 + }, + { + "epoch": 0.47907, + "grad_norm": 0.48636319997722127, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 47907 + }, + { + "epoch": 0.47908, + "grad_norm": 0.4226257152781811, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 47908 + }, + { + "epoch": 0.47909, + "grad_norm": 0.48046126187338967, + "learning_rate": 0.003, + "loss": 3.973, + "step": 47909 + }, + { + "epoch": 0.4791, + "grad_norm": 0.4908484239810384, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 47910 + }, + { + "epoch": 0.47911, + "grad_norm": 0.5734391010618972, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 47911 + }, + { + "epoch": 0.47912, + "grad_norm": 0.6489275455997413, + "learning_rate": 0.003, + "loss": 3.9351, + "step": 47912 + }, + { + "epoch": 0.47913, + "grad_norm": 0.7064497791735769, + "learning_rate": 0.003, + "loss": 3.976, + "step": 47913 + }, + { + "epoch": 0.47914, + "grad_norm": 0.8544350860350383, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 47914 + }, + { + "epoch": 0.47915, + "grad_norm": 0.9615226129485696, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 47915 + }, + { + "epoch": 0.47916, + "grad_norm": 0.8804556075457045, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 47916 + }, + { + "epoch": 0.47917, + "grad_norm": 0.7511641366925776, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 47917 + }, + { + "epoch": 0.47918, + "grad_norm": 0.7876436352108613, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 47918 + }, + { + "epoch": 0.47919, + "grad_norm": 0.9434065461934913, + "learning_rate": 0.003, + "loss": 3.9415, + "step": 47919 + }, + { + "epoch": 0.4792, + "grad_norm": 0.9826235950831105, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 47920 + }, + { + "epoch": 0.47921, + "grad_norm": 0.9779760281653731, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 47921 + }, + { + "epoch": 0.47922, + "grad_norm": 0.9578305493370726, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 47922 + }, + { + "epoch": 0.47923, + "grad_norm": 0.8852556492620557, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 47923 + }, + { + "epoch": 0.47924, + "grad_norm": 0.9316075952581034, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 47924 + }, + { + "epoch": 0.47925, + "grad_norm": 0.9446799924578125, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 47925 + }, + { + "epoch": 0.47926, + "grad_norm": 0.8261897500591754, + "learning_rate": 0.003, + "loss": 3.995, + "step": 47926 + }, + { + "epoch": 0.47927, + "grad_norm": 0.8287322273148158, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 47927 + }, + { + "epoch": 0.47928, + "grad_norm": 0.8279459254803941, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 47928 + }, + { + "epoch": 0.47929, + "grad_norm": 0.9606528379585472, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 47929 + }, + { + "epoch": 0.4793, + "grad_norm": 1.1115854952407884, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 47930 + }, + { + "epoch": 0.47931, + "grad_norm": 0.824680340993057, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 47931 + }, + { + "epoch": 0.47932, + "grad_norm": 0.7805958356601757, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 47932 + }, + { + "epoch": 0.47933, + "grad_norm": 0.7188242445380768, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 47933 + }, + { + "epoch": 0.47934, + "grad_norm": 0.6825480569459224, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 47934 + }, + { + "epoch": 0.47935, + "grad_norm": 0.6916448465352613, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 47935 + }, + { + "epoch": 0.47936, + "grad_norm": 0.6931851651650436, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 47936 + }, + { + "epoch": 0.47937, + "grad_norm": 0.6139043394000578, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 47937 + }, + { + "epoch": 0.47938, + "grad_norm": 0.6044519845160845, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 47938 + }, + { + "epoch": 0.47939, + "grad_norm": 0.7731194200398055, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 47939 + }, + { + "epoch": 0.4794, + "grad_norm": 0.8503672935168892, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 47940 + }, + { + "epoch": 0.47941, + "grad_norm": 0.851502964416735, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 47941 + }, + { + "epoch": 0.47942, + "grad_norm": 0.8464460817890403, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 47942 + }, + { + "epoch": 0.47943, + "grad_norm": 0.9119150745440088, + "learning_rate": 0.003, + "loss": 4.0158, + "step": 47943 + }, + { + "epoch": 0.47944, + "grad_norm": 0.8968084733067556, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 47944 + }, + { + "epoch": 0.47945, + "grad_norm": 0.9503921958166454, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 47945 + }, + { + "epoch": 0.47946, + "grad_norm": 0.8460410463480532, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 47946 + }, + { + "epoch": 0.47947, + "grad_norm": 0.8276239456698116, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 47947 + }, + { + "epoch": 0.47948, + "grad_norm": 0.8459205207267195, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 47948 + }, + { + "epoch": 0.47949, + "grad_norm": 0.7823924724194651, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 47949 + }, + { + "epoch": 0.4795, + "grad_norm": 0.6783648524080181, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 47950 + }, + { + "epoch": 0.47951, + "grad_norm": 0.5716743041294992, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 47951 + }, + { + "epoch": 0.47952, + "grad_norm": 0.6259969474699768, + "learning_rate": 0.003, + "loss": 3.989, + "step": 47952 + }, + { + "epoch": 0.47953, + "grad_norm": 0.5894881726102184, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 47953 + }, + { + "epoch": 0.47954, + "grad_norm": 0.647351688187213, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 47954 + }, + { + "epoch": 0.47955, + "grad_norm": 0.6286422862998443, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 47955 + }, + { + "epoch": 0.47956, + "grad_norm": 0.6830179539698124, + "learning_rate": 0.003, + "loss": 3.987, + "step": 47956 + }, + { + "epoch": 0.47957, + "grad_norm": 0.7721441812040244, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 47957 + }, + { + "epoch": 0.47958, + "grad_norm": 0.8823052469220439, + "learning_rate": 0.003, + "loss": 3.982, + "step": 47958 + }, + { + "epoch": 0.47959, + "grad_norm": 0.9712693717529871, + "learning_rate": 0.003, + "loss": 3.962, + "step": 47959 + }, + { + "epoch": 0.4796, + "grad_norm": 0.9937169952520571, + "learning_rate": 0.003, + "loss": 3.96, + "step": 47960 + }, + { + "epoch": 0.47961, + "grad_norm": 0.9664751067248989, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 47961 + }, + { + "epoch": 0.47962, + "grad_norm": 0.820353520847246, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 47962 + }, + { + "epoch": 0.47963, + "grad_norm": 0.6617769971498649, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 47963 + }, + { + "epoch": 0.47964, + "grad_norm": 0.684425951866006, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 47964 + }, + { + "epoch": 0.47965, + "grad_norm": 0.5570545022859076, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 47965 + }, + { + "epoch": 0.47966, + "grad_norm": 0.6499422918437054, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 47966 + }, + { + "epoch": 0.47967, + "grad_norm": 0.6959509237766738, + "learning_rate": 0.003, + "loss": 3.937, + "step": 47967 + }, + { + "epoch": 0.47968, + "grad_norm": 0.7805815737015017, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 47968 + }, + { + "epoch": 0.47969, + "grad_norm": 0.7644604754608931, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 47969 + }, + { + "epoch": 0.4797, + "grad_norm": 0.7925826820387876, + "learning_rate": 0.003, + "loss": 3.997, + "step": 47970 + }, + { + "epoch": 0.47971, + "grad_norm": 0.9455532725128444, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 47971 + }, + { + "epoch": 0.47972, + "grad_norm": 1.075693667024342, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 47972 + }, + { + "epoch": 0.47973, + "grad_norm": 0.8168138119734322, + "learning_rate": 0.003, + "loss": 3.9495, + "step": 47973 + }, + { + "epoch": 0.47974, + "grad_norm": 0.6391249655275788, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 47974 + }, + { + "epoch": 0.47975, + "grad_norm": 0.6971137971274625, + "learning_rate": 0.003, + "loss": 4.015, + "step": 47975 + }, + { + "epoch": 0.47976, + "grad_norm": 0.6837766183442764, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 47976 + }, + { + "epoch": 0.47977, + "grad_norm": 0.6838423073760364, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 47977 + }, + { + "epoch": 0.47978, + "grad_norm": 0.6707861315887006, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 47978 + }, + { + "epoch": 0.47979, + "grad_norm": 0.6407278026635125, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 47979 + }, + { + "epoch": 0.4798, + "grad_norm": 0.5699965717933545, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 47980 + }, + { + "epoch": 0.47981, + "grad_norm": 0.6374096402293784, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 47981 + }, + { + "epoch": 0.47982, + "grad_norm": 0.7186779119318236, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 47982 + }, + { + "epoch": 0.47983, + "grad_norm": 0.7870658947961525, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 47983 + }, + { + "epoch": 0.47984, + "grad_norm": 0.7229242288018669, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 47984 + }, + { + "epoch": 0.47985, + "grad_norm": 0.7746523068183767, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 47985 + }, + { + "epoch": 0.47986, + "grad_norm": 0.8871794480882175, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 47986 + }, + { + "epoch": 0.47987, + "grad_norm": 0.8427556468672598, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 47987 + }, + { + "epoch": 0.47988, + "grad_norm": 0.8500369837907342, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 47988 + }, + { + "epoch": 0.47989, + "grad_norm": 0.9662969765949352, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 47989 + }, + { + "epoch": 0.4799, + "grad_norm": 1.0770078285463898, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 47990 + }, + { + "epoch": 0.47991, + "grad_norm": 1.0350056836488537, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 47991 + }, + { + "epoch": 0.47992, + "grad_norm": 0.8611554590640487, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 47992 + }, + { + "epoch": 0.47993, + "grad_norm": 0.9527987234872224, + "learning_rate": 0.003, + "loss": 3.997, + "step": 47993 + }, + { + "epoch": 0.47994, + "grad_norm": 0.9523743595094013, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 47994 + }, + { + "epoch": 0.47995, + "grad_norm": 0.8682883359744692, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 47995 + }, + { + "epoch": 0.47996, + "grad_norm": 0.8659335487021588, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 47996 + }, + { + "epoch": 0.47997, + "grad_norm": 0.8742972261313993, + "learning_rate": 0.003, + "loss": 3.987, + "step": 47997 + }, + { + "epoch": 0.47998, + "grad_norm": 0.6694811718520034, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 47998 + }, + { + "epoch": 0.47999, + "grad_norm": 0.6315904980754607, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 47999 + }, + { + "epoch": 0.48, + "grad_norm": 0.6015083847728384, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 48000 + }, + { + "epoch": 0.48001, + "grad_norm": 0.7086273588175196, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 48001 + }, + { + "epoch": 0.48002, + "grad_norm": 0.6107143462454758, + "learning_rate": 0.003, + "loss": 3.987, + "step": 48002 + }, + { + "epoch": 0.48003, + "grad_norm": 0.6135300143555071, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 48003 + }, + { + "epoch": 0.48004, + "grad_norm": 0.6086358468334047, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 48004 + }, + { + "epoch": 0.48005, + "grad_norm": 0.7382200034611264, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 48005 + }, + { + "epoch": 0.48006, + "grad_norm": 0.8072879315305437, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 48006 + }, + { + "epoch": 0.48007, + "grad_norm": 1.035546155461534, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 48007 + }, + { + "epoch": 0.48008, + "grad_norm": 0.9167796907288347, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 48008 + }, + { + "epoch": 0.48009, + "grad_norm": 0.7813921607850831, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 48009 + }, + { + "epoch": 0.4801, + "grad_norm": 0.6592583872213191, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 48010 + }, + { + "epoch": 0.48011, + "grad_norm": 0.5759741390864155, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 48011 + }, + { + "epoch": 0.48012, + "grad_norm": 0.5097716506335841, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 48012 + }, + { + "epoch": 0.48013, + "grad_norm": 0.5852144023726438, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 48013 + }, + { + "epoch": 0.48014, + "grad_norm": 0.6606958260600448, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 48014 + }, + { + "epoch": 0.48015, + "grad_norm": 0.6882391478228577, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 48015 + }, + { + "epoch": 0.48016, + "grad_norm": 0.6195104942704475, + "learning_rate": 0.003, + "loss": 3.9443, + "step": 48016 + }, + { + "epoch": 0.48017, + "grad_norm": 0.6102728266033617, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 48017 + }, + { + "epoch": 0.48018, + "grad_norm": 0.6342390044852221, + "learning_rate": 0.003, + "loss": 3.999, + "step": 48018 + }, + { + "epoch": 0.48019, + "grad_norm": 0.6247209298090324, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 48019 + }, + { + "epoch": 0.4802, + "grad_norm": 0.6049186287599252, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 48020 + }, + { + "epoch": 0.48021, + "grad_norm": 0.7243075796904577, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 48021 + }, + { + "epoch": 0.48022, + "grad_norm": 0.8663292276466785, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 48022 + }, + { + "epoch": 0.48023, + "grad_norm": 0.9249369945861, + "learning_rate": 0.003, + "loss": 3.989, + "step": 48023 + }, + { + "epoch": 0.48024, + "grad_norm": 0.8823039296155984, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 48024 + }, + { + "epoch": 0.48025, + "grad_norm": 0.7733959561810073, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 48025 + }, + { + "epoch": 0.48026, + "grad_norm": 0.940373554051887, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 48026 + }, + { + "epoch": 0.48027, + "grad_norm": 1.1439010863985921, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 48027 + }, + { + "epoch": 0.48028, + "grad_norm": 0.8914790649509753, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 48028 + }, + { + "epoch": 0.48029, + "grad_norm": 0.9019579922904705, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 48029 + }, + { + "epoch": 0.4803, + "grad_norm": 0.8731762560794408, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 48030 + }, + { + "epoch": 0.48031, + "grad_norm": 0.9393475625302379, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 48031 + }, + { + "epoch": 0.48032, + "grad_norm": 0.933004501668756, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 48032 + }, + { + "epoch": 0.48033, + "grad_norm": 0.9895771051557458, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 48033 + }, + { + "epoch": 0.48034, + "grad_norm": 0.973687257618384, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 48034 + }, + { + "epoch": 0.48035, + "grad_norm": 1.098417613999153, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 48035 + }, + { + "epoch": 0.48036, + "grad_norm": 0.9323748537577738, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 48036 + }, + { + "epoch": 0.48037, + "grad_norm": 0.8842963326361756, + "learning_rate": 0.003, + "loss": 3.981, + "step": 48037 + }, + { + "epoch": 0.48038, + "grad_norm": 0.8923645663182808, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 48038 + }, + { + "epoch": 0.48039, + "grad_norm": 0.8488536627859953, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 48039 + }, + { + "epoch": 0.4804, + "grad_norm": 0.7931651834043313, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 48040 + }, + { + "epoch": 0.48041, + "grad_norm": 0.6868904646892197, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 48041 + }, + { + "epoch": 0.48042, + "grad_norm": 0.6574780531143941, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 48042 + }, + { + "epoch": 0.48043, + "grad_norm": 0.6502698250374213, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 48043 + }, + { + "epoch": 0.48044, + "grad_norm": 0.725500991654662, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 48044 + }, + { + "epoch": 0.48045, + "grad_norm": 0.6930499557371256, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 48045 + }, + { + "epoch": 0.48046, + "grad_norm": 0.58435174508349, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 48046 + }, + { + "epoch": 0.48047, + "grad_norm": 0.5872795824403042, + "learning_rate": 0.003, + "loss": 3.973, + "step": 48047 + }, + { + "epoch": 0.48048, + "grad_norm": 0.6364369909473567, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 48048 + }, + { + "epoch": 0.48049, + "grad_norm": 0.6740254296110024, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 48049 + }, + { + "epoch": 0.4805, + "grad_norm": 0.7036621683810359, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 48050 + }, + { + "epoch": 0.48051, + "grad_norm": 0.8046223575725368, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 48051 + }, + { + "epoch": 0.48052, + "grad_norm": 0.7679452566729085, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 48052 + }, + { + "epoch": 0.48053, + "grad_norm": 0.823167765290801, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 48053 + }, + { + "epoch": 0.48054, + "grad_norm": 0.7145560259876242, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 48054 + }, + { + "epoch": 0.48055, + "grad_norm": 0.672774450815709, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 48055 + }, + { + "epoch": 0.48056, + "grad_norm": 0.612531441604702, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 48056 + }, + { + "epoch": 0.48057, + "grad_norm": 0.6061015108881204, + "learning_rate": 0.003, + "loss": 3.9467, + "step": 48057 + }, + { + "epoch": 0.48058, + "grad_norm": 0.7148819452022, + "learning_rate": 0.003, + "loss": 4.0, + "step": 48058 + }, + { + "epoch": 0.48059, + "grad_norm": 0.8449978413947078, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 48059 + }, + { + "epoch": 0.4806, + "grad_norm": 0.9074311577743559, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 48060 + }, + { + "epoch": 0.48061, + "grad_norm": 1.0470308253988536, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 48061 + }, + { + "epoch": 0.48062, + "grad_norm": 0.9025292835177537, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 48062 + }, + { + "epoch": 0.48063, + "grad_norm": 0.837458792359772, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 48063 + }, + { + "epoch": 0.48064, + "grad_norm": 0.7232222120737297, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 48064 + }, + { + "epoch": 0.48065, + "grad_norm": 0.625297396083215, + "learning_rate": 0.003, + "loss": 3.9397, + "step": 48065 + }, + { + "epoch": 0.48066, + "grad_norm": 0.6302484202875714, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 48066 + }, + { + "epoch": 0.48067, + "grad_norm": 0.6950424331476319, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 48067 + }, + { + "epoch": 0.48068, + "grad_norm": 0.69783959348518, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 48068 + }, + { + "epoch": 0.48069, + "grad_norm": 0.6477749314748286, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 48069 + }, + { + "epoch": 0.4807, + "grad_norm": 0.6302131493992874, + "learning_rate": 0.003, + "loss": 3.946, + "step": 48070 + }, + { + "epoch": 0.48071, + "grad_norm": 0.6832459917571493, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 48071 + }, + { + "epoch": 0.48072, + "grad_norm": 0.7251343950283883, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 48072 + }, + { + "epoch": 0.48073, + "grad_norm": 0.842950387612556, + "learning_rate": 0.003, + "loss": 3.979, + "step": 48073 + }, + { + "epoch": 0.48074, + "grad_norm": 0.9222251214503846, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 48074 + }, + { + "epoch": 0.48075, + "grad_norm": 0.842810058734866, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 48075 + }, + { + "epoch": 0.48076, + "grad_norm": 0.7704747190403256, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 48076 + }, + { + "epoch": 0.48077, + "grad_norm": 0.852831348507692, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 48077 + }, + { + "epoch": 0.48078, + "grad_norm": 0.9911869469396022, + "learning_rate": 0.003, + "loss": 4.0171, + "step": 48078 + }, + { + "epoch": 0.48079, + "grad_norm": 0.9196999043264348, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 48079 + }, + { + "epoch": 0.4808, + "grad_norm": 0.9659620068950612, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 48080 + }, + { + "epoch": 0.48081, + "grad_norm": 1.0418516433674367, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 48081 + }, + { + "epoch": 0.48082, + "grad_norm": 0.9079559970959333, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 48082 + }, + { + "epoch": 0.48083, + "grad_norm": 1.0136015065225459, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 48083 + }, + { + "epoch": 0.48084, + "grad_norm": 1.1364828105414884, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 48084 + }, + { + "epoch": 0.48085, + "grad_norm": 1.1266296084692946, + "learning_rate": 0.003, + "loss": 4.025, + "step": 48085 + }, + { + "epoch": 0.48086, + "grad_norm": 0.9746812677561862, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 48086 + }, + { + "epoch": 0.48087, + "grad_norm": 0.8926236660488167, + "learning_rate": 0.003, + "loss": 3.99, + "step": 48087 + }, + { + "epoch": 0.48088, + "grad_norm": 0.8730292356412954, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 48088 + }, + { + "epoch": 0.48089, + "grad_norm": 0.8614599035974776, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 48089 + }, + { + "epoch": 0.4809, + "grad_norm": 0.872289288391776, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 48090 + }, + { + "epoch": 0.48091, + "grad_norm": 0.727833754216597, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 48091 + }, + { + "epoch": 0.48092, + "grad_norm": 0.7545039686315422, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 48092 + }, + { + "epoch": 0.48093, + "grad_norm": 0.7857452853944097, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 48093 + }, + { + "epoch": 0.48094, + "grad_norm": 0.729010264378299, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 48094 + }, + { + "epoch": 0.48095, + "grad_norm": 0.6102056243783522, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 48095 + }, + { + "epoch": 0.48096, + "grad_norm": 0.5854900667099463, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 48096 + }, + { + "epoch": 0.48097, + "grad_norm": 0.5891575887339899, + "learning_rate": 0.003, + "loss": 3.9459, + "step": 48097 + }, + { + "epoch": 0.48098, + "grad_norm": 0.6611957390868957, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 48098 + }, + { + "epoch": 0.48099, + "grad_norm": 0.706434147512304, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 48099 + }, + { + "epoch": 0.481, + "grad_norm": 0.7691350217013088, + "learning_rate": 0.003, + "loss": 3.986, + "step": 48100 + }, + { + "epoch": 0.48101, + "grad_norm": 0.7727984337749904, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 48101 + }, + { + "epoch": 0.48102, + "grad_norm": 0.7608441818045174, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 48102 + }, + { + "epoch": 0.48103, + "grad_norm": 0.7471506449519772, + "learning_rate": 0.003, + "loss": 3.983, + "step": 48103 + }, + { + "epoch": 0.48104, + "grad_norm": 0.6334408116187739, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 48104 + }, + { + "epoch": 0.48105, + "grad_norm": 0.6088974802546959, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 48105 + }, + { + "epoch": 0.48106, + "grad_norm": 0.7203305193072551, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 48106 + }, + { + "epoch": 0.48107, + "grad_norm": 0.799907476397267, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 48107 + }, + { + "epoch": 0.48108, + "grad_norm": 0.8457820629680974, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 48108 + }, + { + "epoch": 0.48109, + "grad_norm": 0.7786459047958626, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 48109 + }, + { + "epoch": 0.4811, + "grad_norm": 0.6310403218508356, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 48110 + }, + { + "epoch": 0.48111, + "grad_norm": 0.574528601390577, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 48111 + }, + { + "epoch": 0.48112, + "grad_norm": 0.5926675388758634, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 48112 + }, + { + "epoch": 0.48113, + "grad_norm": 0.6341352484244718, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 48113 + }, + { + "epoch": 0.48114, + "grad_norm": 0.731546970392096, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 48114 + }, + { + "epoch": 0.48115, + "grad_norm": 0.8643655485097056, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 48115 + }, + { + "epoch": 0.48116, + "grad_norm": 0.9937093713934569, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 48116 + }, + { + "epoch": 0.48117, + "grad_norm": 1.0233960075072148, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 48117 + }, + { + "epoch": 0.48118, + "grad_norm": 0.861912831517287, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 48118 + }, + { + "epoch": 0.48119, + "grad_norm": 0.780486655423662, + "learning_rate": 0.003, + "loss": 4.0, + "step": 48119 + }, + { + "epoch": 0.4812, + "grad_norm": 0.6191108292418839, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 48120 + }, + { + "epoch": 0.48121, + "grad_norm": 0.6917046283317773, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 48121 + }, + { + "epoch": 0.48122, + "grad_norm": 0.8039323058553727, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 48122 + }, + { + "epoch": 0.48123, + "grad_norm": 0.9399192954237902, + "learning_rate": 0.003, + "loss": 3.9498, + "step": 48123 + }, + { + "epoch": 0.48124, + "grad_norm": 1.2023306372667926, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 48124 + }, + { + "epoch": 0.48125, + "grad_norm": 0.88526765817242, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 48125 + }, + { + "epoch": 0.48126, + "grad_norm": 0.7427006339258646, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 48126 + }, + { + "epoch": 0.48127, + "grad_norm": 0.7214865191581135, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 48127 + }, + { + "epoch": 0.48128, + "grad_norm": 0.8250630584797871, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 48128 + }, + { + "epoch": 0.48129, + "grad_norm": 0.8086893595158445, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 48129 + }, + { + "epoch": 0.4813, + "grad_norm": 0.7382901797484872, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 48130 + }, + { + "epoch": 0.48131, + "grad_norm": 0.7415500715245888, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 48131 + }, + { + "epoch": 0.48132, + "grad_norm": 0.7568229137071564, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 48132 + }, + { + "epoch": 0.48133, + "grad_norm": 0.7882839674467452, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 48133 + }, + { + "epoch": 0.48134, + "grad_norm": 0.8160874467340874, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 48134 + }, + { + "epoch": 0.48135, + "grad_norm": 0.7632620756812473, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 48135 + }, + { + "epoch": 0.48136, + "grad_norm": 0.6103210656112077, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 48136 + }, + { + "epoch": 0.48137, + "grad_norm": 0.5902212356891721, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 48137 + }, + { + "epoch": 0.48138, + "grad_norm": 0.6031435951097287, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 48138 + }, + { + "epoch": 0.48139, + "grad_norm": 0.6781001702152044, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 48139 + }, + { + "epoch": 0.4814, + "grad_norm": 0.6743562256469555, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 48140 + }, + { + "epoch": 0.48141, + "grad_norm": 0.7256118044370506, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 48141 + }, + { + "epoch": 0.48142, + "grad_norm": 0.7780131000607995, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 48142 + }, + { + "epoch": 0.48143, + "grad_norm": 0.7080024512734859, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 48143 + }, + { + "epoch": 0.48144, + "grad_norm": 0.7486661377457762, + "learning_rate": 0.003, + "loss": 3.964, + "step": 48144 + }, + { + "epoch": 0.48145, + "grad_norm": 0.9368445339290764, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 48145 + }, + { + "epoch": 0.48146, + "grad_norm": 1.0313947735895375, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 48146 + }, + { + "epoch": 0.48147, + "grad_norm": 1.0079381075552356, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 48147 + }, + { + "epoch": 0.48148, + "grad_norm": 0.9795729194074166, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 48148 + }, + { + "epoch": 0.48149, + "grad_norm": 0.84367584165024, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 48149 + }, + { + "epoch": 0.4815, + "grad_norm": 0.8957530476818234, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 48150 + }, + { + "epoch": 0.48151, + "grad_norm": 0.8596207404301763, + "learning_rate": 0.003, + "loss": 4.0187, + "step": 48151 + }, + { + "epoch": 0.48152, + "grad_norm": 0.8178212975486457, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 48152 + }, + { + "epoch": 0.48153, + "grad_norm": 0.7982638088767104, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 48153 + }, + { + "epoch": 0.48154, + "grad_norm": 0.8354031334222956, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 48154 + }, + { + "epoch": 0.48155, + "grad_norm": 0.8042590093236128, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 48155 + }, + { + "epoch": 0.48156, + "grad_norm": 0.8579591018918652, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 48156 + }, + { + "epoch": 0.48157, + "grad_norm": 0.8307548891352233, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 48157 + }, + { + "epoch": 0.48158, + "grad_norm": 0.732738048379598, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 48158 + }, + { + "epoch": 0.48159, + "grad_norm": 0.7285197220533559, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 48159 + }, + { + "epoch": 0.4816, + "grad_norm": 0.7545391108546813, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 48160 + }, + { + "epoch": 0.48161, + "grad_norm": 0.796567240684381, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 48161 + }, + { + "epoch": 0.48162, + "grad_norm": 0.703620194626958, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 48162 + }, + { + "epoch": 0.48163, + "grad_norm": 0.7538677101669542, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 48163 + }, + { + "epoch": 0.48164, + "grad_norm": 0.8314583617534603, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 48164 + }, + { + "epoch": 0.48165, + "grad_norm": 0.83691452647519, + "learning_rate": 0.003, + "loss": 4.001, + "step": 48165 + }, + { + "epoch": 0.48166, + "grad_norm": 0.7676065474816737, + "learning_rate": 0.003, + "loss": 4.001, + "step": 48166 + }, + { + "epoch": 0.48167, + "grad_norm": 0.7241723416380851, + "learning_rate": 0.003, + "loss": 3.963, + "step": 48167 + }, + { + "epoch": 0.48168, + "grad_norm": 0.7136420884566806, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 48168 + }, + { + "epoch": 0.48169, + "grad_norm": 0.7083451874559297, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 48169 + }, + { + "epoch": 0.4817, + "grad_norm": 0.7644855169432986, + "learning_rate": 0.003, + "loss": 3.979, + "step": 48170 + }, + { + "epoch": 0.48171, + "grad_norm": 0.7178838021071439, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 48171 + }, + { + "epoch": 0.48172, + "grad_norm": 0.6540193988063109, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 48172 + }, + { + "epoch": 0.48173, + "grad_norm": 0.7414429173597804, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 48173 + }, + { + "epoch": 0.48174, + "grad_norm": 0.927549061882745, + "learning_rate": 0.003, + "loss": 3.955, + "step": 48174 + }, + { + "epoch": 0.48175, + "grad_norm": 1.163853410653689, + "learning_rate": 0.003, + "loss": 4.01, + "step": 48175 + }, + { + "epoch": 0.48176, + "grad_norm": 0.9545167729315915, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 48176 + }, + { + "epoch": 0.48177, + "grad_norm": 0.6924963135229057, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 48177 + }, + { + "epoch": 0.48178, + "grad_norm": 0.5865750513553902, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 48178 + }, + { + "epoch": 0.48179, + "grad_norm": 0.618266670706092, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 48179 + }, + { + "epoch": 0.4818, + "grad_norm": 0.6629059824090917, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 48180 + }, + { + "epoch": 0.48181, + "grad_norm": 0.6556284983765205, + "learning_rate": 0.003, + "loss": 3.982, + "step": 48181 + }, + { + "epoch": 0.48182, + "grad_norm": 0.7108677737868653, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 48182 + }, + { + "epoch": 0.48183, + "grad_norm": 0.7960416187260697, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 48183 + }, + { + "epoch": 0.48184, + "grad_norm": 0.7786311491631221, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 48184 + }, + { + "epoch": 0.48185, + "grad_norm": 0.7399748749884673, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 48185 + }, + { + "epoch": 0.48186, + "grad_norm": 0.7534842854798496, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 48186 + }, + { + "epoch": 0.48187, + "grad_norm": 0.7790471195361591, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 48187 + }, + { + "epoch": 0.48188, + "grad_norm": 0.7908625896772276, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 48188 + }, + { + "epoch": 0.48189, + "grad_norm": 0.8436385851190229, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 48189 + }, + { + "epoch": 0.4819, + "grad_norm": 0.9233770394743444, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 48190 + }, + { + "epoch": 0.48191, + "grad_norm": 0.8313643133074152, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 48191 + }, + { + "epoch": 0.48192, + "grad_norm": 0.8065013384672167, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 48192 + }, + { + "epoch": 0.48193, + "grad_norm": 0.8182271335334592, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 48193 + }, + { + "epoch": 0.48194, + "grad_norm": 0.8611504346803289, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 48194 + }, + { + "epoch": 0.48195, + "grad_norm": 0.9024651969359043, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 48195 + }, + { + "epoch": 0.48196, + "grad_norm": 0.9678514924136595, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 48196 + }, + { + "epoch": 0.48197, + "grad_norm": 0.8590012165905284, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 48197 + }, + { + "epoch": 0.48198, + "grad_norm": 0.7873439372651376, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 48198 + }, + { + "epoch": 0.48199, + "grad_norm": 0.8275289976218126, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 48199 + }, + { + "epoch": 0.482, + "grad_norm": 0.8662000347816893, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 48200 + }, + { + "epoch": 0.48201, + "grad_norm": 0.8822905094151854, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 48201 + }, + { + "epoch": 0.48202, + "grad_norm": 0.9215054252036212, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 48202 + }, + { + "epoch": 0.48203, + "grad_norm": 0.912115942159047, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 48203 + }, + { + "epoch": 0.48204, + "grad_norm": 0.924739241431857, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 48204 + }, + { + "epoch": 0.48205, + "grad_norm": 0.8383025848929905, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 48205 + }, + { + "epoch": 0.48206, + "grad_norm": 0.7857736517051221, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 48206 + }, + { + "epoch": 0.48207, + "grad_norm": 0.7648300467308445, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 48207 + }, + { + "epoch": 0.48208, + "grad_norm": 0.7251411080065805, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 48208 + }, + { + "epoch": 0.48209, + "grad_norm": 0.6501765284028297, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 48209 + }, + { + "epoch": 0.4821, + "grad_norm": 0.6572151723788747, + "learning_rate": 0.003, + "loss": 3.997, + "step": 48210 + }, + { + "epoch": 0.48211, + "grad_norm": 0.5746469421251365, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 48211 + }, + { + "epoch": 0.48212, + "grad_norm": 0.6266191471088647, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 48212 + }, + { + "epoch": 0.48213, + "grad_norm": 0.6176478458916953, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 48213 + }, + { + "epoch": 0.48214, + "grad_norm": 0.6241838660466437, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 48214 + }, + { + "epoch": 0.48215, + "grad_norm": 0.6777991961691382, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 48215 + }, + { + "epoch": 0.48216, + "grad_norm": 0.774263124936983, + "learning_rate": 0.003, + "loss": 3.963, + "step": 48216 + }, + { + "epoch": 0.48217, + "grad_norm": 0.7584244969834327, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 48217 + }, + { + "epoch": 0.48218, + "grad_norm": 0.666766563114667, + "learning_rate": 0.003, + "loss": 3.961, + "step": 48218 + }, + { + "epoch": 0.48219, + "grad_norm": 0.5826556309480363, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 48219 + }, + { + "epoch": 0.4822, + "grad_norm": 0.6711873107553123, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 48220 + }, + { + "epoch": 0.48221, + "grad_norm": 0.79772577129622, + "learning_rate": 0.003, + "loss": 3.945, + "step": 48221 + }, + { + "epoch": 0.48222, + "grad_norm": 0.8713948133716392, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 48222 + }, + { + "epoch": 0.48223, + "grad_norm": 0.9977448136929764, + "learning_rate": 0.003, + "loss": 3.97, + "step": 48223 + }, + { + "epoch": 0.48224, + "grad_norm": 0.8964772814420667, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 48224 + }, + { + "epoch": 0.48225, + "grad_norm": 0.702003456412317, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 48225 + }, + { + "epoch": 0.48226, + "grad_norm": 0.629032200316872, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 48226 + }, + { + "epoch": 0.48227, + "grad_norm": 0.7313782937047996, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 48227 + }, + { + "epoch": 0.48228, + "grad_norm": 0.7708292916767151, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 48228 + }, + { + "epoch": 0.48229, + "grad_norm": 0.6777526082711668, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 48229 + }, + { + "epoch": 0.4823, + "grad_norm": 0.5667440027500353, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 48230 + }, + { + "epoch": 0.48231, + "grad_norm": 0.6344791483898353, + "learning_rate": 0.003, + "loss": 3.967, + "step": 48231 + }, + { + "epoch": 0.48232, + "grad_norm": 0.6543800852214613, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 48232 + }, + { + "epoch": 0.48233, + "grad_norm": 0.6688895939022503, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 48233 + }, + { + "epoch": 0.48234, + "grad_norm": 0.6648835459762844, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 48234 + }, + { + "epoch": 0.48235, + "grad_norm": 0.5891832883820138, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 48235 + }, + { + "epoch": 0.48236, + "grad_norm": 0.6114208874847408, + "learning_rate": 0.003, + "loss": 3.997, + "step": 48236 + }, + { + "epoch": 0.48237, + "grad_norm": 0.5782201292086582, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 48237 + }, + { + "epoch": 0.48238, + "grad_norm": 0.654529449286849, + "learning_rate": 0.003, + "loss": 4.011, + "step": 48238 + }, + { + "epoch": 0.48239, + "grad_norm": 0.7664850357335433, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 48239 + }, + { + "epoch": 0.4824, + "grad_norm": 0.8420363679404519, + "learning_rate": 0.003, + "loss": 3.967, + "step": 48240 + }, + { + "epoch": 0.48241, + "grad_norm": 0.9026547026097951, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 48241 + }, + { + "epoch": 0.48242, + "grad_norm": 0.9985156068527159, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 48242 + }, + { + "epoch": 0.48243, + "grad_norm": 1.060635687310205, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 48243 + }, + { + "epoch": 0.48244, + "grad_norm": 0.9171811238375518, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 48244 + }, + { + "epoch": 0.48245, + "grad_norm": 0.8186777547321118, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 48245 + }, + { + "epoch": 0.48246, + "grad_norm": 0.8610098452328979, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 48246 + }, + { + "epoch": 0.48247, + "grad_norm": 0.7986748541951737, + "learning_rate": 0.003, + "loss": 4.0, + "step": 48247 + }, + { + "epoch": 0.48248, + "grad_norm": 0.8820151203587694, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 48248 + }, + { + "epoch": 0.48249, + "grad_norm": 1.0629308609094437, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 48249 + }, + { + "epoch": 0.4825, + "grad_norm": 1.1041081257126164, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 48250 + }, + { + "epoch": 0.48251, + "grad_norm": 1.0384344232179281, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 48251 + }, + { + "epoch": 0.48252, + "grad_norm": 0.9601232548298085, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 48252 + }, + { + "epoch": 0.48253, + "grad_norm": 0.8560899730346088, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 48253 + }, + { + "epoch": 0.48254, + "grad_norm": 0.6547663109890678, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 48254 + }, + { + "epoch": 0.48255, + "grad_norm": 0.5938321130702894, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 48255 + }, + { + "epoch": 0.48256, + "grad_norm": 0.6699332807225664, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 48256 + }, + { + "epoch": 0.48257, + "grad_norm": 0.7980600435788012, + "learning_rate": 0.003, + "loss": 3.957, + "step": 48257 + }, + { + "epoch": 0.48258, + "grad_norm": 0.8727180908475116, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 48258 + }, + { + "epoch": 0.48259, + "grad_norm": 0.9244804004946141, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 48259 + }, + { + "epoch": 0.4826, + "grad_norm": 0.8749419487159644, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 48260 + }, + { + "epoch": 0.48261, + "grad_norm": 0.8095500859296455, + "learning_rate": 0.003, + "loss": 4.0232, + "step": 48261 + }, + { + "epoch": 0.48262, + "grad_norm": 0.8422901656812134, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 48262 + }, + { + "epoch": 0.48263, + "grad_norm": 0.807843259315975, + "learning_rate": 0.003, + "loss": 3.9462, + "step": 48263 + }, + { + "epoch": 0.48264, + "grad_norm": 0.7786279113906065, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 48264 + }, + { + "epoch": 0.48265, + "grad_norm": 0.7844132302618172, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 48265 + }, + { + "epoch": 0.48266, + "grad_norm": 0.6809235080676027, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 48266 + }, + { + "epoch": 0.48267, + "grad_norm": 0.6192435360338383, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 48267 + }, + { + "epoch": 0.48268, + "grad_norm": 0.6266116238277962, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 48268 + }, + { + "epoch": 0.48269, + "grad_norm": 0.7399177331401906, + "learning_rate": 0.003, + "loss": 3.967, + "step": 48269 + }, + { + "epoch": 0.4827, + "grad_norm": 0.836175973110675, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 48270 + }, + { + "epoch": 0.48271, + "grad_norm": 0.8668179063620648, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 48271 + }, + { + "epoch": 0.48272, + "grad_norm": 0.8615080381619262, + "learning_rate": 0.003, + "loss": 4.008, + "step": 48272 + }, + { + "epoch": 0.48273, + "grad_norm": 0.8469664354752775, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 48273 + }, + { + "epoch": 0.48274, + "grad_norm": 0.926340247335285, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 48274 + }, + { + "epoch": 0.48275, + "grad_norm": 0.7997612495858167, + "learning_rate": 0.003, + "loss": 3.986, + "step": 48275 + }, + { + "epoch": 0.48276, + "grad_norm": 0.7827420345459254, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 48276 + }, + { + "epoch": 0.48277, + "grad_norm": 0.7525077911099846, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 48277 + }, + { + "epoch": 0.48278, + "grad_norm": 0.7132986016052354, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 48278 + }, + { + "epoch": 0.48279, + "grad_norm": 0.7050858580802121, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 48279 + }, + { + "epoch": 0.4828, + "grad_norm": 0.7211951692435441, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 48280 + }, + { + "epoch": 0.48281, + "grad_norm": 0.847333835083243, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 48281 + }, + { + "epoch": 0.48282, + "grad_norm": 0.8604183234705903, + "learning_rate": 0.003, + "loss": 3.986, + "step": 48282 + }, + { + "epoch": 0.48283, + "grad_norm": 0.866380595229288, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 48283 + }, + { + "epoch": 0.48284, + "grad_norm": 0.9629717582768562, + "learning_rate": 0.003, + "loss": 3.9534, + "step": 48284 + }, + { + "epoch": 0.48285, + "grad_norm": 0.9208765235933098, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 48285 + }, + { + "epoch": 0.48286, + "grad_norm": 0.7188533688620772, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 48286 + }, + { + "epoch": 0.48287, + "grad_norm": 0.6065371078428896, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 48287 + }, + { + "epoch": 0.48288, + "grad_norm": 0.6543979744774158, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 48288 + }, + { + "epoch": 0.48289, + "grad_norm": 0.6379641772818502, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 48289 + }, + { + "epoch": 0.4829, + "grad_norm": 0.6186470361987387, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 48290 + }, + { + "epoch": 0.48291, + "grad_norm": 0.6481422242112892, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 48291 + }, + { + "epoch": 0.48292, + "grad_norm": 0.6890973201576599, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 48292 + }, + { + "epoch": 0.48293, + "grad_norm": 0.7855290448271244, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 48293 + }, + { + "epoch": 0.48294, + "grad_norm": 0.9109755288785004, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 48294 + }, + { + "epoch": 0.48295, + "grad_norm": 0.7812227532764404, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 48295 + }, + { + "epoch": 0.48296, + "grad_norm": 0.6992721147565776, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 48296 + }, + { + "epoch": 0.48297, + "grad_norm": 0.7546831478026805, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 48297 + }, + { + "epoch": 0.48298, + "grad_norm": 0.8219885474189739, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 48298 + }, + { + "epoch": 0.48299, + "grad_norm": 0.8068898155520982, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 48299 + }, + { + "epoch": 0.483, + "grad_norm": 0.8831197409890206, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 48300 + }, + { + "epoch": 0.48301, + "grad_norm": 0.8637086844022464, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 48301 + }, + { + "epoch": 0.48302, + "grad_norm": 0.8514290037863184, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 48302 + }, + { + "epoch": 0.48303, + "grad_norm": 0.9683885806320384, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 48303 + }, + { + "epoch": 0.48304, + "grad_norm": 0.9650163923105922, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 48304 + }, + { + "epoch": 0.48305, + "grad_norm": 0.9940231888310525, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 48305 + }, + { + "epoch": 0.48306, + "grad_norm": 0.9483722413295732, + "learning_rate": 0.003, + "loss": 3.9417, + "step": 48306 + }, + { + "epoch": 0.48307, + "grad_norm": 0.8479034460943919, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 48307 + }, + { + "epoch": 0.48308, + "grad_norm": 0.8278324270336941, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 48308 + }, + { + "epoch": 0.48309, + "grad_norm": 0.7434414605025902, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 48309 + }, + { + "epoch": 0.4831, + "grad_norm": 0.666310774595407, + "learning_rate": 0.003, + "loss": 4.022, + "step": 48310 + }, + { + "epoch": 0.48311, + "grad_norm": 0.6771897181950806, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 48311 + }, + { + "epoch": 0.48312, + "grad_norm": 0.6853448986548897, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 48312 + }, + { + "epoch": 0.48313, + "grad_norm": 0.631434550312622, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 48313 + }, + { + "epoch": 0.48314, + "grad_norm": 0.600489967101081, + "learning_rate": 0.003, + "loss": 3.9482, + "step": 48314 + }, + { + "epoch": 0.48315, + "grad_norm": 0.5271444912588303, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 48315 + }, + { + "epoch": 0.48316, + "grad_norm": 0.4919331131777379, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 48316 + }, + { + "epoch": 0.48317, + "grad_norm": 0.5290752110722696, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 48317 + }, + { + "epoch": 0.48318, + "grad_norm": 0.5562015217871843, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 48318 + }, + { + "epoch": 0.48319, + "grad_norm": 0.6870169193364246, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 48319 + }, + { + "epoch": 0.4832, + "grad_norm": 0.721745967586984, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 48320 + }, + { + "epoch": 0.48321, + "grad_norm": 0.6535230822805733, + "learning_rate": 0.003, + "loss": 3.985, + "step": 48321 + }, + { + "epoch": 0.48322, + "grad_norm": 0.5970957877865822, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 48322 + }, + { + "epoch": 0.48323, + "grad_norm": 0.719115509019714, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 48323 + }, + { + "epoch": 0.48324, + "grad_norm": 0.9818527840996231, + "learning_rate": 0.003, + "loss": 3.959, + "step": 48324 + }, + { + "epoch": 0.48325, + "grad_norm": 1.1536186691589163, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 48325 + }, + { + "epoch": 0.48326, + "grad_norm": 0.8288485108610945, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 48326 + }, + { + "epoch": 0.48327, + "grad_norm": 0.7242707452698004, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 48327 + }, + { + "epoch": 0.48328, + "grad_norm": 0.7656701318038566, + "learning_rate": 0.003, + "loss": 3.982, + "step": 48328 + }, + { + "epoch": 0.48329, + "grad_norm": 0.8401502985650274, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 48329 + }, + { + "epoch": 0.4833, + "grad_norm": 0.8347659728884799, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 48330 + }, + { + "epoch": 0.48331, + "grad_norm": 0.8465351213988163, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 48331 + }, + { + "epoch": 0.48332, + "grad_norm": 0.9128166877966974, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 48332 + }, + { + "epoch": 0.48333, + "grad_norm": 0.8576926315996574, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 48333 + }, + { + "epoch": 0.48334, + "grad_norm": 0.6736871071786946, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 48334 + }, + { + "epoch": 0.48335, + "grad_norm": 0.7644882920311213, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 48335 + }, + { + "epoch": 0.48336, + "grad_norm": 0.8430404685397849, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 48336 + }, + { + "epoch": 0.48337, + "grad_norm": 0.8656534068909214, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 48337 + }, + { + "epoch": 0.48338, + "grad_norm": 0.7950915230580662, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 48338 + }, + { + "epoch": 0.48339, + "grad_norm": 0.7896391692859939, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 48339 + }, + { + "epoch": 0.4834, + "grad_norm": 0.8363738546740017, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 48340 + }, + { + "epoch": 0.48341, + "grad_norm": 1.0026845535880173, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 48341 + }, + { + "epoch": 0.48342, + "grad_norm": 1.0113892821154473, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 48342 + }, + { + "epoch": 0.48343, + "grad_norm": 0.9143386471698007, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 48343 + }, + { + "epoch": 0.48344, + "grad_norm": 0.8386169796406241, + "learning_rate": 0.003, + "loss": 4.0273, + "step": 48344 + }, + { + "epoch": 0.48345, + "grad_norm": 0.9093065836378329, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 48345 + }, + { + "epoch": 0.48346, + "grad_norm": 1.1394995013743672, + "learning_rate": 0.003, + "loss": 4.0179, + "step": 48346 + }, + { + "epoch": 0.48347, + "grad_norm": 0.9729960902154363, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 48347 + }, + { + "epoch": 0.48348, + "grad_norm": 1.0825534177474125, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 48348 + }, + { + "epoch": 0.48349, + "grad_norm": 0.9630701178411021, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 48349 + }, + { + "epoch": 0.4835, + "grad_norm": 0.9106445009176084, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 48350 + }, + { + "epoch": 0.48351, + "grad_norm": 0.8447505297463852, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 48351 + }, + { + "epoch": 0.48352, + "grad_norm": 0.7591985360298812, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 48352 + }, + { + "epoch": 0.48353, + "grad_norm": 0.8620303770542567, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 48353 + }, + { + "epoch": 0.48354, + "grad_norm": 0.8690526374591153, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 48354 + }, + { + "epoch": 0.48355, + "grad_norm": 0.849416020330688, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 48355 + }, + { + "epoch": 0.48356, + "grad_norm": 0.9337670295194289, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 48356 + }, + { + "epoch": 0.48357, + "grad_norm": 0.8700317718120507, + "learning_rate": 0.003, + "loss": 4.001, + "step": 48357 + }, + { + "epoch": 0.48358, + "grad_norm": 0.8635299005339516, + "learning_rate": 0.003, + "loss": 4.0192, + "step": 48358 + }, + { + "epoch": 0.48359, + "grad_norm": 0.9152436177871267, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 48359 + }, + { + "epoch": 0.4836, + "grad_norm": 0.9175700809164078, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 48360 + }, + { + "epoch": 0.48361, + "grad_norm": 0.8671888490460764, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 48361 + }, + { + "epoch": 0.48362, + "grad_norm": 0.8506176993074428, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 48362 + }, + { + "epoch": 0.48363, + "grad_norm": 0.9115948139174486, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 48363 + }, + { + "epoch": 0.48364, + "grad_norm": 0.882943501799306, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 48364 + }, + { + "epoch": 0.48365, + "grad_norm": 0.8895797097206798, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 48365 + }, + { + "epoch": 0.48366, + "grad_norm": 0.9358718480990946, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 48366 + }, + { + "epoch": 0.48367, + "grad_norm": 0.8714563936761822, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 48367 + }, + { + "epoch": 0.48368, + "grad_norm": 0.8474431891420322, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 48368 + }, + { + "epoch": 0.48369, + "grad_norm": 0.7049435430710334, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 48369 + }, + { + "epoch": 0.4837, + "grad_norm": 0.6262122923004367, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 48370 + }, + { + "epoch": 0.48371, + "grad_norm": 0.7029651741587493, + "learning_rate": 0.003, + "loss": 3.975, + "step": 48371 + }, + { + "epoch": 0.48372, + "grad_norm": 0.6315390830467916, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 48372 + }, + { + "epoch": 0.48373, + "grad_norm": 0.602986793783489, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 48373 + }, + { + "epoch": 0.48374, + "grad_norm": 0.567591187315759, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 48374 + }, + { + "epoch": 0.48375, + "grad_norm": 0.6373072606239307, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 48375 + }, + { + "epoch": 0.48376, + "grad_norm": 0.628963334736688, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 48376 + }, + { + "epoch": 0.48377, + "grad_norm": 0.6406493505974704, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 48377 + }, + { + "epoch": 0.48378, + "grad_norm": 0.6900221946346669, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 48378 + }, + { + "epoch": 0.48379, + "grad_norm": 0.7191420930828278, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 48379 + }, + { + "epoch": 0.4838, + "grad_norm": 0.714338968231839, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 48380 + }, + { + "epoch": 0.48381, + "grad_norm": 0.7101587732553502, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 48381 + }, + { + "epoch": 0.48382, + "grad_norm": 0.6725587330451663, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 48382 + }, + { + "epoch": 0.48383, + "grad_norm": 0.6807929488780436, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 48383 + }, + { + "epoch": 0.48384, + "grad_norm": 0.6899571209833222, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 48384 + }, + { + "epoch": 0.48385, + "grad_norm": 0.6762776797778921, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 48385 + }, + { + "epoch": 0.48386, + "grad_norm": 0.7137717283022605, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 48386 + }, + { + "epoch": 0.48387, + "grad_norm": 0.7982087310881919, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 48387 + }, + { + "epoch": 0.48388, + "grad_norm": 0.8036725813814208, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 48388 + }, + { + "epoch": 0.48389, + "grad_norm": 0.7661020891818524, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 48389 + }, + { + "epoch": 0.4839, + "grad_norm": 0.6672159691573255, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 48390 + }, + { + "epoch": 0.48391, + "grad_norm": 0.5889054880490282, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 48391 + }, + { + "epoch": 0.48392, + "grad_norm": 0.5752240375559989, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 48392 + }, + { + "epoch": 0.48393, + "grad_norm": 0.6228382927147393, + "learning_rate": 0.003, + "loss": 3.9296, + "step": 48393 + }, + { + "epoch": 0.48394, + "grad_norm": 0.5966801299241447, + "learning_rate": 0.003, + "loss": 3.957, + "step": 48394 + }, + { + "epoch": 0.48395, + "grad_norm": 0.6144196751259517, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 48395 + }, + { + "epoch": 0.48396, + "grad_norm": 0.6785372086367139, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 48396 + }, + { + "epoch": 0.48397, + "grad_norm": 0.7460699137946849, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 48397 + }, + { + "epoch": 0.48398, + "grad_norm": 0.826226102988762, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 48398 + }, + { + "epoch": 0.48399, + "grad_norm": 0.9861175037504157, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 48399 + }, + { + "epoch": 0.484, + "grad_norm": 1.1037366505834372, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 48400 + }, + { + "epoch": 0.48401, + "grad_norm": 0.9286467539584411, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 48401 + }, + { + "epoch": 0.48402, + "grad_norm": 0.7723978503212571, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 48402 + }, + { + "epoch": 0.48403, + "grad_norm": 0.6994100575847825, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 48403 + }, + { + "epoch": 0.48404, + "grad_norm": 0.6095029014090306, + "learning_rate": 0.003, + "loss": 3.9571, + "step": 48404 + }, + { + "epoch": 0.48405, + "grad_norm": 0.6491456144663124, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 48405 + }, + { + "epoch": 0.48406, + "grad_norm": 0.6226814602791476, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 48406 + }, + { + "epoch": 0.48407, + "grad_norm": 0.7418508548956537, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 48407 + }, + { + "epoch": 0.48408, + "grad_norm": 0.8442070424361293, + "learning_rate": 0.003, + "loss": 3.981, + "step": 48408 + }, + { + "epoch": 0.48409, + "grad_norm": 0.9421927944482693, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 48409 + }, + { + "epoch": 0.4841, + "grad_norm": 1.0737895383510896, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 48410 + }, + { + "epoch": 0.48411, + "grad_norm": 0.9804124304002829, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 48411 + }, + { + "epoch": 0.48412, + "grad_norm": 0.8853440835672369, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 48412 + }, + { + "epoch": 0.48413, + "grad_norm": 0.9599168033793939, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 48413 + }, + { + "epoch": 0.48414, + "grad_norm": 0.8660540703273135, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 48414 + }, + { + "epoch": 0.48415, + "grad_norm": 0.8752135537049843, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 48415 + }, + { + "epoch": 0.48416, + "grad_norm": 0.8449756083856974, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 48416 + }, + { + "epoch": 0.48417, + "grad_norm": 0.9028638784631416, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 48417 + }, + { + "epoch": 0.48418, + "grad_norm": 0.9419530569231226, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 48418 + }, + { + "epoch": 0.48419, + "grad_norm": 0.899631479219356, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 48419 + }, + { + "epoch": 0.4842, + "grad_norm": 0.8105379917794874, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 48420 + }, + { + "epoch": 0.48421, + "grad_norm": 0.7720487809582168, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 48421 + }, + { + "epoch": 0.48422, + "grad_norm": 0.7434931685629832, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 48422 + }, + { + "epoch": 0.48423, + "grad_norm": 0.7270010566846913, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 48423 + }, + { + "epoch": 0.48424, + "grad_norm": 0.7354145786135781, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 48424 + }, + { + "epoch": 0.48425, + "grad_norm": 0.7259968007645138, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 48425 + }, + { + "epoch": 0.48426, + "grad_norm": 0.7627068821265918, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 48426 + }, + { + "epoch": 0.48427, + "grad_norm": 0.9441533433005401, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 48427 + }, + { + "epoch": 0.48428, + "grad_norm": 1.0187502987521744, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 48428 + }, + { + "epoch": 0.48429, + "grad_norm": 1.0577172046709302, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 48429 + }, + { + "epoch": 0.4843, + "grad_norm": 1.0126765898431527, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 48430 + }, + { + "epoch": 0.48431, + "grad_norm": 1.0297479580451303, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 48431 + }, + { + "epoch": 0.48432, + "grad_norm": 0.9227573338051973, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 48432 + }, + { + "epoch": 0.48433, + "grad_norm": 0.6470872831799049, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 48433 + }, + { + "epoch": 0.48434, + "grad_norm": 0.5891626770955821, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 48434 + }, + { + "epoch": 0.48435, + "grad_norm": 0.5670051257209288, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 48435 + }, + { + "epoch": 0.48436, + "grad_norm": 0.5238435830281312, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 48436 + }, + { + "epoch": 0.48437, + "grad_norm": 0.5362809214185127, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 48437 + }, + { + "epoch": 0.48438, + "grad_norm": 0.5096433361363558, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 48438 + }, + { + "epoch": 0.48439, + "grad_norm": 0.5568700568559242, + "learning_rate": 0.003, + "loss": 4.006, + "step": 48439 + }, + { + "epoch": 0.4844, + "grad_norm": 0.6697730040361051, + "learning_rate": 0.003, + "loss": 3.989, + "step": 48440 + }, + { + "epoch": 0.48441, + "grad_norm": 0.8416427967528882, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 48441 + }, + { + "epoch": 0.48442, + "grad_norm": 0.9401754802435071, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 48442 + }, + { + "epoch": 0.48443, + "grad_norm": 0.8416051861002877, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 48443 + }, + { + "epoch": 0.48444, + "grad_norm": 0.6693484790903635, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 48444 + }, + { + "epoch": 0.48445, + "grad_norm": 0.6053704645841349, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 48445 + }, + { + "epoch": 0.48446, + "grad_norm": 0.6674148040426967, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 48446 + }, + { + "epoch": 0.48447, + "grad_norm": 0.7500927879835559, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 48447 + }, + { + "epoch": 0.48448, + "grad_norm": 0.8139823020299767, + "learning_rate": 0.003, + "loss": 3.9586, + "step": 48448 + }, + { + "epoch": 0.48449, + "grad_norm": 0.8202865049893406, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 48449 + }, + { + "epoch": 0.4845, + "grad_norm": 0.8333680728098632, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 48450 + }, + { + "epoch": 0.48451, + "grad_norm": 0.7664707724752836, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 48451 + }, + { + "epoch": 0.48452, + "grad_norm": 0.617475219025463, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 48452 + }, + { + "epoch": 0.48453, + "grad_norm": 0.6190987536019228, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 48453 + }, + { + "epoch": 0.48454, + "grad_norm": 0.6584917888959723, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 48454 + }, + { + "epoch": 0.48455, + "grad_norm": 0.738290375091726, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 48455 + }, + { + "epoch": 0.48456, + "grad_norm": 0.752377510278999, + "learning_rate": 0.003, + "loss": 3.9475, + "step": 48456 + }, + { + "epoch": 0.48457, + "grad_norm": 0.8491709613870358, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 48457 + }, + { + "epoch": 0.48458, + "grad_norm": 0.8877257769359692, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 48458 + }, + { + "epoch": 0.48459, + "grad_norm": 0.9430202727719448, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 48459 + }, + { + "epoch": 0.4846, + "grad_norm": 0.8217137519610942, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 48460 + }, + { + "epoch": 0.48461, + "grad_norm": 0.8442055673760157, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 48461 + }, + { + "epoch": 0.48462, + "grad_norm": 0.8471014064653355, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 48462 + }, + { + "epoch": 0.48463, + "grad_norm": 0.7561527975919823, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 48463 + }, + { + "epoch": 0.48464, + "grad_norm": 0.7134714543532754, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 48464 + }, + { + "epoch": 0.48465, + "grad_norm": 0.7306945245468949, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 48465 + }, + { + "epoch": 0.48466, + "grad_norm": 0.7648164523095077, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 48466 + }, + { + "epoch": 0.48467, + "grad_norm": 0.7863905049480877, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 48467 + }, + { + "epoch": 0.48468, + "grad_norm": 0.8654050415931513, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 48468 + }, + { + "epoch": 0.48469, + "grad_norm": 0.911331452061812, + "learning_rate": 0.003, + "loss": 4.0214, + "step": 48469 + }, + { + "epoch": 0.4847, + "grad_norm": 0.8929171177661406, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 48470 + }, + { + "epoch": 0.48471, + "grad_norm": 0.9946237405614926, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 48471 + }, + { + "epoch": 0.48472, + "grad_norm": 1.0357105637773276, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 48472 + }, + { + "epoch": 0.48473, + "grad_norm": 1.1826349126571727, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 48473 + }, + { + "epoch": 0.48474, + "grad_norm": 0.9333750162494581, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 48474 + }, + { + "epoch": 0.48475, + "grad_norm": 0.9785520332323782, + "learning_rate": 0.003, + "loss": 3.971, + "step": 48475 + }, + { + "epoch": 0.48476, + "grad_norm": 0.90879553122993, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 48476 + }, + { + "epoch": 0.48477, + "grad_norm": 0.8195859465731625, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 48477 + }, + { + "epoch": 0.48478, + "grad_norm": 0.8990566092712208, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 48478 + }, + { + "epoch": 0.48479, + "grad_norm": 0.9110970728069121, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 48479 + }, + { + "epoch": 0.4848, + "grad_norm": 0.8483751204217904, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 48480 + }, + { + "epoch": 0.48481, + "grad_norm": 0.7155339140183494, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 48481 + }, + { + "epoch": 0.48482, + "grad_norm": 0.5943956435700268, + "learning_rate": 0.003, + "loss": 3.988, + "step": 48482 + }, + { + "epoch": 0.48483, + "grad_norm": 0.5550473994631524, + "learning_rate": 0.003, + "loss": 3.948, + "step": 48483 + }, + { + "epoch": 0.48484, + "grad_norm": 0.5411334582543175, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 48484 + }, + { + "epoch": 0.48485, + "grad_norm": 0.5367711610625281, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 48485 + }, + { + "epoch": 0.48486, + "grad_norm": 0.5483088408643532, + "learning_rate": 0.003, + "loss": 3.972, + "step": 48486 + }, + { + "epoch": 0.48487, + "grad_norm": 0.5923433065650765, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 48487 + }, + { + "epoch": 0.48488, + "grad_norm": 0.6424789233953192, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 48488 + }, + { + "epoch": 0.48489, + "grad_norm": 0.694184822688187, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 48489 + }, + { + "epoch": 0.4849, + "grad_norm": 0.7300922502068183, + "learning_rate": 0.003, + "loss": 3.976, + "step": 48490 + }, + { + "epoch": 0.48491, + "grad_norm": 0.7008382900794456, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 48491 + }, + { + "epoch": 0.48492, + "grad_norm": 0.6188122543582755, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 48492 + }, + { + "epoch": 0.48493, + "grad_norm": 0.636552813174313, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 48493 + }, + { + "epoch": 0.48494, + "grad_norm": 0.8804622415771013, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 48494 + }, + { + "epoch": 0.48495, + "grad_norm": 1.0306058189626341, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 48495 + }, + { + "epoch": 0.48496, + "grad_norm": 1.0669000701884452, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 48496 + }, + { + "epoch": 0.48497, + "grad_norm": 0.7838433706992325, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 48497 + }, + { + "epoch": 0.48498, + "grad_norm": 0.6795795567129165, + "learning_rate": 0.003, + "loss": 3.956, + "step": 48498 + }, + { + "epoch": 0.48499, + "grad_norm": 0.6632410769594081, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 48499 + }, + { + "epoch": 0.485, + "grad_norm": 0.6253602775121025, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 48500 + }, + { + "epoch": 0.48501, + "grad_norm": 0.5918755712374008, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 48501 + }, + { + "epoch": 0.48502, + "grad_norm": 0.5714131270265835, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 48502 + }, + { + "epoch": 0.48503, + "grad_norm": 0.6289090836731683, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 48503 + }, + { + "epoch": 0.48504, + "grad_norm": 0.7007795871606877, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 48504 + }, + { + "epoch": 0.48505, + "grad_norm": 0.7342242989411618, + "learning_rate": 0.003, + "loss": 3.9246, + "step": 48505 + }, + { + "epoch": 0.48506, + "grad_norm": 0.7718249715974974, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 48506 + }, + { + "epoch": 0.48507, + "grad_norm": 0.861500287465662, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 48507 + }, + { + "epoch": 0.48508, + "grad_norm": 0.9239644528232158, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 48508 + }, + { + "epoch": 0.48509, + "grad_norm": 0.9570099668025516, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 48509 + }, + { + "epoch": 0.4851, + "grad_norm": 1.0014791763858488, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 48510 + }, + { + "epoch": 0.48511, + "grad_norm": 0.7804174125297719, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 48511 + }, + { + "epoch": 0.48512, + "grad_norm": 0.7004724178996752, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 48512 + }, + { + "epoch": 0.48513, + "grad_norm": 0.7647613341986788, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 48513 + }, + { + "epoch": 0.48514, + "grad_norm": 0.6876680024403141, + "learning_rate": 0.003, + "loss": 3.994, + "step": 48514 + }, + { + "epoch": 0.48515, + "grad_norm": 0.7703406082202352, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 48515 + }, + { + "epoch": 0.48516, + "grad_norm": 0.8150654390405496, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 48516 + }, + { + "epoch": 0.48517, + "grad_norm": 0.9403692153145529, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 48517 + }, + { + "epoch": 0.48518, + "grad_norm": 1.0221606804227374, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 48518 + }, + { + "epoch": 0.48519, + "grad_norm": 0.9133176163115098, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 48519 + }, + { + "epoch": 0.4852, + "grad_norm": 1.0041042999294156, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 48520 + }, + { + "epoch": 0.48521, + "grad_norm": 1.0814748251952444, + "learning_rate": 0.003, + "loss": 4.009, + "step": 48521 + }, + { + "epoch": 0.48522, + "grad_norm": 0.8856458636253088, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 48522 + }, + { + "epoch": 0.48523, + "grad_norm": 0.6699074527421551, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 48523 + }, + { + "epoch": 0.48524, + "grad_norm": 0.5942381104912335, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 48524 + }, + { + "epoch": 0.48525, + "grad_norm": 0.6566674919127332, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 48525 + }, + { + "epoch": 0.48526, + "grad_norm": 0.6119025761226663, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 48526 + }, + { + "epoch": 0.48527, + "grad_norm": 0.6084556211657891, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 48527 + }, + { + "epoch": 0.48528, + "grad_norm": 0.6621121037855902, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 48528 + }, + { + "epoch": 0.48529, + "grad_norm": 0.7457107210923126, + "learning_rate": 0.003, + "loss": 3.989, + "step": 48529 + }, + { + "epoch": 0.4853, + "grad_norm": 0.8421334408543709, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 48530 + }, + { + "epoch": 0.48531, + "grad_norm": 0.994124758597536, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 48531 + }, + { + "epoch": 0.48532, + "grad_norm": 1.016568308487743, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 48532 + }, + { + "epoch": 0.48533, + "grad_norm": 0.9259726778246239, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 48533 + }, + { + "epoch": 0.48534, + "grad_norm": 0.7144730841776455, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 48534 + }, + { + "epoch": 0.48535, + "grad_norm": 0.6906961730846991, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 48535 + }, + { + "epoch": 0.48536, + "grad_norm": 0.7015255724776784, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 48536 + }, + { + "epoch": 0.48537, + "grad_norm": 0.7577333924795967, + "learning_rate": 0.003, + "loss": 3.998, + "step": 48537 + }, + { + "epoch": 0.48538, + "grad_norm": 0.735016588627576, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 48538 + }, + { + "epoch": 0.48539, + "grad_norm": 0.6829417971224532, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 48539 + }, + { + "epoch": 0.4854, + "grad_norm": 0.6815768063111967, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 48540 + }, + { + "epoch": 0.48541, + "grad_norm": 0.7309150749391048, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 48541 + }, + { + "epoch": 0.48542, + "grad_norm": 0.8450506072187697, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 48542 + }, + { + "epoch": 0.48543, + "grad_norm": 0.8550468683667173, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 48543 + }, + { + "epoch": 0.48544, + "grad_norm": 0.6767200831430409, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 48544 + }, + { + "epoch": 0.48545, + "grad_norm": 0.7103879209152609, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 48545 + }, + { + "epoch": 0.48546, + "grad_norm": 0.7355905365813082, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 48546 + }, + { + "epoch": 0.48547, + "grad_norm": 0.7267986555068295, + "learning_rate": 0.003, + "loss": 4.002, + "step": 48547 + }, + { + "epoch": 0.48548, + "grad_norm": 0.6956872334814206, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 48548 + }, + { + "epoch": 0.48549, + "grad_norm": 0.7067822562616753, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 48549 + }, + { + "epoch": 0.4855, + "grad_norm": 0.7391600011764116, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 48550 + }, + { + "epoch": 0.48551, + "grad_norm": 0.6689074851743065, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 48551 + }, + { + "epoch": 0.48552, + "grad_norm": 0.6129963267940476, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 48552 + }, + { + "epoch": 0.48553, + "grad_norm": 0.7162664830857287, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 48553 + }, + { + "epoch": 0.48554, + "grad_norm": 0.7732083957731217, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 48554 + }, + { + "epoch": 0.48555, + "grad_norm": 0.8904219877435395, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 48555 + }, + { + "epoch": 0.48556, + "grad_norm": 0.8583117595090466, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 48556 + }, + { + "epoch": 0.48557, + "grad_norm": 0.8887241887465257, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 48557 + }, + { + "epoch": 0.48558, + "grad_norm": 0.7832780555372445, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 48558 + }, + { + "epoch": 0.48559, + "grad_norm": 0.7339272852986262, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 48559 + }, + { + "epoch": 0.4856, + "grad_norm": 0.6527883531835446, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 48560 + }, + { + "epoch": 0.48561, + "grad_norm": 0.6255457819646226, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 48561 + }, + { + "epoch": 0.48562, + "grad_norm": 0.6632509681247928, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 48562 + }, + { + "epoch": 0.48563, + "grad_norm": 0.821735428164067, + "learning_rate": 0.003, + "loss": 3.984, + "step": 48563 + }, + { + "epoch": 0.48564, + "grad_norm": 0.9577578218902498, + "learning_rate": 0.003, + "loss": 4.001, + "step": 48564 + }, + { + "epoch": 0.48565, + "grad_norm": 0.9649109536452617, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 48565 + }, + { + "epoch": 0.48566, + "grad_norm": 0.7632418156317284, + "learning_rate": 0.003, + "loss": 4.007, + "step": 48566 + }, + { + "epoch": 0.48567, + "grad_norm": 0.9286014000639791, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 48567 + }, + { + "epoch": 0.48568, + "grad_norm": 0.8961250890089257, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 48568 + }, + { + "epoch": 0.48569, + "grad_norm": 0.9136202498658764, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 48569 + }, + { + "epoch": 0.4857, + "grad_norm": 1.0525371142618876, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 48570 + }, + { + "epoch": 0.48571, + "grad_norm": 1.0268950643311863, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 48571 + }, + { + "epoch": 0.48572, + "grad_norm": 0.9099535143624463, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 48572 + }, + { + "epoch": 0.48573, + "grad_norm": 0.8918892714172908, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 48573 + }, + { + "epoch": 0.48574, + "grad_norm": 0.8550009493530417, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 48574 + }, + { + "epoch": 0.48575, + "grad_norm": 0.8041607190531097, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 48575 + }, + { + "epoch": 0.48576, + "grad_norm": 0.7876968801207498, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 48576 + }, + { + "epoch": 0.48577, + "grad_norm": 0.7314724197058188, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 48577 + }, + { + "epoch": 0.48578, + "grad_norm": 0.7369753084144665, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 48578 + }, + { + "epoch": 0.48579, + "grad_norm": 0.6563080508718762, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 48579 + }, + { + "epoch": 0.4858, + "grad_norm": 0.6421635914025592, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 48580 + }, + { + "epoch": 0.48581, + "grad_norm": 0.6861631470697589, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 48581 + }, + { + "epoch": 0.48582, + "grad_norm": 0.7484670231900008, + "learning_rate": 0.003, + "loss": 3.971, + "step": 48582 + }, + { + "epoch": 0.48583, + "grad_norm": 0.7615115804050171, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 48583 + }, + { + "epoch": 0.48584, + "grad_norm": 0.765481860977513, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 48584 + }, + { + "epoch": 0.48585, + "grad_norm": 0.7370191794529009, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 48585 + }, + { + "epoch": 0.48586, + "grad_norm": 0.7541456378571731, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 48586 + }, + { + "epoch": 0.48587, + "grad_norm": 0.807447207668524, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 48587 + }, + { + "epoch": 0.48588, + "grad_norm": 0.839294099647994, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 48588 + }, + { + "epoch": 0.48589, + "grad_norm": 0.7963139963867619, + "learning_rate": 0.003, + "loss": 3.951, + "step": 48589 + }, + { + "epoch": 0.4859, + "grad_norm": 0.8730469990329288, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 48590 + }, + { + "epoch": 0.48591, + "grad_norm": 0.8930031386608654, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 48591 + }, + { + "epoch": 0.48592, + "grad_norm": 0.7017793315690134, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 48592 + }, + { + "epoch": 0.48593, + "grad_norm": 0.6542148242072363, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 48593 + }, + { + "epoch": 0.48594, + "grad_norm": 0.7080680688378839, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 48594 + }, + { + "epoch": 0.48595, + "grad_norm": 0.8075592168958066, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 48595 + }, + { + "epoch": 0.48596, + "grad_norm": 0.9216780311322655, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 48596 + }, + { + "epoch": 0.48597, + "grad_norm": 1.0494061367709169, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 48597 + }, + { + "epoch": 0.48598, + "grad_norm": 0.8737229230903674, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 48598 + }, + { + "epoch": 0.48599, + "grad_norm": 0.8941271397296934, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 48599 + }, + { + "epoch": 0.486, + "grad_norm": 0.8208256910115613, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 48600 + }, + { + "epoch": 0.48601, + "grad_norm": 0.7597528549669226, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 48601 + }, + { + "epoch": 0.48602, + "grad_norm": 0.7925523259662134, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 48602 + }, + { + "epoch": 0.48603, + "grad_norm": 0.7970816196699886, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 48603 + }, + { + "epoch": 0.48604, + "grad_norm": 0.7343827937637131, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 48604 + }, + { + "epoch": 0.48605, + "grad_norm": 0.6937272633484403, + "learning_rate": 0.003, + "loss": 3.963, + "step": 48605 + }, + { + "epoch": 0.48606, + "grad_norm": 0.6571631344679018, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 48606 + }, + { + "epoch": 0.48607, + "grad_norm": 0.7248511720620853, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 48607 + }, + { + "epoch": 0.48608, + "grad_norm": 0.802786723290701, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 48608 + }, + { + "epoch": 0.48609, + "grad_norm": 0.7890763214590854, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 48609 + }, + { + "epoch": 0.4861, + "grad_norm": 0.8559964852963137, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 48610 + }, + { + "epoch": 0.48611, + "grad_norm": 0.8863622314470867, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 48611 + }, + { + "epoch": 0.48612, + "grad_norm": 0.6816860609441264, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 48612 + }, + { + "epoch": 0.48613, + "grad_norm": 0.5632068608967017, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 48613 + }, + { + "epoch": 0.48614, + "grad_norm": 0.5458686710179261, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 48614 + }, + { + "epoch": 0.48615, + "grad_norm": 0.5438843282964018, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 48615 + }, + { + "epoch": 0.48616, + "grad_norm": 0.5716382117005399, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 48616 + }, + { + "epoch": 0.48617, + "grad_norm": 0.611974145993531, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 48617 + }, + { + "epoch": 0.48618, + "grad_norm": 0.6416413710847685, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 48618 + }, + { + "epoch": 0.48619, + "grad_norm": 0.7204402742867245, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 48619 + }, + { + "epoch": 0.4862, + "grad_norm": 0.7693920079584705, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 48620 + }, + { + "epoch": 0.48621, + "grad_norm": 0.8190010090423546, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 48621 + }, + { + "epoch": 0.48622, + "grad_norm": 0.8636593239502066, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 48622 + }, + { + "epoch": 0.48623, + "grad_norm": 0.9049168035675078, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 48623 + }, + { + "epoch": 0.48624, + "grad_norm": 0.846206838513047, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 48624 + }, + { + "epoch": 0.48625, + "grad_norm": 0.7585384199683509, + "learning_rate": 0.003, + "loss": 4.0088, + "step": 48625 + }, + { + "epoch": 0.48626, + "grad_norm": 0.8286291277775745, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 48626 + }, + { + "epoch": 0.48627, + "grad_norm": 0.7569122281735499, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 48627 + }, + { + "epoch": 0.48628, + "grad_norm": 0.7463156423093075, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 48628 + }, + { + "epoch": 0.48629, + "grad_norm": 0.6597208921949185, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 48629 + }, + { + "epoch": 0.4863, + "grad_norm": 0.763665497591136, + "learning_rate": 0.003, + "loss": 3.99, + "step": 48630 + }, + { + "epoch": 0.48631, + "grad_norm": 0.9545092290403193, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 48631 + }, + { + "epoch": 0.48632, + "grad_norm": 1.289347544638885, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 48632 + }, + { + "epoch": 0.48633, + "grad_norm": 0.6705672894109219, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 48633 + }, + { + "epoch": 0.48634, + "grad_norm": 0.7346998181956336, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 48634 + }, + { + "epoch": 0.48635, + "grad_norm": 0.8182518786438442, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 48635 + }, + { + "epoch": 0.48636, + "grad_norm": 0.828418532360436, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 48636 + }, + { + "epoch": 0.48637, + "grad_norm": 0.8850164375119288, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 48637 + }, + { + "epoch": 0.48638, + "grad_norm": 0.8545912075731832, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 48638 + }, + { + "epoch": 0.48639, + "grad_norm": 0.9717065900944146, + "learning_rate": 0.003, + "loss": 4.0347, + "step": 48639 + }, + { + "epoch": 0.4864, + "grad_norm": 0.9612657130191569, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 48640 + }, + { + "epoch": 0.48641, + "grad_norm": 0.9631353000724874, + "learning_rate": 0.003, + "loss": 4.0208, + "step": 48641 + }, + { + "epoch": 0.48642, + "grad_norm": 0.9894280115330094, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 48642 + }, + { + "epoch": 0.48643, + "grad_norm": 0.9837640476233782, + "learning_rate": 0.003, + "loss": 4.0, + "step": 48643 + }, + { + "epoch": 0.48644, + "grad_norm": 0.9600227490533624, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 48644 + }, + { + "epoch": 0.48645, + "grad_norm": 0.7832980508928531, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 48645 + }, + { + "epoch": 0.48646, + "grad_norm": 0.7161752987320251, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 48646 + }, + { + "epoch": 0.48647, + "grad_norm": 0.6119408911200693, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 48647 + }, + { + "epoch": 0.48648, + "grad_norm": 0.6100580885410682, + "learning_rate": 0.003, + "loss": 3.962, + "step": 48648 + }, + { + "epoch": 0.48649, + "grad_norm": 0.644631451950328, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 48649 + }, + { + "epoch": 0.4865, + "grad_norm": 0.7170482712309793, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 48650 + }, + { + "epoch": 0.48651, + "grad_norm": 0.8742381829605047, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 48651 + }, + { + "epoch": 0.48652, + "grad_norm": 0.8553808533004419, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 48652 + }, + { + "epoch": 0.48653, + "grad_norm": 0.7869368376274783, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 48653 + }, + { + "epoch": 0.48654, + "grad_norm": 0.6721645901713602, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 48654 + }, + { + "epoch": 0.48655, + "grad_norm": 0.6162270664857433, + "learning_rate": 0.003, + "loss": 3.9399, + "step": 48655 + }, + { + "epoch": 0.48656, + "grad_norm": 0.6738293542016224, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 48656 + }, + { + "epoch": 0.48657, + "grad_norm": 0.6659525411027255, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 48657 + }, + { + "epoch": 0.48658, + "grad_norm": 0.7662860062003302, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 48658 + }, + { + "epoch": 0.48659, + "grad_norm": 0.740811702126688, + "learning_rate": 0.003, + "loss": 3.979, + "step": 48659 + }, + { + "epoch": 0.4866, + "grad_norm": 0.7633547465237305, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 48660 + }, + { + "epoch": 0.48661, + "grad_norm": 0.7397417611953442, + "learning_rate": 0.003, + "loss": 3.971, + "step": 48661 + }, + { + "epoch": 0.48662, + "grad_norm": 0.7037465362172711, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 48662 + }, + { + "epoch": 0.48663, + "grad_norm": 0.7870579398535575, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 48663 + }, + { + "epoch": 0.48664, + "grad_norm": 0.9877407072961211, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 48664 + }, + { + "epoch": 0.48665, + "grad_norm": 1.0653216041987228, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 48665 + }, + { + "epoch": 0.48666, + "grad_norm": 0.9199372738557768, + "learning_rate": 0.003, + "loss": 3.992, + "step": 48666 + }, + { + "epoch": 0.48667, + "grad_norm": 0.833864697683132, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 48667 + }, + { + "epoch": 0.48668, + "grad_norm": 0.8207689387274013, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 48668 + }, + { + "epoch": 0.48669, + "grad_norm": 0.7616141197539743, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 48669 + }, + { + "epoch": 0.4867, + "grad_norm": 0.7554995725010595, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 48670 + }, + { + "epoch": 0.48671, + "grad_norm": 0.6891291417058165, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 48671 + }, + { + "epoch": 0.48672, + "grad_norm": 0.7865170723902133, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 48672 + }, + { + "epoch": 0.48673, + "grad_norm": 0.8018063690445847, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 48673 + }, + { + "epoch": 0.48674, + "grad_norm": 0.8644168049356457, + "learning_rate": 0.003, + "loss": 3.979, + "step": 48674 + }, + { + "epoch": 0.48675, + "grad_norm": 0.843881330222281, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 48675 + }, + { + "epoch": 0.48676, + "grad_norm": 0.7650790510522493, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 48676 + }, + { + "epoch": 0.48677, + "grad_norm": 0.7235327217916451, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 48677 + }, + { + "epoch": 0.48678, + "grad_norm": 0.6878249398681564, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 48678 + }, + { + "epoch": 0.48679, + "grad_norm": 0.663068857097657, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 48679 + }, + { + "epoch": 0.4868, + "grad_norm": 0.6881451244925824, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 48680 + }, + { + "epoch": 0.48681, + "grad_norm": 0.721871847442708, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 48681 + }, + { + "epoch": 0.48682, + "grad_norm": 0.6989180105846051, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 48682 + }, + { + "epoch": 0.48683, + "grad_norm": 0.7314469403616435, + "learning_rate": 0.003, + "loss": 3.967, + "step": 48683 + }, + { + "epoch": 0.48684, + "grad_norm": 0.7831888022556297, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 48684 + }, + { + "epoch": 0.48685, + "grad_norm": 0.7642397924224367, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 48685 + }, + { + "epoch": 0.48686, + "grad_norm": 0.7130233740803057, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 48686 + }, + { + "epoch": 0.48687, + "grad_norm": 0.7305612138944193, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 48687 + }, + { + "epoch": 0.48688, + "grad_norm": 0.8669845571519571, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 48688 + }, + { + "epoch": 0.48689, + "grad_norm": 0.9538570085751854, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 48689 + }, + { + "epoch": 0.4869, + "grad_norm": 0.8157241504023326, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 48690 + }, + { + "epoch": 0.48691, + "grad_norm": 0.7545800833075759, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 48691 + }, + { + "epoch": 0.48692, + "grad_norm": 0.7633202323280752, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 48692 + }, + { + "epoch": 0.48693, + "grad_norm": 0.7940830218335581, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 48693 + }, + { + "epoch": 0.48694, + "grad_norm": 0.8729124326488076, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 48694 + }, + { + "epoch": 0.48695, + "grad_norm": 0.6939245322106031, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 48695 + }, + { + "epoch": 0.48696, + "grad_norm": 0.7071313374287111, + "learning_rate": 0.003, + "loss": 3.977, + "step": 48696 + }, + { + "epoch": 0.48697, + "grad_norm": 0.7086961399231894, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 48697 + }, + { + "epoch": 0.48698, + "grad_norm": 0.6680157111308387, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 48698 + }, + { + "epoch": 0.48699, + "grad_norm": 0.6047960292197297, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 48699 + }, + { + "epoch": 0.487, + "grad_norm": 0.6396568669963504, + "learning_rate": 0.003, + "loss": 3.9432, + "step": 48700 + }, + { + "epoch": 0.48701, + "grad_norm": 0.7057818689067155, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 48701 + }, + { + "epoch": 0.48702, + "grad_norm": 0.7311015710069945, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 48702 + }, + { + "epoch": 0.48703, + "grad_norm": 0.6289863247533384, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 48703 + }, + { + "epoch": 0.48704, + "grad_norm": 0.5837977394483146, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 48704 + }, + { + "epoch": 0.48705, + "grad_norm": 0.6176597273954474, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 48705 + }, + { + "epoch": 0.48706, + "grad_norm": 0.6536565455199704, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 48706 + }, + { + "epoch": 0.48707, + "grad_norm": 0.7327245125468402, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 48707 + }, + { + "epoch": 0.48708, + "grad_norm": 0.9379481065099734, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 48708 + }, + { + "epoch": 0.48709, + "grad_norm": 0.9544678051066466, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 48709 + }, + { + "epoch": 0.4871, + "grad_norm": 0.8313322706591023, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 48710 + }, + { + "epoch": 0.48711, + "grad_norm": 0.8673634461369001, + "learning_rate": 0.003, + "loss": 4.006, + "step": 48711 + }, + { + "epoch": 0.48712, + "grad_norm": 1.0007568552670674, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 48712 + }, + { + "epoch": 0.48713, + "grad_norm": 1.012174322526282, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 48713 + }, + { + "epoch": 0.48714, + "grad_norm": 1.0148600082487, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 48714 + }, + { + "epoch": 0.48715, + "grad_norm": 0.987075060866878, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 48715 + }, + { + "epoch": 0.48716, + "grad_norm": 1.0439862405436637, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 48716 + }, + { + "epoch": 0.48717, + "grad_norm": 0.9788326797437896, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 48717 + }, + { + "epoch": 0.48718, + "grad_norm": 0.9831025617065718, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 48718 + }, + { + "epoch": 0.48719, + "grad_norm": 0.8109689905171511, + "learning_rate": 0.003, + "loss": 4.0106, + "step": 48719 + }, + { + "epoch": 0.4872, + "grad_norm": 0.7885978134854984, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 48720 + }, + { + "epoch": 0.48721, + "grad_norm": 0.804403307463379, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 48721 + }, + { + "epoch": 0.48722, + "grad_norm": 0.6998326358321172, + "learning_rate": 0.003, + "loss": 3.9329, + "step": 48722 + }, + { + "epoch": 0.48723, + "grad_norm": 0.728029023303318, + "learning_rate": 0.003, + "loss": 4.013, + "step": 48723 + }, + { + "epoch": 0.48724, + "grad_norm": 0.6932218371632527, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 48724 + }, + { + "epoch": 0.48725, + "grad_norm": 0.6053626910783271, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 48725 + }, + { + "epoch": 0.48726, + "grad_norm": 0.624685620829238, + "learning_rate": 0.003, + "loss": 3.972, + "step": 48726 + }, + { + "epoch": 0.48727, + "grad_norm": 0.7288716544986147, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 48727 + }, + { + "epoch": 0.48728, + "grad_norm": 1.011030427336902, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 48728 + }, + { + "epoch": 0.48729, + "grad_norm": 1.198816988753528, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 48729 + }, + { + "epoch": 0.4873, + "grad_norm": 0.5872648992344623, + "learning_rate": 0.003, + "loss": 4.0112, + "step": 48730 + }, + { + "epoch": 0.48731, + "grad_norm": 0.7361320285552261, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 48731 + }, + { + "epoch": 0.48732, + "grad_norm": 0.8672620411182648, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 48732 + }, + { + "epoch": 0.48733, + "grad_norm": 0.8773052883149535, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 48733 + }, + { + "epoch": 0.48734, + "grad_norm": 0.7713508381906204, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 48734 + }, + { + "epoch": 0.48735, + "grad_norm": 0.6722479715546132, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 48735 + }, + { + "epoch": 0.48736, + "grad_norm": 0.649286000028865, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 48736 + }, + { + "epoch": 0.48737, + "grad_norm": 0.7403812417254715, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 48737 + }, + { + "epoch": 0.48738, + "grad_norm": 0.8973099261133852, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 48738 + }, + { + "epoch": 0.48739, + "grad_norm": 0.9341747258785671, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 48739 + }, + { + "epoch": 0.4874, + "grad_norm": 0.8356673876284376, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 48740 + }, + { + "epoch": 0.48741, + "grad_norm": 0.8504618222889782, + "learning_rate": 0.003, + "loss": 4.0245, + "step": 48741 + }, + { + "epoch": 0.48742, + "grad_norm": 0.8132917423730662, + "learning_rate": 0.003, + "loss": 3.967, + "step": 48742 + }, + { + "epoch": 0.48743, + "grad_norm": 0.7503709541861446, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 48743 + }, + { + "epoch": 0.48744, + "grad_norm": 0.7409214740754273, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 48744 + }, + { + "epoch": 0.48745, + "grad_norm": 0.6305380423293143, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 48745 + }, + { + "epoch": 0.48746, + "grad_norm": 0.5980492611818443, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 48746 + }, + { + "epoch": 0.48747, + "grad_norm": 0.5925985728238126, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 48747 + }, + { + "epoch": 0.48748, + "grad_norm": 0.6595692457157444, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 48748 + }, + { + "epoch": 0.48749, + "grad_norm": 0.6752245644220299, + "learning_rate": 0.003, + "loss": 3.958, + "step": 48749 + }, + { + "epoch": 0.4875, + "grad_norm": 0.7010365966883694, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 48750 + }, + { + "epoch": 0.48751, + "grad_norm": 0.6355353885569804, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 48751 + }, + { + "epoch": 0.48752, + "grad_norm": 0.6267529560489015, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 48752 + }, + { + "epoch": 0.48753, + "grad_norm": 0.5779001855560509, + "learning_rate": 0.003, + "loss": 3.9389, + "step": 48753 + }, + { + "epoch": 0.48754, + "grad_norm": 0.4805953146700279, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 48754 + }, + { + "epoch": 0.48755, + "grad_norm": 0.5307762801895096, + "learning_rate": 0.003, + "loss": 3.9434, + "step": 48755 + }, + { + "epoch": 0.48756, + "grad_norm": 0.5953891367993055, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 48756 + }, + { + "epoch": 0.48757, + "grad_norm": 0.6793163648670958, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 48757 + }, + { + "epoch": 0.48758, + "grad_norm": 0.7808968355007342, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 48758 + }, + { + "epoch": 0.48759, + "grad_norm": 1.0237982990478318, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 48759 + }, + { + "epoch": 0.4876, + "grad_norm": 1.251613662921549, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 48760 + }, + { + "epoch": 0.48761, + "grad_norm": 0.6848764915088493, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 48761 + }, + { + "epoch": 0.48762, + "grad_norm": 0.7266493349465184, + "learning_rate": 0.003, + "loss": 3.97, + "step": 48762 + }, + { + "epoch": 0.48763, + "grad_norm": 0.7178083509075297, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 48763 + }, + { + "epoch": 0.48764, + "grad_norm": 0.7529915257557414, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 48764 + }, + { + "epoch": 0.48765, + "grad_norm": 0.787735913190546, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 48765 + }, + { + "epoch": 0.48766, + "grad_norm": 0.8654116873872467, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 48766 + }, + { + "epoch": 0.48767, + "grad_norm": 1.0291330056138348, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 48767 + }, + { + "epoch": 0.48768, + "grad_norm": 0.861339516393791, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 48768 + }, + { + "epoch": 0.48769, + "grad_norm": 0.7145025200295835, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 48769 + }, + { + "epoch": 0.4877, + "grad_norm": 0.6620962183335727, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 48770 + }, + { + "epoch": 0.48771, + "grad_norm": 0.7365413636643183, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 48771 + }, + { + "epoch": 0.48772, + "grad_norm": 0.9367217240208856, + "learning_rate": 0.003, + "loss": 3.983, + "step": 48772 + }, + { + "epoch": 0.48773, + "grad_norm": 1.003229863495116, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 48773 + }, + { + "epoch": 0.48774, + "grad_norm": 1.018552496760495, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 48774 + }, + { + "epoch": 0.48775, + "grad_norm": 0.9236497560319032, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 48775 + }, + { + "epoch": 0.48776, + "grad_norm": 0.9186679547725068, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 48776 + }, + { + "epoch": 0.48777, + "grad_norm": 0.9546371486809717, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 48777 + }, + { + "epoch": 0.48778, + "grad_norm": 0.9103813620507439, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 48778 + }, + { + "epoch": 0.48779, + "grad_norm": 0.8839901359930374, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 48779 + }, + { + "epoch": 0.4878, + "grad_norm": 0.9581672387708291, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 48780 + }, + { + "epoch": 0.48781, + "grad_norm": 1.0118665846117938, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 48781 + }, + { + "epoch": 0.48782, + "grad_norm": 1.1290367190792419, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 48782 + }, + { + "epoch": 0.48783, + "grad_norm": 0.8739185383664746, + "learning_rate": 0.003, + "loss": 4.0326, + "step": 48783 + }, + { + "epoch": 0.48784, + "grad_norm": 0.9064236122617916, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 48784 + }, + { + "epoch": 0.48785, + "grad_norm": 0.9605868449622808, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 48785 + }, + { + "epoch": 0.48786, + "grad_norm": 0.9798716523116369, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 48786 + }, + { + "epoch": 0.48787, + "grad_norm": 0.8527685162721714, + "learning_rate": 0.003, + "loss": 4.0202, + "step": 48787 + }, + { + "epoch": 0.48788, + "grad_norm": 0.8159989482870175, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 48788 + }, + { + "epoch": 0.48789, + "grad_norm": 0.8462850599501136, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 48789 + }, + { + "epoch": 0.4879, + "grad_norm": 0.8260478166991015, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 48790 + }, + { + "epoch": 0.48791, + "grad_norm": 0.7473358273996288, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 48791 + }, + { + "epoch": 0.48792, + "grad_norm": 0.6595769605365734, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 48792 + }, + { + "epoch": 0.48793, + "grad_norm": 0.6332929637858179, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 48793 + }, + { + "epoch": 0.48794, + "grad_norm": 0.6609465575221088, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 48794 + }, + { + "epoch": 0.48795, + "grad_norm": 0.6565532090606149, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 48795 + }, + { + "epoch": 0.48796, + "grad_norm": 0.6953241817762025, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 48796 + }, + { + "epoch": 0.48797, + "grad_norm": 0.7509275877313206, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 48797 + }, + { + "epoch": 0.48798, + "grad_norm": 0.6499655008360922, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 48798 + }, + { + "epoch": 0.48799, + "grad_norm": 0.539520781303673, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 48799 + }, + { + "epoch": 0.488, + "grad_norm": 0.5235514367110741, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 48800 + }, + { + "epoch": 0.48801, + "grad_norm": 0.5502837377491235, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 48801 + }, + { + "epoch": 0.48802, + "grad_norm": 0.539103522220125, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 48802 + }, + { + "epoch": 0.48803, + "grad_norm": 0.5820076823325279, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 48803 + }, + { + "epoch": 0.48804, + "grad_norm": 0.5981235791623472, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 48804 + }, + { + "epoch": 0.48805, + "grad_norm": 0.5955408260227146, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 48805 + }, + { + "epoch": 0.48806, + "grad_norm": 0.6204480412772554, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 48806 + }, + { + "epoch": 0.48807, + "grad_norm": 0.6729382693044829, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 48807 + }, + { + "epoch": 0.48808, + "grad_norm": 0.7215932810538248, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 48808 + }, + { + "epoch": 0.48809, + "grad_norm": 0.7610970282529899, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 48809 + }, + { + "epoch": 0.4881, + "grad_norm": 0.8969832854891686, + "learning_rate": 0.003, + "loss": 3.9494, + "step": 48810 + }, + { + "epoch": 0.48811, + "grad_norm": 0.9781329913778704, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 48811 + }, + { + "epoch": 0.48812, + "grad_norm": 1.0548613036492536, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 48812 + }, + { + "epoch": 0.48813, + "grad_norm": 0.8271576656656144, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 48813 + }, + { + "epoch": 0.48814, + "grad_norm": 0.7966480180840309, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 48814 + }, + { + "epoch": 0.48815, + "grad_norm": 0.8685279029566282, + "learning_rate": 0.003, + "loss": 3.98, + "step": 48815 + }, + { + "epoch": 0.48816, + "grad_norm": 0.8620945633528637, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 48816 + }, + { + "epoch": 0.48817, + "grad_norm": 0.8793698434000107, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 48817 + }, + { + "epoch": 0.48818, + "grad_norm": 0.761451270193433, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 48818 + }, + { + "epoch": 0.48819, + "grad_norm": 0.6803609477386907, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 48819 + }, + { + "epoch": 0.4882, + "grad_norm": 0.6075082081162229, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 48820 + }, + { + "epoch": 0.48821, + "grad_norm": 0.6482454477286778, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 48821 + }, + { + "epoch": 0.48822, + "grad_norm": 0.7035439457636535, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 48822 + }, + { + "epoch": 0.48823, + "grad_norm": 0.6463757520516251, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 48823 + }, + { + "epoch": 0.48824, + "grad_norm": 0.6607604133031204, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 48824 + }, + { + "epoch": 0.48825, + "grad_norm": 0.7317237947047014, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 48825 + }, + { + "epoch": 0.48826, + "grad_norm": 0.7194018938926098, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 48826 + }, + { + "epoch": 0.48827, + "grad_norm": 0.6851155574360076, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 48827 + }, + { + "epoch": 0.48828, + "grad_norm": 0.7732291331768602, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 48828 + }, + { + "epoch": 0.48829, + "grad_norm": 0.8428758199153419, + "learning_rate": 0.003, + "loss": 3.972, + "step": 48829 + }, + { + "epoch": 0.4883, + "grad_norm": 0.7872269024582564, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 48830 + }, + { + "epoch": 0.48831, + "grad_norm": 0.6912515091035379, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 48831 + }, + { + "epoch": 0.48832, + "grad_norm": 0.7024774673683764, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 48832 + }, + { + "epoch": 0.48833, + "grad_norm": 0.7226554930318729, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 48833 + }, + { + "epoch": 0.48834, + "grad_norm": 0.6074964940011407, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 48834 + }, + { + "epoch": 0.48835, + "grad_norm": 0.6327465292827303, + "learning_rate": 0.003, + "loss": 3.9585, + "step": 48835 + }, + { + "epoch": 0.48836, + "grad_norm": 0.6075614897206305, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 48836 + }, + { + "epoch": 0.48837, + "grad_norm": 0.7174125527577107, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 48837 + }, + { + "epoch": 0.48838, + "grad_norm": 0.8597525004045831, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 48838 + }, + { + "epoch": 0.48839, + "grad_norm": 1.1384472155767846, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 48839 + }, + { + "epoch": 0.4884, + "grad_norm": 0.9700189907806163, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 48840 + }, + { + "epoch": 0.48841, + "grad_norm": 0.792222769458798, + "learning_rate": 0.003, + "loss": 3.995, + "step": 48841 + }, + { + "epoch": 0.48842, + "grad_norm": 0.7699986077591129, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 48842 + }, + { + "epoch": 0.48843, + "grad_norm": 0.9000404444081966, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 48843 + }, + { + "epoch": 0.48844, + "grad_norm": 1.1795570608720487, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 48844 + }, + { + "epoch": 0.48845, + "grad_norm": 0.9388111460333379, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 48845 + }, + { + "epoch": 0.48846, + "grad_norm": 0.8593652315374652, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 48846 + }, + { + "epoch": 0.48847, + "grad_norm": 0.8547962971500136, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 48847 + }, + { + "epoch": 0.48848, + "grad_norm": 0.8492955942926984, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 48848 + }, + { + "epoch": 0.48849, + "grad_norm": 0.6958568295688722, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 48849 + }, + { + "epoch": 0.4885, + "grad_norm": 0.8072990287342897, + "learning_rate": 0.003, + "loss": 3.9591, + "step": 48850 + }, + { + "epoch": 0.48851, + "grad_norm": 0.9711014892890036, + "learning_rate": 0.003, + "loss": 3.974, + "step": 48851 + }, + { + "epoch": 0.48852, + "grad_norm": 1.0558303527220365, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 48852 + }, + { + "epoch": 0.48853, + "grad_norm": 0.8930040742463354, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 48853 + }, + { + "epoch": 0.48854, + "grad_norm": 0.8913525739777821, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 48854 + }, + { + "epoch": 0.48855, + "grad_norm": 0.8238563761515542, + "learning_rate": 0.003, + "loss": 3.998, + "step": 48855 + }, + { + "epoch": 0.48856, + "grad_norm": 0.8461303633367252, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 48856 + }, + { + "epoch": 0.48857, + "grad_norm": 0.9740197151848513, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 48857 + }, + { + "epoch": 0.48858, + "grad_norm": 0.9700442878138802, + "learning_rate": 0.003, + "loss": 4.029, + "step": 48858 + }, + { + "epoch": 0.48859, + "grad_norm": 0.8759665213016342, + "learning_rate": 0.003, + "loss": 4.0299, + "step": 48859 + }, + { + "epoch": 0.4886, + "grad_norm": 0.8539161387699481, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 48860 + }, + { + "epoch": 0.48861, + "grad_norm": 0.7779518510732621, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 48861 + }, + { + "epoch": 0.48862, + "grad_norm": 0.738800598214188, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 48862 + }, + { + "epoch": 0.48863, + "grad_norm": 0.8325999880425838, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 48863 + }, + { + "epoch": 0.48864, + "grad_norm": 1.000539652261361, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 48864 + }, + { + "epoch": 0.48865, + "grad_norm": 1.1905828611013933, + "learning_rate": 0.003, + "loss": 3.994, + "step": 48865 + }, + { + "epoch": 0.48866, + "grad_norm": 0.8673371854793089, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 48866 + }, + { + "epoch": 0.48867, + "grad_norm": 0.7541440617045707, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 48867 + }, + { + "epoch": 0.48868, + "grad_norm": 0.6904376398235863, + "learning_rate": 0.003, + "loss": 4.001, + "step": 48868 + }, + { + "epoch": 0.48869, + "grad_norm": 0.663057777338178, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 48869 + }, + { + "epoch": 0.4887, + "grad_norm": 0.6168090025235583, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 48870 + }, + { + "epoch": 0.48871, + "grad_norm": 0.6694945138832885, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 48871 + }, + { + "epoch": 0.48872, + "grad_norm": 0.6433360968336849, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 48872 + }, + { + "epoch": 0.48873, + "grad_norm": 0.6716419917431379, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 48873 + }, + { + "epoch": 0.48874, + "grad_norm": 0.7083577143829313, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 48874 + }, + { + "epoch": 0.48875, + "grad_norm": 0.6883196011743435, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 48875 + }, + { + "epoch": 0.48876, + "grad_norm": 0.5978132581767727, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 48876 + }, + { + "epoch": 0.48877, + "grad_norm": 0.5100925703208434, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 48877 + }, + { + "epoch": 0.48878, + "grad_norm": 0.46087017572575745, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 48878 + }, + { + "epoch": 0.48879, + "grad_norm": 0.4846385744154429, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 48879 + }, + { + "epoch": 0.4888, + "grad_norm": 0.5464502087571225, + "learning_rate": 0.003, + "loss": 4.004, + "step": 48880 + }, + { + "epoch": 0.48881, + "grad_norm": 0.5484809386555678, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 48881 + }, + { + "epoch": 0.48882, + "grad_norm": 0.5368893385589587, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 48882 + }, + { + "epoch": 0.48883, + "grad_norm": 0.5380939630441203, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 48883 + }, + { + "epoch": 0.48884, + "grad_norm": 0.5532178682168613, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 48884 + }, + { + "epoch": 0.48885, + "grad_norm": 0.5725288862104824, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 48885 + }, + { + "epoch": 0.48886, + "grad_norm": 0.689086784156007, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 48886 + }, + { + "epoch": 0.48887, + "grad_norm": 0.8039679871306862, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 48887 + }, + { + "epoch": 0.48888, + "grad_norm": 0.8843560084284748, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 48888 + }, + { + "epoch": 0.48889, + "grad_norm": 0.9046934791633839, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 48889 + }, + { + "epoch": 0.4889, + "grad_norm": 1.0824722718980633, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 48890 + }, + { + "epoch": 0.48891, + "grad_norm": 0.9487216262708272, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 48891 + }, + { + "epoch": 0.48892, + "grad_norm": 0.7884341182783745, + "learning_rate": 0.003, + "loss": 3.9488, + "step": 48892 + }, + { + "epoch": 0.48893, + "grad_norm": 0.6842088008676066, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 48893 + }, + { + "epoch": 0.48894, + "grad_norm": 0.6410345228161519, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 48894 + }, + { + "epoch": 0.48895, + "grad_norm": 0.8449629626074168, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 48895 + }, + { + "epoch": 0.48896, + "grad_norm": 1.0000802639542334, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 48896 + }, + { + "epoch": 0.48897, + "grad_norm": 1.2428215645687157, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 48897 + }, + { + "epoch": 0.48898, + "grad_norm": 0.8728797474908089, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 48898 + }, + { + "epoch": 0.48899, + "grad_norm": 0.9210908346182853, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 48899 + }, + { + "epoch": 0.489, + "grad_norm": 0.8499610291808227, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 48900 + }, + { + "epoch": 0.48901, + "grad_norm": 0.737182098279379, + "learning_rate": 0.003, + "loss": 3.969, + "step": 48901 + }, + { + "epoch": 0.48902, + "grad_norm": 0.6758995864682165, + "learning_rate": 0.003, + "loss": 4.0297, + "step": 48902 + }, + { + "epoch": 0.48903, + "grad_norm": 0.7471512529506434, + "learning_rate": 0.003, + "loss": 3.997, + "step": 48903 + }, + { + "epoch": 0.48904, + "grad_norm": 0.7581174539058378, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 48904 + }, + { + "epoch": 0.48905, + "grad_norm": 0.737950661181258, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 48905 + }, + { + "epoch": 0.48906, + "grad_norm": 0.6334608191955291, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 48906 + }, + { + "epoch": 0.48907, + "grad_norm": 0.5712301563622411, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 48907 + }, + { + "epoch": 0.48908, + "grad_norm": 0.530864177549952, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 48908 + }, + { + "epoch": 0.48909, + "grad_norm": 0.5738962757850362, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 48909 + }, + { + "epoch": 0.4891, + "grad_norm": 0.6160851208572181, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 48910 + }, + { + "epoch": 0.48911, + "grad_norm": 0.707972199041839, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 48911 + }, + { + "epoch": 0.48912, + "grad_norm": 0.7508885825038679, + "learning_rate": 0.003, + "loss": 4.0152, + "step": 48912 + }, + { + "epoch": 0.48913, + "grad_norm": 0.7469979180540919, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 48913 + }, + { + "epoch": 0.48914, + "grad_norm": 0.8097955447022224, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 48914 + }, + { + "epoch": 0.48915, + "grad_norm": 0.8997667827043595, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 48915 + }, + { + "epoch": 0.48916, + "grad_norm": 0.8520433594274857, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 48916 + }, + { + "epoch": 0.48917, + "grad_norm": 0.9819549814310143, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 48917 + }, + { + "epoch": 0.48918, + "grad_norm": 0.9542615802802891, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 48918 + }, + { + "epoch": 0.48919, + "grad_norm": 0.8661433946922098, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 48919 + }, + { + "epoch": 0.4892, + "grad_norm": 0.8503555792196137, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 48920 + }, + { + "epoch": 0.48921, + "grad_norm": 0.8913460230255739, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 48921 + }, + { + "epoch": 0.48922, + "grad_norm": 0.8771928857393769, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 48922 + }, + { + "epoch": 0.48923, + "grad_norm": 0.8312279866024935, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 48923 + }, + { + "epoch": 0.48924, + "grad_norm": 0.7157391918075804, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 48924 + }, + { + "epoch": 0.48925, + "grad_norm": 0.6849685736986514, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 48925 + }, + { + "epoch": 0.48926, + "grad_norm": 0.737710937738467, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 48926 + }, + { + "epoch": 0.48927, + "grad_norm": 0.7260532069861195, + "learning_rate": 0.003, + "loss": 3.9435, + "step": 48927 + }, + { + "epoch": 0.48928, + "grad_norm": 0.7910676850020939, + "learning_rate": 0.003, + "loss": 4.014, + "step": 48928 + }, + { + "epoch": 0.48929, + "grad_norm": 0.8970275199715736, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 48929 + }, + { + "epoch": 0.4893, + "grad_norm": 1.0604853059174137, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 48930 + }, + { + "epoch": 0.48931, + "grad_norm": 0.9077963699828984, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 48931 + }, + { + "epoch": 0.48932, + "grad_norm": 0.8755825368418969, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 48932 + }, + { + "epoch": 0.48933, + "grad_norm": 0.9081891750285236, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 48933 + }, + { + "epoch": 0.48934, + "grad_norm": 0.9665219571263528, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 48934 + }, + { + "epoch": 0.48935, + "grad_norm": 0.9070614396015856, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 48935 + }, + { + "epoch": 0.48936, + "grad_norm": 0.8522553698971319, + "learning_rate": 0.003, + "loss": 4.004, + "step": 48936 + }, + { + "epoch": 0.48937, + "grad_norm": 0.7279818061140414, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 48937 + }, + { + "epoch": 0.48938, + "grad_norm": 0.6641217214398719, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 48938 + }, + { + "epoch": 0.48939, + "grad_norm": 0.7019744293666514, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 48939 + }, + { + "epoch": 0.4894, + "grad_norm": 0.7265908709791725, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 48940 + }, + { + "epoch": 0.48941, + "grad_norm": 0.7314719681951094, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 48941 + }, + { + "epoch": 0.48942, + "grad_norm": 0.7091992821772948, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 48942 + }, + { + "epoch": 0.48943, + "grad_norm": 0.7386233685500377, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 48943 + }, + { + "epoch": 0.48944, + "grad_norm": 0.9042173887301913, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 48944 + }, + { + "epoch": 0.48945, + "grad_norm": 1.1029491644355442, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 48945 + }, + { + "epoch": 0.48946, + "grad_norm": 1.0657161129675925, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 48946 + }, + { + "epoch": 0.48947, + "grad_norm": 0.8620726374994243, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 48947 + }, + { + "epoch": 0.48948, + "grad_norm": 0.9377003062589501, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 48948 + }, + { + "epoch": 0.48949, + "grad_norm": 0.8364045773030494, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 48949 + }, + { + "epoch": 0.4895, + "grad_norm": 0.7925123202454819, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 48950 + }, + { + "epoch": 0.48951, + "grad_norm": 0.6915791094775536, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 48951 + }, + { + "epoch": 0.48952, + "grad_norm": 0.687032875684965, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 48952 + }, + { + "epoch": 0.48953, + "grad_norm": 0.5717308221553691, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 48953 + }, + { + "epoch": 0.48954, + "grad_norm": 0.6136941072336468, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 48954 + }, + { + "epoch": 0.48955, + "grad_norm": 0.5389322782109461, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 48955 + }, + { + "epoch": 0.48956, + "grad_norm": 0.4961789146100712, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 48956 + }, + { + "epoch": 0.48957, + "grad_norm": 0.5174127844073789, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 48957 + }, + { + "epoch": 0.48958, + "grad_norm": 0.6320700454389855, + "learning_rate": 0.003, + "loss": 3.951, + "step": 48958 + }, + { + "epoch": 0.48959, + "grad_norm": 0.6554721496319412, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 48959 + }, + { + "epoch": 0.4896, + "grad_norm": 0.7310297187444368, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 48960 + }, + { + "epoch": 0.48961, + "grad_norm": 0.8904844049001548, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 48961 + }, + { + "epoch": 0.48962, + "grad_norm": 0.920009433470542, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 48962 + }, + { + "epoch": 0.48963, + "grad_norm": 0.8580769166917965, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 48963 + }, + { + "epoch": 0.48964, + "grad_norm": 0.8105681151781832, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 48964 + }, + { + "epoch": 0.48965, + "grad_norm": 0.7815688485308436, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 48965 + }, + { + "epoch": 0.48966, + "grad_norm": 0.5885976259494801, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 48966 + }, + { + "epoch": 0.48967, + "grad_norm": 0.5771378724402088, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 48967 + }, + { + "epoch": 0.48968, + "grad_norm": 0.6068566003428364, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 48968 + }, + { + "epoch": 0.48969, + "grad_norm": 0.6542970046250064, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 48969 + }, + { + "epoch": 0.4897, + "grad_norm": 0.6940542650245886, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 48970 + }, + { + "epoch": 0.48971, + "grad_norm": 0.6907624988258821, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 48971 + }, + { + "epoch": 0.48972, + "grad_norm": 0.7015783005154057, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 48972 + }, + { + "epoch": 0.48973, + "grad_norm": 0.7694952172162022, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 48973 + }, + { + "epoch": 0.48974, + "grad_norm": 0.8028362404683707, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 48974 + }, + { + "epoch": 0.48975, + "grad_norm": 0.8605117144065381, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 48975 + }, + { + "epoch": 0.48976, + "grad_norm": 0.7387709677631767, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 48976 + }, + { + "epoch": 0.48977, + "grad_norm": 0.6727545373435915, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 48977 + }, + { + "epoch": 0.48978, + "grad_norm": 0.6374207753006202, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 48978 + }, + { + "epoch": 0.48979, + "grad_norm": 0.6467177462380779, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 48979 + }, + { + "epoch": 0.4898, + "grad_norm": 0.7561888642834644, + "learning_rate": 0.003, + "loss": 3.996, + "step": 48980 + }, + { + "epoch": 0.48981, + "grad_norm": 0.8421423481321586, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 48981 + }, + { + "epoch": 0.48982, + "grad_norm": 0.7884218038200338, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 48982 + }, + { + "epoch": 0.48983, + "grad_norm": 0.8429690138668151, + "learning_rate": 0.003, + "loss": 3.997, + "step": 48983 + }, + { + "epoch": 0.48984, + "grad_norm": 0.9269934987154816, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 48984 + }, + { + "epoch": 0.48985, + "grad_norm": 1.1471520825949142, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 48985 + }, + { + "epoch": 0.48986, + "grad_norm": 0.9423112098857325, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 48986 + }, + { + "epoch": 0.48987, + "grad_norm": 0.8429603577982506, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 48987 + }, + { + "epoch": 0.48988, + "grad_norm": 0.8000332146823717, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 48988 + }, + { + "epoch": 0.48989, + "grad_norm": 0.7793181346512466, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 48989 + }, + { + "epoch": 0.4899, + "grad_norm": 0.636461657607414, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 48990 + }, + { + "epoch": 0.48991, + "grad_norm": 0.671969930902124, + "learning_rate": 0.003, + "loss": 3.999, + "step": 48991 + }, + { + "epoch": 0.48992, + "grad_norm": 0.7076320879842749, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 48992 + }, + { + "epoch": 0.48993, + "grad_norm": 0.8250094694949967, + "learning_rate": 0.003, + "loss": 3.972, + "step": 48993 + }, + { + "epoch": 0.48994, + "grad_norm": 1.118820255581653, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 48994 + }, + { + "epoch": 0.48995, + "grad_norm": 1.014283890616246, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 48995 + }, + { + "epoch": 0.48996, + "grad_norm": 0.9007495427303149, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 48996 + }, + { + "epoch": 0.48997, + "grad_norm": 0.7813892585767743, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 48997 + }, + { + "epoch": 0.48998, + "grad_norm": 0.8752613278369712, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 48998 + }, + { + "epoch": 0.48999, + "grad_norm": 0.8887097777239981, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 48999 + }, + { + "epoch": 0.49, + "grad_norm": 0.9621005802818965, + "learning_rate": 0.003, + "loss": 3.998, + "step": 49000 + }, + { + "epoch": 0.49001, + "grad_norm": 0.9378665095332831, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 49001 + }, + { + "epoch": 0.49002, + "grad_norm": 0.7294534985052914, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 49002 + }, + { + "epoch": 0.49003, + "grad_norm": 0.6677823725657066, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 49003 + }, + { + "epoch": 0.49004, + "grad_norm": 0.712124800363768, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 49004 + }, + { + "epoch": 0.49005, + "grad_norm": 0.716823128917643, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 49005 + }, + { + "epoch": 0.49006, + "grad_norm": 0.6724899530093296, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 49006 + }, + { + "epoch": 0.49007, + "grad_norm": 0.7690379418085187, + "learning_rate": 0.003, + "loss": 3.99, + "step": 49007 + }, + { + "epoch": 0.49008, + "grad_norm": 0.8405776958178843, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 49008 + }, + { + "epoch": 0.49009, + "grad_norm": 0.7810167740428827, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 49009 + }, + { + "epoch": 0.4901, + "grad_norm": 0.6620742006115684, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 49010 + }, + { + "epoch": 0.49011, + "grad_norm": 0.6496326658471062, + "learning_rate": 0.003, + "loss": 3.983, + "step": 49011 + }, + { + "epoch": 0.49012, + "grad_norm": 0.6552751326424806, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 49012 + }, + { + "epoch": 0.49013, + "grad_norm": 0.7444393562266506, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 49013 + }, + { + "epoch": 0.49014, + "grad_norm": 0.8003696671572412, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 49014 + }, + { + "epoch": 0.49015, + "grad_norm": 0.8842352517309113, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 49015 + }, + { + "epoch": 0.49016, + "grad_norm": 0.8430065066946607, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 49016 + }, + { + "epoch": 0.49017, + "grad_norm": 0.7891990832052248, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 49017 + }, + { + "epoch": 0.49018, + "grad_norm": 0.8369019900341204, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 49018 + }, + { + "epoch": 0.49019, + "grad_norm": 0.9203294619103871, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 49019 + }, + { + "epoch": 0.4902, + "grad_norm": 1.1360887345050263, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 49020 + }, + { + "epoch": 0.49021, + "grad_norm": 0.8996898152785384, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 49021 + }, + { + "epoch": 0.49022, + "grad_norm": 0.8944600668807611, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 49022 + }, + { + "epoch": 0.49023, + "grad_norm": 0.8577694926036032, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 49023 + }, + { + "epoch": 0.49024, + "grad_norm": 0.8198002722664222, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 49024 + }, + { + "epoch": 0.49025, + "grad_norm": 0.8002198009918728, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 49025 + }, + { + "epoch": 0.49026, + "grad_norm": 0.6322606265654146, + "learning_rate": 0.003, + "loss": 3.966, + "step": 49026 + }, + { + "epoch": 0.49027, + "grad_norm": 0.5928595589613371, + "learning_rate": 0.003, + "loss": 3.962, + "step": 49027 + }, + { + "epoch": 0.49028, + "grad_norm": 0.6234491226542492, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 49028 + }, + { + "epoch": 0.49029, + "grad_norm": 0.5867232548939072, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 49029 + }, + { + "epoch": 0.4903, + "grad_norm": 0.6631285133788194, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 49030 + }, + { + "epoch": 0.49031, + "grad_norm": 0.7341832327076239, + "learning_rate": 0.003, + "loss": 3.95, + "step": 49031 + }, + { + "epoch": 0.49032, + "grad_norm": 0.827451956259037, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 49032 + }, + { + "epoch": 0.49033, + "grad_norm": 0.9462801799846825, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 49033 + }, + { + "epoch": 0.49034, + "grad_norm": 0.8187273622243028, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 49034 + }, + { + "epoch": 0.49035, + "grad_norm": 0.7545392665315425, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 49035 + }, + { + "epoch": 0.49036, + "grad_norm": 0.9728667441095502, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 49036 + }, + { + "epoch": 0.49037, + "grad_norm": 0.9746382747247919, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 49037 + }, + { + "epoch": 0.49038, + "grad_norm": 0.9742151813629462, + "learning_rate": 0.003, + "loss": 3.968, + "step": 49038 + }, + { + "epoch": 0.49039, + "grad_norm": 0.8964486326117875, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 49039 + }, + { + "epoch": 0.4904, + "grad_norm": 0.8580560475211596, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 49040 + }, + { + "epoch": 0.49041, + "grad_norm": 0.7973471607286032, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 49041 + }, + { + "epoch": 0.49042, + "grad_norm": 0.7930032984781304, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 49042 + }, + { + "epoch": 0.49043, + "grad_norm": 0.7764002372503099, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 49043 + }, + { + "epoch": 0.49044, + "grad_norm": 0.6921323927552229, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 49044 + }, + { + "epoch": 0.49045, + "grad_norm": 0.5858788084127058, + "learning_rate": 0.003, + "loss": 3.982, + "step": 49045 + }, + { + "epoch": 0.49046, + "grad_norm": 0.6036866289657961, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 49046 + }, + { + "epoch": 0.49047, + "grad_norm": 0.6673109226732196, + "learning_rate": 0.003, + "loss": 4.008, + "step": 49047 + }, + { + "epoch": 0.49048, + "grad_norm": 0.8952935749372368, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 49048 + }, + { + "epoch": 0.49049, + "grad_norm": 1.184045035099886, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 49049 + }, + { + "epoch": 0.4905, + "grad_norm": 0.9787835087061669, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 49050 + }, + { + "epoch": 0.49051, + "grad_norm": 0.8934278630164448, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 49051 + }, + { + "epoch": 0.49052, + "grad_norm": 0.8093728939926135, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 49052 + }, + { + "epoch": 0.49053, + "grad_norm": 0.800723296089464, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 49053 + }, + { + "epoch": 0.49054, + "grad_norm": 0.8432644651895973, + "learning_rate": 0.003, + "loss": 3.99, + "step": 49054 + }, + { + "epoch": 0.49055, + "grad_norm": 0.8316506317204153, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 49055 + }, + { + "epoch": 0.49056, + "grad_norm": 0.7028862877701658, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 49056 + }, + { + "epoch": 0.49057, + "grad_norm": 0.7246942344379841, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 49057 + }, + { + "epoch": 0.49058, + "grad_norm": 0.6941450081589148, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 49058 + }, + { + "epoch": 0.49059, + "grad_norm": 0.6767261987460532, + "learning_rate": 0.003, + "loss": 3.992, + "step": 49059 + }, + { + "epoch": 0.4906, + "grad_norm": 0.665334099264291, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 49060 + }, + { + "epoch": 0.49061, + "grad_norm": 0.7107311487731965, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 49061 + }, + { + "epoch": 0.49062, + "grad_norm": 0.8356528832087523, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 49062 + }, + { + "epoch": 0.49063, + "grad_norm": 0.8585729490865021, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 49063 + }, + { + "epoch": 0.49064, + "grad_norm": 0.8189420684652872, + "learning_rate": 0.003, + "loss": 3.992, + "step": 49064 + }, + { + "epoch": 0.49065, + "grad_norm": 0.6934225013398638, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 49065 + }, + { + "epoch": 0.49066, + "grad_norm": 0.5645808791341289, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 49066 + }, + { + "epoch": 0.49067, + "grad_norm": 0.6002701677927522, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 49067 + }, + { + "epoch": 0.49068, + "grad_norm": 0.6040578589573664, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 49068 + }, + { + "epoch": 0.49069, + "grad_norm": 0.6283142276416819, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 49069 + }, + { + "epoch": 0.4907, + "grad_norm": 0.542400024109153, + "learning_rate": 0.003, + "loss": 3.982, + "step": 49070 + }, + { + "epoch": 0.49071, + "grad_norm": 0.4866871622764497, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 49071 + }, + { + "epoch": 0.49072, + "grad_norm": 0.4817763958218107, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 49072 + }, + { + "epoch": 0.49073, + "grad_norm": 0.4346064028406011, + "learning_rate": 0.003, + "loss": 3.944, + "step": 49073 + }, + { + "epoch": 0.49074, + "grad_norm": 0.46454179672553253, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 49074 + }, + { + "epoch": 0.49075, + "grad_norm": 0.4733767327595923, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 49075 + }, + { + "epoch": 0.49076, + "grad_norm": 0.6146438260823015, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 49076 + }, + { + "epoch": 0.49077, + "grad_norm": 0.799544959729936, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 49077 + }, + { + "epoch": 0.49078, + "grad_norm": 1.0022315964043218, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 49078 + }, + { + "epoch": 0.49079, + "grad_norm": 0.9812743265395367, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 49079 + }, + { + "epoch": 0.4908, + "grad_norm": 0.9308709386275074, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 49080 + }, + { + "epoch": 0.49081, + "grad_norm": 0.8932491534273171, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 49081 + }, + { + "epoch": 0.49082, + "grad_norm": 0.8281366691391582, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 49082 + }, + { + "epoch": 0.49083, + "grad_norm": 0.8317598921821295, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 49083 + }, + { + "epoch": 0.49084, + "grad_norm": 0.7455061510064408, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 49084 + }, + { + "epoch": 0.49085, + "grad_norm": 0.6410841208044714, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 49085 + }, + { + "epoch": 0.49086, + "grad_norm": 0.7079292042652604, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 49086 + }, + { + "epoch": 0.49087, + "grad_norm": 0.8055537907861634, + "learning_rate": 0.003, + "loss": 3.973, + "step": 49087 + }, + { + "epoch": 0.49088, + "grad_norm": 0.8819887493476332, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 49088 + }, + { + "epoch": 0.49089, + "grad_norm": 0.8152606177060272, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 49089 + }, + { + "epoch": 0.4909, + "grad_norm": 0.846270974829888, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 49090 + }, + { + "epoch": 0.49091, + "grad_norm": 0.9654216926664579, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 49091 + }, + { + "epoch": 0.49092, + "grad_norm": 0.8047581156709833, + "learning_rate": 0.003, + "loss": 3.985, + "step": 49092 + }, + { + "epoch": 0.49093, + "grad_norm": 0.8954785061263741, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 49093 + }, + { + "epoch": 0.49094, + "grad_norm": 0.9907448610291749, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 49094 + }, + { + "epoch": 0.49095, + "grad_norm": 0.9939574223160083, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 49095 + }, + { + "epoch": 0.49096, + "grad_norm": 0.8607483533866201, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 49096 + }, + { + "epoch": 0.49097, + "grad_norm": 0.8543925069459195, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 49097 + }, + { + "epoch": 0.49098, + "grad_norm": 0.7670519399530368, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 49098 + }, + { + "epoch": 0.49099, + "grad_norm": 0.7365413815561053, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 49099 + }, + { + "epoch": 0.491, + "grad_norm": 0.7074560301792491, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 49100 + }, + { + "epoch": 0.49101, + "grad_norm": 0.8521181716044366, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 49101 + }, + { + "epoch": 0.49102, + "grad_norm": 0.9064767023013678, + "learning_rate": 0.003, + "loss": 3.981, + "step": 49102 + }, + { + "epoch": 0.49103, + "grad_norm": 1.0422846957011784, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 49103 + }, + { + "epoch": 0.49104, + "grad_norm": 1.1589350641512903, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 49104 + }, + { + "epoch": 0.49105, + "grad_norm": 1.0925208156870507, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 49105 + }, + { + "epoch": 0.49106, + "grad_norm": 0.8333594787219729, + "learning_rate": 0.003, + "loss": 3.98, + "step": 49106 + }, + { + "epoch": 0.49107, + "grad_norm": 0.7727869720272134, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 49107 + }, + { + "epoch": 0.49108, + "grad_norm": 0.7287122190809577, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 49108 + }, + { + "epoch": 0.49109, + "grad_norm": 0.7291904243233217, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 49109 + }, + { + "epoch": 0.4911, + "grad_norm": 0.7458258432483204, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 49110 + }, + { + "epoch": 0.49111, + "grad_norm": 0.7055043327423745, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 49111 + }, + { + "epoch": 0.49112, + "grad_norm": 0.6438530869312341, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 49112 + }, + { + "epoch": 0.49113, + "grad_norm": 0.6877555062530527, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 49113 + }, + { + "epoch": 0.49114, + "grad_norm": 0.6696477570276599, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 49114 + }, + { + "epoch": 0.49115, + "grad_norm": 0.6363601617641738, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 49115 + }, + { + "epoch": 0.49116, + "grad_norm": 0.621112796957358, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 49116 + }, + { + "epoch": 0.49117, + "grad_norm": 0.6320915149172419, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 49117 + }, + { + "epoch": 0.49118, + "grad_norm": 0.79879413136186, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 49118 + }, + { + "epoch": 0.49119, + "grad_norm": 0.9477276031483954, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 49119 + }, + { + "epoch": 0.4912, + "grad_norm": 1.1055452821584961, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 49120 + }, + { + "epoch": 0.49121, + "grad_norm": 0.8825422290433099, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 49121 + }, + { + "epoch": 0.49122, + "grad_norm": 0.7192625987646205, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 49122 + }, + { + "epoch": 0.49123, + "grad_norm": 0.6022780037096883, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 49123 + }, + { + "epoch": 0.49124, + "grad_norm": 0.5846030314088015, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 49124 + }, + { + "epoch": 0.49125, + "grad_norm": 0.504495909373285, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 49125 + }, + { + "epoch": 0.49126, + "grad_norm": 0.5655795626750769, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 49126 + }, + { + "epoch": 0.49127, + "grad_norm": 0.6111907191764624, + "learning_rate": 0.003, + "loss": 3.967, + "step": 49127 + }, + { + "epoch": 0.49128, + "grad_norm": 0.5408461925599996, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 49128 + }, + { + "epoch": 0.49129, + "grad_norm": 0.6071734732687811, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 49129 + }, + { + "epoch": 0.4913, + "grad_norm": 0.7509295498252577, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 49130 + }, + { + "epoch": 0.49131, + "grad_norm": 1.0272161314911061, + "learning_rate": 0.003, + "loss": 3.994, + "step": 49131 + }, + { + "epoch": 0.49132, + "grad_norm": 1.0461162344723125, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 49132 + }, + { + "epoch": 0.49133, + "grad_norm": 0.796383670114, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 49133 + }, + { + "epoch": 0.49134, + "grad_norm": 0.6477681416176979, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 49134 + }, + { + "epoch": 0.49135, + "grad_norm": 0.6765605643912491, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 49135 + }, + { + "epoch": 0.49136, + "grad_norm": 0.6739030361696264, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 49136 + }, + { + "epoch": 0.49137, + "grad_norm": 0.6727561963949292, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 49137 + }, + { + "epoch": 0.49138, + "grad_norm": 0.6182872870148345, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 49138 + }, + { + "epoch": 0.49139, + "grad_norm": 0.6049504104638567, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 49139 + }, + { + "epoch": 0.4914, + "grad_norm": 0.6960480463096916, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 49140 + }, + { + "epoch": 0.49141, + "grad_norm": 0.719374709083749, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 49141 + }, + { + "epoch": 0.49142, + "grad_norm": 0.7649569781392996, + "learning_rate": 0.003, + "loss": 3.9429, + "step": 49142 + }, + { + "epoch": 0.49143, + "grad_norm": 0.805009996375855, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 49143 + }, + { + "epoch": 0.49144, + "grad_norm": 0.8995711803596708, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 49144 + }, + { + "epoch": 0.49145, + "grad_norm": 0.8983833425035872, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 49145 + }, + { + "epoch": 0.49146, + "grad_norm": 0.920325989746591, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 49146 + }, + { + "epoch": 0.49147, + "grad_norm": 0.9494823988657193, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 49147 + }, + { + "epoch": 0.49148, + "grad_norm": 1.0964147120365337, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 49148 + }, + { + "epoch": 0.49149, + "grad_norm": 0.8988158675659421, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 49149 + }, + { + "epoch": 0.4915, + "grad_norm": 0.9237975568633661, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 49150 + }, + { + "epoch": 0.49151, + "grad_norm": 0.9864448862463704, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 49151 + }, + { + "epoch": 0.49152, + "grad_norm": 1.0734078481125817, + "learning_rate": 0.003, + "loss": 4.0124, + "step": 49152 + }, + { + "epoch": 0.49153, + "grad_norm": 0.8390666871390671, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 49153 + }, + { + "epoch": 0.49154, + "grad_norm": 0.7744854792782244, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 49154 + }, + { + "epoch": 0.49155, + "grad_norm": 0.7058028437167574, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 49155 + }, + { + "epoch": 0.49156, + "grad_norm": 0.6819955494484832, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 49156 + }, + { + "epoch": 0.49157, + "grad_norm": 0.6949602335017032, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 49157 + }, + { + "epoch": 0.49158, + "grad_norm": 0.7751284104829742, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 49158 + }, + { + "epoch": 0.49159, + "grad_norm": 0.8392021206161691, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 49159 + }, + { + "epoch": 0.4916, + "grad_norm": 0.908972567044677, + "learning_rate": 0.003, + "loss": 4.006, + "step": 49160 + }, + { + "epoch": 0.49161, + "grad_norm": 0.8733106517011833, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 49161 + }, + { + "epoch": 0.49162, + "grad_norm": 0.7461032481879111, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 49162 + }, + { + "epoch": 0.49163, + "grad_norm": 0.6789254226404638, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 49163 + }, + { + "epoch": 0.49164, + "grad_norm": 0.7095544135429632, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 49164 + }, + { + "epoch": 0.49165, + "grad_norm": 0.7116390307366189, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 49165 + }, + { + "epoch": 0.49166, + "grad_norm": 0.648817323067954, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 49166 + }, + { + "epoch": 0.49167, + "grad_norm": 0.6681841719905262, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 49167 + }, + { + "epoch": 0.49168, + "grad_norm": 0.6435789603006695, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 49168 + }, + { + "epoch": 0.49169, + "grad_norm": 0.6842530577392774, + "learning_rate": 0.003, + "loss": 3.976, + "step": 49169 + }, + { + "epoch": 0.4917, + "grad_norm": 0.7906370408630644, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 49170 + }, + { + "epoch": 0.49171, + "grad_norm": 0.8109872680031872, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 49171 + }, + { + "epoch": 0.49172, + "grad_norm": 0.7740627910419253, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 49172 + }, + { + "epoch": 0.49173, + "grad_norm": 0.7799955418830381, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 49173 + }, + { + "epoch": 0.49174, + "grad_norm": 0.8663012870145383, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 49174 + }, + { + "epoch": 0.49175, + "grad_norm": 1.0328786215776382, + "learning_rate": 0.003, + "loss": 4.0381, + "step": 49175 + }, + { + "epoch": 0.49176, + "grad_norm": 1.0090920018826763, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 49176 + }, + { + "epoch": 0.49177, + "grad_norm": 1.0716644343722406, + "learning_rate": 0.003, + "loss": 3.9604, + "step": 49177 + }, + { + "epoch": 0.49178, + "grad_norm": 1.0338389828102232, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 49178 + }, + { + "epoch": 0.49179, + "grad_norm": 0.9040220226869348, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 49179 + }, + { + "epoch": 0.4918, + "grad_norm": 0.8396037624502123, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 49180 + }, + { + "epoch": 0.49181, + "grad_norm": 0.7859877917240234, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 49181 + }, + { + "epoch": 0.49182, + "grad_norm": 0.6188265539890937, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 49182 + }, + { + "epoch": 0.49183, + "grad_norm": 0.6223705957409927, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 49183 + }, + { + "epoch": 0.49184, + "grad_norm": 0.7201696856465678, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 49184 + }, + { + "epoch": 0.49185, + "grad_norm": 0.7392004539162242, + "learning_rate": 0.003, + "loss": 3.96, + "step": 49185 + }, + { + "epoch": 0.49186, + "grad_norm": 0.7125350276050703, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 49186 + }, + { + "epoch": 0.49187, + "grad_norm": 0.8134879852413767, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 49187 + }, + { + "epoch": 0.49188, + "grad_norm": 0.8094986796173267, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 49188 + }, + { + "epoch": 0.49189, + "grad_norm": 0.7955983522659701, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 49189 + }, + { + "epoch": 0.4919, + "grad_norm": 0.7464757804637338, + "learning_rate": 0.003, + "loss": 3.944, + "step": 49190 + }, + { + "epoch": 0.49191, + "grad_norm": 0.7493177179823918, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 49191 + }, + { + "epoch": 0.49192, + "grad_norm": 0.6613823974811326, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 49192 + }, + { + "epoch": 0.49193, + "grad_norm": 0.8151815821859468, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 49193 + }, + { + "epoch": 0.49194, + "grad_norm": 0.8117803716846697, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 49194 + }, + { + "epoch": 0.49195, + "grad_norm": 0.7942084847786176, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 49195 + }, + { + "epoch": 0.49196, + "grad_norm": 0.7152530956225738, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 49196 + }, + { + "epoch": 0.49197, + "grad_norm": 0.6642497448603911, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 49197 + }, + { + "epoch": 0.49198, + "grad_norm": 0.6280211831634535, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 49198 + }, + { + "epoch": 0.49199, + "grad_norm": 0.6528360212368943, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 49199 + }, + { + "epoch": 0.492, + "grad_norm": 0.7018217945540349, + "learning_rate": 0.003, + "loss": 3.9421, + "step": 49200 + }, + { + "epoch": 0.49201, + "grad_norm": 0.7459203228161984, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 49201 + }, + { + "epoch": 0.49202, + "grad_norm": 0.7879400782141139, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 49202 + }, + { + "epoch": 0.49203, + "grad_norm": 0.9423060063821553, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 49203 + }, + { + "epoch": 0.49204, + "grad_norm": 1.157374452686444, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 49204 + }, + { + "epoch": 0.49205, + "grad_norm": 0.7242383555742565, + "learning_rate": 0.003, + "loss": 3.979, + "step": 49205 + }, + { + "epoch": 0.49206, + "grad_norm": 0.6820408939608645, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 49206 + }, + { + "epoch": 0.49207, + "grad_norm": 0.7643006458257062, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 49207 + }, + { + "epoch": 0.49208, + "grad_norm": 0.8542083161101947, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 49208 + }, + { + "epoch": 0.49209, + "grad_norm": 0.8630435965337444, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 49209 + }, + { + "epoch": 0.4921, + "grad_norm": 0.7663132128074964, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 49210 + }, + { + "epoch": 0.49211, + "grad_norm": 0.7105154633435795, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 49211 + }, + { + "epoch": 0.49212, + "grad_norm": 0.6817324094084739, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 49212 + }, + { + "epoch": 0.49213, + "grad_norm": 0.6789898224345788, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 49213 + }, + { + "epoch": 0.49214, + "grad_norm": 0.72353926656654, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 49214 + }, + { + "epoch": 0.49215, + "grad_norm": 0.842441181989476, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 49215 + }, + { + "epoch": 0.49216, + "grad_norm": 0.8443849770162253, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 49216 + }, + { + "epoch": 0.49217, + "grad_norm": 0.7877962435076922, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 49217 + }, + { + "epoch": 0.49218, + "grad_norm": 0.8053395104252352, + "learning_rate": 0.003, + "loss": 4.0083, + "step": 49218 + }, + { + "epoch": 0.49219, + "grad_norm": 0.7974012965011932, + "learning_rate": 0.003, + "loss": 3.9365, + "step": 49219 + }, + { + "epoch": 0.4922, + "grad_norm": 0.8445043390778137, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 49220 + }, + { + "epoch": 0.49221, + "grad_norm": 0.8888360090686986, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 49221 + }, + { + "epoch": 0.49222, + "grad_norm": 0.781040415192949, + "learning_rate": 0.003, + "loss": 3.955, + "step": 49222 + }, + { + "epoch": 0.49223, + "grad_norm": 0.7065986290853843, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 49223 + }, + { + "epoch": 0.49224, + "grad_norm": 0.6925506890784885, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 49224 + }, + { + "epoch": 0.49225, + "grad_norm": 0.7327918520510224, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 49225 + }, + { + "epoch": 0.49226, + "grad_norm": 0.6551605819437789, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 49226 + }, + { + "epoch": 0.49227, + "grad_norm": 0.6568196325743426, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 49227 + }, + { + "epoch": 0.49228, + "grad_norm": 0.7111832758899134, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 49228 + }, + { + "epoch": 0.49229, + "grad_norm": 0.8430974212578325, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 49229 + }, + { + "epoch": 0.4923, + "grad_norm": 0.8860048207057714, + "learning_rate": 0.003, + "loss": 3.991, + "step": 49230 + }, + { + "epoch": 0.49231, + "grad_norm": 0.9542636315685306, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 49231 + }, + { + "epoch": 0.49232, + "grad_norm": 0.8751838673107853, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 49232 + }, + { + "epoch": 0.49233, + "grad_norm": 0.817253582479441, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 49233 + }, + { + "epoch": 0.49234, + "grad_norm": 0.7859185093090542, + "learning_rate": 0.003, + "loss": 4.0159, + "step": 49234 + }, + { + "epoch": 0.49235, + "grad_norm": 0.7409391184974393, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 49235 + }, + { + "epoch": 0.49236, + "grad_norm": 0.7882632916637846, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 49236 + }, + { + "epoch": 0.49237, + "grad_norm": 0.8568352826054781, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 49237 + }, + { + "epoch": 0.49238, + "grad_norm": 1.0752579995743605, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 49238 + }, + { + "epoch": 0.49239, + "grad_norm": 1.0867867742973512, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 49239 + }, + { + "epoch": 0.4924, + "grad_norm": 0.9972854109741162, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 49240 + }, + { + "epoch": 0.49241, + "grad_norm": 1.0063676398319146, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 49241 + }, + { + "epoch": 0.49242, + "grad_norm": 0.9078488356457195, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 49242 + }, + { + "epoch": 0.49243, + "grad_norm": 0.7007263744043245, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 49243 + }, + { + "epoch": 0.49244, + "grad_norm": 0.703896864490081, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 49244 + }, + { + "epoch": 0.49245, + "grad_norm": 0.7175455236437628, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 49245 + }, + { + "epoch": 0.49246, + "grad_norm": 0.7849553480503662, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 49246 + }, + { + "epoch": 0.49247, + "grad_norm": 0.80613130378336, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 49247 + }, + { + "epoch": 0.49248, + "grad_norm": 0.7383166927762147, + "learning_rate": 0.003, + "loss": 3.9713, + "step": 49248 + }, + { + "epoch": 0.49249, + "grad_norm": 0.6793478438956368, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 49249 + }, + { + "epoch": 0.4925, + "grad_norm": 0.6760900122415565, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 49250 + }, + { + "epoch": 0.49251, + "grad_norm": 0.7505458458851966, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 49251 + }, + { + "epoch": 0.49252, + "grad_norm": 0.7407034030094896, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 49252 + }, + { + "epoch": 0.49253, + "grad_norm": 0.7138305566498292, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 49253 + }, + { + "epoch": 0.49254, + "grad_norm": 0.6462726938879119, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 49254 + }, + { + "epoch": 0.49255, + "grad_norm": 0.5866470136012425, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 49255 + }, + { + "epoch": 0.49256, + "grad_norm": 0.5683418851625248, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 49256 + }, + { + "epoch": 0.49257, + "grad_norm": 0.4992375601029884, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 49257 + }, + { + "epoch": 0.49258, + "grad_norm": 0.5223906748316176, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 49258 + }, + { + "epoch": 0.49259, + "grad_norm": 0.5728542146957981, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 49259 + }, + { + "epoch": 0.4926, + "grad_norm": 0.6111412345590626, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 49260 + }, + { + "epoch": 0.49261, + "grad_norm": 0.5695888616954904, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 49261 + }, + { + "epoch": 0.49262, + "grad_norm": 0.5573520389348329, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 49262 + }, + { + "epoch": 0.49263, + "grad_norm": 0.6572595114469092, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 49263 + }, + { + "epoch": 0.49264, + "grad_norm": 0.7713569872694462, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 49264 + }, + { + "epoch": 0.49265, + "grad_norm": 0.9530626719508244, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 49265 + }, + { + "epoch": 0.49266, + "grad_norm": 1.168054306261319, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 49266 + }, + { + "epoch": 0.49267, + "grad_norm": 1.0247863655678842, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 49267 + }, + { + "epoch": 0.49268, + "grad_norm": 0.9464632328618945, + "learning_rate": 0.003, + "loss": 3.984, + "step": 49268 + }, + { + "epoch": 0.49269, + "grad_norm": 0.9202047412549698, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 49269 + }, + { + "epoch": 0.4927, + "grad_norm": 0.8707885351187252, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 49270 + }, + { + "epoch": 0.49271, + "grad_norm": 0.9099759747670634, + "learning_rate": 0.003, + "loss": 3.979, + "step": 49271 + }, + { + "epoch": 0.49272, + "grad_norm": 0.9569154089849976, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 49272 + }, + { + "epoch": 0.49273, + "grad_norm": 0.9110770110202726, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 49273 + }, + { + "epoch": 0.49274, + "grad_norm": 0.9461565346220268, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 49274 + }, + { + "epoch": 0.49275, + "grad_norm": 0.8630674084453092, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 49275 + }, + { + "epoch": 0.49276, + "grad_norm": 0.6996502746765375, + "learning_rate": 0.003, + "loss": 3.985, + "step": 49276 + }, + { + "epoch": 0.49277, + "grad_norm": 0.7235373692673809, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 49277 + }, + { + "epoch": 0.49278, + "grad_norm": 0.6949870878632831, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 49278 + }, + { + "epoch": 0.49279, + "grad_norm": 0.741756613931111, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 49279 + }, + { + "epoch": 0.4928, + "grad_norm": 0.7878823445863342, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 49280 + }, + { + "epoch": 0.49281, + "grad_norm": 0.7853107987648804, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 49281 + }, + { + "epoch": 0.49282, + "grad_norm": 0.7329635609038596, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 49282 + }, + { + "epoch": 0.49283, + "grad_norm": 0.7969858494368876, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 49283 + }, + { + "epoch": 0.49284, + "grad_norm": 0.8835976334970546, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 49284 + }, + { + "epoch": 0.49285, + "grad_norm": 0.8151775311616382, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 49285 + }, + { + "epoch": 0.49286, + "grad_norm": 0.887669666658952, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 49286 + }, + { + "epoch": 0.49287, + "grad_norm": 1.074875442452299, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 49287 + }, + { + "epoch": 0.49288, + "grad_norm": 1.1145910138971122, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 49288 + }, + { + "epoch": 0.49289, + "grad_norm": 0.8062007619757225, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 49289 + }, + { + "epoch": 0.4929, + "grad_norm": 0.5799933654227456, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 49290 + }, + { + "epoch": 0.49291, + "grad_norm": 0.6518373662123804, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 49291 + }, + { + "epoch": 0.49292, + "grad_norm": 0.6650837539026114, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 49292 + }, + { + "epoch": 0.49293, + "grad_norm": 0.6808615654481771, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 49293 + }, + { + "epoch": 0.49294, + "grad_norm": 0.6493859715797584, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 49294 + }, + { + "epoch": 0.49295, + "grad_norm": 0.6135712639667402, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 49295 + }, + { + "epoch": 0.49296, + "grad_norm": 0.654003583991863, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 49296 + }, + { + "epoch": 0.49297, + "grad_norm": 0.7706449354990046, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 49297 + }, + { + "epoch": 0.49298, + "grad_norm": 0.8933125001826648, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 49298 + }, + { + "epoch": 0.49299, + "grad_norm": 0.8550224134792251, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 49299 + }, + { + "epoch": 0.493, + "grad_norm": 0.6853454819492235, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 49300 + }, + { + "epoch": 0.49301, + "grad_norm": 0.5374835150502285, + "learning_rate": 0.003, + "loss": 3.9439, + "step": 49301 + }, + { + "epoch": 0.49302, + "grad_norm": 0.5960142388671943, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 49302 + }, + { + "epoch": 0.49303, + "grad_norm": 0.7652916651961119, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 49303 + }, + { + "epoch": 0.49304, + "grad_norm": 0.8084097023006671, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 49304 + }, + { + "epoch": 0.49305, + "grad_norm": 0.7995414032182617, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 49305 + }, + { + "epoch": 0.49306, + "grad_norm": 0.7799451142482356, + "learning_rate": 0.003, + "loss": 3.9418, + "step": 49306 + }, + { + "epoch": 0.49307, + "grad_norm": 0.8077104344362124, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 49307 + }, + { + "epoch": 0.49308, + "grad_norm": 0.9141878690529374, + "learning_rate": 0.003, + "loss": 3.98, + "step": 49308 + }, + { + "epoch": 0.49309, + "grad_norm": 0.8825463423608271, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 49309 + }, + { + "epoch": 0.4931, + "grad_norm": 0.676468688931922, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 49310 + }, + { + "epoch": 0.49311, + "grad_norm": 0.6661662261955409, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 49311 + }, + { + "epoch": 0.49312, + "grad_norm": 0.657892389807028, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 49312 + }, + { + "epoch": 0.49313, + "grad_norm": 0.5745239335476394, + "learning_rate": 0.003, + "loss": 3.982, + "step": 49313 + }, + { + "epoch": 0.49314, + "grad_norm": 0.6313258918674627, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 49314 + }, + { + "epoch": 0.49315, + "grad_norm": 0.6299645505984526, + "learning_rate": 0.003, + "loss": 3.9607, + "step": 49315 + }, + { + "epoch": 0.49316, + "grad_norm": 0.6857689539302303, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 49316 + }, + { + "epoch": 0.49317, + "grad_norm": 0.7704216849384538, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 49317 + }, + { + "epoch": 0.49318, + "grad_norm": 0.8661391181980593, + "learning_rate": 0.003, + "loss": 3.9397, + "step": 49318 + }, + { + "epoch": 0.49319, + "grad_norm": 0.9517679634995729, + "learning_rate": 0.003, + "loss": 4.0133, + "step": 49319 + }, + { + "epoch": 0.4932, + "grad_norm": 0.8177242023298336, + "learning_rate": 0.003, + "loss": 3.9465, + "step": 49320 + }, + { + "epoch": 0.49321, + "grad_norm": 0.5990618104245289, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 49321 + }, + { + "epoch": 0.49322, + "grad_norm": 0.6212148367299216, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 49322 + }, + { + "epoch": 0.49323, + "grad_norm": 0.8451478894042415, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 49323 + }, + { + "epoch": 0.49324, + "grad_norm": 1.024454343002592, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 49324 + }, + { + "epoch": 0.49325, + "grad_norm": 1.015317976703537, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 49325 + }, + { + "epoch": 0.49326, + "grad_norm": 0.9634723546622751, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 49326 + }, + { + "epoch": 0.49327, + "grad_norm": 0.8522733419953021, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 49327 + }, + { + "epoch": 0.49328, + "grad_norm": 0.809189878605239, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 49328 + }, + { + "epoch": 0.49329, + "grad_norm": 0.8290924145846252, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 49329 + }, + { + "epoch": 0.4933, + "grad_norm": 0.7765879358767613, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 49330 + }, + { + "epoch": 0.49331, + "grad_norm": 0.7533786335322825, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 49331 + }, + { + "epoch": 0.49332, + "grad_norm": 0.8070914962492509, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 49332 + }, + { + "epoch": 0.49333, + "grad_norm": 0.8462241758614696, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 49333 + }, + { + "epoch": 0.49334, + "grad_norm": 0.9331594679912439, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 49334 + }, + { + "epoch": 0.49335, + "grad_norm": 0.951349663313021, + "learning_rate": 0.003, + "loss": 4.008, + "step": 49335 + }, + { + "epoch": 0.49336, + "grad_norm": 0.9548503232877471, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 49336 + }, + { + "epoch": 0.49337, + "grad_norm": 0.9871213475840231, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 49337 + }, + { + "epoch": 0.49338, + "grad_norm": 0.9348765315889618, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 49338 + }, + { + "epoch": 0.49339, + "grad_norm": 0.8411957883282629, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 49339 + }, + { + "epoch": 0.4934, + "grad_norm": 0.989616720183131, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 49340 + }, + { + "epoch": 0.49341, + "grad_norm": 0.9671400878655312, + "learning_rate": 0.003, + "loss": 3.9446, + "step": 49341 + }, + { + "epoch": 0.49342, + "grad_norm": 0.986452340558316, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 49342 + }, + { + "epoch": 0.49343, + "grad_norm": 0.9740637442225497, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 49343 + }, + { + "epoch": 0.49344, + "grad_norm": 0.905770632832477, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 49344 + }, + { + "epoch": 0.49345, + "grad_norm": 0.6402897504175454, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 49345 + }, + { + "epoch": 0.49346, + "grad_norm": 0.6371609001332761, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 49346 + }, + { + "epoch": 0.49347, + "grad_norm": 0.7349237892992952, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 49347 + }, + { + "epoch": 0.49348, + "grad_norm": 0.7635218483518208, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 49348 + }, + { + "epoch": 0.49349, + "grad_norm": 0.7456387180326479, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 49349 + }, + { + "epoch": 0.4935, + "grad_norm": 0.7156191055369576, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 49350 + }, + { + "epoch": 0.49351, + "grad_norm": 0.8415042127570259, + "learning_rate": 0.003, + "loss": 3.987, + "step": 49351 + }, + { + "epoch": 0.49352, + "grad_norm": 0.908901691837681, + "learning_rate": 0.003, + "loss": 4.025, + "step": 49352 + }, + { + "epoch": 0.49353, + "grad_norm": 0.8601637115484952, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 49353 + }, + { + "epoch": 0.49354, + "grad_norm": 0.7596760655575436, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 49354 + }, + { + "epoch": 0.49355, + "grad_norm": 0.6543037618221742, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 49355 + }, + { + "epoch": 0.49356, + "grad_norm": 0.7695504184624925, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 49356 + }, + { + "epoch": 0.49357, + "grad_norm": 0.7591021889471691, + "learning_rate": 0.003, + "loss": 3.9425, + "step": 49357 + }, + { + "epoch": 0.49358, + "grad_norm": 0.8128845554946786, + "learning_rate": 0.003, + "loss": 3.96, + "step": 49358 + }, + { + "epoch": 0.49359, + "grad_norm": 0.8216073970385023, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 49359 + }, + { + "epoch": 0.4936, + "grad_norm": 0.7685173338516318, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 49360 + }, + { + "epoch": 0.49361, + "grad_norm": 0.670815110076781, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 49361 + }, + { + "epoch": 0.49362, + "grad_norm": 0.5957713061422248, + "learning_rate": 0.003, + "loss": 3.954, + "step": 49362 + }, + { + "epoch": 0.49363, + "grad_norm": 0.5868437331272962, + "learning_rate": 0.003, + "loss": 3.9191, + "step": 49363 + }, + { + "epoch": 0.49364, + "grad_norm": 0.5355678707673984, + "learning_rate": 0.003, + "loss": 3.9435, + "step": 49364 + }, + { + "epoch": 0.49365, + "grad_norm": 0.539264883577111, + "learning_rate": 0.003, + "loss": 3.9473, + "step": 49365 + }, + { + "epoch": 0.49366, + "grad_norm": 0.5209479880386987, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 49366 + }, + { + "epoch": 0.49367, + "grad_norm": 0.5530749930034619, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 49367 + }, + { + "epoch": 0.49368, + "grad_norm": 0.5509495327149329, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 49368 + }, + { + "epoch": 0.49369, + "grad_norm": 0.610588930293441, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 49369 + }, + { + "epoch": 0.4937, + "grad_norm": 0.6654480700984209, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 49370 + }, + { + "epoch": 0.49371, + "grad_norm": 0.6806468001964094, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 49371 + }, + { + "epoch": 0.49372, + "grad_norm": 0.7542487164896373, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 49372 + }, + { + "epoch": 0.49373, + "grad_norm": 0.8265881274062526, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 49373 + }, + { + "epoch": 0.49374, + "grad_norm": 0.8839739980953626, + "learning_rate": 0.003, + "loss": 3.966, + "step": 49374 + }, + { + "epoch": 0.49375, + "grad_norm": 0.8811882869812382, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 49375 + }, + { + "epoch": 0.49376, + "grad_norm": 0.7689171892234028, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 49376 + }, + { + "epoch": 0.49377, + "grad_norm": 0.713439193163746, + "learning_rate": 0.003, + "loss": 3.9388, + "step": 49377 + }, + { + "epoch": 0.49378, + "grad_norm": 0.6808843557426009, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 49378 + }, + { + "epoch": 0.49379, + "grad_norm": 0.7128087147928874, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 49379 + }, + { + "epoch": 0.4938, + "grad_norm": 0.6774963843488027, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 49380 + }, + { + "epoch": 0.49381, + "grad_norm": 0.6962131163278369, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 49381 + }, + { + "epoch": 0.49382, + "grad_norm": 0.8316159776416948, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 49382 + }, + { + "epoch": 0.49383, + "grad_norm": 1.1198588800258957, + "learning_rate": 0.003, + "loss": 3.961, + "step": 49383 + }, + { + "epoch": 0.49384, + "grad_norm": 0.962366066576059, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 49384 + }, + { + "epoch": 0.49385, + "grad_norm": 0.8652197363210444, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 49385 + }, + { + "epoch": 0.49386, + "grad_norm": 0.814135613390192, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 49386 + }, + { + "epoch": 0.49387, + "grad_norm": 0.8813847685408179, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 49387 + }, + { + "epoch": 0.49388, + "grad_norm": 0.8397318199407351, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 49388 + }, + { + "epoch": 0.49389, + "grad_norm": 0.9150637835007093, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 49389 + }, + { + "epoch": 0.4939, + "grad_norm": 0.9169362646404139, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 49390 + }, + { + "epoch": 0.49391, + "grad_norm": 0.9167063265276643, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 49391 + }, + { + "epoch": 0.49392, + "grad_norm": 0.9270300069478292, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 49392 + }, + { + "epoch": 0.49393, + "grad_norm": 0.9457263123009546, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 49393 + }, + { + "epoch": 0.49394, + "grad_norm": 0.8742014878748038, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 49394 + }, + { + "epoch": 0.49395, + "grad_norm": 0.8862689424847718, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 49395 + }, + { + "epoch": 0.49396, + "grad_norm": 0.8354381853557241, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 49396 + }, + { + "epoch": 0.49397, + "grad_norm": 0.8817735788065698, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 49397 + }, + { + "epoch": 0.49398, + "grad_norm": 0.9859406904615218, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 49398 + }, + { + "epoch": 0.49399, + "grad_norm": 1.0279549683313076, + "learning_rate": 0.003, + "loss": 4.0216, + "step": 49399 + }, + { + "epoch": 0.494, + "grad_norm": 0.9405221714758633, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 49400 + }, + { + "epoch": 0.49401, + "grad_norm": 0.8777306587844351, + "learning_rate": 0.003, + "loss": 4.007, + "step": 49401 + }, + { + "epoch": 0.49402, + "grad_norm": 0.7843244923018877, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 49402 + }, + { + "epoch": 0.49403, + "grad_norm": 0.8264655609235113, + "learning_rate": 0.003, + "loss": 3.997, + "step": 49403 + }, + { + "epoch": 0.49404, + "grad_norm": 0.8433615786541999, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 49404 + }, + { + "epoch": 0.49405, + "grad_norm": 0.8373116574731869, + "learning_rate": 0.003, + "loss": 4.0261, + "step": 49405 + }, + { + "epoch": 0.49406, + "grad_norm": 0.7031456727797187, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 49406 + }, + { + "epoch": 0.49407, + "grad_norm": 0.6417219194875282, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 49407 + }, + { + "epoch": 0.49408, + "grad_norm": 0.6184771390402541, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 49408 + }, + { + "epoch": 0.49409, + "grad_norm": 0.5293391396800802, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 49409 + }, + { + "epoch": 0.4941, + "grad_norm": 0.6463569949608421, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 49410 + }, + { + "epoch": 0.49411, + "grad_norm": 0.7338598870693644, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 49411 + }, + { + "epoch": 0.49412, + "grad_norm": 0.8587878254126061, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 49412 + }, + { + "epoch": 0.49413, + "grad_norm": 0.9613732010457907, + "learning_rate": 0.003, + "loss": 4.0212, + "step": 49413 + }, + { + "epoch": 0.49414, + "grad_norm": 1.0434478377824332, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 49414 + }, + { + "epoch": 0.49415, + "grad_norm": 0.8219710422785773, + "learning_rate": 0.003, + "loss": 3.956, + "step": 49415 + }, + { + "epoch": 0.49416, + "grad_norm": 0.8402417238964305, + "learning_rate": 0.003, + "loss": 3.99, + "step": 49416 + }, + { + "epoch": 0.49417, + "grad_norm": 0.8282254932053115, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 49417 + }, + { + "epoch": 0.49418, + "grad_norm": 0.6427263981651945, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 49418 + }, + { + "epoch": 0.49419, + "grad_norm": 0.5397876690512382, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 49419 + }, + { + "epoch": 0.4942, + "grad_norm": 0.5594236001592496, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 49420 + }, + { + "epoch": 0.49421, + "grad_norm": 0.5969310116505722, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 49421 + }, + { + "epoch": 0.49422, + "grad_norm": 0.6643207137372109, + "learning_rate": 0.003, + "loss": 3.9469, + "step": 49422 + }, + { + "epoch": 0.49423, + "grad_norm": 0.5879156758290294, + "learning_rate": 0.003, + "loss": 3.9453, + "step": 49423 + }, + { + "epoch": 0.49424, + "grad_norm": 0.6138398011183243, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 49424 + }, + { + "epoch": 0.49425, + "grad_norm": 0.5804440462309468, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 49425 + }, + { + "epoch": 0.49426, + "grad_norm": 0.5653006708362458, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 49426 + }, + { + "epoch": 0.49427, + "grad_norm": 0.5464751710954325, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 49427 + }, + { + "epoch": 0.49428, + "grad_norm": 0.6038159094278198, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 49428 + }, + { + "epoch": 0.49429, + "grad_norm": 0.6977711262683878, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 49429 + }, + { + "epoch": 0.4943, + "grad_norm": 0.6482708088305228, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 49430 + }, + { + "epoch": 0.49431, + "grad_norm": 0.6640843094859942, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 49431 + }, + { + "epoch": 0.49432, + "grad_norm": 0.7811937187204765, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 49432 + }, + { + "epoch": 0.49433, + "grad_norm": 0.8350196350692578, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 49433 + }, + { + "epoch": 0.49434, + "grad_norm": 0.7991637323227105, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 49434 + }, + { + "epoch": 0.49435, + "grad_norm": 0.6612267310307649, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 49435 + }, + { + "epoch": 0.49436, + "grad_norm": 0.6799573234402135, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 49436 + }, + { + "epoch": 0.49437, + "grad_norm": 0.7304224337440356, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 49437 + }, + { + "epoch": 0.49438, + "grad_norm": 0.8858690535460798, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 49438 + }, + { + "epoch": 0.49439, + "grad_norm": 0.9528739396027336, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 49439 + }, + { + "epoch": 0.4944, + "grad_norm": 1.0162427707130777, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 49440 + }, + { + "epoch": 0.49441, + "grad_norm": 0.8736324798729432, + "learning_rate": 0.003, + "loss": 3.953, + "step": 49441 + }, + { + "epoch": 0.49442, + "grad_norm": 0.8367675527360888, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 49442 + }, + { + "epoch": 0.49443, + "grad_norm": 0.8114751045845453, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 49443 + }, + { + "epoch": 0.49444, + "grad_norm": 0.8508424616945492, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 49444 + }, + { + "epoch": 0.49445, + "grad_norm": 0.8815594007173712, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 49445 + }, + { + "epoch": 0.49446, + "grad_norm": 0.7420236702015982, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 49446 + }, + { + "epoch": 0.49447, + "grad_norm": 0.7124038435469745, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 49447 + }, + { + "epoch": 0.49448, + "grad_norm": 0.7535450135630615, + "learning_rate": 0.003, + "loss": 4.0, + "step": 49448 + }, + { + "epoch": 0.49449, + "grad_norm": 0.7535349833785606, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 49449 + }, + { + "epoch": 0.4945, + "grad_norm": 0.7477293844223811, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 49450 + }, + { + "epoch": 0.49451, + "grad_norm": 0.8102536324311003, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 49451 + }, + { + "epoch": 0.49452, + "grad_norm": 0.8893239794753606, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 49452 + }, + { + "epoch": 0.49453, + "grad_norm": 0.8789154528590545, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 49453 + }, + { + "epoch": 0.49454, + "grad_norm": 0.8353661605053125, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 49454 + }, + { + "epoch": 0.49455, + "grad_norm": 0.839869891961449, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 49455 + }, + { + "epoch": 0.49456, + "grad_norm": 0.8119263710801724, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 49456 + }, + { + "epoch": 0.49457, + "grad_norm": 0.8885730105587211, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 49457 + }, + { + "epoch": 0.49458, + "grad_norm": 0.8887295488554883, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 49458 + }, + { + "epoch": 0.49459, + "grad_norm": 0.8976613136915144, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 49459 + }, + { + "epoch": 0.4946, + "grad_norm": 0.8790105036793241, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 49460 + }, + { + "epoch": 0.49461, + "grad_norm": 0.8474574315391105, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 49461 + }, + { + "epoch": 0.49462, + "grad_norm": 0.8726081795304641, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 49462 + }, + { + "epoch": 0.49463, + "grad_norm": 0.9023789375176511, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 49463 + }, + { + "epoch": 0.49464, + "grad_norm": 0.874852734941904, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 49464 + }, + { + "epoch": 0.49465, + "grad_norm": 0.8412493820584579, + "learning_rate": 0.003, + "loss": 3.948, + "step": 49465 + }, + { + "epoch": 0.49466, + "grad_norm": 0.8812700539479137, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 49466 + }, + { + "epoch": 0.49467, + "grad_norm": 1.0000214648187011, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 49467 + }, + { + "epoch": 0.49468, + "grad_norm": 0.9288661485013555, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 49468 + }, + { + "epoch": 0.49469, + "grad_norm": 0.8307405779763323, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 49469 + }, + { + "epoch": 0.4947, + "grad_norm": 0.7303140958331359, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 49470 + }, + { + "epoch": 0.49471, + "grad_norm": 0.8302295759921696, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 49471 + }, + { + "epoch": 0.49472, + "grad_norm": 0.9605211459740579, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 49472 + }, + { + "epoch": 0.49473, + "grad_norm": 0.9908382525231825, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 49473 + }, + { + "epoch": 0.49474, + "grad_norm": 0.9626534686627919, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 49474 + }, + { + "epoch": 0.49475, + "grad_norm": 0.9386471347737798, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 49475 + }, + { + "epoch": 0.49476, + "grad_norm": 0.8277984802356569, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 49476 + }, + { + "epoch": 0.49477, + "grad_norm": 0.7782074804596796, + "learning_rate": 0.003, + "loss": 4.0234, + "step": 49477 + }, + { + "epoch": 0.49478, + "grad_norm": 0.6946654218046399, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 49478 + }, + { + "epoch": 0.49479, + "grad_norm": 0.667595154273093, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 49479 + }, + { + "epoch": 0.4948, + "grad_norm": 0.6020814183006546, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 49480 + }, + { + "epoch": 0.49481, + "grad_norm": 0.5586721758774011, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 49481 + }, + { + "epoch": 0.49482, + "grad_norm": 0.5526237347533052, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 49482 + }, + { + "epoch": 0.49483, + "grad_norm": 0.6447898744260446, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 49483 + }, + { + "epoch": 0.49484, + "grad_norm": 0.5614128471103444, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 49484 + }, + { + "epoch": 0.49485, + "grad_norm": 0.5295892184095373, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 49485 + }, + { + "epoch": 0.49486, + "grad_norm": 0.586419116756585, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 49486 + }, + { + "epoch": 0.49487, + "grad_norm": 0.5880237677092974, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 49487 + }, + { + "epoch": 0.49488, + "grad_norm": 0.6770072863655413, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 49488 + }, + { + "epoch": 0.49489, + "grad_norm": 0.6561801423988618, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 49489 + }, + { + "epoch": 0.4949, + "grad_norm": 0.6069377962363025, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 49490 + }, + { + "epoch": 0.49491, + "grad_norm": 0.524040399285762, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 49491 + }, + { + "epoch": 0.49492, + "grad_norm": 0.546033976850662, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 49492 + }, + { + "epoch": 0.49493, + "grad_norm": 0.5000130093787639, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 49493 + }, + { + "epoch": 0.49494, + "grad_norm": 0.5805564636169492, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 49494 + }, + { + "epoch": 0.49495, + "grad_norm": 0.7285973395732307, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 49495 + }, + { + "epoch": 0.49496, + "grad_norm": 0.9695540066368087, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 49496 + }, + { + "epoch": 0.49497, + "grad_norm": 1.1387406802192372, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 49497 + }, + { + "epoch": 0.49498, + "grad_norm": 0.9078172056591093, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 49498 + }, + { + "epoch": 0.49499, + "grad_norm": 0.853519944570276, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 49499 + }, + { + "epoch": 0.495, + "grad_norm": 0.770146257198604, + "learning_rate": 0.003, + "loss": 3.96, + "step": 49500 + }, + { + "epoch": 0.49501, + "grad_norm": 0.7175135334724452, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 49501 + }, + { + "epoch": 0.49502, + "grad_norm": 0.7945252711142887, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 49502 + }, + { + "epoch": 0.49503, + "grad_norm": 0.9256329875700876, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 49503 + }, + { + "epoch": 0.49504, + "grad_norm": 1.0408572836808647, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 49504 + }, + { + "epoch": 0.49505, + "grad_norm": 0.9975067258357748, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 49505 + }, + { + "epoch": 0.49506, + "grad_norm": 0.9114830752676368, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 49506 + }, + { + "epoch": 0.49507, + "grad_norm": 0.8958023720062628, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 49507 + }, + { + "epoch": 0.49508, + "grad_norm": 0.8493026683334223, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 49508 + }, + { + "epoch": 0.49509, + "grad_norm": 0.8832916582098912, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 49509 + }, + { + "epoch": 0.4951, + "grad_norm": 0.8508557418242054, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 49510 + }, + { + "epoch": 0.49511, + "grad_norm": 0.9290525704981739, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 49511 + }, + { + "epoch": 0.49512, + "grad_norm": 0.9442944363442288, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 49512 + }, + { + "epoch": 0.49513, + "grad_norm": 0.9910343866522486, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 49513 + }, + { + "epoch": 0.49514, + "grad_norm": 0.9652872905742624, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 49514 + }, + { + "epoch": 0.49515, + "grad_norm": 0.8347875031705617, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 49515 + }, + { + "epoch": 0.49516, + "grad_norm": 0.8380051085223426, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 49516 + }, + { + "epoch": 0.49517, + "grad_norm": 0.8113746188125383, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 49517 + }, + { + "epoch": 0.49518, + "grad_norm": 0.7867425542621145, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 49518 + }, + { + "epoch": 0.49519, + "grad_norm": 0.707255170024279, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 49519 + }, + { + "epoch": 0.4952, + "grad_norm": 0.6066289619995519, + "learning_rate": 0.003, + "loss": 3.97, + "step": 49520 + }, + { + "epoch": 0.49521, + "grad_norm": 0.5768125572347718, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 49521 + }, + { + "epoch": 0.49522, + "grad_norm": 0.5325945249383711, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 49522 + }, + { + "epoch": 0.49523, + "grad_norm": 0.5858987282493866, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 49523 + }, + { + "epoch": 0.49524, + "grad_norm": 0.6058655780365625, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 49524 + }, + { + "epoch": 0.49525, + "grad_norm": 0.6383039590125005, + "learning_rate": 0.003, + "loss": 3.982, + "step": 49525 + }, + { + "epoch": 0.49526, + "grad_norm": 0.6930118863146774, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 49526 + }, + { + "epoch": 0.49527, + "grad_norm": 0.6687044616608953, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 49527 + }, + { + "epoch": 0.49528, + "grad_norm": 0.6107598343570927, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 49528 + }, + { + "epoch": 0.49529, + "grad_norm": 0.548107224712041, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 49529 + }, + { + "epoch": 0.4953, + "grad_norm": 0.5118481100458886, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 49530 + }, + { + "epoch": 0.49531, + "grad_norm": 0.5586609505838268, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 49531 + }, + { + "epoch": 0.49532, + "grad_norm": 0.5778466502493917, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 49532 + }, + { + "epoch": 0.49533, + "grad_norm": 0.6166647677987576, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 49533 + }, + { + "epoch": 0.49534, + "grad_norm": 0.7400894320045662, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 49534 + }, + { + "epoch": 0.49535, + "grad_norm": 0.7183598869613942, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 49535 + }, + { + "epoch": 0.49536, + "grad_norm": 0.7048408079788051, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 49536 + }, + { + "epoch": 0.49537, + "grad_norm": 0.7698179686512056, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 49537 + }, + { + "epoch": 0.49538, + "grad_norm": 0.798436727345943, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 49538 + }, + { + "epoch": 0.49539, + "grad_norm": 0.9390675301441284, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 49539 + }, + { + "epoch": 0.4954, + "grad_norm": 1.0193342523400732, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 49540 + }, + { + "epoch": 0.49541, + "grad_norm": 1.11293880951261, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 49541 + }, + { + "epoch": 0.49542, + "grad_norm": 0.9739353130175442, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 49542 + }, + { + "epoch": 0.49543, + "grad_norm": 1.1345566012378758, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 49543 + }, + { + "epoch": 0.49544, + "grad_norm": 0.9467264587804266, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 49544 + }, + { + "epoch": 0.49545, + "grad_norm": 0.7892180759200425, + "learning_rate": 0.003, + "loss": 4.0286, + "step": 49545 + }, + { + "epoch": 0.49546, + "grad_norm": 0.711377554758997, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 49546 + }, + { + "epoch": 0.49547, + "grad_norm": 0.68258915356544, + "learning_rate": 0.003, + "loss": 3.9275, + "step": 49547 + }, + { + "epoch": 0.49548, + "grad_norm": 0.68947864560552, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 49548 + }, + { + "epoch": 0.49549, + "grad_norm": 0.7282821653592894, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 49549 + }, + { + "epoch": 0.4955, + "grad_norm": 0.6948833107311281, + "learning_rate": 0.003, + "loss": 3.974, + "step": 49550 + }, + { + "epoch": 0.49551, + "grad_norm": 0.7182088735420595, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 49551 + }, + { + "epoch": 0.49552, + "grad_norm": 0.6664370206075428, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 49552 + }, + { + "epoch": 0.49553, + "grad_norm": 0.6669480706884081, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 49553 + }, + { + "epoch": 0.49554, + "grad_norm": 0.6028722512871368, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 49554 + }, + { + "epoch": 0.49555, + "grad_norm": 0.6095021784483738, + "learning_rate": 0.003, + "loss": 3.9485, + "step": 49555 + }, + { + "epoch": 0.49556, + "grad_norm": 0.691063473926345, + "learning_rate": 0.003, + "loss": 3.993, + "step": 49556 + }, + { + "epoch": 0.49557, + "grad_norm": 0.6781923557816503, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 49557 + }, + { + "epoch": 0.49558, + "grad_norm": 0.6839325364201583, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 49558 + }, + { + "epoch": 0.49559, + "grad_norm": 0.5855324645378105, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 49559 + }, + { + "epoch": 0.4956, + "grad_norm": 0.5788201637316748, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 49560 + }, + { + "epoch": 0.49561, + "grad_norm": 0.5708468121337815, + "learning_rate": 0.003, + "loss": 3.9443, + "step": 49561 + }, + { + "epoch": 0.49562, + "grad_norm": 0.5986432906702048, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 49562 + }, + { + "epoch": 0.49563, + "grad_norm": 0.612456170441652, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 49563 + }, + { + "epoch": 0.49564, + "grad_norm": 0.6442359413154974, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 49564 + }, + { + "epoch": 0.49565, + "grad_norm": 0.7524764185437225, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 49565 + }, + { + "epoch": 0.49566, + "grad_norm": 0.9397852179425186, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 49566 + }, + { + "epoch": 0.49567, + "grad_norm": 1.2203160404161555, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 49567 + }, + { + "epoch": 0.49568, + "grad_norm": 0.952554929588728, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 49568 + }, + { + "epoch": 0.49569, + "grad_norm": 0.8189630428910367, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 49569 + }, + { + "epoch": 0.4957, + "grad_norm": 0.7924770486895654, + "learning_rate": 0.003, + "loss": 3.983, + "step": 49570 + }, + { + "epoch": 0.49571, + "grad_norm": 0.9634202260767221, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 49571 + }, + { + "epoch": 0.49572, + "grad_norm": 1.1238159163945025, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 49572 + }, + { + "epoch": 0.49573, + "grad_norm": 1.030081179830143, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 49573 + }, + { + "epoch": 0.49574, + "grad_norm": 1.0041347990132536, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 49574 + }, + { + "epoch": 0.49575, + "grad_norm": 0.8714353219700011, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 49575 + }, + { + "epoch": 0.49576, + "grad_norm": 0.78730365778666, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 49576 + }, + { + "epoch": 0.49577, + "grad_norm": 0.7702058628460072, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 49577 + }, + { + "epoch": 0.49578, + "grad_norm": 0.859924127212274, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 49578 + }, + { + "epoch": 0.49579, + "grad_norm": 0.9045076740399575, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 49579 + }, + { + "epoch": 0.4958, + "grad_norm": 0.8769275240219789, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 49580 + }, + { + "epoch": 0.49581, + "grad_norm": 0.8375379309819301, + "learning_rate": 0.003, + "loss": 4.0266, + "step": 49581 + }, + { + "epoch": 0.49582, + "grad_norm": 0.816906496566108, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 49582 + }, + { + "epoch": 0.49583, + "grad_norm": 0.721312905797969, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 49583 + }, + { + "epoch": 0.49584, + "grad_norm": 0.6969989408730501, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 49584 + }, + { + "epoch": 0.49585, + "grad_norm": 0.6164044975015633, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 49585 + }, + { + "epoch": 0.49586, + "grad_norm": 0.748899516629871, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 49586 + }, + { + "epoch": 0.49587, + "grad_norm": 0.7759470216792691, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 49587 + }, + { + "epoch": 0.49588, + "grad_norm": 0.6967687988895317, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 49588 + }, + { + "epoch": 0.49589, + "grad_norm": 0.6631393683223828, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 49589 + }, + { + "epoch": 0.4959, + "grad_norm": 0.5830993751049905, + "learning_rate": 0.003, + "loss": 3.972, + "step": 49590 + }, + { + "epoch": 0.49591, + "grad_norm": 0.6440553661537303, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 49591 + }, + { + "epoch": 0.49592, + "grad_norm": 0.7470880496747418, + "learning_rate": 0.003, + "loss": 3.955, + "step": 49592 + }, + { + "epoch": 0.49593, + "grad_norm": 0.7918944234277493, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 49593 + }, + { + "epoch": 0.49594, + "grad_norm": 0.8484985080162353, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 49594 + }, + { + "epoch": 0.49595, + "grad_norm": 0.8938625438508812, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 49595 + }, + { + "epoch": 0.49596, + "grad_norm": 0.8631804970870132, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 49596 + }, + { + "epoch": 0.49597, + "grad_norm": 0.8727113712468334, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 49597 + }, + { + "epoch": 0.49598, + "grad_norm": 0.7808493799952184, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 49598 + }, + { + "epoch": 0.49599, + "grad_norm": 0.7225013940516872, + "learning_rate": 0.003, + "loss": 4.0064, + "step": 49599 + }, + { + "epoch": 0.496, + "grad_norm": 0.6419366928221911, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 49600 + }, + { + "epoch": 0.49601, + "grad_norm": 0.7024782907478995, + "learning_rate": 0.003, + "loss": 3.9448, + "step": 49601 + }, + { + "epoch": 0.49602, + "grad_norm": 0.6708146743999996, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 49602 + }, + { + "epoch": 0.49603, + "grad_norm": 0.5837860455741417, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 49603 + }, + { + "epoch": 0.49604, + "grad_norm": 0.49708020459468594, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 49604 + }, + { + "epoch": 0.49605, + "grad_norm": 0.4776741709613452, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 49605 + }, + { + "epoch": 0.49606, + "grad_norm": 0.46605167961245936, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 49606 + }, + { + "epoch": 0.49607, + "grad_norm": 0.4997803409998639, + "learning_rate": 0.003, + "loss": 3.969, + "step": 49607 + }, + { + "epoch": 0.49608, + "grad_norm": 0.580826188135438, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 49608 + }, + { + "epoch": 0.49609, + "grad_norm": 0.5572974897800762, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 49609 + }, + { + "epoch": 0.4961, + "grad_norm": 0.5510577971648294, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 49610 + }, + { + "epoch": 0.49611, + "grad_norm": 0.6612809229913811, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 49611 + }, + { + "epoch": 0.49612, + "grad_norm": 0.8258891356515339, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 49612 + }, + { + "epoch": 0.49613, + "grad_norm": 0.9019634888709118, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 49613 + }, + { + "epoch": 0.49614, + "grad_norm": 0.7938711962960782, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 49614 + }, + { + "epoch": 0.49615, + "grad_norm": 0.9298162234990056, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 49615 + }, + { + "epoch": 0.49616, + "grad_norm": 1.140529251336779, + "learning_rate": 0.003, + "loss": 3.993, + "step": 49616 + }, + { + "epoch": 0.49617, + "grad_norm": 1.080483714218885, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 49617 + }, + { + "epoch": 0.49618, + "grad_norm": 0.9300713974331617, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 49618 + }, + { + "epoch": 0.49619, + "grad_norm": 0.8127185547415484, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 49619 + }, + { + "epoch": 0.4962, + "grad_norm": 0.8581804203278424, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 49620 + }, + { + "epoch": 0.49621, + "grad_norm": 0.8251128847942067, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 49621 + }, + { + "epoch": 0.49622, + "grad_norm": 0.8843841391859282, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 49622 + }, + { + "epoch": 0.49623, + "grad_norm": 1.0000862238274313, + "learning_rate": 0.003, + "loss": 3.968, + "step": 49623 + }, + { + "epoch": 0.49624, + "grad_norm": 1.182594373250697, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 49624 + }, + { + "epoch": 0.49625, + "grad_norm": 0.9861644043163824, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 49625 + }, + { + "epoch": 0.49626, + "grad_norm": 1.0017977587190683, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 49626 + }, + { + "epoch": 0.49627, + "grad_norm": 0.9122929131963413, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 49627 + }, + { + "epoch": 0.49628, + "grad_norm": 0.9332267866447119, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 49628 + }, + { + "epoch": 0.49629, + "grad_norm": 0.9206953920052376, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 49629 + }, + { + "epoch": 0.4963, + "grad_norm": 0.9733295312510596, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 49630 + }, + { + "epoch": 0.49631, + "grad_norm": 1.0382799357427162, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 49631 + }, + { + "epoch": 0.49632, + "grad_norm": 0.8675662130551575, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 49632 + }, + { + "epoch": 0.49633, + "grad_norm": 0.8224709505609964, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 49633 + }, + { + "epoch": 0.49634, + "grad_norm": 0.7938715875864566, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 49634 + }, + { + "epoch": 0.49635, + "grad_norm": 0.9041685192974198, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 49635 + }, + { + "epoch": 0.49636, + "grad_norm": 0.9766211542015267, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 49636 + }, + { + "epoch": 0.49637, + "grad_norm": 1.0219369718293447, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 49637 + }, + { + "epoch": 0.49638, + "grad_norm": 1.0163677348279836, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 49638 + }, + { + "epoch": 0.49639, + "grad_norm": 0.9069439436995073, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 49639 + }, + { + "epoch": 0.4964, + "grad_norm": 0.7428113955214026, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 49640 + }, + { + "epoch": 0.49641, + "grad_norm": 0.7949779642351421, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 49641 + }, + { + "epoch": 0.49642, + "grad_norm": 0.9363989093150611, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 49642 + }, + { + "epoch": 0.49643, + "grad_norm": 0.9374985497109218, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 49643 + }, + { + "epoch": 0.49644, + "grad_norm": 0.8399324750852571, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 49644 + }, + { + "epoch": 0.49645, + "grad_norm": 0.814443678293366, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 49645 + }, + { + "epoch": 0.49646, + "grad_norm": 0.7254648085839241, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 49646 + }, + { + "epoch": 0.49647, + "grad_norm": 0.8006692544889148, + "learning_rate": 0.003, + "loss": 4.002, + "step": 49647 + }, + { + "epoch": 0.49648, + "grad_norm": 0.9195686689034719, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 49648 + }, + { + "epoch": 0.49649, + "grad_norm": 0.8198788380055538, + "learning_rate": 0.003, + "loss": 3.9528, + "step": 49649 + }, + { + "epoch": 0.4965, + "grad_norm": 0.738803258494361, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 49650 + }, + { + "epoch": 0.49651, + "grad_norm": 0.8344675423399721, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 49651 + }, + { + "epoch": 0.49652, + "grad_norm": 0.7590382467979333, + "learning_rate": 0.003, + "loss": 4.0128, + "step": 49652 + }, + { + "epoch": 0.49653, + "grad_norm": 0.7058574524098702, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 49653 + }, + { + "epoch": 0.49654, + "grad_norm": 0.5604641709695302, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 49654 + }, + { + "epoch": 0.49655, + "grad_norm": 0.5368858788230882, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 49655 + }, + { + "epoch": 0.49656, + "grad_norm": 0.49718517754279024, + "learning_rate": 0.003, + "loss": 3.957, + "step": 49656 + }, + { + "epoch": 0.49657, + "grad_norm": 0.5590114230917856, + "learning_rate": 0.003, + "loss": 3.954, + "step": 49657 + }, + { + "epoch": 0.49658, + "grad_norm": 0.5270587162934712, + "learning_rate": 0.003, + "loss": 3.967, + "step": 49658 + }, + { + "epoch": 0.49659, + "grad_norm": 0.4851626345548323, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 49659 + }, + { + "epoch": 0.4966, + "grad_norm": 0.4606639536433146, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 49660 + }, + { + "epoch": 0.49661, + "grad_norm": 0.4689163970443496, + "learning_rate": 0.003, + "loss": 3.974, + "step": 49661 + }, + { + "epoch": 0.49662, + "grad_norm": 0.46520628625196014, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 49662 + }, + { + "epoch": 0.49663, + "grad_norm": 0.5655437072515885, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 49663 + }, + { + "epoch": 0.49664, + "grad_norm": 0.5891161764268947, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 49664 + }, + { + "epoch": 0.49665, + "grad_norm": 0.6237300082821797, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 49665 + }, + { + "epoch": 0.49666, + "grad_norm": 0.7424448208724882, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 49666 + }, + { + "epoch": 0.49667, + "grad_norm": 1.023635813349304, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 49667 + }, + { + "epoch": 0.49668, + "grad_norm": 1.0280135969716453, + "learning_rate": 0.003, + "loss": 4.034, + "step": 49668 + }, + { + "epoch": 0.49669, + "grad_norm": 0.9118870839328108, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 49669 + }, + { + "epoch": 0.4967, + "grad_norm": 0.9504390896195554, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 49670 + }, + { + "epoch": 0.49671, + "grad_norm": 0.852995045410684, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 49671 + }, + { + "epoch": 0.49672, + "grad_norm": 0.7973159158142501, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 49672 + }, + { + "epoch": 0.49673, + "grad_norm": 0.7915169535693841, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 49673 + }, + { + "epoch": 0.49674, + "grad_norm": 0.817081446657622, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 49674 + }, + { + "epoch": 0.49675, + "grad_norm": 0.8026804010811812, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 49675 + }, + { + "epoch": 0.49676, + "grad_norm": 0.805451021160997, + "learning_rate": 0.003, + "loss": 4.0183, + "step": 49676 + }, + { + "epoch": 0.49677, + "grad_norm": 0.895174347879041, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 49677 + }, + { + "epoch": 0.49678, + "grad_norm": 0.9590943097897638, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 49678 + }, + { + "epoch": 0.49679, + "grad_norm": 0.9286671951978648, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 49679 + }, + { + "epoch": 0.4968, + "grad_norm": 0.7557853045140861, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 49680 + }, + { + "epoch": 0.49681, + "grad_norm": 0.7675178264915329, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 49681 + }, + { + "epoch": 0.49682, + "grad_norm": 0.8382735834797228, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 49682 + }, + { + "epoch": 0.49683, + "grad_norm": 0.8311371035405766, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 49683 + }, + { + "epoch": 0.49684, + "grad_norm": 0.874949294201715, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 49684 + }, + { + "epoch": 0.49685, + "grad_norm": 1.1152143230588198, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 49685 + }, + { + "epoch": 0.49686, + "grad_norm": 0.9065781702117431, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 49686 + }, + { + "epoch": 0.49687, + "grad_norm": 0.7592859383398469, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 49687 + }, + { + "epoch": 0.49688, + "grad_norm": 0.7319441307863342, + "learning_rate": 0.003, + "loss": 3.979, + "step": 49688 + }, + { + "epoch": 0.49689, + "grad_norm": 0.6409571226650541, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 49689 + }, + { + "epoch": 0.4969, + "grad_norm": 0.6425368817386826, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 49690 + }, + { + "epoch": 0.49691, + "grad_norm": 0.7835719373723524, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 49691 + }, + { + "epoch": 0.49692, + "grad_norm": 1.1206407214773781, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 49692 + }, + { + "epoch": 0.49693, + "grad_norm": 1.069855821564123, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 49693 + }, + { + "epoch": 0.49694, + "grad_norm": 0.6923176860558391, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 49694 + }, + { + "epoch": 0.49695, + "grad_norm": 0.5929596595121505, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 49695 + }, + { + "epoch": 0.49696, + "grad_norm": 0.6918311011095225, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 49696 + }, + { + "epoch": 0.49697, + "grad_norm": 0.8403495889108062, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 49697 + }, + { + "epoch": 0.49698, + "grad_norm": 0.8477097410436616, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 49698 + }, + { + "epoch": 0.49699, + "grad_norm": 0.720652475317375, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 49699 + }, + { + "epoch": 0.497, + "grad_norm": 0.6807185874505144, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 49700 + }, + { + "epoch": 0.49701, + "grad_norm": 0.7084778928798477, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 49701 + }, + { + "epoch": 0.49702, + "grad_norm": 0.6289925910521034, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 49702 + }, + { + "epoch": 0.49703, + "grad_norm": 0.6215168473201689, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 49703 + }, + { + "epoch": 0.49704, + "grad_norm": 0.6955741830905207, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 49704 + }, + { + "epoch": 0.49705, + "grad_norm": 0.7717835189884021, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 49705 + }, + { + "epoch": 0.49706, + "grad_norm": 0.7699851015554777, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 49706 + }, + { + "epoch": 0.49707, + "grad_norm": 0.7353960122757804, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 49707 + }, + { + "epoch": 0.49708, + "grad_norm": 0.7084296904174554, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 49708 + }, + { + "epoch": 0.49709, + "grad_norm": 0.7953001592193684, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 49709 + }, + { + "epoch": 0.4971, + "grad_norm": 0.8349952722560587, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 49710 + }, + { + "epoch": 0.49711, + "grad_norm": 0.8846099003700705, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 49711 + }, + { + "epoch": 0.49712, + "grad_norm": 0.7843181438682526, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 49712 + }, + { + "epoch": 0.49713, + "grad_norm": 0.6408150888778922, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 49713 + }, + { + "epoch": 0.49714, + "grad_norm": 0.6218369156185862, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 49714 + }, + { + "epoch": 0.49715, + "grad_norm": 0.6093911040594852, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 49715 + }, + { + "epoch": 0.49716, + "grad_norm": 0.5761709404260201, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 49716 + }, + { + "epoch": 0.49717, + "grad_norm": 0.5772898566029677, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 49717 + }, + { + "epoch": 0.49718, + "grad_norm": 0.6521356535934686, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 49718 + }, + { + "epoch": 0.49719, + "grad_norm": 0.7346007327743863, + "learning_rate": 0.003, + "loss": 3.964, + "step": 49719 + }, + { + "epoch": 0.4972, + "grad_norm": 0.7967954974320288, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 49720 + }, + { + "epoch": 0.49721, + "grad_norm": 0.8090437484127855, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 49721 + }, + { + "epoch": 0.49722, + "grad_norm": 0.8192427030824597, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 49722 + }, + { + "epoch": 0.49723, + "grad_norm": 0.8420255532408417, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 49723 + }, + { + "epoch": 0.49724, + "grad_norm": 0.8394722819381448, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 49724 + }, + { + "epoch": 0.49725, + "grad_norm": 0.8934572312955387, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 49725 + }, + { + "epoch": 0.49726, + "grad_norm": 0.9720627805027274, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 49726 + }, + { + "epoch": 0.49727, + "grad_norm": 0.9263115070936913, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 49727 + }, + { + "epoch": 0.49728, + "grad_norm": 0.8644591593241953, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 49728 + }, + { + "epoch": 0.49729, + "grad_norm": 0.8098026106112083, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 49729 + }, + { + "epoch": 0.4973, + "grad_norm": 0.8306532467052464, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 49730 + }, + { + "epoch": 0.49731, + "grad_norm": 0.7522244003602983, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 49731 + }, + { + "epoch": 0.49732, + "grad_norm": 0.6982942191743396, + "learning_rate": 0.003, + "loss": 3.993, + "step": 49732 + }, + { + "epoch": 0.49733, + "grad_norm": 0.6356545271717962, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 49733 + }, + { + "epoch": 0.49734, + "grad_norm": 0.7392803145653742, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 49734 + }, + { + "epoch": 0.49735, + "grad_norm": 0.7632750198746514, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 49735 + }, + { + "epoch": 0.49736, + "grad_norm": 0.8437045121199976, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 49736 + }, + { + "epoch": 0.49737, + "grad_norm": 0.974240814442135, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 49737 + }, + { + "epoch": 0.49738, + "grad_norm": 0.8802721334030386, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 49738 + }, + { + "epoch": 0.49739, + "grad_norm": 0.6533339891647137, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 49739 + }, + { + "epoch": 0.4974, + "grad_norm": 0.6449363710453923, + "learning_rate": 0.003, + "loss": 4.0073, + "step": 49740 + }, + { + "epoch": 0.49741, + "grad_norm": 0.7495154867774979, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 49741 + }, + { + "epoch": 0.49742, + "grad_norm": 0.8232380256722777, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 49742 + }, + { + "epoch": 0.49743, + "grad_norm": 0.7821235281861143, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 49743 + }, + { + "epoch": 0.49744, + "grad_norm": 0.7267979983043955, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 49744 + }, + { + "epoch": 0.49745, + "grad_norm": 0.7534280569615173, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 49745 + }, + { + "epoch": 0.49746, + "grad_norm": 0.8503419447629148, + "learning_rate": 0.003, + "loss": 3.9358, + "step": 49746 + }, + { + "epoch": 0.49747, + "grad_norm": 0.8441760391972423, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 49747 + }, + { + "epoch": 0.49748, + "grad_norm": 0.8135693847558689, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 49748 + }, + { + "epoch": 0.49749, + "grad_norm": 0.7011588032642992, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 49749 + }, + { + "epoch": 0.4975, + "grad_norm": 0.7911267893600978, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 49750 + }, + { + "epoch": 0.49751, + "grad_norm": 0.8341697004735285, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 49751 + }, + { + "epoch": 0.49752, + "grad_norm": 0.9461535921004595, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 49752 + }, + { + "epoch": 0.49753, + "grad_norm": 0.9498459219268236, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 49753 + }, + { + "epoch": 0.49754, + "grad_norm": 0.8450417713939382, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 49754 + }, + { + "epoch": 0.49755, + "grad_norm": 0.8677374437629368, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 49755 + }, + { + "epoch": 0.49756, + "grad_norm": 0.8927192632759725, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 49756 + }, + { + "epoch": 0.49757, + "grad_norm": 0.882387722625093, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 49757 + }, + { + "epoch": 0.49758, + "grad_norm": 0.9548527281618278, + "learning_rate": 0.003, + "loss": 4.0118, + "step": 49758 + }, + { + "epoch": 0.49759, + "grad_norm": 0.9027214731792388, + "learning_rate": 0.003, + "loss": 4.0115, + "step": 49759 + }, + { + "epoch": 0.4976, + "grad_norm": 0.8771744227938625, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 49760 + }, + { + "epoch": 0.49761, + "grad_norm": 0.8915836428520071, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 49761 + }, + { + "epoch": 0.49762, + "grad_norm": 0.9061963800978958, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 49762 + }, + { + "epoch": 0.49763, + "grad_norm": 0.8361549557741861, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 49763 + }, + { + "epoch": 0.49764, + "grad_norm": 0.7709848257609813, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 49764 + }, + { + "epoch": 0.49765, + "grad_norm": 0.6891260973286243, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 49765 + }, + { + "epoch": 0.49766, + "grad_norm": 0.6635884623733256, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 49766 + }, + { + "epoch": 0.49767, + "grad_norm": 0.9133796744874286, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 49767 + }, + { + "epoch": 0.49768, + "grad_norm": 1.1305506473262081, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 49768 + }, + { + "epoch": 0.49769, + "grad_norm": 0.9067376089945249, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 49769 + }, + { + "epoch": 0.4977, + "grad_norm": 0.7291045076488345, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 49770 + }, + { + "epoch": 0.49771, + "grad_norm": 0.7306598305261789, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 49771 + }, + { + "epoch": 0.49772, + "grad_norm": 0.7723462051210587, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 49772 + }, + { + "epoch": 0.49773, + "grad_norm": 0.7669289382320749, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 49773 + }, + { + "epoch": 0.49774, + "grad_norm": 0.8157187071595955, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 49774 + }, + { + "epoch": 0.49775, + "grad_norm": 0.8197172770791391, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 49775 + }, + { + "epoch": 0.49776, + "grad_norm": 0.8491150913274504, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 49776 + }, + { + "epoch": 0.49777, + "grad_norm": 0.7158001950911602, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 49777 + }, + { + "epoch": 0.49778, + "grad_norm": 0.6075064639717401, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 49778 + }, + { + "epoch": 0.49779, + "grad_norm": 0.5983934750839459, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 49779 + }, + { + "epoch": 0.4978, + "grad_norm": 0.5734449080842662, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 49780 + }, + { + "epoch": 0.49781, + "grad_norm": 0.6300511053674182, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 49781 + }, + { + "epoch": 0.49782, + "grad_norm": 0.5641248691235421, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 49782 + }, + { + "epoch": 0.49783, + "grad_norm": 0.5598067868177119, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 49783 + }, + { + "epoch": 0.49784, + "grad_norm": 0.6144028261298341, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 49784 + }, + { + "epoch": 0.49785, + "grad_norm": 0.6612385020868823, + "learning_rate": 0.003, + "loss": 3.98, + "step": 49785 + }, + { + "epoch": 0.49786, + "grad_norm": 0.7390944210876044, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 49786 + }, + { + "epoch": 0.49787, + "grad_norm": 0.7769034454620081, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 49787 + }, + { + "epoch": 0.49788, + "grad_norm": 0.8077631926569008, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 49788 + }, + { + "epoch": 0.49789, + "grad_norm": 0.9325348894577256, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 49789 + }, + { + "epoch": 0.4979, + "grad_norm": 0.9491147271194863, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 49790 + }, + { + "epoch": 0.49791, + "grad_norm": 0.9369980241894218, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 49791 + }, + { + "epoch": 0.49792, + "grad_norm": 0.816343662653138, + "learning_rate": 0.003, + "loss": 4.0, + "step": 49792 + }, + { + "epoch": 0.49793, + "grad_norm": 0.7092679003904044, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 49793 + }, + { + "epoch": 0.49794, + "grad_norm": 0.7685042727021352, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 49794 + }, + { + "epoch": 0.49795, + "grad_norm": 0.8014751927148402, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 49795 + }, + { + "epoch": 0.49796, + "grad_norm": 0.8254856656601625, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 49796 + }, + { + "epoch": 0.49797, + "grad_norm": 0.8297880130767322, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 49797 + }, + { + "epoch": 0.49798, + "grad_norm": 0.80011699064511, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 49798 + }, + { + "epoch": 0.49799, + "grad_norm": 0.6626215710988599, + "learning_rate": 0.003, + "loss": 3.9461, + "step": 49799 + }, + { + "epoch": 0.498, + "grad_norm": 0.627426099518282, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 49800 + }, + { + "epoch": 0.49801, + "grad_norm": 0.5793426685899582, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 49801 + }, + { + "epoch": 0.49802, + "grad_norm": 0.639996906387327, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 49802 + }, + { + "epoch": 0.49803, + "grad_norm": 0.6202380893722416, + "learning_rate": 0.003, + "loss": 3.994, + "step": 49803 + }, + { + "epoch": 0.49804, + "grad_norm": 0.6973307891204217, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 49804 + }, + { + "epoch": 0.49805, + "grad_norm": 0.7958065580837583, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 49805 + }, + { + "epoch": 0.49806, + "grad_norm": 0.8268886907498684, + "learning_rate": 0.003, + "loss": 3.971, + "step": 49806 + }, + { + "epoch": 0.49807, + "grad_norm": 0.8707818972152325, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 49807 + }, + { + "epoch": 0.49808, + "grad_norm": 0.8509335725487682, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 49808 + }, + { + "epoch": 0.49809, + "grad_norm": 0.7969630543757457, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 49809 + }, + { + "epoch": 0.4981, + "grad_norm": 0.6615434310412908, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 49810 + }, + { + "epoch": 0.49811, + "grad_norm": 0.6935381498828652, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 49811 + }, + { + "epoch": 0.49812, + "grad_norm": 0.6218613747484243, + "learning_rate": 0.003, + "loss": 4.001, + "step": 49812 + }, + { + "epoch": 0.49813, + "grad_norm": 0.622471955306338, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 49813 + }, + { + "epoch": 0.49814, + "grad_norm": 0.6477317709134839, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 49814 + }, + { + "epoch": 0.49815, + "grad_norm": 0.7412191648927806, + "learning_rate": 0.003, + "loss": 3.96, + "step": 49815 + }, + { + "epoch": 0.49816, + "grad_norm": 0.8253086996818706, + "learning_rate": 0.003, + "loss": 3.9467, + "step": 49816 + }, + { + "epoch": 0.49817, + "grad_norm": 0.8545141439767869, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 49817 + }, + { + "epoch": 0.49818, + "grad_norm": 0.884492608831497, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 49818 + }, + { + "epoch": 0.49819, + "grad_norm": 0.8881417183497422, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 49819 + }, + { + "epoch": 0.4982, + "grad_norm": 0.8676351194540617, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 49820 + }, + { + "epoch": 0.49821, + "grad_norm": 0.8106383752747823, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 49821 + }, + { + "epoch": 0.49822, + "grad_norm": 0.8443074184016869, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 49822 + }, + { + "epoch": 0.49823, + "grad_norm": 0.9179528015712145, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 49823 + }, + { + "epoch": 0.49824, + "grad_norm": 0.874193660151208, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 49824 + }, + { + "epoch": 0.49825, + "grad_norm": 0.9490214984470112, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 49825 + }, + { + "epoch": 0.49826, + "grad_norm": 0.8627366329866714, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 49826 + }, + { + "epoch": 0.49827, + "grad_norm": 0.8182872426751439, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 49827 + }, + { + "epoch": 0.49828, + "grad_norm": 0.8593042303094697, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 49828 + }, + { + "epoch": 0.49829, + "grad_norm": 0.8888510768849309, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 49829 + }, + { + "epoch": 0.4983, + "grad_norm": 0.9873350394628063, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 49830 + }, + { + "epoch": 0.49831, + "grad_norm": 1.09442016899906, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 49831 + }, + { + "epoch": 0.49832, + "grad_norm": 0.8392591615334294, + "learning_rate": 0.003, + "loss": 3.985, + "step": 49832 + }, + { + "epoch": 0.49833, + "grad_norm": 0.8074679119253135, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 49833 + }, + { + "epoch": 0.49834, + "grad_norm": 0.7688635708788848, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 49834 + }, + { + "epoch": 0.49835, + "grad_norm": 0.7789563116422172, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 49835 + }, + { + "epoch": 0.49836, + "grad_norm": 0.8300778723804616, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 49836 + }, + { + "epoch": 0.49837, + "grad_norm": 0.7964101313452838, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 49837 + }, + { + "epoch": 0.49838, + "grad_norm": 0.7672214006193558, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 49838 + }, + { + "epoch": 0.49839, + "grad_norm": 0.7359769616780458, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 49839 + }, + { + "epoch": 0.4984, + "grad_norm": 0.8551541724401925, + "learning_rate": 0.003, + "loss": 4.0165, + "step": 49840 + }, + { + "epoch": 0.49841, + "grad_norm": 1.0866661252402172, + "learning_rate": 0.003, + "loss": 4.0322, + "step": 49841 + }, + { + "epoch": 0.49842, + "grad_norm": 0.9122089749234255, + "learning_rate": 0.003, + "loss": 3.9468, + "step": 49842 + }, + { + "epoch": 0.49843, + "grad_norm": 0.7866316084353415, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 49843 + }, + { + "epoch": 0.49844, + "grad_norm": 0.6530161897349552, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 49844 + }, + { + "epoch": 0.49845, + "grad_norm": 0.5856388675618488, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 49845 + }, + { + "epoch": 0.49846, + "grad_norm": 0.6886543813841367, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 49846 + }, + { + "epoch": 0.49847, + "grad_norm": 0.7789236813704921, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 49847 + }, + { + "epoch": 0.49848, + "grad_norm": 0.7530650379034798, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 49848 + }, + { + "epoch": 0.49849, + "grad_norm": 0.6896702810258373, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 49849 + }, + { + "epoch": 0.4985, + "grad_norm": 0.6187985791486607, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 49850 + }, + { + "epoch": 0.49851, + "grad_norm": 0.6543677949593675, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 49851 + }, + { + "epoch": 0.49852, + "grad_norm": 0.6454240742671522, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 49852 + }, + { + "epoch": 0.49853, + "grad_norm": 0.6612915084134412, + "learning_rate": 0.003, + "loss": 3.987, + "step": 49853 + }, + { + "epoch": 0.49854, + "grad_norm": 0.657359222518959, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 49854 + }, + { + "epoch": 0.49855, + "grad_norm": 0.5981159960380538, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 49855 + }, + { + "epoch": 0.49856, + "grad_norm": 0.6504637189370607, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 49856 + }, + { + "epoch": 0.49857, + "grad_norm": 0.7062383894819451, + "learning_rate": 0.003, + "loss": 3.9562, + "step": 49857 + }, + { + "epoch": 0.49858, + "grad_norm": 0.8380085192522475, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 49858 + }, + { + "epoch": 0.49859, + "grad_norm": 0.8866045404082129, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 49859 + }, + { + "epoch": 0.4986, + "grad_norm": 0.9346517121451338, + "learning_rate": 0.003, + "loss": 3.9343, + "step": 49860 + }, + { + "epoch": 0.49861, + "grad_norm": 0.8960737204166199, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 49861 + }, + { + "epoch": 0.49862, + "grad_norm": 0.7518421100882993, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 49862 + }, + { + "epoch": 0.49863, + "grad_norm": 0.7435925677741162, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 49863 + }, + { + "epoch": 0.49864, + "grad_norm": 0.7176908358554607, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 49864 + }, + { + "epoch": 0.49865, + "grad_norm": 0.6398900849102507, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 49865 + }, + { + "epoch": 0.49866, + "grad_norm": 0.6772973377687452, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 49866 + }, + { + "epoch": 0.49867, + "grad_norm": 0.6941448926718289, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 49867 + }, + { + "epoch": 0.49868, + "grad_norm": 0.6075337834873683, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 49868 + }, + { + "epoch": 0.49869, + "grad_norm": 0.5671721160041109, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 49869 + }, + { + "epoch": 0.4987, + "grad_norm": 0.5848470672557975, + "learning_rate": 0.003, + "loss": 3.9405, + "step": 49870 + }, + { + "epoch": 0.49871, + "grad_norm": 0.6741843922925107, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 49871 + }, + { + "epoch": 0.49872, + "grad_norm": 0.8165279381883472, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 49872 + }, + { + "epoch": 0.49873, + "grad_norm": 1.0550698734038202, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 49873 + }, + { + "epoch": 0.49874, + "grad_norm": 1.1219903430695746, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 49874 + }, + { + "epoch": 0.49875, + "grad_norm": 0.7724360283894098, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 49875 + }, + { + "epoch": 0.49876, + "grad_norm": 0.7676747789948108, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 49876 + }, + { + "epoch": 0.49877, + "grad_norm": 0.905582678491774, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 49877 + }, + { + "epoch": 0.49878, + "grad_norm": 0.9642110160163405, + "learning_rate": 0.003, + "loss": 3.991, + "step": 49878 + }, + { + "epoch": 0.49879, + "grad_norm": 0.9392471701775621, + "learning_rate": 0.003, + "loss": 3.981, + "step": 49879 + }, + { + "epoch": 0.4988, + "grad_norm": 0.84259640392833, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 49880 + }, + { + "epoch": 0.49881, + "grad_norm": 0.86553505291734, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 49881 + }, + { + "epoch": 0.49882, + "grad_norm": 0.8103621717960222, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 49882 + }, + { + "epoch": 0.49883, + "grad_norm": 0.8180190869742306, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 49883 + }, + { + "epoch": 0.49884, + "grad_norm": 0.9013741017792858, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 49884 + }, + { + "epoch": 0.49885, + "grad_norm": 0.8770230967384066, + "learning_rate": 0.003, + "loss": 4.0173, + "step": 49885 + }, + { + "epoch": 0.49886, + "grad_norm": 0.8232829592046075, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 49886 + }, + { + "epoch": 0.49887, + "grad_norm": 0.828836195638666, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 49887 + }, + { + "epoch": 0.49888, + "grad_norm": 0.8686063520176457, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 49888 + }, + { + "epoch": 0.49889, + "grad_norm": 0.8738091332206627, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 49889 + }, + { + "epoch": 0.4989, + "grad_norm": 0.8480959416702423, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 49890 + }, + { + "epoch": 0.49891, + "grad_norm": 0.8244734503250004, + "learning_rate": 0.003, + "loss": 4.009, + "step": 49891 + }, + { + "epoch": 0.49892, + "grad_norm": 0.7547247638434464, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 49892 + }, + { + "epoch": 0.49893, + "grad_norm": 0.7574441508392986, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 49893 + }, + { + "epoch": 0.49894, + "grad_norm": 0.7981566615413127, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 49894 + }, + { + "epoch": 0.49895, + "grad_norm": 0.8328379738260797, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 49895 + }, + { + "epoch": 0.49896, + "grad_norm": 0.910064657244823, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 49896 + }, + { + "epoch": 0.49897, + "grad_norm": 0.9760491257053778, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 49897 + }, + { + "epoch": 0.49898, + "grad_norm": 1.0534958513917818, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 49898 + }, + { + "epoch": 0.49899, + "grad_norm": 0.9854811451884695, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 49899 + }, + { + "epoch": 0.499, + "grad_norm": 0.914823604359221, + "learning_rate": 0.003, + "loss": 3.983, + "step": 49900 + }, + { + "epoch": 0.49901, + "grad_norm": 0.7762036948549806, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 49901 + }, + { + "epoch": 0.49902, + "grad_norm": 0.7467703927935835, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 49902 + }, + { + "epoch": 0.49903, + "grad_norm": 0.6649281589379097, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 49903 + }, + { + "epoch": 0.49904, + "grad_norm": 0.5985067407942664, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 49904 + }, + { + "epoch": 0.49905, + "grad_norm": 0.5919592993522507, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 49905 + }, + { + "epoch": 0.49906, + "grad_norm": 0.6359077132761185, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 49906 + }, + { + "epoch": 0.49907, + "grad_norm": 0.6424397174615705, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 49907 + }, + { + "epoch": 0.49908, + "grad_norm": 0.6684374388661961, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 49908 + }, + { + "epoch": 0.49909, + "grad_norm": 0.6142589327204966, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 49909 + }, + { + "epoch": 0.4991, + "grad_norm": 0.5823648479761391, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 49910 + }, + { + "epoch": 0.49911, + "grad_norm": 0.5487885847789007, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 49911 + }, + { + "epoch": 0.49912, + "grad_norm": 0.5547072214093682, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 49912 + }, + { + "epoch": 0.49913, + "grad_norm": 0.5503151531044984, + "learning_rate": 0.003, + "loss": 3.972, + "step": 49913 + }, + { + "epoch": 0.49914, + "grad_norm": 0.589625400753216, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 49914 + }, + { + "epoch": 0.49915, + "grad_norm": 0.5813757050747529, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 49915 + }, + { + "epoch": 0.49916, + "grad_norm": 0.6492489994519236, + "learning_rate": 0.003, + "loss": 3.9429, + "step": 49916 + }, + { + "epoch": 0.49917, + "grad_norm": 0.7904933367874442, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 49917 + }, + { + "epoch": 0.49918, + "grad_norm": 0.9604702036297529, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 49918 + }, + { + "epoch": 0.49919, + "grad_norm": 0.9073271253994987, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 49919 + }, + { + "epoch": 0.4992, + "grad_norm": 0.6948944476011426, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 49920 + }, + { + "epoch": 0.49921, + "grad_norm": 0.6493310680506805, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 49921 + }, + { + "epoch": 0.49922, + "grad_norm": 0.7868632222257177, + "learning_rate": 0.003, + "loss": 3.9477, + "step": 49922 + }, + { + "epoch": 0.49923, + "grad_norm": 0.8012989681240746, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 49923 + }, + { + "epoch": 0.49924, + "grad_norm": 0.6595893192227167, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 49924 + }, + { + "epoch": 0.49925, + "grad_norm": 0.6845643770898152, + "learning_rate": 0.003, + "loss": 3.964, + "step": 49925 + }, + { + "epoch": 0.49926, + "grad_norm": 0.7121091147841472, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 49926 + }, + { + "epoch": 0.49927, + "grad_norm": 0.8723952739080258, + "learning_rate": 0.003, + "loss": 3.946, + "step": 49927 + }, + { + "epoch": 0.49928, + "grad_norm": 0.8730177072616204, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 49928 + }, + { + "epoch": 0.49929, + "grad_norm": 0.8465008118848668, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 49929 + }, + { + "epoch": 0.4993, + "grad_norm": 0.9242632200211918, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 49930 + }, + { + "epoch": 0.49931, + "grad_norm": 0.9709070484554367, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 49931 + }, + { + "epoch": 0.49932, + "grad_norm": 1.0653382995651453, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 49932 + }, + { + "epoch": 0.49933, + "grad_norm": 0.7979694991200601, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 49933 + }, + { + "epoch": 0.49934, + "grad_norm": 0.8675791678706019, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 49934 + }, + { + "epoch": 0.49935, + "grad_norm": 0.9068614303127607, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 49935 + }, + { + "epoch": 0.49936, + "grad_norm": 0.9961781160294638, + "learning_rate": 0.003, + "loss": 3.989, + "step": 49936 + }, + { + "epoch": 0.49937, + "grad_norm": 1.0398250781389748, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 49937 + }, + { + "epoch": 0.49938, + "grad_norm": 0.9875850885829927, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 49938 + }, + { + "epoch": 0.49939, + "grad_norm": 0.9650364749995083, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 49939 + }, + { + "epoch": 0.4994, + "grad_norm": 1.1299114307969913, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 49940 + }, + { + "epoch": 0.49941, + "grad_norm": 0.7903040713045267, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 49941 + }, + { + "epoch": 0.49942, + "grad_norm": 0.7391377793567262, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 49942 + }, + { + "epoch": 0.49943, + "grad_norm": 0.7211737352712176, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 49943 + }, + { + "epoch": 0.49944, + "grad_norm": 0.6994606240840269, + "learning_rate": 0.003, + "loss": 3.98, + "step": 49944 + }, + { + "epoch": 0.49945, + "grad_norm": 0.6883747245308321, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 49945 + }, + { + "epoch": 0.49946, + "grad_norm": 0.6866570495164847, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 49946 + }, + { + "epoch": 0.49947, + "grad_norm": 0.6593705181877773, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 49947 + }, + { + "epoch": 0.49948, + "grad_norm": 0.6211787322676561, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 49948 + }, + { + "epoch": 0.49949, + "grad_norm": 0.598460843806601, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 49949 + }, + { + "epoch": 0.4995, + "grad_norm": 0.5827542382584253, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 49950 + }, + { + "epoch": 0.49951, + "grad_norm": 0.6026253898812166, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 49951 + }, + { + "epoch": 0.49952, + "grad_norm": 0.5695061017820217, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 49952 + }, + { + "epoch": 0.49953, + "grad_norm": 0.5260514903839131, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 49953 + }, + { + "epoch": 0.49954, + "grad_norm": 0.5124155862796819, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 49954 + }, + { + "epoch": 0.49955, + "grad_norm": 0.5888064295081713, + "learning_rate": 0.003, + "loss": 3.977, + "step": 49955 + }, + { + "epoch": 0.49956, + "grad_norm": 0.5530814600859505, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 49956 + }, + { + "epoch": 0.49957, + "grad_norm": 0.5503038853298504, + "learning_rate": 0.003, + "loss": 4.015, + "step": 49957 + }, + { + "epoch": 0.49958, + "grad_norm": 0.5209734909606843, + "learning_rate": 0.003, + "loss": 3.947, + "step": 49958 + }, + { + "epoch": 0.49959, + "grad_norm": 0.5019382003240164, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 49959 + }, + { + "epoch": 0.4996, + "grad_norm": 0.5566218106204515, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 49960 + }, + { + "epoch": 0.49961, + "grad_norm": 0.673495577146729, + "learning_rate": 0.003, + "loss": 3.958, + "step": 49961 + }, + { + "epoch": 0.49962, + "grad_norm": 0.8941582863758557, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 49962 + }, + { + "epoch": 0.49963, + "grad_norm": 1.1403585839886388, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 49963 + }, + { + "epoch": 0.49964, + "grad_norm": 0.8098770111884769, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 49964 + }, + { + "epoch": 0.49965, + "grad_norm": 0.7582072098858024, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 49965 + }, + { + "epoch": 0.49966, + "grad_norm": 0.8653880983368736, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 49966 + }, + { + "epoch": 0.49967, + "grad_norm": 0.9232933649090198, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 49967 + }, + { + "epoch": 0.49968, + "grad_norm": 1.0745778298847124, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 49968 + }, + { + "epoch": 0.49969, + "grad_norm": 0.8769697658521598, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 49969 + }, + { + "epoch": 0.4997, + "grad_norm": 0.7770709879237108, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 49970 + }, + { + "epoch": 0.49971, + "grad_norm": 0.7470857236768341, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 49971 + }, + { + "epoch": 0.49972, + "grad_norm": 0.7012940707234185, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 49972 + }, + { + "epoch": 0.49973, + "grad_norm": 0.7130900727323732, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 49973 + }, + { + "epoch": 0.49974, + "grad_norm": 0.6328076482221544, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 49974 + }, + { + "epoch": 0.49975, + "grad_norm": 0.6435470527571744, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 49975 + }, + { + "epoch": 0.49976, + "grad_norm": 0.690402918855978, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 49976 + }, + { + "epoch": 0.49977, + "grad_norm": 0.7454050502875624, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 49977 + }, + { + "epoch": 0.49978, + "grad_norm": 0.7182720723772958, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 49978 + }, + { + "epoch": 0.49979, + "grad_norm": 0.682877597500722, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 49979 + }, + { + "epoch": 0.4998, + "grad_norm": 0.6675183028659004, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 49980 + }, + { + "epoch": 0.49981, + "grad_norm": 0.6528428146555094, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 49981 + }, + { + "epoch": 0.49982, + "grad_norm": 0.7059058001259407, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 49982 + }, + { + "epoch": 0.49983, + "grad_norm": 0.7302033908514081, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 49983 + }, + { + "epoch": 0.49984, + "grad_norm": 0.6901320677953087, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 49984 + }, + { + "epoch": 0.49985, + "grad_norm": 0.8767600358315821, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 49985 + }, + { + "epoch": 0.49986, + "grad_norm": 0.9191019961904864, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 49986 + }, + { + "epoch": 0.49987, + "grad_norm": 0.9737073514135297, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 49987 + }, + { + "epoch": 0.49988, + "grad_norm": 1.0049713111319283, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 49988 + }, + { + "epoch": 0.49989, + "grad_norm": 0.9561005702314749, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 49989 + }, + { + "epoch": 0.4999, + "grad_norm": 1.0675871407495083, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 49990 + }, + { + "epoch": 0.49991, + "grad_norm": 1.0059876627259503, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 49991 + }, + { + "epoch": 0.49992, + "grad_norm": 1.3002709726144466, + "learning_rate": 0.003, + "loss": 4.014, + "step": 49992 + }, + { + "epoch": 0.49993, + "grad_norm": 0.8488851893437694, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 49993 + }, + { + "epoch": 0.49994, + "grad_norm": 0.8352228292448842, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 49994 + }, + { + "epoch": 0.49995, + "grad_norm": 1.1126523946438318, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 49995 + }, + { + "epoch": 0.49996, + "grad_norm": 1.2085858611644218, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 49996 + }, + { + "epoch": 0.49997, + "grad_norm": 0.7154787221734853, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 49997 + }, + { + "epoch": 0.49998, + "grad_norm": 0.7314081985067743, + "learning_rate": 0.003, + "loss": 3.97, + "step": 49998 + }, + { + "epoch": 0.49999, + "grad_norm": 0.7897391841864382, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 49999 + }, + { + "epoch": 0.5, + "grad_norm": 0.7679855931029036, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 50000 + }, + { + "epoch": 0.50001, + "grad_norm": 0.7623586385596373, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 50001 + }, + { + "epoch": 0.50002, + "grad_norm": 0.685056503742935, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 50002 + }, + { + "epoch": 0.50003, + "grad_norm": 0.6631642142001203, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 50003 + }, + { + "epoch": 0.50004, + "grad_norm": 0.7199287092729216, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 50004 + }, + { + "epoch": 0.50005, + "grad_norm": 0.7544251863033756, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 50005 + }, + { + "epoch": 0.50006, + "grad_norm": 0.8303221886105115, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 50006 + }, + { + "epoch": 0.50007, + "grad_norm": 0.757164153377019, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 50007 + }, + { + "epoch": 0.50008, + "grad_norm": 0.6166425669298627, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 50008 + }, + { + "epoch": 0.50009, + "grad_norm": 0.6000320956691314, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 50009 + }, + { + "epoch": 0.5001, + "grad_norm": 0.5808571161215644, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 50010 + }, + { + "epoch": 0.50011, + "grad_norm": 0.5657757711078455, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 50011 + }, + { + "epoch": 0.50012, + "grad_norm": 0.47880084849840987, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 50012 + }, + { + "epoch": 0.50013, + "grad_norm": 0.5207433693423106, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 50013 + }, + { + "epoch": 0.50014, + "grad_norm": 0.4941209966352517, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 50014 + }, + { + "epoch": 0.50015, + "grad_norm": 0.5758749843922749, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 50015 + }, + { + "epoch": 0.50016, + "grad_norm": 0.6784361459139967, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 50016 + }, + { + "epoch": 0.50017, + "grad_norm": 0.8801614972430335, + "learning_rate": 0.003, + "loss": 3.9409, + "step": 50017 + }, + { + "epoch": 0.50018, + "grad_norm": 1.0264977273845082, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 50018 + }, + { + "epoch": 0.50019, + "grad_norm": 1.113286259091534, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 50019 + }, + { + "epoch": 0.5002, + "grad_norm": 0.7757889698489581, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 50020 + }, + { + "epoch": 0.50021, + "grad_norm": 0.7788278154197763, + "learning_rate": 0.003, + "loss": 4.006, + "step": 50021 + }, + { + "epoch": 0.50022, + "grad_norm": 0.7996752025903495, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 50022 + }, + { + "epoch": 0.50023, + "grad_norm": 0.8299329484564884, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 50023 + }, + { + "epoch": 0.50024, + "grad_norm": 0.7915683480472792, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 50024 + }, + { + "epoch": 0.50025, + "grad_norm": 0.7794690194297537, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 50025 + }, + { + "epoch": 0.50026, + "grad_norm": 0.7453840673863182, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 50026 + }, + { + "epoch": 0.50027, + "grad_norm": 0.6932744020064029, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 50027 + }, + { + "epoch": 0.50028, + "grad_norm": 0.6976479825259083, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 50028 + }, + { + "epoch": 0.50029, + "grad_norm": 0.6523309430637451, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 50029 + }, + { + "epoch": 0.5003, + "grad_norm": 0.6551854846648428, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 50030 + }, + { + "epoch": 0.50031, + "grad_norm": 0.7038319135193019, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 50031 + }, + { + "epoch": 0.50032, + "grad_norm": 0.8491423460781573, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 50032 + }, + { + "epoch": 0.50033, + "grad_norm": 1.0019336552860114, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 50033 + }, + { + "epoch": 0.50034, + "grad_norm": 1.041400427714827, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 50034 + }, + { + "epoch": 0.50035, + "grad_norm": 0.7919709281676712, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 50035 + }, + { + "epoch": 0.50036, + "grad_norm": 0.7935746234823287, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 50036 + }, + { + "epoch": 0.50037, + "grad_norm": 0.9695698499052156, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 50037 + }, + { + "epoch": 0.50038, + "grad_norm": 1.037497959571549, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 50038 + }, + { + "epoch": 0.50039, + "grad_norm": 1.0273982465292708, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 50039 + }, + { + "epoch": 0.5004, + "grad_norm": 0.9453492176842612, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 50040 + }, + { + "epoch": 0.50041, + "grad_norm": 0.9353569358165161, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 50041 + }, + { + "epoch": 0.50042, + "grad_norm": 1.0752337573431536, + "learning_rate": 0.003, + "loss": 4.021, + "step": 50042 + }, + { + "epoch": 0.50043, + "grad_norm": 0.9664098104171489, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 50043 + }, + { + "epoch": 0.50044, + "grad_norm": 0.9528331102480571, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 50044 + }, + { + "epoch": 0.50045, + "grad_norm": 0.9116935460345877, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 50045 + }, + { + "epoch": 0.50046, + "grad_norm": 0.9920555363356554, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 50046 + }, + { + "epoch": 0.50047, + "grad_norm": 0.8900228356394709, + "learning_rate": 0.003, + "loss": 4.0238, + "step": 50047 + }, + { + "epoch": 0.50048, + "grad_norm": 0.8378146316224742, + "learning_rate": 0.003, + "loss": 4.0226, + "step": 50048 + }, + { + "epoch": 0.50049, + "grad_norm": 0.8592370420711609, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 50049 + }, + { + "epoch": 0.5005, + "grad_norm": 0.8899950748898048, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 50050 + }, + { + "epoch": 0.50051, + "grad_norm": 0.6766289320937234, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 50051 + }, + { + "epoch": 0.50052, + "grad_norm": 0.7326076731112234, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 50052 + }, + { + "epoch": 0.50053, + "grad_norm": 0.7944020658431835, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 50053 + }, + { + "epoch": 0.50054, + "grad_norm": 0.9023989709587092, + "learning_rate": 0.003, + "loss": 4.0257, + "step": 50054 + }, + { + "epoch": 0.50055, + "grad_norm": 0.8596215239577251, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 50055 + }, + { + "epoch": 0.50056, + "grad_norm": 0.8214900108781201, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 50056 + }, + { + "epoch": 0.50057, + "grad_norm": 0.7583261755687385, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 50057 + }, + { + "epoch": 0.50058, + "grad_norm": 0.6467140528242501, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 50058 + }, + { + "epoch": 0.50059, + "grad_norm": 0.5270215651055874, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 50059 + }, + { + "epoch": 0.5006, + "grad_norm": 0.4627949225026276, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 50060 + }, + { + "epoch": 0.50061, + "grad_norm": 0.475658497756252, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 50061 + }, + { + "epoch": 0.50062, + "grad_norm": 0.5989370677936016, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 50062 + }, + { + "epoch": 0.50063, + "grad_norm": 0.7443011974100283, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 50063 + }, + { + "epoch": 0.50064, + "grad_norm": 0.8673743227924418, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 50064 + }, + { + "epoch": 0.50065, + "grad_norm": 0.7875807091132745, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 50065 + }, + { + "epoch": 0.50066, + "grad_norm": 0.6662456948988241, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 50066 + }, + { + "epoch": 0.50067, + "grad_norm": 0.6341639678787964, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 50067 + }, + { + "epoch": 0.50068, + "grad_norm": 0.5806615874685362, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 50068 + }, + { + "epoch": 0.50069, + "grad_norm": 0.62577787759511, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 50069 + }, + { + "epoch": 0.5007, + "grad_norm": 0.5939564277773214, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 50070 + }, + { + "epoch": 0.50071, + "grad_norm": 0.5234691109444304, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 50071 + }, + { + "epoch": 0.50072, + "grad_norm": 0.49870109830830256, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 50072 + }, + { + "epoch": 0.50073, + "grad_norm": 0.44920772257471414, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 50073 + }, + { + "epoch": 0.50074, + "grad_norm": 0.42475954216154477, + "learning_rate": 0.003, + "loss": 3.9444, + "step": 50074 + }, + { + "epoch": 0.50075, + "grad_norm": 0.5145668315144304, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 50075 + }, + { + "epoch": 0.50076, + "grad_norm": 0.6878075398495821, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 50076 + }, + { + "epoch": 0.50077, + "grad_norm": 0.8481010457303254, + "learning_rate": 0.003, + "loss": 3.9523, + "step": 50077 + }, + { + "epoch": 0.50078, + "grad_norm": 0.9992277133664456, + "learning_rate": 0.003, + "loss": 3.9597, + "step": 50078 + }, + { + "epoch": 0.50079, + "grad_norm": 0.9148290166110413, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 50079 + }, + { + "epoch": 0.5008, + "grad_norm": 0.8214015338588342, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 50080 + }, + { + "epoch": 0.50081, + "grad_norm": 0.8783767257650271, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 50081 + }, + { + "epoch": 0.50082, + "grad_norm": 0.9601021345390979, + "learning_rate": 0.003, + "loss": 4.021, + "step": 50082 + }, + { + "epoch": 0.50083, + "grad_norm": 0.9729710532677431, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 50083 + }, + { + "epoch": 0.50084, + "grad_norm": 0.7874198722038236, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 50084 + }, + { + "epoch": 0.50085, + "grad_norm": 0.8115677998943065, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 50085 + }, + { + "epoch": 0.50086, + "grad_norm": 0.7491507925219439, + "learning_rate": 0.003, + "loss": 3.976, + "step": 50086 + }, + { + "epoch": 0.50087, + "grad_norm": 0.8046200829310802, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 50087 + }, + { + "epoch": 0.50088, + "grad_norm": 0.8730372437789229, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 50088 + }, + { + "epoch": 0.50089, + "grad_norm": 0.8594830845393785, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 50089 + }, + { + "epoch": 0.5009, + "grad_norm": 0.8936829186432987, + "learning_rate": 0.003, + "loss": 3.983, + "step": 50090 + }, + { + "epoch": 0.50091, + "grad_norm": 0.8618852279494554, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 50091 + }, + { + "epoch": 0.50092, + "grad_norm": 0.9358940244709226, + "learning_rate": 0.003, + "loss": 4.0177, + "step": 50092 + }, + { + "epoch": 0.50093, + "grad_norm": 0.8850610758993028, + "learning_rate": 0.003, + "loss": 3.983, + "step": 50093 + }, + { + "epoch": 0.50094, + "grad_norm": 1.0051224366129836, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 50094 + }, + { + "epoch": 0.50095, + "grad_norm": 1.0034546793785353, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 50095 + }, + { + "epoch": 0.50096, + "grad_norm": 0.8595329998803205, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 50096 + }, + { + "epoch": 0.50097, + "grad_norm": 0.8531332131247069, + "learning_rate": 0.003, + "loss": 3.9874, + "step": 50097 + }, + { + "epoch": 0.50098, + "grad_norm": 0.9578013758433759, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 50098 + }, + { + "epoch": 0.50099, + "grad_norm": 1.0343851195327851, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 50099 + }, + { + "epoch": 0.501, + "grad_norm": 0.9277818675456981, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 50100 + }, + { + "epoch": 0.50101, + "grad_norm": 1.0045792395350361, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 50101 + }, + { + "epoch": 0.50102, + "grad_norm": 1.113075820175179, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 50102 + }, + { + "epoch": 0.50103, + "grad_norm": 0.8128769565969971, + "learning_rate": 0.003, + "loss": 3.9468, + "step": 50103 + }, + { + "epoch": 0.50104, + "grad_norm": 0.7000565844758414, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 50104 + }, + { + "epoch": 0.50105, + "grad_norm": 0.5861211004555327, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 50105 + }, + { + "epoch": 0.50106, + "grad_norm": 0.6781772277262332, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 50106 + }, + { + "epoch": 0.50107, + "grad_norm": 0.6877443968734038, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 50107 + }, + { + "epoch": 0.50108, + "grad_norm": 0.7674104746303929, + "learning_rate": 0.003, + "loss": 3.996, + "step": 50108 + }, + { + "epoch": 0.50109, + "grad_norm": 0.8971444561247843, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 50109 + }, + { + "epoch": 0.5011, + "grad_norm": 0.8658307194166206, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 50110 + }, + { + "epoch": 0.50111, + "grad_norm": 0.7638164855998485, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 50111 + }, + { + "epoch": 0.50112, + "grad_norm": 0.7061999367767525, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 50112 + }, + { + "epoch": 0.50113, + "grad_norm": 0.6294912891083863, + "learning_rate": 0.003, + "loss": 3.978, + "step": 50113 + }, + { + "epoch": 0.50114, + "grad_norm": 0.6298620263649258, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 50114 + }, + { + "epoch": 0.50115, + "grad_norm": 0.582095734116846, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 50115 + }, + { + "epoch": 0.50116, + "grad_norm": 0.6596923706282065, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 50116 + }, + { + "epoch": 0.50117, + "grad_norm": 0.5854443633258631, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 50117 + }, + { + "epoch": 0.50118, + "grad_norm": 0.6263397056742536, + "learning_rate": 0.003, + "loss": 3.957, + "step": 50118 + }, + { + "epoch": 0.50119, + "grad_norm": 0.6812407012852755, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 50119 + }, + { + "epoch": 0.5012, + "grad_norm": 0.7078451766856853, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 50120 + }, + { + "epoch": 0.50121, + "grad_norm": 0.7429332946594496, + "learning_rate": 0.003, + "loss": 4.0, + "step": 50121 + }, + { + "epoch": 0.50122, + "grad_norm": 0.6794772054225123, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 50122 + }, + { + "epoch": 0.50123, + "grad_norm": 0.5863675293327353, + "learning_rate": 0.003, + "loss": 3.966, + "step": 50123 + }, + { + "epoch": 0.50124, + "grad_norm": 0.6109406266339168, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 50124 + }, + { + "epoch": 0.50125, + "grad_norm": 0.5373279007280773, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 50125 + }, + { + "epoch": 0.50126, + "grad_norm": 0.5677061153429314, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 50126 + }, + { + "epoch": 0.50127, + "grad_norm": 0.6534222104360997, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 50127 + }, + { + "epoch": 0.50128, + "grad_norm": 0.7112422384194078, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 50128 + }, + { + "epoch": 0.50129, + "grad_norm": 0.7654024898639715, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 50129 + }, + { + "epoch": 0.5013, + "grad_norm": 0.9523332423011565, + "learning_rate": 0.003, + "loss": 3.971, + "step": 50130 + }, + { + "epoch": 0.50131, + "grad_norm": 1.0479052095403547, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 50131 + }, + { + "epoch": 0.50132, + "grad_norm": 0.9368202407946662, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 50132 + }, + { + "epoch": 0.50133, + "grad_norm": 0.8210955103421836, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 50133 + }, + { + "epoch": 0.50134, + "grad_norm": 0.7917369374117054, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 50134 + }, + { + "epoch": 0.50135, + "grad_norm": 0.7166278078529987, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 50135 + }, + { + "epoch": 0.50136, + "grad_norm": 0.7185596709372022, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 50136 + }, + { + "epoch": 0.50137, + "grad_norm": 0.7813176363691976, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 50137 + }, + { + "epoch": 0.50138, + "grad_norm": 0.8428827816873747, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 50138 + }, + { + "epoch": 0.50139, + "grad_norm": 0.7891876752576517, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 50139 + }, + { + "epoch": 0.5014, + "grad_norm": 0.7308369658948232, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 50140 + }, + { + "epoch": 0.50141, + "grad_norm": 0.7584780136190885, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 50141 + }, + { + "epoch": 0.50142, + "grad_norm": 0.7478631927488106, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 50142 + }, + { + "epoch": 0.50143, + "grad_norm": 0.7030858388995285, + "learning_rate": 0.003, + "loss": 3.9525, + "step": 50143 + }, + { + "epoch": 0.50144, + "grad_norm": 0.7642792945356262, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 50144 + }, + { + "epoch": 0.50145, + "grad_norm": 0.9341725688309157, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 50145 + }, + { + "epoch": 0.50146, + "grad_norm": 0.9885449201587992, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 50146 + }, + { + "epoch": 0.50147, + "grad_norm": 0.8540538070047736, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 50147 + }, + { + "epoch": 0.50148, + "grad_norm": 0.6240596922612441, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 50148 + }, + { + "epoch": 0.50149, + "grad_norm": 0.6365426754199109, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 50149 + }, + { + "epoch": 0.5015, + "grad_norm": 0.6434948383684486, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 50150 + }, + { + "epoch": 0.50151, + "grad_norm": 0.627314681349029, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 50151 + }, + { + "epoch": 0.50152, + "grad_norm": 0.6145801696414848, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 50152 + }, + { + "epoch": 0.50153, + "grad_norm": 0.5420202628411614, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 50153 + }, + { + "epoch": 0.50154, + "grad_norm": 0.5763825085489732, + "learning_rate": 0.003, + "loss": 3.981, + "step": 50154 + }, + { + "epoch": 0.50155, + "grad_norm": 0.5390928438606362, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 50155 + }, + { + "epoch": 0.50156, + "grad_norm": 0.546684857460534, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 50156 + }, + { + "epoch": 0.50157, + "grad_norm": 0.6449449280452907, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 50157 + }, + { + "epoch": 0.50158, + "grad_norm": 0.7938614553931893, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 50158 + }, + { + "epoch": 0.50159, + "grad_norm": 0.8673068610917536, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 50159 + }, + { + "epoch": 0.5016, + "grad_norm": 0.9048895848699575, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 50160 + }, + { + "epoch": 0.50161, + "grad_norm": 0.9860769319251201, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 50161 + }, + { + "epoch": 0.50162, + "grad_norm": 0.9321840344404724, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 50162 + }, + { + "epoch": 0.50163, + "grad_norm": 0.914657356368252, + "learning_rate": 0.003, + "loss": 3.979, + "step": 50163 + }, + { + "epoch": 0.50164, + "grad_norm": 0.9047809037922726, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 50164 + }, + { + "epoch": 0.50165, + "grad_norm": 0.7700983154781792, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 50165 + }, + { + "epoch": 0.50166, + "grad_norm": 0.7098728332508831, + "learning_rate": 0.003, + "loss": 3.982, + "step": 50166 + }, + { + "epoch": 0.50167, + "grad_norm": 0.7004928734546638, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 50167 + }, + { + "epoch": 0.50168, + "grad_norm": 0.6692286815899382, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 50168 + }, + { + "epoch": 0.50169, + "grad_norm": 0.7831503940863448, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 50169 + }, + { + "epoch": 0.5017, + "grad_norm": 0.806384805285984, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 50170 + }, + { + "epoch": 0.50171, + "grad_norm": 0.881991778589839, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 50171 + }, + { + "epoch": 0.50172, + "grad_norm": 0.8941909468707778, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 50172 + }, + { + "epoch": 0.50173, + "grad_norm": 0.792873470574516, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 50173 + }, + { + "epoch": 0.50174, + "grad_norm": 0.8381788936617829, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 50174 + }, + { + "epoch": 0.50175, + "grad_norm": 0.8416506620842777, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 50175 + }, + { + "epoch": 0.50176, + "grad_norm": 0.892952836063056, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 50176 + }, + { + "epoch": 0.50177, + "grad_norm": 0.9158471611740373, + "learning_rate": 0.003, + "loss": 4.0264, + "step": 50177 + }, + { + "epoch": 0.50178, + "grad_norm": 0.8322744265694842, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 50178 + }, + { + "epoch": 0.50179, + "grad_norm": 0.8897198249467934, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 50179 + }, + { + "epoch": 0.5018, + "grad_norm": 0.9099115591611157, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 50180 + }, + { + "epoch": 0.50181, + "grad_norm": 0.7841545984221235, + "learning_rate": 0.003, + "loss": 4.0253, + "step": 50181 + }, + { + "epoch": 0.50182, + "grad_norm": 0.8235811642398715, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 50182 + }, + { + "epoch": 0.50183, + "grad_norm": 0.9565389364056485, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 50183 + }, + { + "epoch": 0.50184, + "grad_norm": 1.132557807827732, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 50184 + }, + { + "epoch": 0.50185, + "grad_norm": 0.901977378326229, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 50185 + }, + { + "epoch": 0.50186, + "grad_norm": 0.8336798065925156, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 50186 + }, + { + "epoch": 0.50187, + "grad_norm": 0.7189865488542981, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 50187 + }, + { + "epoch": 0.50188, + "grad_norm": 0.7011409271356025, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 50188 + }, + { + "epoch": 0.50189, + "grad_norm": 0.7273530336584496, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 50189 + }, + { + "epoch": 0.5019, + "grad_norm": 0.7451594078340082, + "learning_rate": 0.003, + "loss": 4.0119, + "step": 50190 + }, + { + "epoch": 0.50191, + "grad_norm": 0.6420132741165723, + "learning_rate": 0.003, + "loss": 4.0198, + "step": 50191 + }, + { + "epoch": 0.50192, + "grad_norm": 0.6777368993046141, + "learning_rate": 0.003, + "loss": 3.979, + "step": 50192 + }, + { + "epoch": 0.50193, + "grad_norm": 0.6575527583041472, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 50193 + }, + { + "epoch": 0.50194, + "grad_norm": 0.6654789502284156, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 50194 + }, + { + "epoch": 0.50195, + "grad_norm": 0.623592422631797, + "learning_rate": 0.003, + "loss": 3.9348, + "step": 50195 + }, + { + "epoch": 0.50196, + "grad_norm": 0.6125282739742596, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 50196 + }, + { + "epoch": 0.50197, + "grad_norm": 0.6821630816468504, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 50197 + }, + { + "epoch": 0.50198, + "grad_norm": 0.7061004992885954, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 50198 + }, + { + "epoch": 0.50199, + "grad_norm": 0.6955805788476549, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 50199 + }, + { + "epoch": 0.502, + "grad_norm": 0.6741314358413347, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 50200 + }, + { + "epoch": 0.50201, + "grad_norm": 0.6278883509499404, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 50201 + }, + { + "epoch": 0.50202, + "grad_norm": 0.602335364456136, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 50202 + }, + { + "epoch": 0.50203, + "grad_norm": 0.593608982958171, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 50203 + }, + { + "epoch": 0.50204, + "grad_norm": 0.7018114323743045, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 50204 + }, + { + "epoch": 0.50205, + "grad_norm": 0.8198456827565828, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 50205 + }, + { + "epoch": 0.50206, + "grad_norm": 0.8958392971562408, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 50206 + }, + { + "epoch": 0.50207, + "grad_norm": 0.9309813877280052, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 50207 + }, + { + "epoch": 0.50208, + "grad_norm": 0.8151384878434108, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 50208 + }, + { + "epoch": 0.50209, + "grad_norm": 0.7451772082129511, + "learning_rate": 0.003, + "loss": 3.999, + "step": 50209 + }, + { + "epoch": 0.5021, + "grad_norm": 0.7768483602391618, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 50210 + }, + { + "epoch": 0.50211, + "grad_norm": 0.8383559604243882, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 50211 + }, + { + "epoch": 0.50212, + "grad_norm": 0.835560433106676, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 50212 + }, + { + "epoch": 0.50213, + "grad_norm": 0.9793075693514643, + "learning_rate": 0.003, + "loss": 3.962, + "step": 50213 + }, + { + "epoch": 0.50214, + "grad_norm": 1.1353884118743172, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 50214 + }, + { + "epoch": 0.50215, + "grad_norm": 0.7897836456292696, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 50215 + }, + { + "epoch": 0.50216, + "grad_norm": 0.7181468824663818, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 50216 + }, + { + "epoch": 0.50217, + "grad_norm": 0.7543485140347592, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 50217 + }, + { + "epoch": 0.50218, + "grad_norm": 0.7357722824283137, + "learning_rate": 0.003, + "loss": 3.994, + "step": 50218 + }, + { + "epoch": 0.50219, + "grad_norm": 0.6673139721379437, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 50219 + }, + { + "epoch": 0.5022, + "grad_norm": 0.5830363606498741, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 50220 + }, + { + "epoch": 0.50221, + "grad_norm": 0.6189794949047722, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 50221 + }, + { + "epoch": 0.50222, + "grad_norm": 0.6331974387720805, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 50222 + }, + { + "epoch": 0.50223, + "grad_norm": 0.6857338015988662, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 50223 + }, + { + "epoch": 0.50224, + "grad_norm": 0.67741657707062, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 50224 + }, + { + "epoch": 0.50225, + "grad_norm": 0.8355202724954741, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 50225 + }, + { + "epoch": 0.50226, + "grad_norm": 0.8955255025288477, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 50226 + }, + { + "epoch": 0.50227, + "grad_norm": 0.8614515279003667, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 50227 + }, + { + "epoch": 0.50228, + "grad_norm": 0.8364918201477296, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 50228 + }, + { + "epoch": 0.50229, + "grad_norm": 0.9154873977752123, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 50229 + }, + { + "epoch": 0.5023, + "grad_norm": 0.7592750497809352, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 50230 + }, + { + "epoch": 0.50231, + "grad_norm": 0.6880748510661351, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 50231 + }, + { + "epoch": 0.50232, + "grad_norm": 0.7401430460192773, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 50232 + }, + { + "epoch": 0.50233, + "grad_norm": 0.8340081204822344, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 50233 + }, + { + "epoch": 0.50234, + "grad_norm": 0.9845189754680377, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 50234 + }, + { + "epoch": 0.50235, + "grad_norm": 1.2454388970199781, + "learning_rate": 0.003, + "loss": 4.025, + "step": 50235 + }, + { + "epoch": 0.50236, + "grad_norm": 0.6858269529123568, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 50236 + }, + { + "epoch": 0.50237, + "grad_norm": 0.5911101016931563, + "learning_rate": 0.003, + "loss": 3.992, + "step": 50237 + }, + { + "epoch": 0.50238, + "grad_norm": 0.6071689288934585, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 50238 + }, + { + "epoch": 0.50239, + "grad_norm": 0.6906220182007197, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 50239 + }, + { + "epoch": 0.5024, + "grad_norm": 0.7874759671931738, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 50240 + }, + { + "epoch": 0.50241, + "grad_norm": 0.6912674074568117, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 50241 + }, + { + "epoch": 0.50242, + "grad_norm": 0.5911679382423572, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 50242 + }, + { + "epoch": 0.50243, + "grad_norm": 0.6533143394235975, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 50243 + }, + { + "epoch": 0.50244, + "grad_norm": 0.6112394926420037, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 50244 + }, + { + "epoch": 0.50245, + "grad_norm": 0.6905860290685644, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 50245 + }, + { + "epoch": 0.50246, + "grad_norm": 0.7331378779618027, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 50246 + }, + { + "epoch": 0.50247, + "grad_norm": 0.8051547760939093, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 50247 + }, + { + "epoch": 0.50248, + "grad_norm": 0.8249644989261953, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 50248 + }, + { + "epoch": 0.50249, + "grad_norm": 0.8673015431065244, + "learning_rate": 0.003, + "loss": 4.0, + "step": 50249 + }, + { + "epoch": 0.5025, + "grad_norm": 0.9040635262517676, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 50250 + }, + { + "epoch": 0.50251, + "grad_norm": 0.8715806896294909, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 50251 + }, + { + "epoch": 0.50252, + "grad_norm": 0.847636205954293, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 50252 + }, + { + "epoch": 0.50253, + "grad_norm": 0.8005254915379227, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 50253 + }, + { + "epoch": 0.50254, + "grad_norm": 0.7851708085683632, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 50254 + }, + { + "epoch": 0.50255, + "grad_norm": 0.757553335989691, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 50255 + }, + { + "epoch": 0.50256, + "grad_norm": 0.7303790058176693, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 50256 + }, + { + "epoch": 0.50257, + "grad_norm": 0.8246315923192457, + "learning_rate": 0.003, + "loss": 3.961, + "step": 50257 + }, + { + "epoch": 0.50258, + "grad_norm": 0.9605145423916125, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 50258 + }, + { + "epoch": 0.50259, + "grad_norm": 0.9545341030026342, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 50259 + }, + { + "epoch": 0.5026, + "grad_norm": 0.985824846403322, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 50260 + }, + { + "epoch": 0.50261, + "grad_norm": 0.9252118470772873, + "learning_rate": 0.003, + "loss": 3.991, + "step": 50261 + }, + { + "epoch": 0.50262, + "grad_norm": 0.9244035982931855, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 50262 + }, + { + "epoch": 0.50263, + "grad_norm": 0.9593843277616637, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 50263 + }, + { + "epoch": 0.50264, + "grad_norm": 1.0285214449187579, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 50264 + }, + { + "epoch": 0.50265, + "grad_norm": 1.0495977173243975, + "learning_rate": 0.003, + "loss": 4.0078, + "step": 50265 + }, + { + "epoch": 0.50266, + "grad_norm": 0.9325004359265927, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 50266 + }, + { + "epoch": 0.50267, + "grad_norm": 0.7912755777782208, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 50267 + }, + { + "epoch": 0.50268, + "grad_norm": 0.8710086661030579, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 50268 + }, + { + "epoch": 0.50269, + "grad_norm": 0.8682103041349502, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 50269 + }, + { + "epoch": 0.5027, + "grad_norm": 0.8097642813250852, + "learning_rate": 0.003, + "loss": 3.973, + "step": 50270 + }, + { + "epoch": 0.50271, + "grad_norm": 0.7060875331167754, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 50271 + }, + { + "epoch": 0.50272, + "grad_norm": 0.7273235191338295, + "learning_rate": 0.003, + "loss": 4.001, + "step": 50272 + }, + { + "epoch": 0.50273, + "grad_norm": 0.6768988477596689, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 50273 + }, + { + "epoch": 0.50274, + "grad_norm": 0.6335478667133735, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 50274 + }, + { + "epoch": 0.50275, + "grad_norm": 0.649415747868845, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 50275 + }, + { + "epoch": 0.50276, + "grad_norm": 0.7162424044292095, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 50276 + }, + { + "epoch": 0.50277, + "grad_norm": 0.75758919932559, + "learning_rate": 0.003, + "loss": 4.0084, + "step": 50277 + }, + { + "epoch": 0.50278, + "grad_norm": 0.6256758252020621, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 50278 + }, + { + "epoch": 0.50279, + "grad_norm": 0.6295080674045024, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 50279 + }, + { + "epoch": 0.5028, + "grad_norm": 0.6481583795421313, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 50280 + }, + { + "epoch": 0.50281, + "grad_norm": 0.6751293223792164, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 50281 + }, + { + "epoch": 0.50282, + "grad_norm": 0.6988738312241011, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 50282 + }, + { + "epoch": 0.50283, + "grad_norm": 0.7007576121654923, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 50283 + }, + { + "epoch": 0.50284, + "grad_norm": 0.7259289687378468, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 50284 + }, + { + "epoch": 0.50285, + "grad_norm": 0.6986541880273898, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 50285 + }, + { + "epoch": 0.50286, + "grad_norm": 0.7443616043673742, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 50286 + }, + { + "epoch": 0.50287, + "grad_norm": 0.734853313854986, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 50287 + }, + { + "epoch": 0.50288, + "grad_norm": 0.6855528671897801, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 50288 + }, + { + "epoch": 0.50289, + "grad_norm": 0.6784365086969416, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 50289 + }, + { + "epoch": 0.5029, + "grad_norm": 0.6856729361574629, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 50290 + }, + { + "epoch": 0.50291, + "grad_norm": 0.7598283374170678, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 50291 + }, + { + "epoch": 0.50292, + "grad_norm": 0.904600984867347, + "learning_rate": 0.003, + "loss": 4.037, + "step": 50292 + }, + { + "epoch": 0.50293, + "grad_norm": 0.9528042316268119, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 50293 + }, + { + "epoch": 0.50294, + "grad_norm": 0.8139462734588908, + "learning_rate": 0.003, + "loss": 3.9418, + "step": 50294 + }, + { + "epoch": 0.50295, + "grad_norm": 0.7609205955129154, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 50295 + }, + { + "epoch": 0.50296, + "grad_norm": 0.8079970011414787, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 50296 + }, + { + "epoch": 0.50297, + "grad_norm": 0.7649711938854163, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 50297 + }, + { + "epoch": 0.50298, + "grad_norm": 0.817455255834617, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 50298 + }, + { + "epoch": 0.50299, + "grad_norm": 0.8895658366488242, + "learning_rate": 0.003, + "loss": 4.008, + "step": 50299 + }, + { + "epoch": 0.503, + "grad_norm": 0.9475401590342013, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 50300 + }, + { + "epoch": 0.50301, + "grad_norm": 0.8739877805060765, + "learning_rate": 0.003, + "loss": 4.0, + "step": 50301 + }, + { + "epoch": 0.50302, + "grad_norm": 0.9173464139187777, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 50302 + }, + { + "epoch": 0.50303, + "grad_norm": 1.070831523326177, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 50303 + }, + { + "epoch": 0.50304, + "grad_norm": 1.0284094666674841, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 50304 + }, + { + "epoch": 0.50305, + "grad_norm": 1.05248171030249, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 50305 + }, + { + "epoch": 0.50306, + "grad_norm": 0.8939638066876641, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 50306 + }, + { + "epoch": 0.50307, + "grad_norm": 0.8901054150982844, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 50307 + }, + { + "epoch": 0.50308, + "grad_norm": 0.8034504747196477, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 50308 + }, + { + "epoch": 0.50309, + "grad_norm": 0.7243893574764144, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 50309 + }, + { + "epoch": 0.5031, + "grad_norm": 0.8239026290772908, + "learning_rate": 0.003, + "loss": 4.0113, + "step": 50310 + }, + { + "epoch": 0.50311, + "grad_norm": 0.9853170786264253, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 50311 + }, + { + "epoch": 0.50312, + "grad_norm": 0.8946938797372082, + "learning_rate": 0.003, + "loss": 3.987, + "step": 50312 + }, + { + "epoch": 0.50313, + "grad_norm": 0.756876312474718, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 50313 + }, + { + "epoch": 0.50314, + "grad_norm": 0.7867947118010774, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 50314 + }, + { + "epoch": 0.50315, + "grad_norm": 0.7974197591353478, + "learning_rate": 0.003, + "loss": 4.0222, + "step": 50315 + }, + { + "epoch": 0.50316, + "grad_norm": 0.8269737099853212, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 50316 + }, + { + "epoch": 0.50317, + "grad_norm": 0.8203067823556692, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 50317 + }, + { + "epoch": 0.50318, + "grad_norm": 0.7968220375972905, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 50318 + }, + { + "epoch": 0.50319, + "grad_norm": 0.6682216622651139, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 50319 + }, + { + "epoch": 0.5032, + "grad_norm": 0.6604801862627422, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 50320 + }, + { + "epoch": 0.50321, + "grad_norm": 0.6226876338484651, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 50321 + }, + { + "epoch": 0.50322, + "grad_norm": 0.6404787407298245, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 50322 + }, + { + "epoch": 0.50323, + "grad_norm": 0.5759337032919836, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 50323 + }, + { + "epoch": 0.50324, + "grad_norm": 0.5496756958676557, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 50324 + }, + { + "epoch": 0.50325, + "grad_norm": 0.5535154798130726, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 50325 + }, + { + "epoch": 0.50326, + "grad_norm": 0.5686175818201554, + "learning_rate": 0.003, + "loss": 3.97, + "step": 50326 + }, + { + "epoch": 0.50327, + "grad_norm": 0.6327036911904611, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 50327 + }, + { + "epoch": 0.50328, + "grad_norm": 0.7058909505420512, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 50328 + }, + { + "epoch": 0.50329, + "grad_norm": 0.7287734758732032, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 50329 + }, + { + "epoch": 0.5033, + "grad_norm": 0.6979760036426941, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 50330 + }, + { + "epoch": 0.50331, + "grad_norm": 0.7460233332765306, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 50331 + }, + { + "epoch": 0.50332, + "grad_norm": 0.6828059371845682, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 50332 + }, + { + "epoch": 0.50333, + "grad_norm": 0.6688508037363833, + "learning_rate": 0.003, + "loss": 3.968, + "step": 50333 + }, + { + "epoch": 0.50334, + "grad_norm": 0.768329738164016, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 50334 + }, + { + "epoch": 0.50335, + "grad_norm": 0.8510796507338069, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 50335 + }, + { + "epoch": 0.50336, + "grad_norm": 0.8757222998729955, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 50336 + }, + { + "epoch": 0.50337, + "grad_norm": 0.8778451885016889, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 50337 + }, + { + "epoch": 0.50338, + "grad_norm": 0.8656817874661352, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 50338 + }, + { + "epoch": 0.50339, + "grad_norm": 0.8671230802279676, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 50339 + }, + { + "epoch": 0.5034, + "grad_norm": 0.9101370120764558, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 50340 + }, + { + "epoch": 0.50341, + "grad_norm": 0.8943218345034414, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 50341 + }, + { + "epoch": 0.50342, + "grad_norm": 0.8799962047515876, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 50342 + }, + { + "epoch": 0.50343, + "grad_norm": 0.7285911625310374, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 50343 + }, + { + "epoch": 0.50344, + "grad_norm": 0.6983702192354257, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 50344 + }, + { + "epoch": 0.50345, + "grad_norm": 0.7726587297241567, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 50345 + }, + { + "epoch": 0.50346, + "grad_norm": 0.8117061655328709, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 50346 + }, + { + "epoch": 0.50347, + "grad_norm": 0.8126887388198994, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 50347 + }, + { + "epoch": 0.50348, + "grad_norm": 0.9112329261939787, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 50348 + }, + { + "epoch": 0.50349, + "grad_norm": 0.8930866463560002, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 50349 + }, + { + "epoch": 0.5035, + "grad_norm": 0.8527939089910542, + "learning_rate": 0.003, + "loss": 3.974, + "step": 50350 + }, + { + "epoch": 0.50351, + "grad_norm": 0.7026981322932372, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 50351 + }, + { + "epoch": 0.50352, + "grad_norm": 0.6242587752590149, + "learning_rate": 0.003, + "loss": 3.968, + "step": 50352 + }, + { + "epoch": 0.50353, + "grad_norm": 0.6328779822705016, + "learning_rate": 0.003, + "loss": 3.98, + "step": 50353 + }, + { + "epoch": 0.50354, + "grad_norm": 0.7057886154296956, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 50354 + }, + { + "epoch": 0.50355, + "grad_norm": 0.7622773167866256, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 50355 + }, + { + "epoch": 0.50356, + "grad_norm": 0.8275059522019699, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 50356 + }, + { + "epoch": 0.50357, + "grad_norm": 0.8340464715543342, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 50357 + }, + { + "epoch": 0.50358, + "grad_norm": 0.7979673959784577, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 50358 + }, + { + "epoch": 0.50359, + "grad_norm": 0.7221968957776881, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 50359 + }, + { + "epoch": 0.5036, + "grad_norm": 0.6124873050641522, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 50360 + }, + { + "epoch": 0.50361, + "grad_norm": 0.7269217112563764, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 50361 + }, + { + "epoch": 0.50362, + "grad_norm": 0.765974795622556, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 50362 + }, + { + "epoch": 0.50363, + "grad_norm": 0.8170155049580037, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 50363 + }, + { + "epoch": 0.50364, + "grad_norm": 0.6654611541771909, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 50364 + }, + { + "epoch": 0.50365, + "grad_norm": 0.5875535475240824, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 50365 + }, + { + "epoch": 0.50366, + "grad_norm": 0.6019902828769922, + "learning_rate": 0.003, + "loss": 3.9401, + "step": 50366 + }, + { + "epoch": 0.50367, + "grad_norm": 0.6946725350758894, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 50367 + }, + { + "epoch": 0.50368, + "grad_norm": 0.7829873028597714, + "learning_rate": 0.003, + "loss": 3.966, + "step": 50368 + }, + { + "epoch": 0.50369, + "grad_norm": 0.9125987341292734, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 50369 + }, + { + "epoch": 0.5037, + "grad_norm": 1.0859535764449884, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 50370 + }, + { + "epoch": 0.50371, + "grad_norm": 1.0164697505385554, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 50371 + }, + { + "epoch": 0.50372, + "grad_norm": 0.784457663444835, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 50372 + }, + { + "epoch": 0.50373, + "grad_norm": 0.8008013674187823, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 50373 + }, + { + "epoch": 0.50374, + "grad_norm": 0.9594271261569643, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 50374 + }, + { + "epoch": 0.50375, + "grad_norm": 0.9786153621472146, + "learning_rate": 0.003, + "loss": 4.0221, + "step": 50375 + }, + { + "epoch": 0.50376, + "grad_norm": 0.8933560956369181, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 50376 + }, + { + "epoch": 0.50377, + "grad_norm": 0.8854119945095129, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 50377 + }, + { + "epoch": 0.50378, + "grad_norm": 0.9650094993652896, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 50378 + }, + { + "epoch": 0.50379, + "grad_norm": 0.9747530870078709, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 50379 + }, + { + "epoch": 0.5038, + "grad_norm": 0.8767916890770049, + "learning_rate": 0.003, + "loss": 4.0178, + "step": 50380 + }, + { + "epoch": 0.50381, + "grad_norm": 1.075337349058012, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 50381 + }, + { + "epoch": 0.50382, + "grad_norm": 1.014584541202576, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 50382 + }, + { + "epoch": 0.50383, + "grad_norm": 1.1272574884340516, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 50383 + }, + { + "epoch": 0.50384, + "grad_norm": 0.8817892866997911, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 50384 + }, + { + "epoch": 0.50385, + "grad_norm": 0.8438114562355975, + "learning_rate": 0.003, + "loss": 4.007, + "step": 50385 + }, + { + "epoch": 0.50386, + "grad_norm": 0.8103519379083429, + "learning_rate": 0.003, + "loss": 3.998, + "step": 50386 + }, + { + "epoch": 0.50387, + "grad_norm": 0.6760064442717598, + "learning_rate": 0.003, + "loss": 4.028, + "step": 50387 + }, + { + "epoch": 0.50388, + "grad_norm": 0.6954402922221029, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 50388 + }, + { + "epoch": 0.50389, + "grad_norm": 0.643580164192441, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 50389 + }, + { + "epoch": 0.5039, + "grad_norm": 0.6348656263595409, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 50390 + }, + { + "epoch": 0.50391, + "grad_norm": 0.5600599948738959, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 50391 + }, + { + "epoch": 0.50392, + "grad_norm": 0.5285351068998311, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 50392 + }, + { + "epoch": 0.50393, + "grad_norm": 0.5171318244571153, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 50393 + }, + { + "epoch": 0.50394, + "grad_norm": 0.5259638632659367, + "learning_rate": 0.003, + "loss": 3.96, + "step": 50394 + }, + { + "epoch": 0.50395, + "grad_norm": 0.4970949027009854, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 50395 + }, + { + "epoch": 0.50396, + "grad_norm": 0.5330356599171971, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 50396 + }, + { + "epoch": 0.50397, + "grad_norm": 0.5690574865358408, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 50397 + }, + { + "epoch": 0.50398, + "grad_norm": 0.6002345545263793, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 50398 + }, + { + "epoch": 0.50399, + "grad_norm": 0.6362696260263513, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 50399 + }, + { + "epoch": 0.504, + "grad_norm": 0.6189973836681161, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 50400 + }, + { + "epoch": 0.50401, + "grad_norm": 0.6264031509693931, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 50401 + }, + { + "epoch": 0.50402, + "grad_norm": 0.7667187337498028, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 50402 + }, + { + "epoch": 0.50403, + "grad_norm": 0.9443309488923053, + "learning_rate": 0.003, + "loss": 3.9559, + "step": 50403 + }, + { + "epoch": 0.50404, + "grad_norm": 0.9723979750932691, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 50404 + }, + { + "epoch": 0.50405, + "grad_norm": 0.883003853579508, + "learning_rate": 0.003, + "loss": 4.001, + "step": 50405 + }, + { + "epoch": 0.50406, + "grad_norm": 0.8382360586571492, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 50406 + }, + { + "epoch": 0.50407, + "grad_norm": 0.8475144491201749, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 50407 + }, + { + "epoch": 0.50408, + "grad_norm": 0.8277446023813473, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 50408 + }, + { + "epoch": 0.50409, + "grad_norm": 0.7703488360477038, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 50409 + }, + { + "epoch": 0.5041, + "grad_norm": 0.8609523825313415, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 50410 + }, + { + "epoch": 0.50411, + "grad_norm": 0.9802826090189349, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 50411 + }, + { + "epoch": 0.50412, + "grad_norm": 0.7991137816095426, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 50412 + }, + { + "epoch": 0.50413, + "grad_norm": 0.770787774772071, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 50413 + }, + { + "epoch": 0.50414, + "grad_norm": 0.8404429981208488, + "learning_rate": 0.003, + "loss": 4.0218, + "step": 50414 + }, + { + "epoch": 0.50415, + "grad_norm": 0.9034571768475129, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 50415 + }, + { + "epoch": 0.50416, + "grad_norm": 0.8409374754971355, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 50416 + }, + { + "epoch": 0.50417, + "grad_norm": 0.7040788391317401, + "learning_rate": 0.003, + "loss": 3.985, + "step": 50417 + }, + { + "epoch": 0.50418, + "grad_norm": 0.7978347269841818, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 50418 + }, + { + "epoch": 0.50419, + "grad_norm": 0.8522371138893478, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 50419 + }, + { + "epoch": 0.5042, + "grad_norm": 0.9876527310634402, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 50420 + }, + { + "epoch": 0.50421, + "grad_norm": 1.1621007078977494, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 50421 + }, + { + "epoch": 0.50422, + "grad_norm": 0.8726038325993145, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 50422 + }, + { + "epoch": 0.50423, + "grad_norm": 0.9285317456936572, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 50423 + }, + { + "epoch": 0.50424, + "grad_norm": 1.0999793480489042, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 50424 + }, + { + "epoch": 0.50425, + "grad_norm": 0.9768393647635566, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 50425 + }, + { + "epoch": 0.50426, + "grad_norm": 1.0177733872025119, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 50426 + }, + { + "epoch": 0.50427, + "grad_norm": 0.7481925915848743, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 50427 + }, + { + "epoch": 0.50428, + "grad_norm": 0.7083789689217361, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 50428 + }, + { + "epoch": 0.50429, + "grad_norm": 0.6953768600705498, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 50429 + }, + { + "epoch": 0.5043, + "grad_norm": 0.7806544806864104, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 50430 + }, + { + "epoch": 0.50431, + "grad_norm": 0.8347214524325137, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 50431 + }, + { + "epoch": 0.50432, + "grad_norm": 0.7259711194600541, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 50432 + }, + { + "epoch": 0.50433, + "grad_norm": 0.6157343551696867, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 50433 + }, + { + "epoch": 0.50434, + "grad_norm": 0.5435381226964134, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 50434 + }, + { + "epoch": 0.50435, + "grad_norm": 0.508889669183709, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 50435 + }, + { + "epoch": 0.50436, + "grad_norm": 0.5128637065174301, + "learning_rate": 0.003, + "loss": 3.9471, + "step": 50436 + }, + { + "epoch": 0.50437, + "grad_norm": 0.48781944917382114, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 50437 + }, + { + "epoch": 0.50438, + "grad_norm": 0.5147448337049229, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 50438 + }, + { + "epoch": 0.50439, + "grad_norm": 0.5333955424686327, + "learning_rate": 0.003, + "loss": 3.964, + "step": 50439 + }, + { + "epoch": 0.5044, + "grad_norm": 0.5257439560745237, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 50440 + }, + { + "epoch": 0.50441, + "grad_norm": 0.636779066378584, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 50441 + }, + { + "epoch": 0.50442, + "grad_norm": 0.6779914416051126, + "learning_rate": 0.003, + "loss": 3.9437, + "step": 50442 + }, + { + "epoch": 0.50443, + "grad_norm": 0.6906626402203241, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 50443 + }, + { + "epoch": 0.50444, + "grad_norm": 0.6220644949569953, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 50444 + }, + { + "epoch": 0.50445, + "grad_norm": 0.5324049828293446, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 50445 + }, + { + "epoch": 0.50446, + "grad_norm": 0.574692378211473, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 50446 + }, + { + "epoch": 0.50447, + "grad_norm": 0.5629166474807548, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 50447 + }, + { + "epoch": 0.50448, + "grad_norm": 0.47722732887663816, + "learning_rate": 0.003, + "loss": 3.9423, + "step": 50448 + }, + { + "epoch": 0.50449, + "grad_norm": 0.5514212854097142, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 50449 + }, + { + "epoch": 0.5045, + "grad_norm": 0.5595294510842845, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 50450 + }, + { + "epoch": 0.50451, + "grad_norm": 0.5724789515185449, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 50451 + }, + { + "epoch": 0.50452, + "grad_norm": 0.6931407352235652, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 50452 + }, + { + "epoch": 0.50453, + "grad_norm": 0.821982103670903, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 50453 + }, + { + "epoch": 0.50454, + "grad_norm": 0.9683224770333616, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 50454 + }, + { + "epoch": 0.50455, + "grad_norm": 1.2006830741550396, + "learning_rate": 0.003, + "loss": 4.001, + "step": 50455 + }, + { + "epoch": 0.50456, + "grad_norm": 0.8612497647002403, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 50456 + }, + { + "epoch": 0.50457, + "grad_norm": 0.8717590518764305, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 50457 + }, + { + "epoch": 0.50458, + "grad_norm": 0.9332945464525867, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 50458 + }, + { + "epoch": 0.50459, + "grad_norm": 0.9863352464194466, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 50459 + }, + { + "epoch": 0.5046, + "grad_norm": 1.0755108408674185, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 50460 + }, + { + "epoch": 0.50461, + "grad_norm": 0.894313851553204, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 50461 + }, + { + "epoch": 0.50462, + "grad_norm": 1.0045571531583235, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 50462 + }, + { + "epoch": 0.50463, + "grad_norm": 0.9125836500208595, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 50463 + }, + { + "epoch": 0.50464, + "grad_norm": 0.9101733361780406, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 50464 + }, + { + "epoch": 0.50465, + "grad_norm": 0.969118162251416, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 50465 + }, + { + "epoch": 0.50466, + "grad_norm": 0.8649324036602591, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 50466 + }, + { + "epoch": 0.50467, + "grad_norm": 0.8104330067067009, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 50467 + }, + { + "epoch": 0.50468, + "grad_norm": 0.8313104373365771, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 50468 + }, + { + "epoch": 0.50469, + "grad_norm": 0.8590905509265145, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 50469 + }, + { + "epoch": 0.5047, + "grad_norm": 0.8297137807613599, + "learning_rate": 0.003, + "loss": 4.019, + "step": 50470 + }, + { + "epoch": 0.50471, + "grad_norm": 0.8992357628724271, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 50471 + }, + { + "epoch": 0.50472, + "grad_norm": 0.9084461307344595, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 50472 + }, + { + "epoch": 0.50473, + "grad_norm": 0.869821810864637, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 50473 + }, + { + "epoch": 0.50474, + "grad_norm": 0.8525598436035458, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 50474 + }, + { + "epoch": 0.50475, + "grad_norm": 0.835189634301587, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 50475 + }, + { + "epoch": 0.50476, + "grad_norm": 0.8148884255839874, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 50476 + }, + { + "epoch": 0.50477, + "grad_norm": 0.6550501002412159, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 50477 + }, + { + "epoch": 0.50478, + "grad_norm": 0.6721434140590066, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 50478 + }, + { + "epoch": 0.50479, + "grad_norm": 0.6977980582488127, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 50479 + }, + { + "epoch": 0.5048, + "grad_norm": 0.8046812641885118, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 50480 + }, + { + "epoch": 0.50481, + "grad_norm": 0.8250311570740307, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 50481 + }, + { + "epoch": 0.50482, + "grad_norm": 0.6862452579558961, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 50482 + }, + { + "epoch": 0.50483, + "grad_norm": 0.6331680562370141, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 50483 + }, + { + "epoch": 0.50484, + "grad_norm": 0.7071963429513141, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 50484 + }, + { + "epoch": 0.50485, + "grad_norm": 0.7973148797531189, + "learning_rate": 0.003, + "loss": 4.008, + "step": 50485 + }, + { + "epoch": 0.50486, + "grad_norm": 0.9163803944188184, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 50486 + }, + { + "epoch": 0.50487, + "grad_norm": 0.9152209199194334, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 50487 + }, + { + "epoch": 0.50488, + "grad_norm": 0.8107029696210057, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 50488 + }, + { + "epoch": 0.50489, + "grad_norm": 0.5932532073272673, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 50489 + }, + { + "epoch": 0.5049, + "grad_norm": 0.5461500002047456, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 50490 + }, + { + "epoch": 0.50491, + "grad_norm": 0.6807561302127881, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 50491 + }, + { + "epoch": 0.50492, + "grad_norm": 0.742253151514157, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 50492 + }, + { + "epoch": 0.50493, + "grad_norm": 0.7101536604457998, + "learning_rate": 0.003, + "loss": 3.9558, + "step": 50493 + }, + { + "epoch": 0.50494, + "grad_norm": 0.6307223637302786, + "learning_rate": 0.003, + "loss": 3.953, + "step": 50494 + }, + { + "epoch": 0.50495, + "grad_norm": 0.6964346027697408, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 50495 + }, + { + "epoch": 0.50496, + "grad_norm": 0.7134744701837796, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 50496 + }, + { + "epoch": 0.50497, + "grad_norm": 0.5912857175720857, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 50497 + }, + { + "epoch": 0.50498, + "grad_norm": 0.6712784459098323, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 50498 + }, + { + "epoch": 0.50499, + "grad_norm": 0.8252859447571023, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 50499 + }, + { + "epoch": 0.505, + "grad_norm": 0.9301007123191939, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 50500 + }, + { + "epoch": 0.50501, + "grad_norm": 0.8408786067324755, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 50501 + }, + { + "epoch": 0.50502, + "grad_norm": 0.7331749360625731, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 50502 + }, + { + "epoch": 0.50503, + "grad_norm": 0.69533866064616, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 50503 + }, + { + "epoch": 0.50504, + "grad_norm": 0.7501657238603232, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 50504 + }, + { + "epoch": 0.50505, + "grad_norm": 0.8265342099052894, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 50505 + }, + { + "epoch": 0.50506, + "grad_norm": 0.8679963553097517, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 50506 + }, + { + "epoch": 0.50507, + "grad_norm": 0.8860873243463634, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 50507 + }, + { + "epoch": 0.50508, + "grad_norm": 1.069523636126382, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 50508 + }, + { + "epoch": 0.50509, + "grad_norm": 0.9349047721790469, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 50509 + }, + { + "epoch": 0.5051, + "grad_norm": 0.819666121110816, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 50510 + }, + { + "epoch": 0.50511, + "grad_norm": 0.7489326338071212, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 50511 + }, + { + "epoch": 0.50512, + "grad_norm": 0.6980278627311575, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 50512 + }, + { + "epoch": 0.50513, + "grad_norm": 0.7375036340988623, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 50513 + }, + { + "epoch": 0.50514, + "grad_norm": 0.7054696092288621, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 50514 + }, + { + "epoch": 0.50515, + "grad_norm": 0.6856175641108478, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 50515 + }, + { + "epoch": 0.50516, + "grad_norm": 0.6217396640464673, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 50516 + }, + { + "epoch": 0.50517, + "grad_norm": 0.5911999620352429, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 50517 + }, + { + "epoch": 0.50518, + "grad_norm": 0.6295816704722673, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 50518 + }, + { + "epoch": 0.50519, + "grad_norm": 0.6370308197816743, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 50519 + }, + { + "epoch": 0.5052, + "grad_norm": 0.7009994620324883, + "learning_rate": 0.003, + "loss": 4.001, + "step": 50520 + }, + { + "epoch": 0.50521, + "grad_norm": 0.7769955684202888, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 50521 + }, + { + "epoch": 0.50522, + "grad_norm": 0.71735061705011, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 50522 + }, + { + "epoch": 0.50523, + "grad_norm": 0.8805577350295173, + "learning_rate": 0.003, + "loss": 3.967, + "step": 50523 + }, + { + "epoch": 0.50524, + "grad_norm": 0.8907544148094141, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 50524 + }, + { + "epoch": 0.50525, + "grad_norm": 0.8519287430313583, + "learning_rate": 0.003, + "loss": 3.982, + "step": 50525 + }, + { + "epoch": 0.50526, + "grad_norm": 0.8682896280796422, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 50526 + }, + { + "epoch": 0.50527, + "grad_norm": 0.8686852512553046, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 50527 + }, + { + "epoch": 0.50528, + "grad_norm": 0.9020862128001541, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 50528 + }, + { + "epoch": 0.50529, + "grad_norm": 0.9833043438348518, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 50529 + }, + { + "epoch": 0.5053, + "grad_norm": 1.1690850758933795, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 50530 + }, + { + "epoch": 0.50531, + "grad_norm": 0.955568495530057, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 50531 + }, + { + "epoch": 0.50532, + "grad_norm": 0.7521721588380037, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 50532 + }, + { + "epoch": 0.50533, + "grad_norm": 0.590102831780164, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 50533 + }, + { + "epoch": 0.50534, + "grad_norm": 0.553434944719, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 50534 + }, + { + "epoch": 0.50535, + "grad_norm": 0.6076355782010423, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 50535 + }, + { + "epoch": 0.50536, + "grad_norm": 0.7072895143362523, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 50536 + }, + { + "epoch": 0.50537, + "grad_norm": 0.6950664795615535, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 50537 + }, + { + "epoch": 0.50538, + "grad_norm": 0.6347598301114772, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 50538 + }, + { + "epoch": 0.50539, + "grad_norm": 0.7032571897420956, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 50539 + }, + { + "epoch": 0.5054, + "grad_norm": 0.8379488689264398, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 50540 + }, + { + "epoch": 0.50541, + "grad_norm": 0.9061168384885316, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 50541 + }, + { + "epoch": 0.50542, + "grad_norm": 0.8993215585924955, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 50542 + }, + { + "epoch": 0.50543, + "grad_norm": 0.888891662476811, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 50543 + }, + { + "epoch": 0.50544, + "grad_norm": 0.9171452844258731, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 50544 + }, + { + "epoch": 0.50545, + "grad_norm": 0.8744189048456783, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 50545 + }, + { + "epoch": 0.50546, + "grad_norm": 0.8052263052567662, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 50546 + }, + { + "epoch": 0.50547, + "grad_norm": 0.8338999381760768, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 50547 + }, + { + "epoch": 0.50548, + "grad_norm": 0.8278287163035045, + "learning_rate": 0.003, + "loss": 4.0263, + "step": 50548 + }, + { + "epoch": 0.50549, + "grad_norm": 0.8812584805762662, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 50549 + }, + { + "epoch": 0.5055, + "grad_norm": 1.0106673591193367, + "learning_rate": 0.003, + "loss": 4.0172, + "step": 50550 + }, + { + "epoch": 0.50551, + "grad_norm": 1.0950567735833094, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 50551 + }, + { + "epoch": 0.50552, + "grad_norm": 0.7098766857444659, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 50552 + }, + { + "epoch": 0.50553, + "grad_norm": 0.7796923939978181, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 50553 + }, + { + "epoch": 0.50554, + "grad_norm": 0.8540810104996731, + "learning_rate": 0.003, + "loss": 4.003, + "step": 50554 + }, + { + "epoch": 0.50555, + "grad_norm": 0.7957533151138703, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 50555 + }, + { + "epoch": 0.50556, + "grad_norm": 0.6676378627159716, + "learning_rate": 0.003, + "loss": 3.983, + "step": 50556 + }, + { + "epoch": 0.50557, + "grad_norm": 0.5754882768172692, + "learning_rate": 0.003, + "loss": 3.972, + "step": 50557 + }, + { + "epoch": 0.50558, + "grad_norm": 0.5929760072206035, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 50558 + }, + { + "epoch": 0.50559, + "grad_norm": 0.7387169426091773, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 50559 + }, + { + "epoch": 0.5056, + "grad_norm": 0.8621180229703251, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 50560 + }, + { + "epoch": 0.50561, + "grad_norm": 0.9195613005321819, + "learning_rate": 0.003, + "loss": 4.0111, + "step": 50561 + }, + { + "epoch": 0.50562, + "grad_norm": 0.8448175804760745, + "learning_rate": 0.003, + "loss": 4.007, + "step": 50562 + }, + { + "epoch": 0.50563, + "grad_norm": 0.7218246501765914, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 50563 + }, + { + "epoch": 0.50564, + "grad_norm": 0.6781129269078422, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 50564 + }, + { + "epoch": 0.50565, + "grad_norm": 0.6251398168679801, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 50565 + }, + { + "epoch": 0.50566, + "grad_norm": 0.5964077683852799, + "learning_rate": 0.003, + "loss": 3.968, + "step": 50566 + }, + { + "epoch": 0.50567, + "grad_norm": 0.6496973503748075, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 50567 + }, + { + "epoch": 0.50568, + "grad_norm": 0.6394183001394024, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 50568 + }, + { + "epoch": 0.50569, + "grad_norm": 0.6647097065798103, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 50569 + }, + { + "epoch": 0.5057, + "grad_norm": 0.6364908812284582, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 50570 + }, + { + "epoch": 0.50571, + "grad_norm": 0.7214837165093312, + "learning_rate": 0.003, + "loss": 4.0186, + "step": 50571 + }, + { + "epoch": 0.50572, + "grad_norm": 0.8718288928554833, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 50572 + }, + { + "epoch": 0.50573, + "grad_norm": 0.9941826142332599, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 50573 + }, + { + "epoch": 0.50574, + "grad_norm": 0.9774053898237554, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 50574 + }, + { + "epoch": 0.50575, + "grad_norm": 0.9728710579871745, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 50575 + }, + { + "epoch": 0.50576, + "grad_norm": 0.9267725209417017, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 50576 + }, + { + "epoch": 0.50577, + "grad_norm": 0.8838246449022827, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 50577 + }, + { + "epoch": 0.50578, + "grad_norm": 0.838701888992364, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 50578 + }, + { + "epoch": 0.50579, + "grad_norm": 0.7105416127257013, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 50579 + }, + { + "epoch": 0.5058, + "grad_norm": 0.7829801573276877, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 50580 + }, + { + "epoch": 0.50581, + "grad_norm": 0.7630175592133541, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 50581 + }, + { + "epoch": 0.50582, + "grad_norm": 0.6525292012192698, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 50582 + }, + { + "epoch": 0.50583, + "grad_norm": 0.6371921664239287, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 50583 + }, + { + "epoch": 0.50584, + "grad_norm": 0.6389862248696199, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 50584 + }, + { + "epoch": 0.50585, + "grad_norm": 0.6862519406225462, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 50585 + }, + { + "epoch": 0.50586, + "grad_norm": 0.7372607115426223, + "learning_rate": 0.003, + "loss": 3.966, + "step": 50586 + }, + { + "epoch": 0.50587, + "grad_norm": 0.7849488325578932, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 50587 + }, + { + "epoch": 0.50588, + "grad_norm": 0.7452463564252725, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 50588 + }, + { + "epoch": 0.50589, + "grad_norm": 0.8070470288678101, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 50589 + }, + { + "epoch": 0.5059, + "grad_norm": 0.77023074498308, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 50590 + }, + { + "epoch": 0.50591, + "grad_norm": 0.7573885328333863, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 50591 + }, + { + "epoch": 0.50592, + "grad_norm": 0.7540888938779833, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 50592 + }, + { + "epoch": 0.50593, + "grad_norm": 0.7512934826497961, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 50593 + }, + { + "epoch": 0.50594, + "grad_norm": 0.9167298748882279, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 50594 + }, + { + "epoch": 0.50595, + "grad_norm": 0.9442198266885518, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 50595 + }, + { + "epoch": 0.50596, + "grad_norm": 0.9402915667467571, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 50596 + }, + { + "epoch": 0.50597, + "grad_norm": 1.05288756865729, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 50597 + }, + { + "epoch": 0.50598, + "grad_norm": 0.9619980470631979, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 50598 + }, + { + "epoch": 0.50599, + "grad_norm": 0.8521999261289126, + "learning_rate": 0.003, + "loss": 3.995, + "step": 50599 + }, + { + "epoch": 0.506, + "grad_norm": 0.7031896178257282, + "learning_rate": 0.003, + "loss": 4.005, + "step": 50600 + }, + { + "epoch": 0.50601, + "grad_norm": 0.6793367547788792, + "learning_rate": 0.003, + "loss": 3.963, + "step": 50601 + }, + { + "epoch": 0.50602, + "grad_norm": 0.6789965097909894, + "learning_rate": 0.003, + "loss": 3.9278, + "step": 50602 + }, + { + "epoch": 0.50603, + "grad_norm": 0.6822309184573722, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 50603 + }, + { + "epoch": 0.50604, + "grad_norm": 0.727934129094122, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 50604 + }, + { + "epoch": 0.50605, + "grad_norm": 0.7133160103800027, + "learning_rate": 0.003, + "loss": 3.983, + "step": 50605 + }, + { + "epoch": 0.50606, + "grad_norm": 0.6312774502010267, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 50606 + }, + { + "epoch": 0.50607, + "grad_norm": 0.6419388034153878, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 50607 + }, + { + "epoch": 0.50608, + "grad_norm": 0.7407044214191946, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 50608 + }, + { + "epoch": 0.50609, + "grad_norm": 0.7695154076764766, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 50609 + }, + { + "epoch": 0.5061, + "grad_norm": 0.8379025333448417, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 50610 + }, + { + "epoch": 0.50611, + "grad_norm": 0.9540719724982476, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 50611 + }, + { + "epoch": 0.50612, + "grad_norm": 0.9233118034173589, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 50612 + }, + { + "epoch": 0.50613, + "grad_norm": 0.7918981849747606, + "learning_rate": 0.003, + "loss": 4.0126, + "step": 50613 + }, + { + "epoch": 0.50614, + "grad_norm": 0.7841464476742438, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 50614 + }, + { + "epoch": 0.50615, + "grad_norm": 0.7744006435585125, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 50615 + }, + { + "epoch": 0.50616, + "grad_norm": 0.7300883663697116, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 50616 + }, + { + "epoch": 0.50617, + "grad_norm": 0.7766921394188782, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 50617 + }, + { + "epoch": 0.50618, + "grad_norm": 0.8047723015478259, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 50618 + }, + { + "epoch": 0.50619, + "grad_norm": 0.904752766172502, + "learning_rate": 0.003, + "loss": 3.986, + "step": 50619 + }, + { + "epoch": 0.5062, + "grad_norm": 0.9453356505732634, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 50620 + }, + { + "epoch": 0.50621, + "grad_norm": 0.9429430476680625, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 50621 + }, + { + "epoch": 0.50622, + "grad_norm": 0.8433313380023528, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 50622 + }, + { + "epoch": 0.50623, + "grad_norm": 0.8191983218619353, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 50623 + }, + { + "epoch": 0.50624, + "grad_norm": 0.6959930879993153, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 50624 + }, + { + "epoch": 0.50625, + "grad_norm": 0.8332799512460618, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 50625 + }, + { + "epoch": 0.50626, + "grad_norm": 1.0014033234592465, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 50626 + }, + { + "epoch": 0.50627, + "grad_norm": 1.0522049113127918, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 50627 + }, + { + "epoch": 0.50628, + "grad_norm": 0.8976825313283784, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 50628 + }, + { + "epoch": 0.50629, + "grad_norm": 0.7852569317861552, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 50629 + }, + { + "epoch": 0.5063, + "grad_norm": 0.7971673017492673, + "learning_rate": 0.003, + "loss": 3.9994, + "step": 50630 + }, + { + "epoch": 0.50631, + "grad_norm": 0.7573652204840421, + "learning_rate": 0.003, + "loss": 3.975, + "step": 50631 + }, + { + "epoch": 0.50632, + "grad_norm": 0.6833814932158305, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 50632 + }, + { + "epoch": 0.50633, + "grad_norm": 0.6967631396940193, + "learning_rate": 0.003, + "loss": 3.969, + "step": 50633 + }, + { + "epoch": 0.50634, + "grad_norm": 0.7060451267571379, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 50634 + }, + { + "epoch": 0.50635, + "grad_norm": 0.830691334175916, + "learning_rate": 0.003, + "loss": 4.004, + "step": 50635 + }, + { + "epoch": 0.50636, + "grad_norm": 0.8784459939673199, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 50636 + }, + { + "epoch": 0.50637, + "grad_norm": 0.971451472547427, + "learning_rate": 0.003, + "loss": 3.9944, + "step": 50637 + }, + { + "epoch": 0.50638, + "grad_norm": 0.9968985734219566, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 50638 + }, + { + "epoch": 0.50639, + "grad_norm": 0.907572177659367, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 50639 + }, + { + "epoch": 0.5064, + "grad_norm": 1.0103849629743091, + "learning_rate": 0.003, + "loss": 4.0251, + "step": 50640 + }, + { + "epoch": 0.50641, + "grad_norm": 0.796180827104208, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 50641 + }, + { + "epoch": 0.50642, + "grad_norm": 0.7302577076404084, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 50642 + }, + { + "epoch": 0.50643, + "grad_norm": 0.6962536177610152, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 50643 + }, + { + "epoch": 0.50644, + "grad_norm": 0.7607443725536529, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 50644 + }, + { + "epoch": 0.50645, + "grad_norm": 0.8294049496858331, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 50645 + }, + { + "epoch": 0.50646, + "grad_norm": 0.830203489958256, + "learning_rate": 0.003, + "loss": 3.9512, + "step": 50646 + }, + { + "epoch": 0.50647, + "grad_norm": 0.7208171169567166, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 50647 + }, + { + "epoch": 0.50648, + "grad_norm": 0.5734045168260183, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 50648 + }, + { + "epoch": 0.50649, + "grad_norm": 0.6049443168007975, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 50649 + }, + { + "epoch": 0.5065, + "grad_norm": 0.5493090622368147, + "learning_rate": 0.003, + "loss": 3.958, + "step": 50650 + }, + { + "epoch": 0.50651, + "grad_norm": 0.5231760188855412, + "learning_rate": 0.003, + "loss": 3.9411, + "step": 50651 + }, + { + "epoch": 0.50652, + "grad_norm": 0.5033072588549807, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 50652 + }, + { + "epoch": 0.50653, + "grad_norm": 0.5810659074889647, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 50653 + }, + { + "epoch": 0.50654, + "grad_norm": 0.667064157035949, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 50654 + }, + { + "epoch": 0.50655, + "grad_norm": 0.6552339747263551, + "learning_rate": 0.003, + "loss": 3.9367, + "step": 50655 + }, + { + "epoch": 0.50656, + "grad_norm": 0.6184924373598291, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 50656 + }, + { + "epoch": 0.50657, + "grad_norm": 0.5806307516993227, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 50657 + }, + { + "epoch": 0.50658, + "grad_norm": 0.6548382299038614, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 50658 + }, + { + "epoch": 0.50659, + "grad_norm": 0.6669707191186333, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 50659 + }, + { + "epoch": 0.5066, + "grad_norm": 0.6896972371948727, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 50660 + }, + { + "epoch": 0.50661, + "grad_norm": 0.8299828168404664, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 50661 + }, + { + "epoch": 0.50662, + "grad_norm": 0.9653975353705351, + "learning_rate": 0.003, + "loss": 3.9427, + "step": 50662 + }, + { + "epoch": 0.50663, + "grad_norm": 0.9528928544940889, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 50663 + }, + { + "epoch": 0.50664, + "grad_norm": 1.0577743920246658, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 50664 + }, + { + "epoch": 0.50665, + "grad_norm": 0.8891241822905633, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 50665 + }, + { + "epoch": 0.50666, + "grad_norm": 0.6395262823002703, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 50666 + }, + { + "epoch": 0.50667, + "grad_norm": 0.6606392135052648, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 50667 + }, + { + "epoch": 0.50668, + "grad_norm": 0.7630082919342128, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 50668 + }, + { + "epoch": 0.50669, + "grad_norm": 0.8788640991811193, + "learning_rate": 0.003, + "loss": 3.993, + "step": 50669 + }, + { + "epoch": 0.5067, + "grad_norm": 0.9026264481106396, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 50670 + }, + { + "epoch": 0.50671, + "grad_norm": 0.8824693285832161, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 50671 + }, + { + "epoch": 0.50672, + "grad_norm": 0.7453071805642719, + "learning_rate": 0.003, + "loss": 3.98, + "step": 50672 + }, + { + "epoch": 0.50673, + "grad_norm": 0.6756597663607109, + "learning_rate": 0.003, + "loss": 4.0306, + "step": 50673 + }, + { + "epoch": 0.50674, + "grad_norm": 0.7277158937905527, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 50674 + }, + { + "epoch": 0.50675, + "grad_norm": 0.7744177409490237, + "learning_rate": 0.003, + "loss": 3.978, + "step": 50675 + }, + { + "epoch": 0.50676, + "grad_norm": 0.7899412136988261, + "learning_rate": 0.003, + "loss": 3.968, + "step": 50676 + }, + { + "epoch": 0.50677, + "grad_norm": 0.7274381941821388, + "learning_rate": 0.003, + "loss": 3.974, + "step": 50677 + }, + { + "epoch": 0.50678, + "grad_norm": 0.6792786634448864, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 50678 + }, + { + "epoch": 0.50679, + "grad_norm": 0.7215955523147675, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 50679 + }, + { + "epoch": 0.5068, + "grad_norm": 0.7605079187557532, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 50680 + }, + { + "epoch": 0.50681, + "grad_norm": 0.7892054102608814, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 50681 + }, + { + "epoch": 0.50682, + "grad_norm": 0.711371909309745, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 50682 + }, + { + "epoch": 0.50683, + "grad_norm": 0.7550684207067083, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 50683 + }, + { + "epoch": 0.50684, + "grad_norm": 0.7761101216432539, + "learning_rate": 0.003, + "loss": 3.9533, + "step": 50684 + }, + { + "epoch": 0.50685, + "grad_norm": 0.7937211967490386, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 50685 + }, + { + "epoch": 0.50686, + "grad_norm": 0.9119729751050185, + "learning_rate": 0.003, + "loss": 3.979, + "step": 50686 + }, + { + "epoch": 0.50687, + "grad_norm": 1.092569664892945, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 50687 + }, + { + "epoch": 0.50688, + "grad_norm": 0.9985022290355868, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 50688 + }, + { + "epoch": 0.50689, + "grad_norm": 1.1275386079264782, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 50689 + }, + { + "epoch": 0.5069, + "grad_norm": 0.8063580919301492, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 50690 + }, + { + "epoch": 0.50691, + "grad_norm": 0.8334391194917581, + "learning_rate": 0.003, + "loss": 3.997, + "step": 50691 + }, + { + "epoch": 0.50692, + "grad_norm": 0.8015311508455009, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 50692 + }, + { + "epoch": 0.50693, + "grad_norm": 0.7882809598076573, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 50693 + }, + { + "epoch": 0.50694, + "grad_norm": 0.6488160650234445, + "learning_rate": 0.003, + "loss": 3.978, + "step": 50694 + }, + { + "epoch": 0.50695, + "grad_norm": 0.5799193620886376, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 50695 + }, + { + "epoch": 0.50696, + "grad_norm": 0.6066066191873526, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 50696 + }, + { + "epoch": 0.50697, + "grad_norm": 0.6837433099848139, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 50697 + }, + { + "epoch": 0.50698, + "grad_norm": 0.712228588741148, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 50698 + }, + { + "epoch": 0.50699, + "grad_norm": 0.6522570748929392, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 50699 + }, + { + "epoch": 0.507, + "grad_norm": 0.6355298103795526, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 50700 + }, + { + "epoch": 0.50701, + "grad_norm": 0.5926207480382103, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 50701 + }, + { + "epoch": 0.50702, + "grad_norm": 0.5637851975644161, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 50702 + }, + { + "epoch": 0.50703, + "grad_norm": 0.6127147026962859, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 50703 + }, + { + "epoch": 0.50704, + "grad_norm": 0.6752377047441579, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 50704 + }, + { + "epoch": 0.50705, + "grad_norm": 0.7570254430124086, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 50705 + }, + { + "epoch": 0.50706, + "grad_norm": 0.7891258092604605, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 50706 + }, + { + "epoch": 0.50707, + "grad_norm": 0.8371508732068101, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 50707 + }, + { + "epoch": 0.50708, + "grad_norm": 0.8324845236020627, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 50708 + }, + { + "epoch": 0.50709, + "grad_norm": 0.820015697333407, + "learning_rate": 0.003, + "loss": 4.0333, + "step": 50709 + }, + { + "epoch": 0.5071, + "grad_norm": 0.8468717445175666, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 50710 + }, + { + "epoch": 0.50711, + "grad_norm": 0.7745278934271236, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 50711 + }, + { + "epoch": 0.50712, + "grad_norm": 0.8657239014435797, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 50712 + }, + { + "epoch": 0.50713, + "grad_norm": 0.9565186906438441, + "learning_rate": 0.003, + "loss": 3.9479, + "step": 50713 + }, + { + "epoch": 0.50714, + "grad_norm": 0.940552575171535, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 50714 + }, + { + "epoch": 0.50715, + "grad_norm": 0.8533262233091045, + "learning_rate": 0.003, + "loss": 4.0107, + "step": 50715 + }, + { + "epoch": 0.50716, + "grad_norm": 0.8640805885956838, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 50716 + }, + { + "epoch": 0.50717, + "grad_norm": 0.9568453241149765, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 50717 + }, + { + "epoch": 0.50718, + "grad_norm": 0.9355171793670338, + "learning_rate": 0.003, + "loss": 3.982, + "step": 50718 + }, + { + "epoch": 0.50719, + "grad_norm": 0.8128152632145315, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 50719 + }, + { + "epoch": 0.5072, + "grad_norm": 0.8673110369830366, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 50720 + }, + { + "epoch": 0.50721, + "grad_norm": 0.922171690539008, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 50721 + }, + { + "epoch": 0.50722, + "grad_norm": 0.9298913980307022, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 50722 + }, + { + "epoch": 0.50723, + "grad_norm": 0.9355431508532085, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 50723 + }, + { + "epoch": 0.50724, + "grad_norm": 1.0602964705726257, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 50724 + }, + { + "epoch": 0.50725, + "grad_norm": 1.1113031281383419, + "learning_rate": 0.003, + "loss": 4.0317, + "step": 50725 + }, + { + "epoch": 0.50726, + "grad_norm": 0.6863671486956526, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 50726 + }, + { + "epoch": 0.50727, + "grad_norm": 0.7118341078121417, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 50727 + }, + { + "epoch": 0.50728, + "grad_norm": 0.8292831905977481, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 50728 + }, + { + "epoch": 0.50729, + "grad_norm": 0.8631901360181663, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 50729 + }, + { + "epoch": 0.5073, + "grad_norm": 0.8368422547105031, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 50730 + }, + { + "epoch": 0.50731, + "grad_norm": 0.8487158318896065, + "learning_rate": 0.003, + "loss": 4.0072, + "step": 50731 + }, + { + "epoch": 0.50732, + "grad_norm": 0.8544152367267545, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 50732 + }, + { + "epoch": 0.50733, + "grad_norm": 0.7481435839759519, + "learning_rate": 0.003, + "loss": 3.9808, + "step": 50733 + }, + { + "epoch": 0.50734, + "grad_norm": 0.6793485968244325, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 50734 + }, + { + "epoch": 0.50735, + "grad_norm": 0.7873745874185116, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 50735 + }, + { + "epoch": 0.50736, + "grad_norm": 0.7276682048068692, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 50736 + }, + { + "epoch": 0.50737, + "grad_norm": 0.7036456280353224, + "learning_rate": 0.003, + "loss": 3.994, + "step": 50737 + }, + { + "epoch": 0.50738, + "grad_norm": 0.692978468589967, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 50738 + }, + { + "epoch": 0.50739, + "grad_norm": 0.7123107071576081, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 50739 + }, + { + "epoch": 0.5074, + "grad_norm": 0.6921072941612474, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 50740 + }, + { + "epoch": 0.50741, + "grad_norm": 0.7207030525647221, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 50741 + }, + { + "epoch": 0.50742, + "grad_norm": 0.7232751527947769, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 50742 + }, + { + "epoch": 0.50743, + "grad_norm": 0.7240461448830453, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 50743 + }, + { + "epoch": 0.50744, + "grad_norm": 0.8030366234407388, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 50744 + }, + { + "epoch": 0.50745, + "grad_norm": 0.8401584959791684, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 50745 + }, + { + "epoch": 0.50746, + "grad_norm": 1.0075376517078094, + "learning_rate": 0.003, + "loss": 4.0189, + "step": 50746 + }, + { + "epoch": 0.50747, + "grad_norm": 1.004470169656025, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 50747 + }, + { + "epoch": 0.50748, + "grad_norm": 0.9018392414525163, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 50748 + }, + { + "epoch": 0.50749, + "grad_norm": 0.7636834145927917, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 50749 + }, + { + "epoch": 0.5075, + "grad_norm": 0.7156884573487744, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 50750 + }, + { + "epoch": 0.50751, + "grad_norm": 0.6220966714351555, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 50751 + }, + { + "epoch": 0.50752, + "grad_norm": 0.6945157702997781, + "learning_rate": 0.003, + "loss": 3.9926, + "step": 50752 + }, + { + "epoch": 0.50753, + "grad_norm": 0.8033065682091064, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 50753 + }, + { + "epoch": 0.50754, + "grad_norm": 0.7338674323812125, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 50754 + }, + { + "epoch": 0.50755, + "grad_norm": 0.7114433948128214, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 50755 + }, + { + "epoch": 0.50756, + "grad_norm": 0.7010401177997357, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 50756 + }, + { + "epoch": 0.50757, + "grad_norm": 0.6677842044676722, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 50757 + }, + { + "epoch": 0.50758, + "grad_norm": 0.6685411401335292, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 50758 + }, + { + "epoch": 0.50759, + "grad_norm": 0.6089123511533348, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 50759 + }, + { + "epoch": 0.5076, + "grad_norm": 0.6187176206522312, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 50760 + }, + { + "epoch": 0.50761, + "grad_norm": 0.6300818691662659, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 50761 + }, + { + "epoch": 0.50762, + "grad_norm": 0.6528528158532523, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 50762 + }, + { + "epoch": 0.50763, + "grad_norm": 0.6025775513801181, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 50763 + }, + { + "epoch": 0.50764, + "grad_norm": 0.5747431937714081, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 50764 + }, + { + "epoch": 0.50765, + "grad_norm": 0.5542492686144264, + "learning_rate": 0.003, + "loss": 3.988, + "step": 50765 + }, + { + "epoch": 0.50766, + "grad_norm": 0.6965410913491815, + "learning_rate": 0.003, + "loss": 3.9747, + "step": 50766 + }, + { + "epoch": 0.50767, + "grad_norm": 0.7853227402118741, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 50767 + }, + { + "epoch": 0.50768, + "grad_norm": 0.7628981220651634, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 50768 + }, + { + "epoch": 0.50769, + "grad_norm": 0.6933467220902163, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 50769 + }, + { + "epoch": 0.5077, + "grad_norm": 0.7242744422218438, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 50770 + }, + { + "epoch": 0.50771, + "grad_norm": 0.7496274324897164, + "learning_rate": 0.003, + "loss": 3.976, + "step": 50771 + }, + { + "epoch": 0.50772, + "grad_norm": 0.6576932155291965, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 50772 + }, + { + "epoch": 0.50773, + "grad_norm": 0.6614697930808986, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 50773 + }, + { + "epoch": 0.50774, + "grad_norm": 0.678748544869682, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 50774 + }, + { + "epoch": 0.50775, + "grad_norm": 0.5707473122689196, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 50775 + }, + { + "epoch": 0.50776, + "grad_norm": 0.6010670157481665, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 50776 + }, + { + "epoch": 0.50777, + "grad_norm": 0.7381673244647519, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 50777 + }, + { + "epoch": 0.50778, + "grad_norm": 0.9455941224644955, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 50778 + }, + { + "epoch": 0.50779, + "grad_norm": 1.0118679614030333, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 50779 + }, + { + "epoch": 0.5078, + "grad_norm": 0.9360573285075656, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 50780 + }, + { + "epoch": 0.50781, + "grad_norm": 0.8914669009369669, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 50781 + }, + { + "epoch": 0.50782, + "grad_norm": 0.8570772871904544, + "learning_rate": 0.003, + "loss": 3.989, + "step": 50782 + }, + { + "epoch": 0.50783, + "grad_norm": 0.9129131439165251, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 50783 + }, + { + "epoch": 0.50784, + "grad_norm": 0.8718427551484128, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 50784 + }, + { + "epoch": 0.50785, + "grad_norm": 0.8402942663486238, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 50785 + }, + { + "epoch": 0.50786, + "grad_norm": 0.7939912686116593, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 50786 + }, + { + "epoch": 0.50787, + "grad_norm": 0.8282718250559564, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 50787 + }, + { + "epoch": 0.50788, + "grad_norm": 0.8202018598606088, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 50788 + }, + { + "epoch": 0.50789, + "grad_norm": 0.9627747102914166, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 50789 + }, + { + "epoch": 0.5079, + "grad_norm": 0.9500397824227307, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 50790 + }, + { + "epoch": 0.50791, + "grad_norm": 0.9260368443772891, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 50791 + }, + { + "epoch": 0.50792, + "grad_norm": 0.8539404205778879, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 50792 + }, + { + "epoch": 0.50793, + "grad_norm": 0.7790139997499034, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 50793 + }, + { + "epoch": 0.50794, + "grad_norm": 0.7690448618684464, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 50794 + }, + { + "epoch": 0.50795, + "grad_norm": 0.779799902471736, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 50795 + }, + { + "epoch": 0.50796, + "grad_norm": 0.8229144697298268, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 50796 + }, + { + "epoch": 0.50797, + "grad_norm": 1.0677778665164583, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 50797 + }, + { + "epoch": 0.50798, + "grad_norm": 1.1547272808961666, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 50798 + }, + { + "epoch": 0.50799, + "grad_norm": 0.7754468133806591, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 50799 + }, + { + "epoch": 0.508, + "grad_norm": 0.7012934551979293, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 50800 + }, + { + "epoch": 0.50801, + "grad_norm": 0.6648570824452609, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 50801 + }, + { + "epoch": 0.50802, + "grad_norm": 0.6219994339869613, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 50802 + }, + { + "epoch": 0.50803, + "grad_norm": 0.6387169797374231, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 50803 + }, + { + "epoch": 0.50804, + "grad_norm": 0.6692959104131417, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 50804 + }, + { + "epoch": 0.50805, + "grad_norm": 0.7515651162549004, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 50805 + }, + { + "epoch": 0.50806, + "grad_norm": 0.73126265750875, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 50806 + }, + { + "epoch": 0.50807, + "grad_norm": 0.7652134068827192, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 50807 + }, + { + "epoch": 0.50808, + "grad_norm": 0.7508618513074482, + "learning_rate": 0.003, + "loss": 4.0058, + "step": 50808 + }, + { + "epoch": 0.50809, + "grad_norm": 0.6722293821397725, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 50809 + }, + { + "epoch": 0.5081, + "grad_norm": 0.6051901800098224, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 50810 + }, + { + "epoch": 0.50811, + "grad_norm": 0.5710170004090334, + "learning_rate": 0.003, + "loss": 3.96, + "step": 50811 + }, + { + "epoch": 0.50812, + "grad_norm": 0.5718535146951913, + "learning_rate": 0.003, + "loss": 3.9555, + "step": 50812 + }, + { + "epoch": 0.50813, + "grad_norm": 0.5536778994105193, + "learning_rate": 0.003, + "loss": 3.983, + "step": 50813 + }, + { + "epoch": 0.50814, + "grad_norm": 0.5856220459061902, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 50814 + }, + { + "epoch": 0.50815, + "grad_norm": 0.6811514329494436, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 50815 + }, + { + "epoch": 0.50816, + "grad_norm": 0.877970053819876, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 50816 + }, + { + "epoch": 0.50817, + "grad_norm": 1.105091672236333, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 50817 + }, + { + "epoch": 0.50818, + "grad_norm": 0.97799591319747, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 50818 + }, + { + "epoch": 0.50819, + "grad_norm": 0.8052647684389846, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 50819 + }, + { + "epoch": 0.5082, + "grad_norm": 0.7786492960107843, + "learning_rate": 0.003, + "loss": 3.968, + "step": 50820 + }, + { + "epoch": 0.50821, + "grad_norm": 0.7272018482594371, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 50821 + }, + { + "epoch": 0.50822, + "grad_norm": 0.8100028729734997, + "learning_rate": 0.003, + "loss": 4.001, + "step": 50822 + }, + { + "epoch": 0.50823, + "grad_norm": 0.7289496436631621, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 50823 + }, + { + "epoch": 0.50824, + "grad_norm": 0.7658322140106798, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 50824 + }, + { + "epoch": 0.50825, + "grad_norm": 0.7844197744567032, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 50825 + }, + { + "epoch": 0.50826, + "grad_norm": 0.7968159076903286, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 50826 + }, + { + "epoch": 0.50827, + "grad_norm": 0.7403841292265735, + "learning_rate": 0.003, + "loss": 4.0224, + "step": 50827 + }, + { + "epoch": 0.50828, + "grad_norm": 0.691126818739896, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 50828 + }, + { + "epoch": 0.50829, + "grad_norm": 0.6060580578749872, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 50829 + }, + { + "epoch": 0.5083, + "grad_norm": 0.5265005320012981, + "learning_rate": 0.003, + "loss": 3.9684, + "step": 50830 + }, + { + "epoch": 0.50831, + "grad_norm": 0.5693498607292793, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 50831 + }, + { + "epoch": 0.50832, + "grad_norm": 0.5633651749912693, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 50832 + }, + { + "epoch": 0.50833, + "grad_norm": 0.6427057789847069, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 50833 + }, + { + "epoch": 0.50834, + "grad_norm": 0.6417081241676664, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 50834 + }, + { + "epoch": 0.50835, + "grad_norm": 0.6722552554559206, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 50835 + }, + { + "epoch": 0.50836, + "grad_norm": 0.7294609507258526, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 50836 + }, + { + "epoch": 0.50837, + "grad_norm": 0.7967480847324644, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 50837 + }, + { + "epoch": 0.50838, + "grad_norm": 0.8847939989070821, + "learning_rate": 0.003, + "loss": 4.0082, + "step": 50838 + }, + { + "epoch": 0.50839, + "grad_norm": 1.02798366503521, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 50839 + }, + { + "epoch": 0.5084, + "grad_norm": 1.1747742256888316, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 50840 + }, + { + "epoch": 0.50841, + "grad_norm": 0.8892896580567423, + "learning_rate": 0.003, + "loss": 4.0048, + "step": 50841 + }, + { + "epoch": 0.50842, + "grad_norm": 0.9505332568044855, + "learning_rate": 0.003, + "loss": 3.987, + "step": 50842 + }, + { + "epoch": 0.50843, + "grad_norm": 0.9885275054889292, + "learning_rate": 0.003, + "loss": 4.005, + "step": 50843 + }, + { + "epoch": 0.50844, + "grad_norm": 1.1094535331898354, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 50844 + }, + { + "epoch": 0.50845, + "grad_norm": 1.0622080712411988, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 50845 + }, + { + "epoch": 0.50846, + "grad_norm": 1.0051350381180415, + "learning_rate": 0.003, + "loss": 3.976, + "step": 50846 + }, + { + "epoch": 0.50847, + "grad_norm": 1.0113348234175885, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 50847 + }, + { + "epoch": 0.50848, + "grad_norm": 0.9585164477502203, + "learning_rate": 0.003, + "loss": 4.0146, + "step": 50848 + }, + { + "epoch": 0.50849, + "grad_norm": 0.8493327916013889, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 50849 + }, + { + "epoch": 0.5085, + "grad_norm": 0.845377551550557, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 50850 + }, + { + "epoch": 0.50851, + "grad_norm": 0.7783041448627644, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 50851 + }, + { + "epoch": 0.50852, + "grad_norm": 0.7289108293111403, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 50852 + }, + { + "epoch": 0.50853, + "grad_norm": 0.7405067007663173, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 50853 + }, + { + "epoch": 0.50854, + "grad_norm": 0.715638004329644, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 50854 + }, + { + "epoch": 0.50855, + "grad_norm": 0.6633289019997052, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 50855 + }, + { + "epoch": 0.50856, + "grad_norm": 0.6530802465535203, + "learning_rate": 0.003, + "loss": 4.0014, + "step": 50856 + }, + { + "epoch": 0.50857, + "grad_norm": 0.6862955632147286, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 50857 + }, + { + "epoch": 0.50858, + "grad_norm": 0.8190004769199114, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 50858 + }, + { + "epoch": 0.50859, + "grad_norm": 1.0328702672874484, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 50859 + }, + { + "epoch": 0.5086, + "grad_norm": 1.1564766529944912, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 50860 + }, + { + "epoch": 0.50861, + "grad_norm": 0.7255927363291033, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 50861 + }, + { + "epoch": 0.50862, + "grad_norm": 0.6891196758899124, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 50862 + }, + { + "epoch": 0.50863, + "grad_norm": 0.5998418408805047, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 50863 + }, + { + "epoch": 0.50864, + "grad_norm": 0.6079447660267409, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 50864 + }, + { + "epoch": 0.50865, + "grad_norm": 0.6548037915880546, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 50865 + }, + { + "epoch": 0.50866, + "grad_norm": 0.6849053468677249, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 50866 + }, + { + "epoch": 0.50867, + "grad_norm": 0.7691727072041378, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 50867 + }, + { + "epoch": 0.50868, + "grad_norm": 0.7812072218554071, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 50868 + }, + { + "epoch": 0.50869, + "grad_norm": 0.8827037749803232, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 50869 + }, + { + "epoch": 0.5087, + "grad_norm": 0.8827021227401792, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 50870 + }, + { + "epoch": 0.50871, + "grad_norm": 0.9115443499351498, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 50871 + }, + { + "epoch": 0.50872, + "grad_norm": 1.0171270097122822, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 50872 + }, + { + "epoch": 0.50873, + "grad_norm": 0.8024189798887165, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 50873 + }, + { + "epoch": 0.50874, + "grad_norm": 0.7262773501439757, + "learning_rate": 0.003, + "loss": 3.965, + "step": 50874 + }, + { + "epoch": 0.50875, + "grad_norm": 0.9106353330509165, + "learning_rate": 0.003, + "loss": 4.009, + "step": 50875 + }, + { + "epoch": 0.50876, + "grad_norm": 0.926376130869098, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 50876 + }, + { + "epoch": 0.50877, + "grad_norm": 0.8447462359293739, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 50877 + }, + { + "epoch": 0.50878, + "grad_norm": 0.8712887662262985, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 50878 + }, + { + "epoch": 0.50879, + "grad_norm": 0.7917277601432142, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 50879 + }, + { + "epoch": 0.5088, + "grad_norm": 0.6301405287874832, + "learning_rate": 0.003, + "loss": 3.991, + "step": 50880 + }, + { + "epoch": 0.50881, + "grad_norm": 0.6890790660037901, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 50881 + }, + { + "epoch": 0.50882, + "grad_norm": 0.7167727646098383, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 50882 + }, + { + "epoch": 0.50883, + "grad_norm": 0.6131438413751364, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 50883 + }, + { + "epoch": 0.50884, + "grad_norm": 0.5020398467093489, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 50884 + }, + { + "epoch": 0.50885, + "grad_norm": 0.6174312780067386, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 50885 + }, + { + "epoch": 0.50886, + "grad_norm": 0.6268886928536515, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 50886 + }, + { + "epoch": 0.50887, + "grad_norm": 0.5496954168215139, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 50887 + }, + { + "epoch": 0.50888, + "grad_norm": 0.5284757373593952, + "learning_rate": 0.003, + "loss": 3.9419, + "step": 50888 + }, + { + "epoch": 0.50889, + "grad_norm": 0.5353386777608339, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 50889 + }, + { + "epoch": 0.5089, + "grad_norm": 0.5515981459221245, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 50890 + }, + { + "epoch": 0.50891, + "grad_norm": 0.5925295824949621, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 50891 + }, + { + "epoch": 0.50892, + "grad_norm": 0.6350182094315722, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 50892 + }, + { + "epoch": 0.50893, + "grad_norm": 0.6695560753301115, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 50893 + }, + { + "epoch": 0.50894, + "grad_norm": 0.6925565111324453, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 50894 + }, + { + "epoch": 0.50895, + "grad_norm": 0.7058102827154867, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 50895 + }, + { + "epoch": 0.50896, + "grad_norm": 0.6933732601246263, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 50896 + }, + { + "epoch": 0.50897, + "grad_norm": 0.7651651638939466, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 50897 + }, + { + "epoch": 0.50898, + "grad_norm": 0.8325480056437456, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 50898 + }, + { + "epoch": 0.50899, + "grad_norm": 0.9189501561261928, + "learning_rate": 0.003, + "loss": 3.9481, + "step": 50899 + }, + { + "epoch": 0.509, + "grad_norm": 0.9801323384797099, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 50900 + }, + { + "epoch": 0.50901, + "grad_norm": 1.033015030233585, + "learning_rate": 0.003, + "loss": 3.9612, + "step": 50901 + }, + { + "epoch": 0.50902, + "grad_norm": 0.8463918186610947, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 50902 + }, + { + "epoch": 0.50903, + "grad_norm": 0.7453221050333071, + "learning_rate": 0.003, + "loss": 3.97, + "step": 50903 + }, + { + "epoch": 0.50904, + "grad_norm": 0.8089608403331211, + "learning_rate": 0.003, + "loss": 3.9671, + "step": 50904 + }, + { + "epoch": 0.50905, + "grad_norm": 0.9397546225184061, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 50905 + }, + { + "epoch": 0.50906, + "grad_norm": 1.0657003231021127, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 50906 + }, + { + "epoch": 0.50907, + "grad_norm": 0.9311042370350401, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 50907 + }, + { + "epoch": 0.50908, + "grad_norm": 0.8436360390924657, + "learning_rate": 0.003, + "loss": 3.959, + "step": 50908 + }, + { + "epoch": 0.50909, + "grad_norm": 0.7040611355312054, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 50909 + }, + { + "epoch": 0.5091, + "grad_norm": 0.6919978076712393, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 50910 + }, + { + "epoch": 0.50911, + "grad_norm": 0.6377126429385049, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 50911 + }, + { + "epoch": 0.50912, + "grad_norm": 0.5433069832249623, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 50912 + }, + { + "epoch": 0.50913, + "grad_norm": 0.6512810751746808, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 50913 + }, + { + "epoch": 0.50914, + "grad_norm": 0.8735154009604801, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 50914 + }, + { + "epoch": 0.50915, + "grad_norm": 1.019056696891147, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 50915 + }, + { + "epoch": 0.50916, + "grad_norm": 1.1031082098597758, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 50916 + }, + { + "epoch": 0.50917, + "grad_norm": 1.0037018654773084, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 50917 + }, + { + "epoch": 0.50918, + "grad_norm": 1.026830895719171, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 50918 + }, + { + "epoch": 0.50919, + "grad_norm": 0.9918225851948625, + "learning_rate": 0.003, + "loss": 4.016, + "step": 50919 + }, + { + "epoch": 0.5092, + "grad_norm": 1.0296317626419544, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 50920 + }, + { + "epoch": 0.50921, + "grad_norm": 1.0321867484065477, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 50921 + }, + { + "epoch": 0.50922, + "grad_norm": 0.8738998607546139, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 50922 + }, + { + "epoch": 0.50923, + "grad_norm": 0.8536340355856109, + "learning_rate": 0.003, + "loss": 4.0276, + "step": 50923 + }, + { + "epoch": 0.50924, + "grad_norm": 0.85643438210706, + "learning_rate": 0.003, + "loss": 4.005, + "step": 50924 + }, + { + "epoch": 0.50925, + "grad_norm": 0.8210083390340204, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 50925 + }, + { + "epoch": 0.50926, + "grad_norm": 0.852205817795156, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 50926 + }, + { + "epoch": 0.50927, + "grad_norm": 0.9213925277751396, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 50927 + }, + { + "epoch": 0.50928, + "grad_norm": 0.9706304695617062, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 50928 + }, + { + "epoch": 0.50929, + "grad_norm": 0.9900774865034085, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 50929 + }, + { + "epoch": 0.5093, + "grad_norm": 0.9533272455612569, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 50930 + }, + { + "epoch": 0.50931, + "grad_norm": 0.8244541502856376, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 50931 + }, + { + "epoch": 0.50932, + "grad_norm": 0.793277178741254, + "learning_rate": 0.003, + "loss": 3.977, + "step": 50932 + }, + { + "epoch": 0.50933, + "grad_norm": 0.6915787350068058, + "learning_rate": 0.003, + "loss": 3.99, + "step": 50933 + }, + { + "epoch": 0.50934, + "grad_norm": 0.6236502415598715, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 50934 + }, + { + "epoch": 0.50935, + "grad_norm": 0.6246924626499415, + "learning_rate": 0.003, + "loss": 3.989, + "step": 50935 + }, + { + "epoch": 0.50936, + "grad_norm": 0.5906772494781143, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 50936 + }, + { + "epoch": 0.50937, + "grad_norm": 0.5376446771051162, + "learning_rate": 0.003, + "loss": 3.979, + "step": 50937 + }, + { + "epoch": 0.50938, + "grad_norm": 0.45150862507254, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 50938 + }, + { + "epoch": 0.50939, + "grad_norm": 0.47597031479422836, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 50939 + }, + { + "epoch": 0.5094, + "grad_norm": 0.5278043986803703, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 50940 + }, + { + "epoch": 0.50941, + "grad_norm": 0.5674943322734983, + "learning_rate": 0.003, + "loss": 3.971, + "step": 50941 + }, + { + "epoch": 0.50942, + "grad_norm": 0.6091369962397764, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 50942 + }, + { + "epoch": 0.50943, + "grad_norm": 0.6898618055882108, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 50943 + }, + { + "epoch": 0.50944, + "grad_norm": 0.6337246988533853, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 50944 + }, + { + "epoch": 0.50945, + "grad_norm": 0.5461157340068095, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 50945 + }, + { + "epoch": 0.50946, + "grad_norm": 0.4909622088223379, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 50946 + }, + { + "epoch": 0.50947, + "grad_norm": 0.4931251943291577, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 50947 + }, + { + "epoch": 0.50948, + "grad_norm": 0.5176799133005295, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 50948 + }, + { + "epoch": 0.50949, + "grad_norm": 0.5789327738370325, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 50949 + }, + { + "epoch": 0.5095, + "grad_norm": 0.6448467472884105, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 50950 + }, + { + "epoch": 0.50951, + "grad_norm": 0.7129550864579425, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 50951 + }, + { + "epoch": 0.50952, + "grad_norm": 0.8450722828792117, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 50952 + }, + { + "epoch": 0.50953, + "grad_norm": 1.0292559677389934, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 50953 + }, + { + "epoch": 0.50954, + "grad_norm": 0.9702375435366356, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 50954 + }, + { + "epoch": 0.50955, + "grad_norm": 0.9899269798862583, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 50955 + }, + { + "epoch": 0.50956, + "grad_norm": 1.0428677812112852, + "learning_rate": 0.003, + "loss": 4.0, + "step": 50956 + }, + { + "epoch": 0.50957, + "grad_norm": 0.9252215602246219, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 50957 + }, + { + "epoch": 0.50958, + "grad_norm": 1.0271580660938378, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 50958 + }, + { + "epoch": 0.50959, + "grad_norm": 1.0256631244538377, + "learning_rate": 0.003, + "loss": 3.982, + "step": 50959 + }, + { + "epoch": 0.5096, + "grad_norm": 0.9544536195847511, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 50960 + }, + { + "epoch": 0.50961, + "grad_norm": 1.1271991765078968, + "learning_rate": 0.003, + "loss": 4.041, + "step": 50961 + }, + { + "epoch": 0.50962, + "grad_norm": 0.9222070336581085, + "learning_rate": 0.003, + "loss": 3.999, + "step": 50962 + }, + { + "epoch": 0.50963, + "grad_norm": 0.8523824970254625, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 50963 + }, + { + "epoch": 0.50964, + "grad_norm": 0.9115575495287759, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 50964 + }, + { + "epoch": 0.50965, + "grad_norm": 1.0841906179478378, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 50965 + }, + { + "epoch": 0.50966, + "grad_norm": 0.8991481825701283, + "learning_rate": 0.003, + "loss": 4.0215, + "step": 50966 + }, + { + "epoch": 0.50967, + "grad_norm": 0.776578080302532, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 50967 + }, + { + "epoch": 0.50968, + "grad_norm": 0.6622695568189899, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 50968 + }, + { + "epoch": 0.50969, + "grad_norm": 0.68228798766324, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 50969 + }, + { + "epoch": 0.5097, + "grad_norm": 0.7151932158767678, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 50970 + }, + { + "epoch": 0.50971, + "grad_norm": 0.6891139643727895, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 50971 + }, + { + "epoch": 0.50972, + "grad_norm": 0.6162668375337456, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 50972 + }, + { + "epoch": 0.50973, + "grad_norm": 0.5807875786220507, + "learning_rate": 0.003, + "loss": 3.992, + "step": 50973 + }, + { + "epoch": 0.50974, + "grad_norm": 0.629908204373315, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 50974 + }, + { + "epoch": 0.50975, + "grad_norm": 0.652087094255046, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 50975 + }, + { + "epoch": 0.50976, + "grad_norm": 0.6990485185107196, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 50976 + }, + { + "epoch": 0.50977, + "grad_norm": 0.6671531064075434, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 50977 + }, + { + "epoch": 0.50978, + "grad_norm": 0.6935950383037333, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 50978 + }, + { + "epoch": 0.50979, + "grad_norm": 0.6291577100935615, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 50979 + }, + { + "epoch": 0.5098, + "grad_norm": 0.570901515222065, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 50980 + }, + { + "epoch": 0.50981, + "grad_norm": 0.6204452500819182, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 50981 + }, + { + "epoch": 0.50982, + "grad_norm": 0.7186319287474234, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 50982 + }, + { + "epoch": 0.50983, + "grad_norm": 0.7030927993038141, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 50983 + }, + { + "epoch": 0.50984, + "grad_norm": 0.7541975736092715, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 50984 + }, + { + "epoch": 0.50985, + "grad_norm": 0.8218975125610833, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 50985 + }, + { + "epoch": 0.50986, + "grad_norm": 0.8595280166953181, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 50986 + }, + { + "epoch": 0.50987, + "grad_norm": 0.9347820000621355, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 50987 + }, + { + "epoch": 0.50988, + "grad_norm": 0.9513063248052632, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 50988 + }, + { + "epoch": 0.50989, + "grad_norm": 1.0387508351279167, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 50989 + }, + { + "epoch": 0.5099, + "grad_norm": 1.0590458187375853, + "learning_rate": 0.003, + "loss": 3.949, + "step": 50990 + }, + { + "epoch": 0.50991, + "grad_norm": 0.8544198992803815, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 50991 + }, + { + "epoch": 0.50992, + "grad_norm": 0.797239098252571, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 50992 + }, + { + "epoch": 0.50993, + "grad_norm": 0.7509495897939041, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 50993 + }, + { + "epoch": 0.50994, + "grad_norm": 0.714196757198261, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 50994 + }, + { + "epoch": 0.50995, + "grad_norm": 0.7949384224723542, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 50995 + }, + { + "epoch": 0.50996, + "grad_norm": 0.8840142087358716, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 50996 + }, + { + "epoch": 0.50997, + "grad_norm": 1.0919427075167984, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 50997 + }, + { + "epoch": 0.50998, + "grad_norm": 0.9258576068753334, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 50998 + }, + { + "epoch": 0.50999, + "grad_norm": 0.8144514419627328, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 50999 + }, + { + "epoch": 0.51, + "grad_norm": 0.7247404670424312, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 51000 + }, + { + "epoch": 0.51001, + "grad_norm": 0.6771644380548416, + "learning_rate": 0.003, + "loss": 4.0241, + "step": 51001 + }, + { + "epoch": 0.51002, + "grad_norm": 0.5237496676341588, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 51002 + }, + { + "epoch": 0.51003, + "grad_norm": 0.5097525656146157, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 51003 + }, + { + "epoch": 0.51004, + "grad_norm": 0.47960103300656987, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 51004 + }, + { + "epoch": 0.51005, + "grad_norm": 0.42835059700353, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 51005 + }, + { + "epoch": 0.51006, + "grad_norm": 0.48036936742697084, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 51006 + }, + { + "epoch": 0.51007, + "grad_norm": 0.5725950421945806, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 51007 + }, + { + "epoch": 0.51008, + "grad_norm": 0.6802502509577513, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 51008 + }, + { + "epoch": 0.51009, + "grad_norm": 0.7326903989804967, + "learning_rate": 0.003, + "loss": 3.964, + "step": 51009 + }, + { + "epoch": 0.5101, + "grad_norm": 0.7585414483241699, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 51010 + }, + { + "epoch": 0.51011, + "grad_norm": 0.7422473734972208, + "learning_rate": 0.003, + "loss": 3.962, + "step": 51011 + }, + { + "epoch": 0.51012, + "grad_norm": 0.7107488987035968, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 51012 + }, + { + "epoch": 0.51013, + "grad_norm": 0.7859559072403162, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 51013 + }, + { + "epoch": 0.51014, + "grad_norm": 0.8085160133032256, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 51014 + }, + { + "epoch": 0.51015, + "grad_norm": 0.7668605253931061, + "learning_rate": 0.003, + "loss": 3.9753, + "step": 51015 + }, + { + "epoch": 0.51016, + "grad_norm": 0.7743133505168076, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 51016 + }, + { + "epoch": 0.51017, + "grad_norm": 0.8536971304182335, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 51017 + }, + { + "epoch": 0.51018, + "grad_norm": 0.8763901048525077, + "learning_rate": 0.003, + "loss": 3.972, + "step": 51018 + }, + { + "epoch": 0.51019, + "grad_norm": 0.9011318606744475, + "learning_rate": 0.003, + "loss": 3.942, + "step": 51019 + }, + { + "epoch": 0.5102, + "grad_norm": 0.9243066407091923, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 51020 + }, + { + "epoch": 0.51021, + "grad_norm": 0.9108749494252203, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 51021 + }, + { + "epoch": 0.51022, + "grad_norm": 0.8443614368132032, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 51022 + }, + { + "epoch": 0.51023, + "grad_norm": 0.8107086361106145, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 51023 + }, + { + "epoch": 0.51024, + "grad_norm": 0.6822163440747359, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 51024 + }, + { + "epoch": 0.51025, + "grad_norm": 0.6831360222031061, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 51025 + }, + { + "epoch": 0.51026, + "grad_norm": 0.735606854269466, + "learning_rate": 0.003, + "loss": 4.0026, + "step": 51026 + }, + { + "epoch": 0.51027, + "grad_norm": 0.7506403226999453, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 51027 + }, + { + "epoch": 0.51028, + "grad_norm": 0.8156982785912424, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 51028 + }, + { + "epoch": 0.51029, + "grad_norm": 0.7079585770890064, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 51029 + }, + { + "epoch": 0.5103, + "grad_norm": 0.6809677266622284, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 51030 + }, + { + "epoch": 0.51031, + "grad_norm": 0.696853880582485, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 51031 + }, + { + "epoch": 0.51032, + "grad_norm": 0.7832951754880426, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 51032 + }, + { + "epoch": 0.51033, + "grad_norm": 0.8533632358805775, + "learning_rate": 0.003, + "loss": 4.0, + "step": 51033 + }, + { + "epoch": 0.51034, + "grad_norm": 0.8557607832430804, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 51034 + }, + { + "epoch": 0.51035, + "grad_norm": 0.9704910438723988, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 51035 + }, + { + "epoch": 0.51036, + "grad_norm": 0.9964257303476927, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 51036 + }, + { + "epoch": 0.51037, + "grad_norm": 0.9819742465633738, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 51037 + }, + { + "epoch": 0.51038, + "grad_norm": 1.1217812675343888, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 51038 + }, + { + "epoch": 0.51039, + "grad_norm": 0.8643312786312314, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 51039 + }, + { + "epoch": 0.5104, + "grad_norm": 0.7906271162519403, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 51040 + }, + { + "epoch": 0.51041, + "grad_norm": 0.7160316500302555, + "learning_rate": 0.003, + "loss": 3.9629, + "step": 51041 + }, + { + "epoch": 0.51042, + "grad_norm": 0.6305085574860003, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 51042 + }, + { + "epoch": 0.51043, + "grad_norm": 0.6424887856347529, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 51043 + }, + { + "epoch": 0.51044, + "grad_norm": 0.6001411768697175, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 51044 + }, + { + "epoch": 0.51045, + "grad_norm": 0.5994479321887232, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 51045 + }, + { + "epoch": 0.51046, + "grad_norm": 0.6839814629874111, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 51046 + }, + { + "epoch": 0.51047, + "grad_norm": 0.783776642073756, + "learning_rate": 0.003, + "loss": 3.978, + "step": 51047 + }, + { + "epoch": 0.51048, + "grad_norm": 0.9545151907911723, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 51048 + }, + { + "epoch": 0.51049, + "grad_norm": 1.1034849169758958, + "learning_rate": 0.003, + "loss": 4.0242, + "step": 51049 + }, + { + "epoch": 0.5105, + "grad_norm": 0.9161020735807008, + "learning_rate": 0.003, + "loss": 3.9823, + "step": 51050 + }, + { + "epoch": 0.51051, + "grad_norm": 0.8273862787447234, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 51051 + }, + { + "epoch": 0.51052, + "grad_norm": 0.7678657771522502, + "learning_rate": 0.003, + "loss": 3.989, + "step": 51052 + }, + { + "epoch": 0.51053, + "grad_norm": 0.7423699110581355, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 51053 + }, + { + "epoch": 0.51054, + "grad_norm": 0.7795897308928783, + "learning_rate": 0.003, + "loss": 3.97, + "step": 51054 + }, + { + "epoch": 0.51055, + "grad_norm": 0.8769412080972491, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 51055 + }, + { + "epoch": 0.51056, + "grad_norm": 0.8673332999899178, + "learning_rate": 0.003, + "loss": 3.976, + "step": 51056 + }, + { + "epoch": 0.51057, + "grad_norm": 0.8444152503107337, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 51057 + }, + { + "epoch": 0.51058, + "grad_norm": 0.8659009912832955, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 51058 + }, + { + "epoch": 0.51059, + "grad_norm": 0.8169207336875339, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 51059 + }, + { + "epoch": 0.5106, + "grad_norm": 0.8394057789283875, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 51060 + }, + { + "epoch": 0.51061, + "grad_norm": 0.786106323764004, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 51061 + }, + { + "epoch": 0.51062, + "grad_norm": 0.750118679266891, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 51062 + }, + { + "epoch": 0.51063, + "grad_norm": 0.6431233272174908, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 51063 + }, + { + "epoch": 0.51064, + "grad_norm": 0.7058502513773348, + "learning_rate": 0.003, + "loss": 3.9645, + "step": 51064 + }, + { + "epoch": 0.51065, + "grad_norm": 0.6780956939136032, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 51065 + }, + { + "epoch": 0.51066, + "grad_norm": 0.7248422504927196, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 51066 + }, + { + "epoch": 0.51067, + "grad_norm": 0.7932611752280735, + "learning_rate": 0.003, + "loss": 3.937, + "step": 51067 + }, + { + "epoch": 0.51068, + "grad_norm": 0.8087795144908138, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 51068 + }, + { + "epoch": 0.51069, + "grad_norm": 0.7564074436331083, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 51069 + }, + { + "epoch": 0.5107, + "grad_norm": 0.8125438068412775, + "learning_rate": 0.003, + "loss": 3.979, + "step": 51070 + }, + { + "epoch": 0.51071, + "grad_norm": 0.8000465859426356, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 51071 + }, + { + "epoch": 0.51072, + "grad_norm": 0.7371692717264088, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 51072 + }, + { + "epoch": 0.51073, + "grad_norm": 0.6533597333101542, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 51073 + }, + { + "epoch": 0.51074, + "grad_norm": 0.6353395008609249, + "learning_rate": 0.003, + "loss": 3.9605, + "step": 51074 + }, + { + "epoch": 0.51075, + "grad_norm": 0.6230499209065835, + "learning_rate": 0.003, + "loss": 3.959, + "step": 51075 + }, + { + "epoch": 0.51076, + "grad_norm": 0.5740807692391674, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 51076 + }, + { + "epoch": 0.51077, + "grad_norm": 0.5592948944121535, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 51077 + }, + { + "epoch": 0.51078, + "grad_norm": 0.5826045362939198, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 51078 + }, + { + "epoch": 0.51079, + "grad_norm": 0.6796450648895122, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 51079 + }, + { + "epoch": 0.5108, + "grad_norm": 0.681826872256678, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 51080 + }, + { + "epoch": 0.51081, + "grad_norm": 0.8077813955419034, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 51081 + }, + { + "epoch": 0.51082, + "grad_norm": 1.0944435852600287, + "learning_rate": 0.003, + "loss": 3.969, + "step": 51082 + }, + { + "epoch": 0.51083, + "grad_norm": 1.04510337766355, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 51083 + }, + { + "epoch": 0.51084, + "grad_norm": 0.8253485520099673, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 51084 + }, + { + "epoch": 0.51085, + "grad_norm": 0.7053862305826152, + "learning_rate": 0.003, + "loss": 3.9599, + "step": 51085 + }, + { + "epoch": 0.51086, + "grad_norm": 0.7491378079560073, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 51086 + }, + { + "epoch": 0.51087, + "grad_norm": 0.8283768204702667, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 51087 + }, + { + "epoch": 0.51088, + "grad_norm": 0.7554151281510463, + "learning_rate": 0.003, + "loss": 3.989, + "step": 51088 + }, + { + "epoch": 0.51089, + "grad_norm": 0.7468597684829792, + "learning_rate": 0.003, + "loss": 3.9772, + "step": 51089 + }, + { + "epoch": 0.5109, + "grad_norm": 0.7619495337398067, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 51090 + }, + { + "epoch": 0.51091, + "grad_norm": 0.7443797372860458, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 51091 + }, + { + "epoch": 0.51092, + "grad_norm": 0.7710241909156383, + "learning_rate": 0.003, + "loss": 3.9608, + "step": 51092 + }, + { + "epoch": 0.51093, + "grad_norm": 0.7784639919681647, + "learning_rate": 0.003, + "loss": 3.9484, + "step": 51093 + }, + { + "epoch": 0.51094, + "grad_norm": 0.8651103422032362, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 51094 + }, + { + "epoch": 0.51095, + "grad_norm": 1.007575196435331, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 51095 + }, + { + "epoch": 0.51096, + "grad_norm": 1.0675674777259396, + "learning_rate": 0.003, + "loss": 3.981, + "step": 51096 + }, + { + "epoch": 0.51097, + "grad_norm": 0.9855897095068814, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 51097 + }, + { + "epoch": 0.51098, + "grad_norm": 0.7866547598283155, + "learning_rate": 0.003, + "loss": 4.0076, + "step": 51098 + }, + { + "epoch": 0.51099, + "grad_norm": 0.6581862627022672, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 51099 + }, + { + "epoch": 0.511, + "grad_norm": 0.6244221107382689, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 51100 + }, + { + "epoch": 0.51101, + "grad_norm": 0.6929849400575864, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 51101 + }, + { + "epoch": 0.51102, + "grad_norm": 0.7988919045288653, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 51102 + }, + { + "epoch": 0.51103, + "grad_norm": 0.8761644059068336, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 51103 + }, + { + "epoch": 0.51104, + "grad_norm": 0.8793762027841395, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 51104 + }, + { + "epoch": 0.51105, + "grad_norm": 1.1006343695542642, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 51105 + }, + { + "epoch": 0.51106, + "grad_norm": 1.0111667778219122, + "learning_rate": 0.003, + "loss": 3.99, + "step": 51106 + }, + { + "epoch": 0.51107, + "grad_norm": 0.7679485275720345, + "learning_rate": 0.003, + "loss": 3.9451, + "step": 51107 + }, + { + "epoch": 0.51108, + "grad_norm": 0.6203526046649747, + "learning_rate": 0.003, + "loss": 4.0272, + "step": 51108 + }, + { + "epoch": 0.51109, + "grad_norm": 0.6145271534067426, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 51109 + }, + { + "epoch": 0.5111, + "grad_norm": 0.6484163851034991, + "learning_rate": 0.003, + "loss": 3.9521, + "step": 51110 + }, + { + "epoch": 0.51111, + "grad_norm": 0.5847718047888363, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 51111 + }, + { + "epoch": 0.51112, + "grad_norm": 0.5997452034778653, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 51112 + }, + { + "epoch": 0.51113, + "grad_norm": 0.5540049237447738, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 51113 + }, + { + "epoch": 0.51114, + "grad_norm": 0.5075574795485055, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 51114 + }, + { + "epoch": 0.51115, + "grad_norm": 0.5135986969915565, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 51115 + }, + { + "epoch": 0.51116, + "grad_norm": 0.5848668009199469, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 51116 + }, + { + "epoch": 0.51117, + "grad_norm": 0.6661391483699192, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 51117 + }, + { + "epoch": 0.51118, + "grad_norm": 0.7466050114718514, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 51118 + }, + { + "epoch": 0.51119, + "grad_norm": 0.8759716759918693, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 51119 + }, + { + "epoch": 0.5112, + "grad_norm": 0.9978656864743067, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 51120 + }, + { + "epoch": 0.51121, + "grad_norm": 0.994049210857807, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 51121 + }, + { + "epoch": 0.51122, + "grad_norm": 0.9808627369607159, + "learning_rate": 0.003, + "loss": 3.975, + "step": 51122 + }, + { + "epoch": 0.51123, + "grad_norm": 0.9099409345104155, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 51123 + }, + { + "epoch": 0.51124, + "grad_norm": 0.8383589864131771, + "learning_rate": 0.003, + "loss": 4.0125, + "step": 51124 + }, + { + "epoch": 0.51125, + "grad_norm": 0.8291255528576024, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 51125 + }, + { + "epoch": 0.51126, + "grad_norm": 0.7911928312666938, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 51126 + }, + { + "epoch": 0.51127, + "grad_norm": 0.8699763967802729, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 51127 + }, + { + "epoch": 0.51128, + "grad_norm": 0.8362354318102615, + "learning_rate": 0.003, + "loss": 3.9988, + "step": 51128 + }, + { + "epoch": 0.51129, + "grad_norm": 0.6937453909829462, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 51129 + }, + { + "epoch": 0.5113, + "grad_norm": 0.6494073963047838, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 51130 + }, + { + "epoch": 0.51131, + "grad_norm": 0.7400895077417821, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 51131 + }, + { + "epoch": 0.51132, + "grad_norm": 0.8194082557974794, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 51132 + }, + { + "epoch": 0.51133, + "grad_norm": 0.8125102710173098, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 51133 + }, + { + "epoch": 0.51134, + "grad_norm": 0.8169201959872572, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 51134 + }, + { + "epoch": 0.51135, + "grad_norm": 0.9435424053723743, + "learning_rate": 0.003, + "loss": 3.983, + "step": 51135 + }, + { + "epoch": 0.51136, + "grad_norm": 0.8694226926269472, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 51136 + }, + { + "epoch": 0.51137, + "grad_norm": 0.9373876566848105, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 51137 + }, + { + "epoch": 0.51138, + "grad_norm": 1.0681774754217517, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 51138 + }, + { + "epoch": 0.51139, + "grad_norm": 0.9889460981192755, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 51139 + }, + { + "epoch": 0.5114, + "grad_norm": 0.8391611757307199, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 51140 + }, + { + "epoch": 0.51141, + "grad_norm": 0.949719999398525, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 51141 + }, + { + "epoch": 0.51142, + "grad_norm": 1.0908231216477562, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 51142 + }, + { + "epoch": 0.51143, + "grad_norm": 1.028713584744525, + "learning_rate": 0.003, + "loss": 4.0233, + "step": 51143 + }, + { + "epoch": 0.51144, + "grad_norm": 1.0542081609433385, + "learning_rate": 0.003, + "loss": 4.015, + "step": 51144 + }, + { + "epoch": 0.51145, + "grad_norm": 0.8712683900239601, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 51145 + }, + { + "epoch": 0.51146, + "grad_norm": 0.8263594696305395, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 51146 + }, + { + "epoch": 0.51147, + "grad_norm": 0.8594281948604129, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 51147 + }, + { + "epoch": 0.51148, + "grad_norm": 0.7766778208274892, + "learning_rate": 0.003, + "loss": 4.0239, + "step": 51148 + }, + { + "epoch": 0.51149, + "grad_norm": 0.7823969608770532, + "learning_rate": 0.003, + "loss": 4.007, + "step": 51149 + }, + { + "epoch": 0.5115, + "grad_norm": 0.8348795330431766, + "learning_rate": 0.003, + "loss": 4.004, + "step": 51150 + }, + { + "epoch": 0.51151, + "grad_norm": 0.7383239824192851, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 51151 + }, + { + "epoch": 0.51152, + "grad_norm": 0.6474617200391453, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 51152 + }, + { + "epoch": 0.51153, + "grad_norm": 0.5656245139216246, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 51153 + }, + { + "epoch": 0.51154, + "grad_norm": 0.5484515448767082, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 51154 + }, + { + "epoch": 0.51155, + "grad_norm": 0.4968404018164355, + "learning_rate": 0.003, + "loss": 4.0246, + "step": 51155 + }, + { + "epoch": 0.51156, + "grad_norm": 0.5731980499073437, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 51156 + }, + { + "epoch": 0.51157, + "grad_norm": 0.5422099686369801, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 51157 + }, + { + "epoch": 0.51158, + "grad_norm": 0.5160932043807934, + "learning_rate": 0.003, + "loss": 3.9625, + "step": 51158 + }, + { + "epoch": 0.51159, + "grad_norm": 0.49316916179702597, + "learning_rate": 0.003, + "loss": 3.958, + "step": 51159 + }, + { + "epoch": 0.5116, + "grad_norm": 0.5540709720187007, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 51160 + }, + { + "epoch": 0.51161, + "grad_norm": 0.5653226480866641, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 51161 + }, + { + "epoch": 0.51162, + "grad_norm": 0.6380274533098337, + "learning_rate": 0.003, + "loss": 3.9665, + "step": 51162 + }, + { + "epoch": 0.51163, + "grad_norm": 0.609482194232398, + "learning_rate": 0.003, + "loss": 3.9557, + "step": 51163 + }, + { + "epoch": 0.51164, + "grad_norm": 0.6021487026866279, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 51164 + }, + { + "epoch": 0.51165, + "grad_norm": 0.735448825466734, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 51165 + }, + { + "epoch": 0.51166, + "grad_norm": 0.9078134476868968, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 51166 + }, + { + "epoch": 0.51167, + "grad_norm": 0.9639545649239203, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 51167 + }, + { + "epoch": 0.51168, + "grad_norm": 0.9409894829869953, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 51168 + }, + { + "epoch": 0.51169, + "grad_norm": 0.9137938591902642, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 51169 + }, + { + "epoch": 0.5117, + "grad_norm": 0.8515578275747648, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 51170 + }, + { + "epoch": 0.51171, + "grad_norm": 0.7140357627387922, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 51171 + }, + { + "epoch": 0.51172, + "grad_norm": 0.886632600471122, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 51172 + }, + { + "epoch": 0.51173, + "grad_norm": 0.8399678429749755, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 51173 + }, + { + "epoch": 0.51174, + "grad_norm": 0.816315688606813, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 51174 + }, + { + "epoch": 0.51175, + "grad_norm": 0.7158301914936177, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 51175 + }, + { + "epoch": 0.51176, + "grad_norm": 0.7341564712791022, + "learning_rate": 0.003, + "loss": 3.996, + "step": 51176 + }, + { + "epoch": 0.51177, + "grad_norm": 0.648178544089813, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 51177 + }, + { + "epoch": 0.51178, + "grad_norm": 0.7007139821549084, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 51178 + }, + { + "epoch": 0.51179, + "grad_norm": 0.6758370133989184, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 51179 + }, + { + "epoch": 0.5118, + "grad_norm": 0.6207495116796026, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 51180 + }, + { + "epoch": 0.51181, + "grad_norm": 0.6573308909842972, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 51181 + }, + { + "epoch": 0.51182, + "grad_norm": 0.6179257218124556, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 51182 + }, + { + "epoch": 0.51183, + "grad_norm": 0.5889601320127967, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 51183 + }, + { + "epoch": 0.51184, + "grad_norm": 0.622453807435086, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 51184 + }, + { + "epoch": 0.51185, + "grad_norm": 0.6707941658387004, + "learning_rate": 0.003, + "loss": 3.9673, + "step": 51185 + }, + { + "epoch": 0.51186, + "grad_norm": 0.7497846852888697, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 51186 + }, + { + "epoch": 0.51187, + "grad_norm": 0.713734370230936, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 51187 + }, + { + "epoch": 0.51188, + "grad_norm": 0.7180049318362015, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 51188 + }, + { + "epoch": 0.51189, + "grad_norm": 0.78960028510014, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 51189 + }, + { + "epoch": 0.5119, + "grad_norm": 0.9442759181353435, + "learning_rate": 0.003, + "loss": 3.9282, + "step": 51190 + }, + { + "epoch": 0.51191, + "grad_norm": 1.1937477833753711, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 51191 + }, + { + "epoch": 0.51192, + "grad_norm": 0.8460884347049096, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 51192 + }, + { + "epoch": 0.51193, + "grad_norm": 0.8364119705279074, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 51193 + }, + { + "epoch": 0.51194, + "grad_norm": 0.7048975947239672, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 51194 + }, + { + "epoch": 0.51195, + "grad_norm": 0.6455887669261463, + "learning_rate": 0.003, + "loss": 4.005, + "step": 51195 + }, + { + "epoch": 0.51196, + "grad_norm": 0.6340892400035995, + "learning_rate": 0.003, + "loss": 3.945, + "step": 51196 + }, + { + "epoch": 0.51197, + "grad_norm": 0.6622783580774465, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 51197 + }, + { + "epoch": 0.51198, + "grad_norm": 0.8202270140074047, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 51198 + }, + { + "epoch": 0.51199, + "grad_norm": 0.9292076878577303, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 51199 + }, + { + "epoch": 0.512, + "grad_norm": 0.8191785272194393, + "learning_rate": 0.003, + "loss": 3.9938, + "step": 51200 + }, + { + "epoch": 0.51201, + "grad_norm": 0.7260022124163119, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 51201 + }, + { + "epoch": 0.51202, + "grad_norm": 0.7574831824069646, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 51202 + }, + { + "epoch": 0.51203, + "grad_norm": 0.9661324755644775, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 51203 + }, + { + "epoch": 0.51204, + "grad_norm": 1.1115502740913734, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 51204 + }, + { + "epoch": 0.51205, + "grad_norm": 0.7986225093819062, + "learning_rate": 0.003, + "loss": 3.9921, + "step": 51205 + }, + { + "epoch": 0.51206, + "grad_norm": 0.7797382031312199, + "learning_rate": 0.003, + "loss": 3.9472, + "step": 51206 + }, + { + "epoch": 0.51207, + "grad_norm": 0.7174313854095893, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 51207 + }, + { + "epoch": 0.51208, + "grad_norm": 0.7379403986486809, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 51208 + }, + { + "epoch": 0.51209, + "grad_norm": 0.7054925234097352, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 51209 + }, + { + "epoch": 0.5121, + "grad_norm": 0.6907589680466856, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 51210 + }, + { + "epoch": 0.51211, + "grad_norm": 0.702392650338888, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 51211 + }, + { + "epoch": 0.51212, + "grad_norm": 0.7606062979866507, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 51212 + }, + { + "epoch": 0.51213, + "grad_norm": 0.6839213223524991, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 51213 + }, + { + "epoch": 0.51214, + "grad_norm": 0.6587838984393273, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 51214 + }, + { + "epoch": 0.51215, + "grad_norm": 0.7304141846111003, + "learning_rate": 0.003, + "loss": 3.9456, + "step": 51215 + }, + { + "epoch": 0.51216, + "grad_norm": 0.7357982945231959, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 51216 + }, + { + "epoch": 0.51217, + "grad_norm": 0.6873970199198691, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 51217 + }, + { + "epoch": 0.51218, + "grad_norm": 0.5847025767762808, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 51218 + }, + { + "epoch": 0.51219, + "grad_norm": 0.6870457640684734, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 51219 + }, + { + "epoch": 0.5122, + "grad_norm": 0.7615479467710454, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 51220 + }, + { + "epoch": 0.51221, + "grad_norm": 0.8232141361787892, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 51221 + }, + { + "epoch": 0.51222, + "grad_norm": 0.9615636531205658, + "learning_rate": 0.003, + "loss": 3.953, + "step": 51222 + }, + { + "epoch": 0.51223, + "grad_norm": 1.0580312505929954, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 51223 + }, + { + "epoch": 0.51224, + "grad_norm": 0.8007125295725059, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 51224 + }, + { + "epoch": 0.51225, + "grad_norm": 0.7706769546610861, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 51225 + }, + { + "epoch": 0.51226, + "grad_norm": 0.8180868756871588, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 51226 + }, + { + "epoch": 0.51227, + "grad_norm": 0.9194933674995415, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 51227 + }, + { + "epoch": 0.51228, + "grad_norm": 0.970663411926486, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 51228 + }, + { + "epoch": 0.51229, + "grad_norm": 0.8825365186998588, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 51229 + }, + { + "epoch": 0.5123, + "grad_norm": 0.7710431307777839, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 51230 + }, + { + "epoch": 0.51231, + "grad_norm": 0.728612965018188, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 51231 + }, + { + "epoch": 0.51232, + "grad_norm": 0.712205359302459, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 51232 + }, + { + "epoch": 0.51233, + "grad_norm": 0.6987798411072453, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 51233 + }, + { + "epoch": 0.51234, + "grad_norm": 0.7302372574729594, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 51234 + }, + { + "epoch": 0.51235, + "grad_norm": 0.7097318236276753, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 51235 + }, + { + "epoch": 0.51236, + "grad_norm": 0.692694987123007, + "learning_rate": 0.003, + "loss": 3.9542, + "step": 51236 + }, + { + "epoch": 0.51237, + "grad_norm": 0.6691931322604702, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 51237 + }, + { + "epoch": 0.51238, + "grad_norm": 0.6672269234084124, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 51238 + }, + { + "epoch": 0.51239, + "grad_norm": 0.6256883621600868, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 51239 + }, + { + "epoch": 0.5124, + "grad_norm": 0.5820334762777304, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 51240 + }, + { + "epoch": 0.51241, + "grad_norm": 0.6590430374417409, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 51241 + }, + { + "epoch": 0.51242, + "grad_norm": 0.7101775437505723, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 51242 + }, + { + "epoch": 0.51243, + "grad_norm": 0.6887625888521154, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 51243 + }, + { + "epoch": 0.51244, + "grad_norm": 0.7257042191510865, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 51244 + }, + { + "epoch": 0.51245, + "grad_norm": 0.7134307886045483, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 51245 + }, + { + "epoch": 0.51246, + "grad_norm": 0.8109082124026386, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 51246 + }, + { + "epoch": 0.51247, + "grad_norm": 0.8699533551229985, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 51247 + }, + { + "epoch": 0.51248, + "grad_norm": 0.9402979611766524, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 51248 + }, + { + "epoch": 0.51249, + "grad_norm": 1.031066902167467, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 51249 + }, + { + "epoch": 0.5125, + "grad_norm": 0.9568407833720319, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 51250 + }, + { + "epoch": 0.51251, + "grad_norm": 0.884372401023486, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 51251 + }, + { + "epoch": 0.51252, + "grad_norm": 1.047607908445142, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 51252 + }, + { + "epoch": 0.51253, + "grad_norm": 1.2186714644851864, + "learning_rate": 0.003, + "loss": 4.017, + "step": 51253 + }, + { + "epoch": 0.51254, + "grad_norm": 0.8990365199692782, + "learning_rate": 0.003, + "loss": 4.007, + "step": 51254 + }, + { + "epoch": 0.51255, + "grad_norm": 0.8380973928789709, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 51255 + }, + { + "epoch": 0.51256, + "grad_norm": 0.8217204714988121, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 51256 + }, + { + "epoch": 0.51257, + "grad_norm": 1.0203354922225492, + "learning_rate": 0.003, + "loss": 4.0331, + "step": 51257 + }, + { + "epoch": 0.51258, + "grad_norm": 1.120135194842158, + "learning_rate": 0.003, + "loss": 4.0359, + "step": 51258 + }, + { + "epoch": 0.51259, + "grad_norm": 1.0642416127879417, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 51259 + }, + { + "epoch": 0.5126, + "grad_norm": 0.8634624385752929, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 51260 + }, + { + "epoch": 0.51261, + "grad_norm": 0.760538864643842, + "learning_rate": 0.003, + "loss": 4.0044, + "step": 51261 + }, + { + "epoch": 0.51262, + "grad_norm": 0.7813413513704974, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 51262 + }, + { + "epoch": 0.51263, + "grad_norm": 0.8677016701186359, + "learning_rate": 0.003, + "loss": 4.0168, + "step": 51263 + }, + { + "epoch": 0.51264, + "grad_norm": 0.9608503019494967, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 51264 + }, + { + "epoch": 0.51265, + "grad_norm": 1.0645366087931487, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 51265 + }, + { + "epoch": 0.51266, + "grad_norm": 0.8835133780587102, + "learning_rate": 0.003, + "loss": 3.996, + "step": 51266 + }, + { + "epoch": 0.51267, + "grad_norm": 0.7482930420123717, + "learning_rate": 0.003, + "loss": 4.0003, + "step": 51267 + }, + { + "epoch": 0.51268, + "grad_norm": 0.5914868130033376, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 51268 + }, + { + "epoch": 0.51269, + "grad_norm": 0.6232025776493376, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 51269 + }, + { + "epoch": 0.5127, + "grad_norm": 0.6098756519587193, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 51270 + }, + { + "epoch": 0.51271, + "grad_norm": 0.6243007037594174, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 51271 + }, + { + "epoch": 0.51272, + "grad_norm": 0.5888468404383809, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 51272 + }, + { + "epoch": 0.51273, + "grad_norm": 0.5462678197080827, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 51273 + }, + { + "epoch": 0.51274, + "grad_norm": 0.5472753134517903, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 51274 + }, + { + "epoch": 0.51275, + "grad_norm": 0.5798705791475314, + "learning_rate": 0.003, + "loss": 3.978, + "step": 51275 + }, + { + "epoch": 0.51276, + "grad_norm": 0.6280643013210663, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 51276 + }, + { + "epoch": 0.51277, + "grad_norm": 0.558769449991999, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 51277 + }, + { + "epoch": 0.51278, + "grad_norm": 0.48480884660310475, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 51278 + }, + { + "epoch": 0.51279, + "grad_norm": 0.5163084687097538, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 51279 + }, + { + "epoch": 0.5128, + "grad_norm": 0.6350168578249716, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 51280 + }, + { + "epoch": 0.51281, + "grad_norm": 0.7879740333568472, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 51281 + }, + { + "epoch": 0.51282, + "grad_norm": 0.8657595377434871, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 51282 + }, + { + "epoch": 0.51283, + "grad_norm": 0.8479589944755954, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 51283 + }, + { + "epoch": 0.51284, + "grad_norm": 0.7353107649011448, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 51284 + }, + { + "epoch": 0.51285, + "grad_norm": 0.704764897767817, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 51285 + }, + { + "epoch": 0.51286, + "grad_norm": 0.7504678550870852, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 51286 + }, + { + "epoch": 0.51287, + "grad_norm": 0.8774846646138675, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 51287 + }, + { + "epoch": 0.51288, + "grad_norm": 0.8854057143628749, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 51288 + }, + { + "epoch": 0.51289, + "grad_norm": 0.8552726144876802, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 51289 + }, + { + "epoch": 0.5129, + "grad_norm": 0.8838667057645675, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 51290 + }, + { + "epoch": 0.51291, + "grad_norm": 0.8834984170102971, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 51291 + }, + { + "epoch": 0.51292, + "grad_norm": 0.9379581420424998, + "learning_rate": 0.003, + "loss": 4.025, + "step": 51292 + }, + { + "epoch": 0.51293, + "grad_norm": 0.8882192020580943, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 51293 + }, + { + "epoch": 0.51294, + "grad_norm": 0.7963373542512882, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 51294 + }, + { + "epoch": 0.51295, + "grad_norm": 0.8071265676630094, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 51295 + }, + { + "epoch": 0.51296, + "grad_norm": 0.6736932024780239, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 51296 + }, + { + "epoch": 0.51297, + "grad_norm": 0.5667883898000422, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 51297 + }, + { + "epoch": 0.51298, + "grad_norm": 0.549364868884712, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 51298 + }, + { + "epoch": 0.51299, + "grad_norm": 0.5062939025989215, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 51299 + }, + { + "epoch": 0.513, + "grad_norm": 0.4863871337032384, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 51300 + }, + { + "epoch": 0.51301, + "grad_norm": 0.47438721307095244, + "learning_rate": 0.003, + "loss": 3.966, + "step": 51301 + }, + { + "epoch": 0.51302, + "grad_norm": 0.523800672216348, + "learning_rate": 0.003, + "loss": 3.9406, + "step": 51302 + }, + { + "epoch": 0.51303, + "grad_norm": 0.5829934626125672, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 51303 + }, + { + "epoch": 0.51304, + "grad_norm": 0.8521650453816608, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 51304 + }, + { + "epoch": 0.51305, + "grad_norm": 1.1838325625639676, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 51305 + }, + { + "epoch": 0.51306, + "grad_norm": 0.9482601721869319, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 51306 + }, + { + "epoch": 0.51307, + "grad_norm": 0.7978684289655642, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 51307 + }, + { + "epoch": 0.51308, + "grad_norm": 0.8341445227812795, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 51308 + }, + { + "epoch": 0.51309, + "grad_norm": 0.8920739255112304, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 51309 + }, + { + "epoch": 0.5131, + "grad_norm": 0.7652683934882132, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 51310 + }, + { + "epoch": 0.51311, + "grad_norm": 0.7476636702404638, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 51311 + }, + { + "epoch": 0.51312, + "grad_norm": 0.8976957772774243, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 51312 + }, + { + "epoch": 0.51313, + "grad_norm": 0.9916262094548144, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 51313 + }, + { + "epoch": 0.51314, + "grad_norm": 0.9339235304463029, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 51314 + }, + { + "epoch": 0.51315, + "grad_norm": 0.8583811228068174, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 51315 + }, + { + "epoch": 0.51316, + "grad_norm": 0.9388786216210565, + "learning_rate": 0.003, + "loss": 4.008, + "step": 51316 + }, + { + "epoch": 0.51317, + "grad_norm": 0.8837038087351075, + "learning_rate": 0.003, + "loss": 3.9511, + "step": 51317 + }, + { + "epoch": 0.51318, + "grad_norm": 0.8634776000496125, + "learning_rate": 0.003, + "loss": 3.9686, + "step": 51318 + }, + { + "epoch": 0.51319, + "grad_norm": 0.8938721819100478, + "learning_rate": 0.003, + "loss": 4.0175, + "step": 51319 + }, + { + "epoch": 0.5132, + "grad_norm": 0.8571647698388255, + "learning_rate": 0.003, + "loss": 3.971, + "step": 51320 + }, + { + "epoch": 0.51321, + "grad_norm": 0.8208157393240844, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 51321 + }, + { + "epoch": 0.51322, + "grad_norm": 0.8274680953105564, + "learning_rate": 0.003, + "loss": 4.0145, + "step": 51322 + }, + { + "epoch": 0.51323, + "grad_norm": 0.8469721251947816, + "learning_rate": 0.003, + "loss": 3.968, + "step": 51323 + }, + { + "epoch": 0.51324, + "grad_norm": 0.7650649115116589, + "learning_rate": 0.003, + "loss": 4.0248, + "step": 51324 + }, + { + "epoch": 0.51325, + "grad_norm": 0.8900447278613528, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 51325 + }, + { + "epoch": 0.51326, + "grad_norm": 1.0381592928889916, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 51326 + }, + { + "epoch": 0.51327, + "grad_norm": 1.0459813297981404, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 51327 + }, + { + "epoch": 0.51328, + "grad_norm": 0.8571707756527807, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 51328 + }, + { + "epoch": 0.51329, + "grad_norm": 0.9832144807015823, + "learning_rate": 0.003, + "loss": 4.0031, + "step": 51329 + }, + { + "epoch": 0.5133, + "grad_norm": 1.2328825601535254, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 51330 + }, + { + "epoch": 0.51331, + "grad_norm": 0.8238230618458109, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 51331 + }, + { + "epoch": 0.51332, + "grad_norm": 0.667757797593433, + "learning_rate": 0.003, + "loss": 4.0199, + "step": 51332 + }, + { + "epoch": 0.51333, + "grad_norm": 0.7249104154839368, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 51333 + }, + { + "epoch": 0.51334, + "grad_norm": 0.668378165431279, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 51334 + }, + { + "epoch": 0.51335, + "grad_norm": 0.6987144480488083, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 51335 + }, + { + "epoch": 0.51336, + "grad_norm": 0.5704333607792142, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 51336 + }, + { + "epoch": 0.51337, + "grad_norm": 0.5025727985754206, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 51337 + }, + { + "epoch": 0.51338, + "grad_norm": 0.4998098400221593, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 51338 + }, + { + "epoch": 0.51339, + "grad_norm": 0.5916416867204707, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 51339 + }, + { + "epoch": 0.5134, + "grad_norm": 0.542275711015676, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 51340 + }, + { + "epoch": 0.51341, + "grad_norm": 0.5661870602131323, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 51341 + }, + { + "epoch": 0.51342, + "grad_norm": 0.5837643004722817, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 51342 + }, + { + "epoch": 0.51343, + "grad_norm": 0.5595402674396703, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 51343 + }, + { + "epoch": 0.51344, + "grad_norm": 0.6351742111348845, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 51344 + }, + { + "epoch": 0.51345, + "grad_norm": 0.6800374199098392, + "learning_rate": 0.003, + "loss": 3.992, + "step": 51345 + }, + { + "epoch": 0.51346, + "grad_norm": 0.5776490210540685, + "learning_rate": 0.003, + "loss": 4.0075, + "step": 51346 + }, + { + "epoch": 0.51347, + "grad_norm": 0.5552762986806864, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 51347 + }, + { + "epoch": 0.51348, + "grad_norm": 0.5790531589143755, + "learning_rate": 0.003, + "loss": 3.9601, + "step": 51348 + }, + { + "epoch": 0.51349, + "grad_norm": 0.584395263114656, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 51349 + }, + { + "epoch": 0.5135, + "grad_norm": 0.6128857272960576, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 51350 + }, + { + "epoch": 0.51351, + "grad_norm": 0.6548069350858272, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 51351 + }, + { + "epoch": 0.51352, + "grad_norm": 0.5880891442758466, + "learning_rate": 0.003, + "loss": 3.9515, + "step": 51352 + }, + { + "epoch": 0.51353, + "grad_norm": 0.6267073238156855, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 51353 + }, + { + "epoch": 0.51354, + "grad_norm": 0.8702955572066937, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 51354 + }, + { + "epoch": 0.51355, + "grad_norm": 0.9951484563457628, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 51355 + }, + { + "epoch": 0.51356, + "grad_norm": 1.0060276259389338, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 51356 + }, + { + "epoch": 0.51357, + "grad_norm": 0.8984006315884837, + "learning_rate": 0.003, + "loss": 3.957, + "step": 51357 + }, + { + "epoch": 0.51358, + "grad_norm": 0.9508198473289183, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 51358 + }, + { + "epoch": 0.51359, + "grad_norm": 1.0085527222755581, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 51359 + }, + { + "epoch": 0.5136, + "grad_norm": 0.8561805540305685, + "learning_rate": 0.003, + "loss": 4.0109, + "step": 51360 + }, + { + "epoch": 0.51361, + "grad_norm": 0.8836817530470159, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 51361 + }, + { + "epoch": 0.51362, + "grad_norm": 0.865961191270125, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 51362 + }, + { + "epoch": 0.51363, + "grad_norm": 0.862367126014425, + "learning_rate": 0.003, + "loss": 4.0151, + "step": 51363 + }, + { + "epoch": 0.51364, + "grad_norm": 0.8516508623142295, + "learning_rate": 0.003, + "loss": 4.0035, + "step": 51364 + }, + { + "epoch": 0.51365, + "grad_norm": 0.8563281201812327, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 51365 + }, + { + "epoch": 0.51366, + "grad_norm": 0.7954967811057899, + "learning_rate": 0.003, + "loss": 4.0167, + "step": 51366 + }, + { + "epoch": 0.51367, + "grad_norm": 0.9166745491438353, + "learning_rate": 0.003, + "loss": 4.0098, + "step": 51367 + }, + { + "epoch": 0.51368, + "grad_norm": 1.0669574094535663, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 51368 + }, + { + "epoch": 0.51369, + "grad_norm": 1.0179964800453893, + "learning_rate": 0.003, + "loss": 4.0134, + "step": 51369 + }, + { + "epoch": 0.5137, + "grad_norm": 0.9465512400365833, + "learning_rate": 0.003, + "loss": 3.9937, + "step": 51370 + }, + { + "epoch": 0.51371, + "grad_norm": 0.9061939385906327, + "learning_rate": 0.003, + "loss": 3.983, + "step": 51371 + }, + { + "epoch": 0.51372, + "grad_norm": 0.8589731976437459, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 51372 + }, + { + "epoch": 0.51373, + "grad_norm": 0.744881584211614, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 51373 + }, + { + "epoch": 0.51374, + "grad_norm": 0.7354241235825326, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 51374 + }, + { + "epoch": 0.51375, + "grad_norm": 0.7237324328534003, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 51375 + }, + { + "epoch": 0.51376, + "grad_norm": 0.6993288125698046, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 51376 + }, + { + "epoch": 0.51377, + "grad_norm": 0.702789654604794, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 51377 + }, + { + "epoch": 0.51378, + "grad_norm": 0.7170307180178447, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 51378 + }, + { + "epoch": 0.51379, + "grad_norm": 0.6537686297595766, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 51379 + }, + { + "epoch": 0.5138, + "grad_norm": 0.6793241152723334, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 51380 + }, + { + "epoch": 0.51381, + "grad_norm": 0.6542455932760473, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 51381 + }, + { + "epoch": 0.51382, + "grad_norm": 0.6796539951520975, + "learning_rate": 0.003, + "loss": 3.994, + "step": 51382 + }, + { + "epoch": 0.51383, + "grad_norm": 0.6457533629731215, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 51383 + }, + { + "epoch": 0.51384, + "grad_norm": 0.683169576809019, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 51384 + }, + { + "epoch": 0.51385, + "grad_norm": 0.7232486054433879, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 51385 + }, + { + "epoch": 0.51386, + "grad_norm": 0.8243862563699426, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 51386 + }, + { + "epoch": 0.51387, + "grad_norm": 0.9351863741317278, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 51387 + }, + { + "epoch": 0.51388, + "grad_norm": 0.9317763859757953, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 51388 + }, + { + "epoch": 0.51389, + "grad_norm": 0.9090127685904373, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 51389 + }, + { + "epoch": 0.5139, + "grad_norm": 0.7231463089430227, + "learning_rate": 0.003, + "loss": 3.9567, + "step": 51390 + }, + { + "epoch": 0.51391, + "grad_norm": 0.7488494917532033, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 51391 + }, + { + "epoch": 0.51392, + "grad_norm": 0.6557123021698624, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 51392 + }, + { + "epoch": 0.51393, + "grad_norm": 0.5611207235771698, + "learning_rate": 0.003, + "loss": 3.996, + "step": 51393 + }, + { + "epoch": 0.51394, + "grad_norm": 0.5330844813699024, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 51394 + }, + { + "epoch": 0.51395, + "grad_norm": 0.5270412472137722, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 51395 + }, + { + "epoch": 0.51396, + "grad_norm": 0.5261937488778657, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 51396 + }, + { + "epoch": 0.51397, + "grad_norm": 0.528400790122094, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 51397 + }, + { + "epoch": 0.51398, + "grad_norm": 0.6883164127143799, + "learning_rate": 0.003, + "loss": 3.964, + "step": 51398 + }, + { + "epoch": 0.51399, + "grad_norm": 0.9338252685112025, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 51399 + }, + { + "epoch": 0.514, + "grad_norm": 1.1701418327356559, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 51400 + }, + { + "epoch": 0.51401, + "grad_norm": 0.6041262690898671, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 51401 + }, + { + "epoch": 0.51402, + "grad_norm": 0.6002838180775787, + "learning_rate": 0.003, + "loss": 3.9681, + "step": 51402 + }, + { + "epoch": 0.51403, + "grad_norm": 0.7706504054362766, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 51403 + }, + { + "epoch": 0.51404, + "grad_norm": 0.8323052945989228, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 51404 + }, + { + "epoch": 0.51405, + "grad_norm": 0.8115561735199839, + "learning_rate": 0.003, + "loss": 3.9579, + "step": 51405 + }, + { + "epoch": 0.51406, + "grad_norm": 0.7648812737623187, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 51406 + }, + { + "epoch": 0.51407, + "grad_norm": 0.7828543254808359, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 51407 + }, + { + "epoch": 0.51408, + "grad_norm": 0.7993968715000109, + "learning_rate": 0.003, + "loss": 3.9504, + "step": 51408 + }, + { + "epoch": 0.51409, + "grad_norm": 0.8832519680006893, + "learning_rate": 0.003, + "loss": 3.953, + "step": 51409 + }, + { + "epoch": 0.5141, + "grad_norm": 0.8613831131872123, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 51410 + }, + { + "epoch": 0.51411, + "grad_norm": 0.8426180219244523, + "learning_rate": 0.003, + "loss": 3.9981, + "step": 51411 + }, + { + "epoch": 0.51412, + "grad_norm": 0.9456810614501102, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 51412 + }, + { + "epoch": 0.51413, + "grad_norm": 1.0172405889645586, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 51413 + }, + { + "epoch": 0.51414, + "grad_norm": 0.9193394375441191, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 51414 + }, + { + "epoch": 0.51415, + "grad_norm": 0.8651901123257891, + "learning_rate": 0.003, + "loss": 3.987, + "step": 51415 + }, + { + "epoch": 0.51416, + "grad_norm": 0.860610070528282, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 51416 + }, + { + "epoch": 0.51417, + "grad_norm": 0.8324262314119525, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 51417 + }, + { + "epoch": 0.51418, + "grad_norm": 0.9053082344731458, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 51418 + }, + { + "epoch": 0.51419, + "grad_norm": 0.9597822654292405, + "learning_rate": 0.003, + "loss": 3.9668, + "step": 51419 + }, + { + "epoch": 0.5142, + "grad_norm": 0.9293588257581181, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 51420 + }, + { + "epoch": 0.51421, + "grad_norm": 0.825362318125531, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 51421 + }, + { + "epoch": 0.51422, + "grad_norm": 0.7691157064291151, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 51422 + }, + { + "epoch": 0.51423, + "grad_norm": 0.6728469410210292, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 51423 + }, + { + "epoch": 0.51424, + "grad_norm": 0.7797707795052528, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 51424 + }, + { + "epoch": 0.51425, + "grad_norm": 0.9707566162544811, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 51425 + }, + { + "epoch": 0.51426, + "grad_norm": 0.9997139745954086, + "learning_rate": 0.003, + "loss": 3.9911, + "step": 51426 + }, + { + "epoch": 0.51427, + "grad_norm": 0.9327989320638044, + "learning_rate": 0.003, + "loss": 3.996, + "step": 51427 + }, + { + "epoch": 0.51428, + "grad_norm": 0.9126814681386842, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 51428 + }, + { + "epoch": 0.51429, + "grad_norm": 0.7369628431607037, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 51429 + }, + { + "epoch": 0.5143, + "grad_norm": 0.6578093924681091, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 51430 + }, + { + "epoch": 0.51431, + "grad_norm": 0.5999218074592704, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 51431 + }, + { + "epoch": 0.51432, + "grad_norm": 0.6325908363208993, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 51432 + }, + { + "epoch": 0.51433, + "grad_norm": 0.5231017108605102, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 51433 + }, + { + "epoch": 0.51434, + "grad_norm": 0.5167998386429182, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 51434 + }, + { + "epoch": 0.51435, + "grad_norm": 0.4726982880950176, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 51435 + }, + { + "epoch": 0.51436, + "grad_norm": 0.5180241024907303, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 51436 + }, + { + "epoch": 0.51437, + "grad_norm": 0.6916562498449187, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 51437 + }, + { + "epoch": 0.51438, + "grad_norm": 0.8065580065934781, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 51438 + }, + { + "epoch": 0.51439, + "grad_norm": 0.8317592380691806, + "learning_rate": 0.003, + "loss": 3.975, + "step": 51439 + }, + { + "epoch": 0.5144, + "grad_norm": 0.7322340610587289, + "learning_rate": 0.003, + "loss": 3.956, + "step": 51440 + }, + { + "epoch": 0.51441, + "grad_norm": 0.7867486484343831, + "learning_rate": 0.003, + "loss": 3.9992, + "step": 51441 + }, + { + "epoch": 0.51442, + "grad_norm": 0.710278667671223, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 51442 + }, + { + "epoch": 0.51443, + "grad_norm": 0.6537229247270057, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 51443 + }, + { + "epoch": 0.51444, + "grad_norm": 0.5975283834790049, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 51444 + }, + { + "epoch": 0.51445, + "grad_norm": 0.6278060893244669, + "learning_rate": 0.003, + "loss": 3.9407, + "step": 51445 + }, + { + "epoch": 0.51446, + "grad_norm": 0.670819289374955, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 51446 + }, + { + "epoch": 0.51447, + "grad_norm": 0.7571517316435863, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 51447 + }, + { + "epoch": 0.51448, + "grad_norm": 0.7983742486496777, + "learning_rate": 0.003, + "loss": 3.988, + "step": 51448 + }, + { + "epoch": 0.51449, + "grad_norm": 0.8330550653525864, + "learning_rate": 0.003, + "loss": 3.9809, + "step": 51449 + }, + { + "epoch": 0.5145, + "grad_norm": 0.8394053225235136, + "learning_rate": 0.003, + "loss": 3.9636, + "step": 51450 + }, + { + "epoch": 0.51451, + "grad_norm": 0.7194926237787976, + "learning_rate": 0.003, + "loss": 3.9501, + "step": 51451 + }, + { + "epoch": 0.51452, + "grad_norm": 0.6243900617515761, + "learning_rate": 0.003, + "loss": 3.9581, + "step": 51452 + }, + { + "epoch": 0.51453, + "grad_norm": 0.6269428870067287, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 51453 + }, + { + "epoch": 0.51454, + "grad_norm": 0.6170757265267436, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 51454 + }, + { + "epoch": 0.51455, + "grad_norm": 0.5677034955493907, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 51455 + }, + { + "epoch": 0.51456, + "grad_norm": 0.6146836687265478, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 51456 + }, + { + "epoch": 0.51457, + "grad_norm": 0.6711797087498093, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 51457 + }, + { + "epoch": 0.51458, + "grad_norm": 0.7074184051413591, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 51458 + }, + { + "epoch": 0.51459, + "grad_norm": 0.8083132429260682, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 51459 + }, + { + "epoch": 0.5146, + "grad_norm": 0.8378773415987708, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 51460 + }, + { + "epoch": 0.51461, + "grad_norm": 0.9744492580619015, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 51461 + }, + { + "epoch": 0.51462, + "grad_norm": 1.0894077285587938, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 51462 + }, + { + "epoch": 0.51463, + "grad_norm": 0.8229392089527536, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 51463 + }, + { + "epoch": 0.51464, + "grad_norm": 0.7725273136897929, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 51464 + }, + { + "epoch": 0.51465, + "grad_norm": 0.829674838823198, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 51465 + }, + { + "epoch": 0.51466, + "grad_norm": 0.9881357485885114, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 51466 + }, + { + "epoch": 0.51467, + "grad_norm": 0.9545553751092041, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 51467 + }, + { + "epoch": 0.51468, + "grad_norm": 0.8285319742535359, + "learning_rate": 0.003, + "loss": 3.992, + "step": 51468 + }, + { + "epoch": 0.51469, + "grad_norm": 0.8419519774662994, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 51469 + }, + { + "epoch": 0.5147, + "grad_norm": 0.823359009380834, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 51470 + }, + { + "epoch": 0.51471, + "grad_norm": 0.8564807164737764, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 51471 + }, + { + "epoch": 0.51472, + "grad_norm": 0.8178135704982614, + "learning_rate": 0.003, + "loss": 3.953, + "step": 51472 + }, + { + "epoch": 0.51473, + "grad_norm": 0.8271374095419761, + "learning_rate": 0.003, + "loss": 3.9897, + "step": 51473 + }, + { + "epoch": 0.51474, + "grad_norm": 0.9633161775028166, + "learning_rate": 0.003, + "loss": 4.01, + "step": 51474 + }, + { + "epoch": 0.51475, + "grad_norm": 0.9992320706406218, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 51475 + }, + { + "epoch": 0.51476, + "grad_norm": 0.9067267673236671, + "learning_rate": 0.003, + "loss": 4.0105, + "step": 51476 + }, + { + "epoch": 0.51477, + "grad_norm": 0.9266837850042109, + "learning_rate": 0.003, + "loss": 3.965, + "step": 51477 + }, + { + "epoch": 0.51478, + "grad_norm": 0.9167479127843726, + "learning_rate": 0.003, + "loss": 3.97, + "step": 51478 + }, + { + "epoch": 0.51479, + "grad_norm": 0.7531798273357143, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 51479 + }, + { + "epoch": 0.5148, + "grad_norm": 0.6756304401772445, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 51480 + }, + { + "epoch": 0.51481, + "grad_norm": 0.6337498751277229, + "learning_rate": 0.003, + "loss": 3.982, + "step": 51481 + }, + { + "epoch": 0.51482, + "grad_norm": 0.5411955255345452, + "learning_rate": 0.003, + "loss": 3.9527, + "step": 51482 + }, + { + "epoch": 0.51483, + "grad_norm": 0.5682382583897212, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 51483 + }, + { + "epoch": 0.51484, + "grad_norm": 0.5456523959322442, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 51484 + }, + { + "epoch": 0.51485, + "grad_norm": 0.629198721691138, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 51485 + }, + { + "epoch": 0.51486, + "grad_norm": 0.7831765729434486, + "learning_rate": 0.003, + "loss": 4.009, + "step": 51486 + }, + { + "epoch": 0.51487, + "grad_norm": 0.9543218050231378, + "learning_rate": 0.003, + "loss": 4.0027, + "step": 51487 + }, + { + "epoch": 0.51488, + "grad_norm": 1.0002056144441274, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 51488 + }, + { + "epoch": 0.51489, + "grad_norm": 0.9962683294060543, + "learning_rate": 0.003, + "loss": 3.9906, + "step": 51489 + }, + { + "epoch": 0.5149, + "grad_norm": 0.9395694130355011, + "learning_rate": 0.003, + "loss": 3.983, + "step": 51490 + }, + { + "epoch": 0.51491, + "grad_norm": 0.9524065117153426, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 51491 + }, + { + "epoch": 0.51492, + "grad_norm": 0.9580016583668572, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 51492 + }, + { + "epoch": 0.51493, + "grad_norm": 0.8200232136838281, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 51493 + }, + { + "epoch": 0.51494, + "grad_norm": 0.7058796948708015, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 51494 + }, + { + "epoch": 0.51495, + "grad_norm": 0.5974813773602032, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 51495 + }, + { + "epoch": 0.51496, + "grad_norm": 0.6761304728270204, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 51496 + }, + { + "epoch": 0.51497, + "grad_norm": 0.7296114915829582, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 51497 + }, + { + "epoch": 0.51498, + "grad_norm": 0.7628422733732871, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 51498 + }, + { + "epoch": 0.51499, + "grad_norm": 0.8424706957193547, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 51499 + }, + { + "epoch": 0.515, + "grad_norm": 0.8305597051831739, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 51500 + }, + { + "epoch": 0.51501, + "grad_norm": 0.8195431620949861, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 51501 + }, + { + "epoch": 0.51502, + "grad_norm": 0.7695635005742614, + "learning_rate": 0.003, + "loss": 3.982, + "step": 51502 + }, + { + "epoch": 0.51503, + "grad_norm": 0.6815661876772221, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 51503 + }, + { + "epoch": 0.51504, + "grad_norm": 0.6683192556461274, + "learning_rate": 0.003, + "loss": 3.9962, + "step": 51504 + }, + { + "epoch": 0.51505, + "grad_norm": 0.6056238346573398, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 51505 + }, + { + "epoch": 0.51506, + "grad_norm": 0.6240692118207462, + "learning_rate": 0.003, + "loss": 3.956, + "step": 51506 + }, + { + "epoch": 0.51507, + "grad_norm": 0.6740691824670252, + "learning_rate": 0.003, + "loss": 3.9697, + "step": 51507 + }, + { + "epoch": 0.51508, + "grad_norm": 0.6240818442611454, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 51508 + }, + { + "epoch": 0.51509, + "grad_norm": 0.5835305158796813, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 51509 + }, + { + "epoch": 0.5151, + "grad_norm": 0.6511776402599981, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 51510 + }, + { + "epoch": 0.51511, + "grad_norm": 0.6527075071721005, + "learning_rate": 0.003, + "loss": 3.9969, + "step": 51511 + }, + { + "epoch": 0.51512, + "grad_norm": 0.6446943075602071, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 51512 + }, + { + "epoch": 0.51513, + "grad_norm": 0.5927190874281869, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 51513 + }, + { + "epoch": 0.51514, + "grad_norm": 0.5899831316096363, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 51514 + }, + { + "epoch": 0.51515, + "grad_norm": 0.6025581482379223, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 51515 + }, + { + "epoch": 0.51516, + "grad_norm": 0.6330030840613599, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 51516 + }, + { + "epoch": 0.51517, + "grad_norm": 0.6307705578905521, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 51517 + }, + { + "epoch": 0.51518, + "grad_norm": 0.6628266432422928, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 51518 + }, + { + "epoch": 0.51519, + "grad_norm": 0.6465726379468771, + "learning_rate": 0.003, + "loss": 3.9431, + "step": 51519 + }, + { + "epoch": 0.5152, + "grad_norm": 0.7813074072185168, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 51520 + }, + { + "epoch": 0.51521, + "grad_norm": 1.0034424972302125, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 51521 + }, + { + "epoch": 0.51522, + "grad_norm": 1.190843997989485, + "learning_rate": 0.003, + "loss": 4.014, + "step": 51522 + }, + { + "epoch": 0.51523, + "grad_norm": 0.6901010877644516, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 51523 + }, + { + "epoch": 0.51524, + "grad_norm": 0.6639618946285104, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 51524 + }, + { + "epoch": 0.51525, + "grad_norm": 0.6947991793227376, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 51525 + }, + { + "epoch": 0.51526, + "grad_norm": 0.8910130025308944, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 51526 + }, + { + "epoch": 0.51527, + "grad_norm": 1.202194042991902, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 51527 + }, + { + "epoch": 0.51528, + "grad_norm": 0.792884002342357, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 51528 + }, + { + "epoch": 0.51529, + "grad_norm": 0.8535983191289565, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 51529 + }, + { + "epoch": 0.5153, + "grad_norm": 0.9361779457966586, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 51530 + }, + { + "epoch": 0.51531, + "grad_norm": 0.9724271695784583, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 51531 + }, + { + "epoch": 0.51532, + "grad_norm": 1.0639972815881737, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 51532 + }, + { + "epoch": 0.51533, + "grad_norm": 0.9080544395004768, + "learning_rate": 0.003, + "loss": 4.0351, + "step": 51533 + }, + { + "epoch": 0.51534, + "grad_norm": 0.9950287778188042, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 51534 + }, + { + "epoch": 0.51535, + "grad_norm": 1.0028120463043502, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 51535 + }, + { + "epoch": 0.51536, + "grad_norm": 0.9212667488203594, + "learning_rate": 0.003, + "loss": 3.9953, + "step": 51536 + }, + { + "epoch": 0.51537, + "grad_norm": 0.8992712766070639, + "learning_rate": 0.003, + "loss": 4.0211, + "step": 51537 + }, + { + "epoch": 0.51538, + "grad_norm": 0.8876155333070135, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 51538 + }, + { + "epoch": 0.51539, + "grad_norm": 0.874461080066948, + "learning_rate": 0.003, + "loss": 3.9968, + "step": 51539 + }, + { + "epoch": 0.5154, + "grad_norm": 0.9101734688025218, + "learning_rate": 0.003, + "loss": 4.0409, + "step": 51540 + }, + { + "epoch": 0.51541, + "grad_norm": 0.8376657782138038, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 51541 + }, + { + "epoch": 0.51542, + "grad_norm": 0.8793030514870257, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 51542 + }, + { + "epoch": 0.51543, + "grad_norm": 1.1670931416423662, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 51543 + }, + { + "epoch": 0.51544, + "grad_norm": 0.99843116446494, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 51544 + }, + { + "epoch": 0.51545, + "grad_norm": 1.437664532578025, + "learning_rate": 0.003, + "loss": 3.991, + "step": 51545 + }, + { + "epoch": 0.51546, + "grad_norm": 2.4666857746075923, + "learning_rate": 0.003, + "loss": 4.0617, + "step": 51546 + }, + { + "epoch": 0.51547, + "grad_norm": 2.0953336038068704, + "learning_rate": 0.003, + "loss": 4.052, + "step": 51547 + }, + { + "epoch": 0.51548, + "grad_norm": 2.535534365982424, + "learning_rate": 0.003, + "loss": 4.057, + "step": 51548 + }, + { + "epoch": 0.51549, + "grad_norm": 0.8466953656518451, + "learning_rate": 0.003, + "loss": 4.0404, + "step": 51549 + }, + { + "epoch": 0.5155, + "grad_norm": 1.2450752586845997, + "learning_rate": 0.003, + "loss": 4.0572, + "step": 51550 + }, + { + "epoch": 0.51551, + "grad_norm": 1.6659763529251788, + "learning_rate": 0.003, + "loss": 4.0837, + "step": 51551 + }, + { + "epoch": 0.51552, + "grad_norm": 1.783051201667157, + "learning_rate": 0.003, + "loss": 4.0807, + "step": 51552 + }, + { + "epoch": 0.51553, + "grad_norm": 1.2729788746541983, + "learning_rate": 0.003, + "loss": 4.0594, + "step": 51553 + }, + { + "epoch": 0.51554, + "grad_norm": 0.8350960261418835, + "learning_rate": 0.003, + "loss": 4.0291, + "step": 51554 + }, + { + "epoch": 0.51555, + "grad_norm": 1.085893197840016, + "learning_rate": 0.003, + "loss": 4.0439, + "step": 51555 + }, + { + "epoch": 0.51556, + "grad_norm": 1.213062275175509, + "learning_rate": 0.003, + "loss": 4.042, + "step": 51556 + }, + { + "epoch": 0.51557, + "grad_norm": 0.99146352997922, + "learning_rate": 0.003, + "loss": 4.011, + "step": 51557 + }, + { + "epoch": 0.51558, + "grad_norm": 1.1028815434546761, + "learning_rate": 0.003, + "loss": 4.0244, + "step": 51558 + }, + { + "epoch": 0.51559, + "grad_norm": 1.0530375436616948, + "learning_rate": 0.003, + "loss": 4.0484, + "step": 51559 + }, + { + "epoch": 0.5156, + "grad_norm": 1.0237584328916436, + "learning_rate": 0.003, + "loss": 4.0323, + "step": 51560 + }, + { + "epoch": 0.51561, + "grad_norm": 1.415720557074784, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 51561 + }, + { + "epoch": 0.51562, + "grad_norm": 0.9083005812487185, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 51562 + }, + { + "epoch": 0.51563, + "grad_norm": 1.0802634198589351, + "learning_rate": 0.003, + "loss": 4.0432, + "step": 51563 + }, + { + "epoch": 0.51564, + "grad_norm": 0.8723335630560145, + "learning_rate": 0.003, + "loss": 3.9811, + "step": 51564 + }, + { + "epoch": 0.51565, + "grad_norm": 0.9833010097547858, + "learning_rate": 0.003, + "loss": 4.0418, + "step": 51565 + }, + { + "epoch": 0.51566, + "grad_norm": 0.9465174844437951, + "learning_rate": 0.003, + "loss": 4.0258, + "step": 51566 + }, + { + "epoch": 0.51567, + "grad_norm": 0.8956211554819647, + "learning_rate": 0.003, + "loss": 4.0363, + "step": 51567 + }, + { + "epoch": 0.51568, + "grad_norm": 0.8713712861771444, + "learning_rate": 0.003, + "loss": 4.0457, + "step": 51568 + }, + { + "epoch": 0.51569, + "grad_norm": 0.9952585977757774, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 51569 + }, + { + "epoch": 0.5157, + "grad_norm": 1.1632547274487044, + "learning_rate": 0.003, + "loss": 4.0054, + "step": 51570 + }, + { + "epoch": 0.51571, + "grad_norm": 0.9323213753299892, + "learning_rate": 0.003, + "loss": 4.0022, + "step": 51571 + }, + { + "epoch": 0.51572, + "grad_norm": 0.8216964287862886, + "learning_rate": 0.003, + "loss": 4.0247, + "step": 51572 + }, + { + "epoch": 0.51573, + "grad_norm": 0.762509638437527, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 51573 + }, + { + "epoch": 0.51574, + "grad_norm": 0.7355564212113491, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 51574 + }, + { + "epoch": 0.51575, + "grad_norm": 0.65927206433418, + "learning_rate": 0.003, + "loss": 4.0116, + "step": 51575 + }, + { + "epoch": 0.51576, + "grad_norm": 0.7399707588929911, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 51576 + }, + { + "epoch": 0.51577, + "grad_norm": 0.91170950062289, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 51577 + }, + { + "epoch": 0.51578, + "grad_norm": 1.0232819200584236, + "learning_rate": 0.003, + "loss": 4.0428, + "step": 51578 + }, + { + "epoch": 0.51579, + "grad_norm": 0.7562680886362227, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 51579 + }, + { + "epoch": 0.5158, + "grad_norm": 0.6663430962167615, + "learning_rate": 0.003, + "loss": 3.977, + "step": 51580 + }, + { + "epoch": 0.51581, + "grad_norm": 0.6672682574773638, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 51581 + }, + { + "epoch": 0.51582, + "grad_norm": 0.7395741910295062, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 51582 + }, + { + "epoch": 0.51583, + "grad_norm": 0.7368296859599703, + "learning_rate": 0.003, + "loss": 4.0325, + "step": 51583 + }, + { + "epoch": 0.51584, + "grad_norm": 0.6142356215529803, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 51584 + }, + { + "epoch": 0.51585, + "grad_norm": 0.6266979966937054, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 51585 + }, + { + "epoch": 0.51586, + "grad_norm": 0.7399080355550964, + "learning_rate": 0.003, + "loss": 3.9952, + "step": 51586 + }, + { + "epoch": 0.51587, + "grad_norm": 0.9364038086265823, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 51587 + }, + { + "epoch": 0.51588, + "grad_norm": 0.9186461934507189, + "learning_rate": 0.003, + "loss": 4.0101, + "step": 51588 + }, + { + "epoch": 0.51589, + "grad_norm": 0.8721367138009474, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 51589 + }, + { + "epoch": 0.5159, + "grad_norm": 0.63683559693252, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 51590 + }, + { + "epoch": 0.51591, + "grad_norm": 0.5263355416634246, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 51591 + }, + { + "epoch": 0.51592, + "grad_norm": 0.5133367979377039, + "learning_rate": 0.003, + "loss": 3.983, + "step": 51592 + }, + { + "epoch": 0.51593, + "grad_norm": 0.5727183646032039, + "learning_rate": 0.003, + "loss": 3.981, + "step": 51593 + }, + { + "epoch": 0.51594, + "grad_norm": 0.6174066539466548, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 51594 + }, + { + "epoch": 0.51595, + "grad_norm": 0.7458621808085796, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 51595 + }, + { + "epoch": 0.51596, + "grad_norm": 0.815912077543887, + "learning_rate": 0.003, + "loss": 3.9677, + "step": 51596 + }, + { + "epoch": 0.51597, + "grad_norm": 0.7761075829824697, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 51597 + }, + { + "epoch": 0.51598, + "grad_norm": 0.677753371752118, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 51598 + }, + { + "epoch": 0.51599, + "grad_norm": 0.6490793895768254, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 51599 + }, + { + "epoch": 0.516, + "grad_norm": 0.6696577074579758, + "learning_rate": 0.003, + "loss": 4.0144, + "step": 51600 + }, + { + "epoch": 0.51601, + "grad_norm": 0.6505781557317233, + "learning_rate": 0.003, + "loss": 3.988, + "step": 51601 + }, + { + "epoch": 0.51602, + "grad_norm": 0.6250673237900964, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 51602 + }, + { + "epoch": 0.51603, + "grad_norm": 0.5122329920548172, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 51603 + }, + { + "epoch": 0.51604, + "grad_norm": 0.5256981231135908, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 51604 + }, + { + "epoch": 0.51605, + "grad_norm": 0.5648321173072942, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 51605 + }, + { + "epoch": 0.51606, + "grad_norm": 0.6730220862737585, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 51606 + }, + { + "epoch": 0.51607, + "grad_norm": 0.8632239875627875, + "learning_rate": 0.003, + "loss": 4.01, + "step": 51607 + }, + { + "epoch": 0.51608, + "grad_norm": 0.9368534800567226, + "learning_rate": 0.003, + "loss": 4.0052, + "step": 51608 + }, + { + "epoch": 0.51609, + "grad_norm": 0.8843022741445591, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 51609 + }, + { + "epoch": 0.5161, + "grad_norm": 0.8357444138260423, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 51610 + }, + { + "epoch": 0.51611, + "grad_norm": 0.780164008754788, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 51611 + }, + { + "epoch": 0.51612, + "grad_norm": 0.6469542943295249, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 51612 + }, + { + "epoch": 0.51613, + "grad_norm": 0.5599139123942328, + "learning_rate": 0.003, + "loss": 3.941, + "step": 51613 + }, + { + "epoch": 0.51614, + "grad_norm": 0.5570527209420498, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 51614 + }, + { + "epoch": 0.51615, + "grad_norm": 0.5412273854104255, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 51615 + }, + { + "epoch": 0.51616, + "grad_norm": 0.5084417218437792, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 51616 + }, + { + "epoch": 0.51617, + "grad_norm": 0.48862030409707385, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 51617 + }, + { + "epoch": 0.51618, + "grad_norm": 0.5500521172362178, + "learning_rate": 0.003, + "loss": 3.954, + "step": 51618 + }, + { + "epoch": 0.51619, + "grad_norm": 0.6131969577222486, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 51619 + }, + { + "epoch": 0.5162, + "grad_norm": 0.6598725200035056, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 51620 + }, + { + "epoch": 0.51621, + "grad_norm": 0.7791673799085855, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 51621 + }, + { + "epoch": 0.51622, + "grad_norm": 0.9235510075706279, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 51622 + }, + { + "epoch": 0.51623, + "grad_norm": 0.8610003256768307, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 51623 + }, + { + "epoch": 0.51624, + "grad_norm": 0.6729442220639226, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 51624 + }, + { + "epoch": 0.51625, + "grad_norm": 0.7035529828768425, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 51625 + }, + { + "epoch": 0.51626, + "grad_norm": 0.7566544882969886, + "learning_rate": 0.003, + "loss": 3.9535, + "step": 51626 + }, + { + "epoch": 0.51627, + "grad_norm": 0.8820951756580069, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 51627 + }, + { + "epoch": 0.51628, + "grad_norm": 0.9861815260795065, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 51628 + }, + { + "epoch": 0.51629, + "grad_norm": 0.9349476882153555, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 51629 + }, + { + "epoch": 0.5163, + "grad_norm": 0.7900814128232164, + "learning_rate": 0.003, + "loss": 3.963, + "step": 51630 + }, + { + "epoch": 0.51631, + "grad_norm": 0.7183681611950903, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 51631 + }, + { + "epoch": 0.51632, + "grad_norm": 0.760688204702673, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 51632 + }, + { + "epoch": 0.51633, + "grad_norm": 0.846108493561872, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 51633 + }, + { + "epoch": 0.51634, + "grad_norm": 0.8633480267084324, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 51634 + }, + { + "epoch": 0.51635, + "grad_norm": 0.8356624711570227, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 51635 + }, + { + "epoch": 0.51636, + "grad_norm": 0.8563329604744642, + "learning_rate": 0.003, + "loss": 3.974, + "step": 51636 + }, + { + "epoch": 0.51637, + "grad_norm": 0.8011785214042461, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 51637 + }, + { + "epoch": 0.51638, + "grad_norm": 0.8648354102273493, + "learning_rate": 0.003, + "loss": 4.0074, + "step": 51638 + }, + { + "epoch": 0.51639, + "grad_norm": 0.920404638921309, + "learning_rate": 0.003, + "loss": 3.989, + "step": 51639 + }, + { + "epoch": 0.5164, + "grad_norm": 0.836043083913232, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 51640 + }, + { + "epoch": 0.51641, + "grad_norm": 0.7309992307411756, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 51641 + }, + { + "epoch": 0.51642, + "grad_norm": 0.7230094415029009, + "learning_rate": 0.003, + "loss": 4.0123, + "step": 51642 + }, + { + "epoch": 0.51643, + "grad_norm": 0.7087866169845772, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 51643 + }, + { + "epoch": 0.51644, + "grad_norm": 0.7654761712638196, + "learning_rate": 0.003, + "loss": 3.995, + "step": 51644 + }, + { + "epoch": 0.51645, + "grad_norm": 0.8406875535960686, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 51645 + }, + { + "epoch": 0.51646, + "grad_norm": 0.8903344720547877, + "learning_rate": 0.003, + "loss": 3.978, + "step": 51646 + }, + { + "epoch": 0.51647, + "grad_norm": 0.8460103976536295, + "learning_rate": 0.003, + "loss": 4.0077, + "step": 51647 + }, + { + "epoch": 0.51648, + "grad_norm": 0.9124664990250777, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 51648 + }, + { + "epoch": 0.51649, + "grad_norm": 0.9151625097284033, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 51649 + }, + { + "epoch": 0.5165, + "grad_norm": 0.8887677287621805, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 51650 + }, + { + "epoch": 0.51651, + "grad_norm": 0.7437502899008787, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 51651 + }, + { + "epoch": 0.51652, + "grad_norm": 0.7959849238507196, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 51652 + }, + { + "epoch": 0.51653, + "grad_norm": 0.7539716412431904, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 51653 + }, + { + "epoch": 0.51654, + "grad_norm": 0.6305831912999829, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 51654 + }, + { + "epoch": 0.51655, + "grad_norm": 0.6286393162548128, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 51655 + }, + { + "epoch": 0.51656, + "grad_norm": 0.6541604302939554, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 51656 + }, + { + "epoch": 0.51657, + "grad_norm": 0.7040546908716394, + "learning_rate": 0.003, + "loss": 3.9505, + "step": 51657 + }, + { + "epoch": 0.51658, + "grad_norm": 0.8832801448087997, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 51658 + }, + { + "epoch": 0.51659, + "grad_norm": 1.1059085794596013, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 51659 + }, + { + "epoch": 0.5166, + "grad_norm": 1.0144519334808708, + "learning_rate": 0.003, + "loss": 4.0161, + "step": 51660 + }, + { + "epoch": 0.51661, + "grad_norm": 0.7220753775494055, + "learning_rate": 0.003, + "loss": 3.9532, + "step": 51661 + }, + { + "epoch": 0.51662, + "grad_norm": 0.6787362886340997, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 51662 + }, + { + "epoch": 0.51663, + "grad_norm": 0.7532335442889234, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 51663 + }, + { + "epoch": 0.51664, + "grad_norm": 0.7893192718487508, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 51664 + }, + { + "epoch": 0.51665, + "grad_norm": 0.7148803701232884, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 51665 + }, + { + "epoch": 0.51666, + "grad_norm": 0.5963291805166654, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 51666 + }, + { + "epoch": 0.51667, + "grad_norm": 0.5294100262657371, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 51667 + }, + { + "epoch": 0.51668, + "grad_norm": 0.5040276097484239, + "learning_rate": 0.003, + "loss": 3.9663, + "step": 51668 + }, + { + "epoch": 0.51669, + "grad_norm": 0.5715809020715737, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 51669 + }, + { + "epoch": 0.5167, + "grad_norm": 0.5390450666410906, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 51670 + }, + { + "epoch": 0.51671, + "grad_norm": 0.5140251122629903, + "learning_rate": 0.003, + "loss": 3.9727, + "step": 51671 + }, + { + "epoch": 0.51672, + "grad_norm": 0.5354019723569092, + "learning_rate": 0.003, + "loss": 3.9509, + "step": 51672 + }, + { + "epoch": 0.51673, + "grad_norm": 0.6008353297560957, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 51673 + }, + { + "epoch": 0.51674, + "grad_norm": 0.657685863737571, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 51674 + }, + { + "epoch": 0.51675, + "grad_norm": 0.7233273289372839, + "learning_rate": 0.003, + "loss": 3.9264, + "step": 51675 + }, + { + "epoch": 0.51676, + "grad_norm": 0.7601515566659329, + "learning_rate": 0.003, + "loss": 3.971, + "step": 51676 + }, + { + "epoch": 0.51677, + "grad_norm": 0.7905344982800729, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 51677 + }, + { + "epoch": 0.51678, + "grad_norm": 0.7414617542780567, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 51678 + }, + { + "epoch": 0.51679, + "grad_norm": 0.6102244626318477, + "learning_rate": 0.003, + "loss": 3.9575, + "step": 51679 + }, + { + "epoch": 0.5168, + "grad_norm": 0.5771767178983795, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 51680 + }, + { + "epoch": 0.51681, + "grad_norm": 0.5862447339424667, + "learning_rate": 0.003, + "loss": 3.9704, + "step": 51681 + }, + { + "epoch": 0.51682, + "grad_norm": 0.6429343984955836, + "learning_rate": 0.003, + "loss": 3.9409, + "step": 51682 + }, + { + "epoch": 0.51683, + "grad_norm": 0.6432528307215504, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 51683 + }, + { + "epoch": 0.51684, + "grad_norm": 0.7353770428580133, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 51684 + }, + { + "epoch": 0.51685, + "grad_norm": 0.8632939785674222, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 51685 + }, + { + "epoch": 0.51686, + "grad_norm": 1.0358027113641866, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 51686 + }, + { + "epoch": 0.51687, + "grad_norm": 0.9581518995673065, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 51687 + }, + { + "epoch": 0.51688, + "grad_norm": 0.8922050360336932, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 51688 + }, + { + "epoch": 0.51689, + "grad_norm": 0.9669596416968719, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 51689 + }, + { + "epoch": 0.5169, + "grad_norm": 0.8960084895624538, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 51690 + }, + { + "epoch": 0.51691, + "grad_norm": 0.7787522876562045, + "learning_rate": 0.003, + "loss": 3.9898, + "step": 51691 + }, + { + "epoch": 0.51692, + "grad_norm": 0.8405831895689515, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 51692 + }, + { + "epoch": 0.51693, + "grad_norm": 0.9205867896804953, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 51693 + }, + { + "epoch": 0.51694, + "grad_norm": 0.7914907835927542, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 51694 + }, + { + "epoch": 0.51695, + "grad_norm": 0.7922078863367059, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 51695 + }, + { + "epoch": 0.51696, + "grad_norm": 0.7944256362621213, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 51696 + }, + { + "epoch": 0.51697, + "grad_norm": 0.8625597870113909, + "learning_rate": 0.003, + "loss": 4.0065, + "step": 51697 + }, + { + "epoch": 0.51698, + "grad_norm": 0.9858848676924152, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 51698 + }, + { + "epoch": 0.51699, + "grad_norm": 0.9841290654245658, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 51699 + }, + { + "epoch": 0.517, + "grad_norm": 0.9772622749774661, + "learning_rate": 0.003, + "loss": 4.006, + "step": 51700 + }, + { + "epoch": 0.51701, + "grad_norm": 0.8977713940730121, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 51701 + }, + { + "epoch": 0.51702, + "grad_norm": 0.9082856384998252, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 51702 + }, + { + "epoch": 0.51703, + "grad_norm": 0.8928195888507405, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 51703 + }, + { + "epoch": 0.51704, + "grad_norm": 0.8857080901236236, + "learning_rate": 0.003, + "loss": 3.9795, + "step": 51704 + }, + { + "epoch": 0.51705, + "grad_norm": 0.9407455228987266, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 51705 + }, + { + "epoch": 0.51706, + "grad_norm": 1.0206087510923012, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 51706 + }, + { + "epoch": 0.51707, + "grad_norm": 1.0731519037463464, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 51707 + }, + { + "epoch": 0.51708, + "grad_norm": 0.8322490481034582, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 51708 + }, + { + "epoch": 0.51709, + "grad_norm": 0.7008037547929898, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 51709 + }, + { + "epoch": 0.5171, + "grad_norm": 0.648149888441771, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 51710 + }, + { + "epoch": 0.51711, + "grad_norm": 0.7466391029666951, + "learning_rate": 0.003, + "loss": 4.0069, + "step": 51711 + }, + { + "epoch": 0.51712, + "grad_norm": 0.7712742249766797, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 51712 + }, + { + "epoch": 0.51713, + "grad_norm": 0.684472228111992, + "learning_rate": 0.003, + "loss": 3.9908, + "step": 51713 + }, + { + "epoch": 0.51714, + "grad_norm": 0.625133477906682, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 51714 + }, + { + "epoch": 0.51715, + "grad_norm": 0.6406331028726969, + "learning_rate": 0.003, + "loss": 3.9628, + "step": 51715 + }, + { + "epoch": 0.51716, + "grad_norm": 0.5503632352132284, + "learning_rate": 0.003, + "loss": 3.978, + "step": 51716 + }, + { + "epoch": 0.51717, + "grad_norm": 0.5176682819005177, + "learning_rate": 0.003, + "loss": 3.9554, + "step": 51717 + }, + { + "epoch": 0.51718, + "grad_norm": 0.515363770040694, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 51718 + }, + { + "epoch": 0.51719, + "grad_norm": 0.5669937809525778, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 51719 + }, + { + "epoch": 0.5172, + "grad_norm": 0.650523994190254, + "learning_rate": 0.003, + "loss": 3.975, + "step": 51720 + }, + { + "epoch": 0.51721, + "grad_norm": 0.7048873121300172, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 51721 + }, + { + "epoch": 0.51722, + "grad_norm": 0.7538270983237549, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 51722 + }, + { + "epoch": 0.51723, + "grad_norm": 0.6961510504689172, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 51723 + }, + { + "epoch": 0.51724, + "grad_norm": 0.593095419780037, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 51724 + }, + { + "epoch": 0.51725, + "grad_norm": 0.5823016888022895, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 51725 + }, + { + "epoch": 0.51726, + "grad_norm": 0.6178069447903363, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 51726 + }, + { + "epoch": 0.51727, + "grad_norm": 0.6819523599954067, + "learning_rate": 0.003, + "loss": 3.9718, + "step": 51727 + }, + { + "epoch": 0.51728, + "grad_norm": 0.7822429240899363, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 51728 + }, + { + "epoch": 0.51729, + "grad_norm": 0.8854681982303703, + "learning_rate": 0.003, + "loss": 3.9578, + "step": 51729 + }, + { + "epoch": 0.5173, + "grad_norm": 0.9184457395791242, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 51730 + }, + { + "epoch": 0.51731, + "grad_norm": 0.9440708965423007, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 51731 + }, + { + "epoch": 0.51732, + "grad_norm": 0.8530282521104475, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 51732 + }, + { + "epoch": 0.51733, + "grad_norm": 0.6821892806416744, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 51733 + }, + { + "epoch": 0.51734, + "grad_norm": 0.6491764847228128, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 51734 + }, + { + "epoch": 0.51735, + "grad_norm": 0.7020647960768372, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 51735 + }, + { + "epoch": 0.51736, + "grad_norm": 0.7094934939782249, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 51736 + }, + { + "epoch": 0.51737, + "grad_norm": 0.7326847708077269, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 51737 + }, + { + "epoch": 0.51738, + "grad_norm": 0.6575946899619334, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 51738 + }, + { + "epoch": 0.51739, + "grad_norm": 0.7815311580845143, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 51739 + }, + { + "epoch": 0.5174, + "grad_norm": 0.963368332289039, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 51740 + }, + { + "epoch": 0.51741, + "grad_norm": 1.0283826165814534, + "learning_rate": 0.003, + "loss": 3.9741, + "step": 51741 + }, + { + "epoch": 0.51742, + "grad_norm": 0.9264308083072705, + "learning_rate": 0.003, + "loss": 3.962, + "step": 51742 + }, + { + "epoch": 0.51743, + "grad_norm": 0.909874477333108, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 51743 + }, + { + "epoch": 0.51744, + "grad_norm": 0.9315856727145226, + "learning_rate": 0.003, + "loss": 4.002, + "step": 51744 + }, + { + "epoch": 0.51745, + "grad_norm": 1.0657956281419823, + "learning_rate": 0.003, + "loss": 4.0096, + "step": 51745 + }, + { + "epoch": 0.51746, + "grad_norm": 0.8485152209541225, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 51746 + }, + { + "epoch": 0.51747, + "grad_norm": 0.833804176449439, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 51747 + }, + { + "epoch": 0.51748, + "grad_norm": 0.7084664282876968, + "learning_rate": 0.003, + "loss": 4.0071, + "step": 51748 + }, + { + "epoch": 0.51749, + "grad_norm": 0.7738673364874242, + "learning_rate": 0.003, + "loss": 3.9877, + "step": 51749 + }, + { + "epoch": 0.5175, + "grad_norm": 0.7122252709496345, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 51750 + }, + { + "epoch": 0.51751, + "grad_norm": 0.7580638256587878, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 51751 + }, + { + "epoch": 0.51752, + "grad_norm": 0.7838436433482825, + "learning_rate": 0.003, + "loss": 3.9615, + "step": 51752 + }, + { + "epoch": 0.51753, + "grad_norm": 0.7882761818135544, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 51753 + }, + { + "epoch": 0.51754, + "grad_norm": 0.8244333778450162, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 51754 + }, + { + "epoch": 0.51755, + "grad_norm": 0.7570763853662437, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 51755 + }, + { + "epoch": 0.51756, + "grad_norm": 0.6508904637005306, + "learning_rate": 0.003, + "loss": 3.977, + "step": 51756 + }, + { + "epoch": 0.51757, + "grad_norm": 0.6460807685589924, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 51757 + }, + { + "epoch": 0.51758, + "grad_norm": 0.7176205588215793, + "learning_rate": 0.003, + "loss": 3.9755, + "step": 51758 + }, + { + "epoch": 0.51759, + "grad_norm": 0.8351303236992628, + "learning_rate": 0.003, + "loss": 3.9378, + "step": 51759 + }, + { + "epoch": 0.5176, + "grad_norm": 0.8093350179751272, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 51760 + }, + { + "epoch": 0.51761, + "grad_norm": 0.7498844576720215, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 51761 + }, + { + "epoch": 0.51762, + "grad_norm": 0.8579516459318358, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 51762 + }, + { + "epoch": 0.51763, + "grad_norm": 0.7999573218524096, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 51763 + }, + { + "epoch": 0.51764, + "grad_norm": 0.7095101637836941, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 51764 + }, + { + "epoch": 0.51765, + "grad_norm": 0.6756116363574338, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 51765 + }, + { + "epoch": 0.51766, + "grad_norm": 0.6767634249237536, + "learning_rate": 0.003, + "loss": 4.0013, + "step": 51766 + }, + { + "epoch": 0.51767, + "grad_norm": 0.6391993032276456, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 51767 + }, + { + "epoch": 0.51768, + "grad_norm": 0.6168970006501698, + "learning_rate": 0.003, + "loss": 3.9524, + "step": 51768 + }, + { + "epoch": 0.51769, + "grad_norm": 0.6625967909623339, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 51769 + }, + { + "epoch": 0.5177, + "grad_norm": 0.7232406312546269, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 51770 + }, + { + "epoch": 0.51771, + "grad_norm": 0.810582225242596, + "learning_rate": 0.003, + "loss": 4.0163, + "step": 51771 + }, + { + "epoch": 0.51772, + "grad_norm": 0.9252810054993742, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 51772 + }, + { + "epoch": 0.51773, + "grad_norm": 0.9256410378002826, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 51773 + }, + { + "epoch": 0.51774, + "grad_norm": 0.8657576455052094, + "learning_rate": 0.003, + "loss": 3.9656, + "step": 51774 + }, + { + "epoch": 0.51775, + "grad_norm": 0.7642828983748639, + "learning_rate": 0.003, + "loss": 3.965, + "step": 51775 + }, + { + "epoch": 0.51776, + "grad_norm": 0.705428854689088, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 51776 + }, + { + "epoch": 0.51777, + "grad_norm": 0.6969101043958806, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 51777 + }, + { + "epoch": 0.51778, + "grad_norm": 0.638444972117712, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 51778 + }, + { + "epoch": 0.51779, + "grad_norm": 0.6257215261997956, + "learning_rate": 0.003, + "loss": 3.9666, + "step": 51779 + }, + { + "epoch": 0.5178, + "grad_norm": 0.5907786551095244, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 51780 + }, + { + "epoch": 0.51781, + "grad_norm": 0.6906544606525329, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 51781 + }, + { + "epoch": 0.51782, + "grad_norm": 0.821582110628168, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 51782 + }, + { + "epoch": 0.51783, + "grad_norm": 0.8409755277420164, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 51783 + }, + { + "epoch": 0.51784, + "grad_norm": 0.8636925338550546, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 51784 + }, + { + "epoch": 0.51785, + "grad_norm": 0.9777860500136567, + "learning_rate": 0.003, + "loss": 3.9929, + "step": 51785 + }, + { + "epoch": 0.51786, + "grad_norm": 1.0664952490423065, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 51786 + }, + { + "epoch": 0.51787, + "grad_norm": 0.8415671273709084, + "learning_rate": 0.003, + "loss": 3.9705, + "step": 51787 + }, + { + "epoch": 0.51788, + "grad_norm": 0.8711868476763146, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 51788 + }, + { + "epoch": 0.51789, + "grad_norm": 0.9848552331962803, + "learning_rate": 0.003, + "loss": 4.019, + "step": 51789 + }, + { + "epoch": 0.5179, + "grad_norm": 1.038454308533228, + "learning_rate": 0.003, + "loss": 4.0184, + "step": 51790 + }, + { + "epoch": 0.51791, + "grad_norm": 1.0829134279186325, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 51791 + }, + { + "epoch": 0.51792, + "grad_norm": 0.8734825821285229, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 51792 + }, + { + "epoch": 0.51793, + "grad_norm": 0.9117441237274114, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 51793 + }, + { + "epoch": 0.51794, + "grad_norm": 0.7556853120861301, + "learning_rate": 0.003, + "loss": 4.0137, + "step": 51794 + }, + { + "epoch": 0.51795, + "grad_norm": 0.7449081554185574, + "learning_rate": 0.003, + "loss": 4.007, + "step": 51795 + }, + { + "epoch": 0.51796, + "grad_norm": 0.9046545756818869, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 51796 + }, + { + "epoch": 0.51797, + "grad_norm": 0.9560367307345552, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 51797 + }, + { + "epoch": 0.51798, + "grad_norm": 0.9000705133429108, + "learning_rate": 0.003, + "loss": 4.0045, + "step": 51798 + }, + { + "epoch": 0.51799, + "grad_norm": 0.7219825406348269, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 51799 + }, + { + "epoch": 0.518, + "grad_norm": 0.56761529495382, + "learning_rate": 0.003, + "loss": 3.976, + "step": 51800 + }, + { + "epoch": 0.51801, + "grad_norm": 0.6383951117928837, + "learning_rate": 0.003, + "loss": 3.974, + "step": 51801 + }, + { + "epoch": 0.51802, + "grad_norm": 0.7310599109685225, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 51802 + }, + { + "epoch": 0.51803, + "grad_norm": 0.8110570404049973, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 51803 + }, + { + "epoch": 0.51804, + "grad_norm": 0.7236201783461381, + "learning_rate": 0.003, + "loss": 3.979, + "step": 51804 + }, + { + "epoch": 0.51805, + "grad_norm": 0.695131405824199, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 51805 + }, + { + "epoch": 0.51806, + "grad_norm": 0.6871535081570403, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 51806 + }, + { + "epoch": 0.51807, + "grad_norm": 0.6675537802242619, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 51807 + }, + { + "epoch": 0.51808, + "grad_norm": 0.6401857062468755, + "learning_rate": 0.003, + "loss": 3.9989, + "step": 51808 + }, + { + "epoch": 0.51809, + "grad_norm": 0.6390163941325426, + "learning_rate": 0.003, + "loss": 3.998, + "step": 51809 + }, + { + "epoch": 0.5181, + "grad_norm": 0.6533115638361607, + "learning_rate": 0.003, + "loss": 3.9778, + "step": 51810 + }, + { + "epoch": 0.51811, + "grad_norm": 0.6982074815595852, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 51811 + }, + { + "epoch": 0.51812, + "grad_norm": 0.6824369807558288, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 51812 + }, + { + "epoch": 0.51813, + "grad_norm": 0.6276122002344419, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 51813 + }, + { + "epoch": 0.51814, + "grad_norm": 0.5345677861011231, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 51814 + }, + { + "epoch": 0.51815, + "grad_norm": 0.45495944000704375, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 51815 + }, + { + "epoch": 0.51816, + "grad_norm": 0.46041254972362916, + "learning_rate": 0.003, + "loss": 3.9676, + "step": 51816 + }, + { + "epoch": 0.51817, + "grad_norm": 0.5006351751771156, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 51817 + }, + { + "epoch": 0.51818, + "grad_norm": 0.625243643010236, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 51818 + }, + { + "epoch": 0.51819, + "grad_norm": 0.798762652937279, + "learning_rate": 0.003, + "loss": 3.977, + "step": 51819 + }, + { + "epoch": 0.5182, + "grad_norm": 1.0108993634697814, + "learning_rate": 0.003, + "loss": 3.962, + "step": 51820 + }, + { + "epoch": 0.51821, + "grad_norm": 1.1371456910298878, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 51821 + }, + { + "epoch": 0.51822, + "grad_norm": 0.7458706672287564, + "learning_rate": 0.003, + "loss": 3.9466, + "step": 51822 + }, + { + "epoch": 0.51823, + "grad_norm": 0.6603111077599176, + "learning_rate": 0.003, + "loss": 3.984, + "step": 51823 + }, + { + "epoch": 0.51824, + "grad_norm": 0.8351217115927094, + "learning_rate": 0.003, + "loss": 4.0099, + "step": 51824 + }, + { + "epoch": 0.51825, + "grad_norm": 0.9134226742095398, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 51825 + }, + { + "epoch": 0.51826, + "grad_norm": 0.9528570404457779, + "learning_rate": 0.003, + "loss": 3.973, + "step": 51826 + }, + { + "epoch": 0.51827, + "grad_norm": 0.9060695633431347, + "learning_rate": 0.003, + "loss": 4.0252, + "step": 51827 + }, + { + "epoch": 0.51828, + "grad_norm": 0.8697284546628213, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 51828 + }, + { + "epoch": 0.51829, + "grad_norm": 0.7419706985765053, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 51829 + }, + { + "epoch": 0.5183, + "grad_norm": 0.7152262605137021, + "learning_rate": 0.003, + "loss": 3.979, + "step": 51830 + }, + { + "epoch": 0.51831, + "grad_norm": 0.7182013201124754, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 51831 + }, + { + "epoch": 0.51832, + "grad_norm": 0.7350398877891637, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 51832 + }, + { + "epoch": 0.51833, + "grad_norm": 0.8264847840040969, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 51833 + }, + { + "epoch": 0.51834, + "grad_norm": 0.9376784987117568, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 51834 + }, + { + "epoch": 0.51835, + "grad_norm": 0.7996285410385979, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 51835 + }, + { + "epoch": 0.51836, + "grad_norm": 0.8362763417729069, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 51836 + }, + { + "epoch": 0.51837, + "grad_norm": 0.8840788277087863, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 51837 + }, + { + "epoch": 0.51838, + "grad_norm": 0.7850382905330726, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 51838 + }, + { + "epoch": 0.51839, + "grad_norm": 0.8043763114882462, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 51839 + }, + { + "epoch": 0.5184, + "grad_norm": 0.8388700828272775, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 51840 + }, + { + "epoch": 0.51841, + "grad_norm": 0.8117167706780708, + "learning_rate": 0.003, + "loss": 3.9541, + "step": 51841 + }, + { + "epoch": 0.51842, + "grad_norm": 0.7033074205616489, + "learning_rate": 0.003, + "loss": 3.9893, + "step": 51842 + }, + { + "epoch": 0.51843, + "grad_norm": 0.586729967702143, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 51843 + }, + { + "epoch": 0.51844, + "grad_norm": 0.5499951181979728, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 51844 + }, + { + "epoch": 0.51845, + "grad_norm": 0.549160481565223, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 51845 + }, + { + "epoch": 0.51846, + "grad_norm": 0.5528247609102442, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 51846 + }, + { + "epoch": 0.51847, + "grad_norm": 0.5822204651734589, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 51847 + }, + { + "epoch": 0.51848, + "grad_norm": 0.6525500188808837, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 51848 + }, + { + "epoch": 0.51849, + "grad_norm": 0.7625667383049682, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 51849 + }, + { + "epoch": 0.5185, + "grad_norm": 0.841922604339925, + "learning_rate": 0.003, + "loss": 3.929, + "step": 51850 + }, + { + "epoch": 0.51851, + "grad_norm": 0.938792150908825, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 51851 + }, + { + "epoch": 0.51852, + "grad_norm": 0.8967779568828078, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 51852 + }, + { + "epoch": 0.51853, + "grad_norm": 0.8775482260227266, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 51853 + }, + { + "epoch": 0.51854, + "grad_norm": 0.8673588707575738, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 51854 + }, + { + "epoch": 0.51855, + "grad_norm": 0.9146602435194361, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 51855 + }, + { + "epoch": 0.51856, + "grad_norm": 0.9424796812425362, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 51856 + }, + { + "epoch": 0.51857, + "grad_norm": 0.8069105517576557, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 51857 + }, + { + "epoch": 0.51858, + "grad_norm": 0.786941735415705, + "learning_rate": 0.003, + "loss": 3.999, + "step": 51858 + }, + { + "epoch": 0.51859, + "grad_norm": 0.8421180569750901, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 51859 + }, + { + "epoch": 0.5186, + "grad_norm": 0.8364510586080558, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 51860 + }, + { + "epoch": 0.51861, + "grad_norm": 0.7852642148672394, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 51861 + }, + { + "epoch": 0.51862, + "grad_norm": 0.7463836731233644, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 51862 + }, + { + "epoch": 0.51863, + "grad_norm": 0.7622291579326919, + "learning_rate": 0.003, + "loss": 3.9582, + "step": 51863 + }, + { + "epoch": 0.51864, + "grad_norm": 0.7018452656797798, + "learning_rate": 0.003, + "loss": 3.9544, + "step": 51864 + }, + { + "epoch": 0.51865, + "grad_norm": 0.6271902045981746, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 51865 + }, + { + "epoch": 0.51866, + "grad_norm": 0.5774140846355438, + "learning_rate": 0.003, + "loss": 3.9955, + "step": 51866 + }, + { + "epoch": 0.51867, + "grad_norm": 0.6151486650186709, + "learning_rate": 0.003, + "loss": 3.9707, + "step": 51867 + }, + { + "epoch": 0.51868, + "grad_norm": 0.6312537496762728, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 51868 + }, + { + "epoch": 0.51869, + "grad_norm": 0.7105882619805013, + "learning_rate": 0.003, + "loss": 3.9919, + "step": 51869 + }, + { + "epoch": 0.5187, + "grad_norm": 0.8243043909534566, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 51870 + }, + { + "epoch": 0.51871, + "grad_norm": 0.9151474832152865, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 51871 + }, + { + "epoch": 0.51872, + "grad_norm": 0.9839217206505001, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 51872 + }, + { + "epoch": 0.51873, + "grad_norm": 1.0107905664475252, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 51873 + }, + { + "epoch": 0.51874, + "grad_norm": 0.829075892496912, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 51874 + }, + { + "epoch": 0.51875, + "grad_norm": 0.639880906184434, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 51875 + }, + { + "epoch": 0.51876, + "grad_norm": 0.7181566229248199, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 51876 + }, + { + "epoch": 0.51877, + "grad_norm": 0.8317101698077481, + "learning_rate": 0.003, + "loss": 3.9821, + "step": 51877 + }, + { + "epoch": 0.51878, + "grad_norm": 0.8790575542900965, + "learning_rate": 0.003, + "loss": 3.9733, + "step": 51878 + }, + { + "epoch": 0.51879, + "grad_norm": 0.8287280831726072, + "learning_rate": 0.003, + "loss": 3.9497, + "step": 51879 + }, + { + "epoch": 0.5188, + "grad_norm": 0.8522767316575139, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 51880 + }, + { + "epoch": 0.51881, + "grad_norm": 0.6595708081373047, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 51881 + }, + { + "epoch": 0.51882, + "grad_norm": 0.6723081755903653, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 51882 + }, + { + "epoch": 0.51883, + "grad_norm": 0.643646217486745, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 51883 + }, + { + "epoch": 0.51884, + "grad_norm": 0.6116306602099747, + "learning_rate": 0.003, + "loss": 3.9873, + "step": 51884 + }, + { + "epoch": 0.51885, + "grad_norm": 0.5764430906291969, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 51885 + }, + { + "epoch": 0.51886, + "grad_norm": 0.5827939355897825, + "learning_rate": 0.003, + "loss": 3.975, + "step": 51886 + }, + { + "epoch": 0.51887, + "grad_norm": 0.5713190338145179, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 51887 + }, + { + "epoch": 0.51888, + "grad_norm": 0.7031366412292911, + "learning_rate": 0.003, + "loss": 4.0016, + "step": 51888 + }, + { + "epoch": 0.51889, + "grad_norm": 0.7611436135451904, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 51889 + }, + { + "epoch": 0.5189, + "grad_norm": 0.8167277426473196, + "learning_rate": 0.003, + "loss": 3.9712, + "step": 51890 + }, + { + "epoch": 0.51891, + "grad_norm": 0.9439701274850391, + "learning_rate": 0.003, + "loss": 3.9917, + "step": 51891 + }, + { + "epoch": 0.51892, + "grad_norm": 1.072860650535593, + "learning_rate": 0.003, + "loss": 3.983, + "step": 51892 + }, + { + "epoch": 0.51893, + "grad_norm": 0.6682218579561572, + "learning_rate": 0.003, + "loss": 4.0195, + "step": 51893 + }, + { + "epoch": 0.51894, + "grad_norm": 0.741724153857823, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 51894 + }, + { + "epoch": 0.51895, + "grad_norm": 0.9546546497168417, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 51895 + }, + { + "epoch": 0.51896, + "grad_norm": 1.0799535312008361, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 51896 + }, + { + "epoch": 0.51897, + "grad_norm": 0.8833279557278892, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 51897 + }, + { + "epoch": 0.51898, + "grad_norm": 0.7596721901055014, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 51898 + }, + { + "epoch": 0.51899, + "grad_norm": 0.6360167808570458, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 51899 + }, + { + "epoch": 0.519, + "grad_norm": 0.6318037343188216, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 51900 + }, + { + "epoch": 0.51901, + "grad_norm": 0.6632421073494559, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 51901 + }, + { + "epoch": 0.51902, + "grad_norm": 0.6535522463547111, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 51902 + }, + { + "epoch": 0.51903, + "grad_norm": 0.5577009176915814, + "learning_rate": 0.003, + "loss": 4.0059, + "step": 51903 + }, + { + "epoch": 0.51904, + "grad_norm": 0.5541017524923979, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 51904 + }, + { + "epoch": 0.51905, + "grad_norm": 0.6326469287325085, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 51905 + }, + { + "epoch": 0.51906, + "grad_norm": 0.6930039653837314, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 51906 + }, + { + "epoch": 0.51907, + "grad_norm": 0.7043578204291506, + "learning_rate": 0.003, + "loss": 3.9369, + "step": 51907 + }, + { + "epoch": 0.51908, + "grad_norm": 0.7451051412385344, + "learning_rate": 0.003, + "loss": 3.9674, + "step": 51908 + }, + { + "epoch": 0.51909, + "grad_norm": 0.7714400238780544, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 51909 + }, + { + "epoch": 0.5191, + "grad_norm": 0.810244547670471, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 51910 + }, + { + "epoch": 0.51911, + "grad_norm": 0.851109513321277, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 51911 + }, + { + "epoch": 0.51912, + "grad_norm": 0.781962608225294, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 51912 + }, + { + "epoch": 0.51913, + "grad_norm": 0.69266231436518, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 51913 + }, + { + "epoch": 0.51914, + "grad_norm": 0.7588479450120643, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 51914 + }, + { + "epoch": 0.51915, + "grad_norm": 0.7789413885362364, + "learning_rate": 0.003, + "loss": 3.9383, + "step": 51915 + }, + { + "epoch": 0.51916, + "grad_norm": 0.794248508000926, + "learning_rate": 0.003, + "loss": 3.986, + "step": 51916 + }, + { + "epoch": 0.51917, + "grad_norm": 0.9230815576321045, + "learning_rate": 0.003, + "loss": 3.983, + "step": 51917 + }, + { + "epoch": 0.51918, + "grad_norm": 1.138807415151138, + "learning_rate": 0.003, + "loss": 3.9529, + "step": 51918 + }, + { + "epoch": 0.51919, + "grad_norm": 0.7350740203431835, + "learning_rate": 0.003, + "loss": 3.994, + "step": 51919 + }, + { + "epoch": 0.5192, + "grad_norm": 0.5950472855231738, + "learning_rate": 0.003, + "loss": 3.972, + "step": 51920 + }, + { + "epoch": 0.51921, + "grad_norm": 0.7380370426437153, + "learning_rate": 0.003, + "loss": 3.9566, + "step": 51921 + }, + { + "epoch": 0.51922, + "grad_norm": 0.926107876833672, + "learning_rate": 0.003, + "loss": 4.0001, + "step": 51922 + }, + { + "epoch": 0.51923, + "grad_norm": 0.9783066738367028, + "learning_rate": 0.003, + "loss": 3.9966, + "step": 51923 + }, + { + "epoch": 0.51924, + "grad_norm": 0.7490205159824369, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 51924 + }, + { + "epoch": 0.51925, + "grad_norm": 0.6641501423416588, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 51925 + }, + { + "epoch": 0.51926, + "grad_norm": 0.7716695570127745, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 51926 + }, + { + "epoch": 0.51927, + "grad_norm": 0.7179299526069237, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 51927 + }, + { + "epoch": 0.51928, + "grad_norm": 0.6699353862345798, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 51928 + }, + { + "epoch": 0.51929, + "grad_norm": 0.7801900581360446, + "learning_rate": 0.003, + "loss": 3.9996, + "step": 51929 + }, + { + "epoch": 0.5193, + "grad_norm": 0.7584086217775626, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 51930 + }, + { + "epoch": 0.51931, + "grad_norm": 0.7533770031485931, + "learning_rate": 0.003, + "loss": 3.9609, + "step": 51931 + }, + { + "epoch": 0.51932, + "grad_norm": 0.7845205717404821, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 51932 + }, + { + "epoch": 0.51933, + "grad_norm": 0.8064589323163127, + "learning_rate": 0.003, + "loss": 3.992, + "step": 51933 + }, + { + "epoch": 0.51934, + "grad_norm": 0.8418592375647564, + "learning_rate": 0.003, + "loss": 3.993, + "step": 51934 + }, + { + "epoch": 0.51935, + "grad_norm": 0.870441193489294, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 51935 + }, + { + "epoch": 0.51936, + "grad_norm": 0.7659683486212349, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 51936 + }, + { + "epoch": 0.51937, + "grad_norm": 0.7630969948599332, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 51937 + }, + { + "epoch": 0.51938, + "grad_norm": 0.5854114697480619, + "learning_rate": 0.003, + "loss": 3.9551, + "step": 51938 + }, + { + "epoch": 0.51939, + "grad_norm": 0.5541631269865643, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 51939 + }, + { + "epoch": 0.5194, + "grad_norm": 0.5640059582355047, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 51940 + }, + { + "epoch": 0.51941, + "grad_norm": 0.6029864873124332, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 51941 + }, + { + "epoch": 0.51942, + "grad_norm": 0.6725102286727681, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 51942 + }, + { + "epoch": 0.51943, + "grad_norm": 0.796116915925004, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 51943 + }, + { + "epoch": 0.51944, + "grad_norm": 0.7825544672862613, + "learning_rate": 0.003, + "loss": 4.0063, + "step": 51944 + }, + { + "epoch": 0.51945, + "grad_norm": 0.7682237359595213, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 51945 + }, + { + "epoch": 0.51946, + "grad_norm": 0.9135191172143505, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 51946 + }, + { + "epoch": 0.51947, + "grad_norm": 0.9022216117182758, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 51947 + }, + { + "epoch": 0.51948, + "grad_norm": 0.8091530881077761, + "learning_rate": 0.003, + "loss": 3.9589, + "step": 51948 + }, + { + "epoch": 0.51949, + "grad_norm": 0.7280552416394159, + "learning_rate": 0.003, + "loss": 3.9617, + "step": 51949 + }, + { + "epoch": 0.5195, + "grad_norm": 0.8047823983139522, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 51950 + }, + { + "epoch": 0.51951, + "grad_norm": 0.766375307714979, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 51951 + }, + { + "epoch": 0.51952, + "grad_norm": 0.7438494956808913, + "learning_rate": 0.003, + "loss": 3.9503, + "step": 51952 + }, + { + "epoch": 0.51953, + "grad_norm": 0.8126833414831044, + "learning_rate": 0.003, + "loss": 3.9995, + "step": 51953 + }, + { + "epoch": 0.51954, + "grad_norm": 0.9380198683448188, + "learning_rate": 0.003, + "loss": 3.9976, + "step": 51954 + }, + { + "epoch": 0.51955, + "grad_norm": 0.9500346594952013, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 51955 + }, + { + "epoch": 0.51956, + "grad_norm": 1.041607604485403, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 51956 + }, + { + "epoch": 0.51957, + "grad_norm": 0.9443856255090332, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 51957 + }, + { + "epoch": 0.51958, + "grad_norm": 0.8854570348900014, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 51958 + }, + { + "epoch": 0.51959, + "grad_norm": 0.806624836678369, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 51959 + }, + { + "epoch": 0.5196, + "grad_norm": 0.7467987259355644, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 51960 + }, + { + "epoch": 0.51961, + "grad_norm": 0.5952740826903208, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 51961 + }, + { + "epoch": 0.51962, + "grad_norm": 0.5260193876500191, + "learning_rate": 0.003, + "loss": 3.9789, + "step": 51962 + }, + { + "epoch": 0.51963, + "grad_norm": 0.5573117801825584, + "learning_rate": 0.003, + "loss": 3.9594, + "step": 51963 + }, + { + "epoch": 0.51964, + "grad_norm": 0.5424737211982574, + "learning_rate": 0.003, + "loss": 3.9439, + "step": 51964 + }, + { + "epoch": 0.51965, + "grad_norm": 0.5465020219087757, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 51965 + }, + { + "epoch": 0.51966, + "grad_norm": 0.6336775323401918, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 51966 + }, + { + "epoch": 0.51967, + "grad_norm": 0.6862051982309522, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 51967 + }, + { + "epoch": 0.51968, + "grad_norm": 0.7243899739678009, + "learning_rate": 0.003, + "loss": 3.9548, + "step": 51968 + }, + { + "epoch": 0.51969, + "grad_norm": 0.8180873266258261, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 51969 + }, + { + "epoch": 0.5197, + "grad_norm": 1.0192514651985307, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 51970 + }, + { + "epoch": 0.51971, + "grad_norm": 0.9825032079710563, + "learning_rate": 0.003, + "loss": 4.005, + "step": 51971 + }, + { + "epoch": 0.51972, + "grad_norm": 0.8261363636007394, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 51972 + }, + { + "epoch": 0.51973, + "grad_norm": 0.9655322178774183, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 51973 + }, + { + "epoch": 0.51974, + "grad_norm": 0.9317231463169602, + "learning_rate": 0.003, + "loss": 4.0141, + "step": 51974 + }, + { + "epoch": 0.51975, + "grad_norm": 1.1219302853454254, + "learning_rate": 0.003, + "loss": 4.0139, + "step": 51975 + }, + { + "epoch": 0.51976, + "grad_norm": 0.8164338114347032, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 51976 + }, + { + "epoch": 0.51977, + "grad_norm": 0.7063226950228784, + "learning_rate": 0.003, + "loss": 3.9723, + "step": 51977 + }, + { + "epoch": 0.51978, + "grad_norm": 0.8533211740526279, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 51978 + }, + { + "epoch": 0.51979, + "grad_norm": 0.9571066722599213, + "learning_rate": 0.003, + "loss": 3.9758, + "step": 51979 + }, + { + "epoch": 0.5198, + "grad_norm": 1.1222393052836874, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 51980 + }, + { + "epoch": 0.51981, + "grad_norm": 1.1313749959263018, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 51981 + }, + { + "epoch": 0.51982, + "grad_norm": 0.8263278547368188, + "learning_rate": 0.003, + "loss": 4.0004, + "step": 51982 + }, + { + "epoch": 0.51983, + "grad_norm": 0.7613837166866526, + "learning_rate": 0.003, + "loss": 4.0089, + "step": 51983 + }, + { + "epoch": 0.51984, + "grad_norm": 0.7752092777087681, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 51984 + }, + { + "epoch": 0.51985, + "grad_norm": 0.8684629498343971, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 51985 + }, + { + "epoch": 0.51986, + "grad_norm": 0.8779815587097826, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 51986 + }, + { + "epoch": 0.51987, + "grad_norm": 0.9531647913389276, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 51987 + }, + { + "epoch": 0.51988, + "grad_norm": 0.8325035205610689, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 51988 + }, + { + "epoch": 0.51989, + "grad_norm": 0.7535134478652673, + "learning_rate": 0.003, + "loss": 3.978, + "step": 51989 + }, + { + "epoch": 0.5199, + "grad_norm": 0.7296338426803612, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 51990 + }, + { + "epoch": 0.51991, + "grad_norm": 0.6471742575427003, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 51991 + }, + { + "epoch": 0.51992, + "grad_norm": 0.5104800289232859, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 51992 + }, + { + "epoch": 0.51993, + "grad_norm": 0.5189434795418348, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 51993 + }, + { + "epoch": 0.51994, + "grad_norm": 0.4878430798808478, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 51994 + }, + { + "epoch": 0.51995, + "grad_norm": 0.5017909598938274, + "learning_rate": 0.003, + "loss": 3.999, + "step": 51995 + }, + { + "epoch": 0.51996, + "grad_norm": 0.5358328223180124, + "learning_rate": 0.003, + "loss": 3.9606, + "step": 51996 + }, + { + "epoch": 0.51997, + "grad_norm": 0.6227512743151494, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 51997 + }, + { + "epoch": 0.51998, + "grad_norm": 0.654160660687204, + "learning_rate": 0.003, + "loss": 3.9791, + "step": 51998 + }, + { + "epoch": 0.51999, + "grad_norm": 0.6652048252822802, + "learning_rate": 0.003, + "loss": 3.9927, + "step": 51999 + }, + { + "epoch": 0.52, + "grad_norm": 0.6914597214051434, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 52000 + }, + { + "epoch": 0.52001, + "grad_norm": 0.6737284113335676, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 52001 + }, + { + "epoch": 0.52002, + "grad_norm": 0.6168571704160905, + "learning_rate": 0.003, + "loss": 3.9967, + "step": 52002 + }, + { + "epoch": 0.52003, + "grad_norm": 0.6340448648572329, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 52003 + }, + { + "epoch": 0.52004, + "grad_norm": 0.7612963148019969, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 52004 + }, + { + "epoch": 0.52005, + "grad_norm": 0.7483132602997221, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 52005 + }, + { + "epoch": 0.52006, + "grad_norm": 0.7781268938539513, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 52006 + }, + { + "epoch": 0.52007, + "grad_norm": 0.901442158863086, + "learning_rate": 0.003, + "loss": 3.9345, + "step": 52007 + }, + { + "epoch": 0.52008, + "grad_norm": 0.9837296107334972, + "learning_rate": 0.003, + "loss": 4.0028, + "step": 52008 + }, + { + "epoch": 0.52009, + "grad_norm": 0.8593932192433272, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 52009 + }, + { + "epoch": 0.5201, + "grad_norm": 0.7964269010866292, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 52010 + }, + { + "epoch": 0.52011, + "grad_norm": 0.7851801540031703, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 52011 + }, + { + "epoch": 0.52012, + "grad_norm": 0.7731508087401878, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 52012 + }, + { + "epoch": 0.52013, + "grad_norm": 0.7607884992885887, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 52013 + }, + { + "epoch": 0.52014, + "grad_norm": 0.9019817498594416, + "learning_rate": 0.003, + "loss": 4.0174, + "step": 52014 + }, + { + "epoch": 0.52015, + "grad_norm": 0.9175278359402774, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 52015 + }, + { + "epoch": 0.52016, + "grad_norm": 1.0499591052507444, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 52016 + }, + { + "epoch": 0.52017, + "grad_norm": 1.0057861361295324, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 52017 + }, + { + "epoch": 0.52018, + "grad_norm": 0.8353067203361173, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 52018 + }, + { + "epoch": 0.52019, + "grad_norm": 0.9228684426529957, + "learning_rate": 0.003, + "loss": 4.0138, + "step": 52019 + }, + { + "epoch": 0.5202, + "grad_norm": 0.9161694202946296, + "learning_rate": 0.003, + "loss": 3.9946, + "step": 52020 + }, + { + "epoch": 0.52021, + "grad_norm": 0.8797863129957574, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 52021 + }, + { + "epoch": 0.52022, + "grad_norm": 0.8299937745272692, + "learning_rate": 0.003, + "loss": 4.0135, + "step": 52022 + }, + { + "epoch": 0.52023, + "grad_norm": 0.8160592729812226, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 52023 + }, + { + "epoch": 0.52024, + "grad_norm": 0.8288420794354393, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 52024 + }, + { + "epoch": 0.52025, + "grad_norm": 0.8177752700800583, + "learning_rate": 0.003, + "loss": 3.987, + "step": 52025 + }, + { + "epoch": 0.52026, + "grad_norm": 0.8642738669454805, + "learning_rate": 0.003, + "loss": 3.9974, + "step": 52026 + }, + { + "epoch": 0.52027, + "grad_norm": 0.8133882219306489, + "learning_rate": 0.003, + "loss": 3.9709, + "step": 52027 + }, + { + "epoch": 0.52028, + "grad_norm": 0.7969945245722502, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 52028 + }, + { + "epoch": 0.52029, + "grad_norm": 0.8486902634506438, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 52029 + }, + { + "epoch": 0.5203, + "grad_norm": 0.8117486814298951, + "learning_rate": 0.003, + "loss": 3.9696, + "step": 52030 + }, + { + "epoch": 0.52031, + "grad_norm": 0.6670986507435058, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 52031 + }, + { + "epoch": 0.52032, + "grad_norm": 0.6962967895616355, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 52032 + }, + { + "epoch": 0.52033, + "grad_norm": 0.6172640886118004, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 52033 + }, + { + "epoch": 0.52034, + "grad_norm": 0.6652956257996331, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 52034 + }, + { + "epoch": 0.52035, + "grad_norm": 0.7324649349181412, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 52035 + }, + { + "epoch": 0.52036, + "grad_norm": 0.7346704099740289, + "learning_rate": 0.003, + "loss": 3.9767, + "step": 52036 + }, + { + "epoch": 0.52037, + "grad_norm": 0.6308282075823058, + "learning_rate": 0.003, + "loss": 4.0079, + "step": 52037 + }, + { + "epoch": 0.52038, + "grad_norm": 0.6245323002297063, + "learning_rate": 0.003, + "loss": 3.9816, + "step": 52038 + }, + { + "epoch": 0.52039, + "grad_norm": 0.6443310242583651, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 52039 + }, + { + "epoch": 0.5204, + "grad_norm": 0.726706337822137, + "learning_rate": 0.003, + "loss": 3.9803, + "step": 52040 + }, + { + "epoch": 0.52041, + "grad_norm": 0.6856063761546743, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 52041 + }, + { + "epoch": 0.52042, + "grad_norm": 0.7608052270529845, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 52042 + }, + { + "epoch": 0.52043, + "grad_norm": 0.7290298010347429, + "learning_rate": 0.003, + "loss": 3.9537, + "step": 52043 + }, + { + "epoch": 0.52044, + "grad_norm": 0.6926763312173285, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 52044 + }, + { + "epoch": 0.52045, + "grad_norm": 0.6758576912446962, + "learning_rate": 0.003, + "loss": 3.959, + "step": 52045 + }, + { + "epoch": 0.52046, + "grad_norm": 0.7201039912216556, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 52046 + }, + { + "epoch": 0.52047, + "grad_norm": 0.7200351931851872, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 52047 + }, + { + "epoch": 0.52048, + "grad_norm": 0.7732731648543983, + "learning_rate": 0.003, + "loss": 3.9435, + "step": 52048 + }, + { + "epoch": 0.52049, + "grad_norm": 0.7478093958668965, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 52049 + }, + { + "epoch": 0.5205, + "grad_norm": 0.6838078714091101, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 52050 + }, + { + "epoch": 0.52051, + "grad_norm": 0.6934835468138125, + "learning_rate": 0.003, + "loss": 3.9467, + "step": 52051 + }, + { + "epoch": 0.52052, + "grad_norm": 0.8079782336627072, + "learning_rate": 0.003, + "loss": 3.9458, + "step": 52052 + }, + { + "epoch": 0.52053, + "grad_norm": 0.9003262385102178, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 52053 + }, + { + "epoch": 0.52054, + "grad_norm": 0.9966729540929675, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 52054 + }, + { + "epoch": 0.52055, + "grad_norm": 1.239538265036228, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 52055 + }, + { + "epoch": 0.52056, + "grad_norm": 0.8068936008892348, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 52056 + }, + { + "epoch": 0.52057, + "grad_norm": 0.6687759819884439, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 52057 + }, + { + "epoch": 0.52058, + "grad_norm": 0.6629577137118152, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 52058 + }, + { + "epoch": 0.52059, + "grad_norm": 0.7015193203071528, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 52059 + }, + { + "epoch": 0.5206, + "grad_norm": 0.7116106352432595, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 52060 + }, + { + "epoch": 0.52061, + "grad_norm": 0.7584104960017769, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 52061 + }, + { + "epoch": 0.52062, + "grad_norm": 0.7774087442529978, + "learning_rate": 0.003, + "loss": 3.9598, + "step": 52062 + }, + { + "epoch": 0.52063, + "grad_norm": 0.7560585611587973, + "learning_rate": 0.003, + "loss": 3.9714, + "step": 52063 + }, + { + "epoch": 0.52064, + "grad_norm": 0.7514487304929118, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 52064 + }, + { + "epoch": 0.52065, + "grad_norm": 0.6781000760819191, + "learning_rate": 0.003, + "loss": 4.0019, + "step": 52065 + }, + { + "epoch": 0.52066, + "grad_norm": 0.6983662051235562, + "learning_rate": 0.003, + "loss": 4.0018, + "step": 52066 + }, + { + "epoch": 0.52067, + "grad_norm": 0.7194978146925981, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 52067 + }, + { + "epoch": 0.52068, + "grad_norm": 0.6474149658686436, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 52068 + }, + { + "epoch": 0.52069, + "grad_norm": 0.6238278831382219, + "learning_rate": 0.003, + "loss": 3.982, + "step": 52069 + }, + { + "epoch": 0.5207, + "grad_norm": 0.6562501444879035, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 52070 + }, + { + "epoch": 0.52071, + "grad_norm": 0.7825004637162778, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 52071 + }, + { + "epoch": 0.52072, + "grad_norm": 0.8885071167239256, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 52072 + }, + { + "epoch": 0.52073, + "grad_norm": 0.75426578119237, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 52073 + }, + { + "epoch": 0.52074, + "grad_norm": 0.6715107863204757, + "learning_rate": 0.003, + "loss": 3.997, + "step": 52074 + }, + { + "epoch": 0.52075, + "grad_norm": 0.7076397246183713, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 52075 + }, + { + "epoch": 0.52076, + "grad_norm": 0.6726462384982809, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 52076 + }, + { + "epoch": 0.52077, + "grad_norm": 0.6983059232433451, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 52077 + }, + { + "epoch": 0.52078, + "grad_norm": 0.7175441237479375, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 52078 + }, + { + "epoch": 0.52079, + "grad_norm": 0.6844072558251034, + "learning_rate": 0.003, + "loss": 4.0093, + "step": 52079 + }, + { + "epoch": 0.5208, + "grad_norm": 0.7544088204820878, + "learning_rate": 0.003, + "loss": 3.986, + "step": 52080 + }, + { + "epoch": 0.52081, + "grad_norm": 0.8495116684411597, + "learning_rate": 0.003, + "loss": 3.957, + "step": 52081 + }, + { + "epoch": 0.52082, + "grad_norm": 0.8590651680818598, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 52082 + }, + { + "epoch": 0.52083, + "grad_norm": 0.8195399386581383, + "learning_rate": 0.003, + "loss": 3.9397, + "step": 52083 + }, + { + "epoch": 0.52084, + "grad_norm": 0.9963911017648819, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 52084 + }, + { + "epoch": 0.52085, + "grad_norm": 0.9292791227628499, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 52085 + }, + { + "epoch": 0.52086, + "grad_norm": 0.8736037940763188, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 52086 + }, + { + "epoch": 0.52087, + "grad_norm": 0.9219270849915074, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 52087 + }, + { + "epoch": 0.52088, + "grad_norm": 0.9449803690766128, + "learning_rate": 0.003, + "loss": 3.9864, + "step": 52088 + }, + { + "epoch": 0.52089, + "grad_norm": 0.9971959490227302, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 52089 + }, + { + "epoch": 0.5209, + "grad_norm": 0.9111414633071356, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 52090 + }, + { + "epoch": 0.52091, + "grad_norm": 0.9485709251498757, + "learning_rate": 0.003, + "loss": 4.0197, + "step": 52091 + }, + { + "epoch": 0.52092, + "grad_norm": 0.8665732611327225, + "learning_rate": 0.003, + "loss": 3.9958, + "step": 52092 + }, + { + "epoch": 0.52093, + "grad_norm": 0.8278585000773122, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 52093 + }, + { + "epoch": 0.52094, + "grad_norm": 0.866718359714595, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 52094 + }, + { + "epoch": 0.52095, + "grad_norm": 0.9050882654238327, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 52095 + }, + { + "epoch": 0.52096, + "grad_norm": 1.1154813707277882, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 52096 + }, + { + "epoch": 0.52097, + "grad_norm": 0.9250788354854881, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 52097 + }, + { + "epoch": 0.52098, + "grad_norm": 0.9711479583230664, + "learning_rate": 0.003, + "loss": 4.022, + "step": 52098 + }, + { + "epoch": 0.52099, + "grad_norm": 1.1156481261871039, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 52099 + }, + { + "epoch": 0.521, + "grad_norm": 1.1168430757513481, + "learning_rate": 0.003, + "loss": 3.9883, + "step": 52100 + }, + { + "epoch": 0.52101, + "grad_norm": 0.8394331122959791, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 52101 + }, + { + "epoch": 0.52102, + "grad_norm": 0.8237857651368506, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 52102 + }, + { + "epoch": 0.52103, + "grad_norm": 0.7853199836692467, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 52103 + }, + { + "epoch": 0.52104, + "grad_norm": 0.6717106242061, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 52104 + }, + { + "epoch": 0.52105, + "grad_norm": 0.6705373736733383, + "learning_rate": 0.003, + "loss": 3.974, + "step": 52105 + }, + { + "epoch": 0.52106, + "grad_norm": 0.723929703505044, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 52106 + }, + { + "epoch": 0.52107, + "grad_norm": 0.7267132558263587, + "learning_rate": 0.003, + "loss": 3.9915, + "step": 52107 + }, + { + "epoch": 0.52108, + "grad_norm": 0.7481959812669222, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 52108 + }, + { + "epoch": 0.52109, + "grad_norm": 0.8173937804051011, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 52109 + }, + { + "epoch": 0.5211, + "grad_norm": 0.8986934445738085, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 52110 + }, + { + "epoch": 0.52111, + "grad_norm": 0.7646550651598556, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 52111 + }, + { + "epoch": 0.52112, + "grad_norm": 0.5891435035219611, + "learning_rate": 0.003, + "loss": 3.9761, + "step": 52112 + }, + { + "epoch": 0.52113, + "grad_norm": 0.5849670602962049, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 52113 + }, + { + "epoch": 0.52114, + "grad_norm": 0.6345374258434293, + "learning_rate": 0.003, + "loss": 3.977, + "step": 52114 + }, + { + "epoch": 0.52115, + "grad_norm": 0.7003238898433732, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 52115 + }, + { + "epoch": 0.52116, + "grad_norm": 0.6529352938975752, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 52116 + }, + { + "epoch": 0.52117, + "grad_norm": 0.7348109403368315, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 52117 + }, + { + "epoch": 0.52118, + "grad_norm": 0.8204589578641344, + "learning_rate": 0.003, + "loss": 3.9655, + "step": 52118 + }, + { + "epoch": 0.52119, + "grad_norm": 0.8387278523137377, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 52119 + }, + { + "epoch": 0.5212, + "grad_norm": 0.6807349786990352, + "learning_rate": 0.003, + "loss": 3.9467, + "step": 52120 + }, + { + "epoch": 0.52121, + "grad_norm": 0.5776020277253089, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 52121 + }, + { + "epoch": 0.52122, + "grad_norm": 0.5850369388557511, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 52122 + }, + { + "epoch": 0.52123, + "grad_norm": 0.49531170092308197, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 52123 + }, + { + "epoch": 0.52124, + "grad_norm": 0.5443416743984982, + "learning_rate": 0.003, + "loss": 3.9683, + "step": 52124 + }, + { + "epoch": 0.52125, + "grad_norm": 0.5199278060799137, + "learning_rate": 0.003, + "loss": 3.965, + "step": 52125 + }, + { + "epoch": 0.52126, + "grad_norm": 0.5662441031301458, + "learning_rate": 0.003, + "loss": 3.9738, + "step": 52126 + }, + { + "epoch": 0.52127, + "grad_norm": 0.6425452170974095, + "learning_rate": 0.003, + "loss": 4.0051, + "step": 52127 + }, + { + "epoch": 0.52128, + "grad_norm": 0.7618234521098288, + "learning_rate": 0.003, + "loss": 4.0068, + "step": 52128 + }, + { + "epoch": 0.52129, + "grad_norm": 0.7880790139634736, + "learning_rate": 0.003, + "loss": 3.9383, + "step": 52129 + }, + { + "epoch": 0.5213, + "grad_norm": 0.7228101399482745, + "learning_rate": 0.003, + "loss": 4.0034, + "step": 52130 + }, + { + "epoch": 0.52131, + "grad_norm": 0.6753180828516462, + "learning_rate": 0.003, + "loss": 3.9845, + "step": 52131 + }, + { + "epoch": 0.52132, + "grad_norm": 0.6358113195608155, + "learning_rate": 0.003, + "loss": 3.9813, + "step": 52132 + }, + { + "epoch": 0.52133, + "grad_norm": 0.6918983740520528, + "learning_rate": 0.003, + "loss": 3.9441, + "step": 52133 + }, + { + "epoch": 0.52134, + "grad_norm": 0.7685014246321527, + "learning_rate": 0.003, + "loss": 3.9913, + "step": 52134 + }, + { + "epoch": 0.52135, + "grad_norm": 0.7800817075149306, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 52135 + }, + { + "epoch": 0.52136, + "grad_norm": 0.7444029804942031, + "learning_rate": 0.003, + "loss": 3.9499, + "step": 52136 + }, + { + "epoch": 0.52137, + "grad_norm": 0.8061016043707534, + "learning_rate": 0.003, + "loss": 3.9478, + "step": 52137 + }, + { + "epoch": 0.52138, + "grad_norm": 0.8538065369005102, + "learning_rate": 0.003, + "loss": 4.0011, + "step": 52138 + }, + { + "epoch": 0.52139, + "grad_norm": 0.9526219927623311, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 52139 + }, + { + "epoch": 0.5214, + "grad_norm": 0.9107341543324665, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 52140 + }, + { + "epoch": 0.52141, + "grad_norm": 0.8176962221999209, + "learning_rate": 0.003, + "loss": 4.0025, + "step": 52141 + }, + { + "epoch": 0.52142, + "grad_norm": 0.8493550788827345, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 52142 + }, + { + "epoch": 0.52143, + "grad_norm": 0.9609784931842624, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 52143 + }, + { + "epoch": 0.52144, + "grad_norm": 1.001632325774111, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 52144 + }, + { + "epoch": 0.52145, + "grad_norm": 0.9286013111944756, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 52145 + }, + { + "epoch": 0.52146, + "grad_norm": 0.9315709877738102, + "learning_rate": 0.003, + "loss": 3.9756, + "step": 52146 + }, + { + "epoch": 0.52147, + "grad_norm": 0.8572706366578698, + "learning_rate": 0.003, + "loss": 3.983, + "step": 52147 + }, + { + "epoch": 0.52148, + "grad_norm": 1.0279817540212461, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 52148 + }, + { + "epoch": 0.52149, + "grad_norm": 0.9789079535016494, + "learning_rate": 0.003, + "loss": 3.9886, + "step": 52149 + }, + { + "epoch": 0.5215, + "grad_norm": 0.8361172542665091, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 52150 + }, + { + "epoch": 0.52151, + "grad_norm": 0.8146145501130114, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 52151 + }, + { + "epoch": 0.52152, + "grad_norm": 0.8250154507794233, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 52152 + }, + { + "epoch": 0.52153, + "grad_norm": 0.9969681843964062, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 52153 + }, + { + "epoch": 0.52154, + "grad_norm": 1.0091734105851222, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 52154 + }, + { + "epoch": 0.52155, + "grad_norm": 0.9288582685742931, + "learning_rate": 0.003, + "loss": 4.0156, + "step": 52155 + }, + { + "epoch": 0.52156, + "grad_norm": 0.924944015101579, + "learning_rate": 0.003, + "loss": 3.986, + "step": 52156 + }, + { + "epoch": 0.52157, + "grad_norm": 0.7671533357837564, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 52157 + }, + { + "epoch": 0.52158, + "grad_norm": 0.67765820095138, + "learning_rate": 0.003, + "loss": 3.9657, + "step": 52158 + }, + { + "epoch": 0.52159, + "grad_norm": 0.6617797023516969, + "learning_rate": 0.003, + "loss": 3.9918, + "step": 52159 + }, + { + "epoch": 0.5216, + "grad_norm": 0.6342258770458663, + "learning_rate": 0.003, + "loss": 3.9644, + "step": 52160 + }, + { + "epoch": 0.52161, + "grad_norm": 0.7047529520135738, + "learning_rate": 0.003, + "loss": 3.9804, + "step": 52161 + }, + { + "epoch": 0.52162, + "grad_norm": 0.7403961664284212, + "learning_rate": 0.003, + "loss": 3.9759, + "step": 52162 + }, + { + "epoch": 0.52163, + "grad_norm": 0.6903244300046212, + "learning_rate": 0.003, + "loss": 3.9788, + "step": 52163 + }, + { + "epoch": 0.52164, + "grad_norm": 0.6205365483290292, + "learning_rate": 0.003, + "loss": 3.9776, + "step": 52164 + }, + { + "epoch": 0.52165, + "grad_norm": 0.6278658184448606, + "learning_rate": 0.003, + "loss": 4.0066, + "step": 52165 + }, + { + "epoch": 0.52166, + "grad_norm": 0.5960119323624292, + "learning_rate": 0.003, + "loss": 3.9568, + "step": 52166 + }, + { + "epoch": 0.52167, + "grad_norm": 0.5395543977769958, + "learning_rate": 0.003, + "loss": 3.9725, + "step": 52167 + }, + { + "epoch": 0.52168, + "grad_norm": 0.5189225271290022, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 52168 + }, + { + "epoch": 0.52169, + "grad_norm": 0.5063087355780638, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 52169 + }, + { + "epoch": 0.5217, + "grad_norm": 0.5293733956546373, + "learning_rate": 0.003, + "loss": 3.9729, + "step": 52170 + }, + { + "epoch": 0.52171, + "grad_norm": 0.5091305108331604, + "learning_rate": 0.003, + "loss": 3.9872, + "step": 52171 + }, + { + "epoch": 0.52172, + "grad_norm": 0.5572872070006373, + "learning_rate": 0.003, + "loss": 3.9595, + "step": 52172 + }, + { + "epoch": 0.52173, + "grad_norm": 0.5610537961721185, + "learning_rate": 0.003, + "loss": 3.9157, + "step": 52173 + }, + { + "epoch": 0.52174, + "grad_norm": 0.6106125238547649, + "learning_rate": 0.003, + "loss": 3.9623, + "step": 52174 + }, + { + "epoch": 0.52175, + "grad_norm": 0.6815246569198729, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 52175 + }, + { + "epoch": 0.52176, + "grad_norm": 0.7003502100905263, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 52176 + }, + { + "epoch": 0.52177, + "grad_norm": 0.7484519142973101, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 52177 + }, + { + "epoch": 0.52178, + "grad_norm": 0.8547412018184071, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 52178 + }, + { + "epoch": 0.52179, + "grad_norm": 0.9050345228882603, + "learning_rate": 0.003, + "loss": 3.952, + "step": 52179 + }, + { + "epoch": 0.5218, + "grad_norm": 0.9544873253746536, + "learning_rate": 0.003, + "loss": 3.9838, + "step": 52180 + }, + { + "epoch": 0.52181, + "grad_norm": 0.910966591386536, + "learning_rate": 0.003, + "loss": 3.9409, + "step": 52181 + }, + { + "epoch": 0.52182, + "grad_norm": 0.8496408927551251, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 52182 + }, + { + "epoch": 0.52183, + "grad_norm": 0.6879964364029644, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 52183 + }, + { + "epoch": 0.52184, + "grad_norm": 0.7590994534838952, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 52184 + }, + { + "epoch": 0.52185, + "grad_norm": 0.7176001463907895, + "learning_rate": 0.003, + "loss": 3.966, + "step": 52185 + }, + { + "epoch": 0.52186, + "grad_norm": 0.7695560931163284, + "learning_rate": 0.003, + "loss": 3.9935, + "step": 52186 + }, + { + "epoch": 0.52187, + "grad_norm": 0.8763510540503106, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 52187 + }, + { + "epoch": 0.52188, + "grad_norm": 1.006964157551516, + "learning_rate": 0.003, + "loss": 3.9997, + "step": 52188 + }, + { + "epoch": 0.52189, + "grad_norm": 0.9411117383698798, + "learning_rate": 0.003, + "loss": 4.0007, + "step": 52189 + }, + { + "epoch": 0.5219, + "grad_norm": 0.7251587015732048, + "learning_rate": 0.003, + "loss": 3.9732, + "step": 52190 + }, + { + "epoch": 0.52191, + "grad_norm": 0.7298159353953293, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 52191 + }, + { + "epoch": 0.52192, + "grad_norm": 0.7797822336235287, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 52192 + }, + { + "epoch": 0.52193, + "grad_norm": 0.9494580272314097, + "learning_rate": 0.003, + "loss": 3.973, + "step": 52193 + }, + { + "epoch": 0.52194, + "grad_norm": 0.9955302691799278, + "learning_rate": 0.003, + "loss": 3.9987, + "step": 52194 + }, + { + "epoch": 0.52195, + "grad_norm": 0.8839369485952505, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 52195 + }, + { + "epoch": 0.52196, + "grad_norm": 0.7905055952525933, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 52196 + }, + { + "epoch": 0.52197, + "grad_norm": 0.8858748889830819, + "learning_rate": 0.003, + "loss": 3.9983, + "step": 52197 + }, + { + "epoch": 0.52198, + "grad_norm": 0.7937921032792483, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 52198 + }, + { + "epoch": 0.52199, + "grad_norm": 0.758120929918194, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 52199 + }, + { + "epoch": 0.522, + "grad_norm": 0.7642432028040033, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 52200 + }, + { + "epoch": 0.52201, + "grad_norm": 0.8273799447373058, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 52201 + }, + { + "epoch": 0.52202, + "grad_norm": 0.8275253529171501, + "learning_rate": 0.003, + "loss": 3.9863, + "step": 52202 + }, + { + "epoch": 0.52203, + "grad_norm": 0.8269829547440747, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 52203 + }, + { + "epoch": 0.52204, + "grad_norm": 0.678877321024153, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 52204 + }, + { + "epoch": 0.52205, + "grad_norm": 0.7471880143635052, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 52205 + }, + { + "epoch": 0.52206, + "grad_norm": 0.7575820214831106, + "learning_rate": 0.003, + "loss": 3.9802, + "step": 52206 + }, + { + "epoch": 0.52207, + "grad_norm": 0.7418011525606533, + "learning_rate": 0.003, + "loss": 4.0155, + "step": 52207 + }, + { + "epoch": 0.52208, + "grad_norm": 0.7430766308029968, + "learning_rate": 0.003, + "loss": 3.9622, + "step": 52208 + }, + { + "epoch": 0.52209, + "grad_norm": 0.7544300483491874, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 52209 + }, + { + "epoch": 0.5221, + "grad_norm": 0.8505405944768304, + "learning_rate": 0.003, + "loss": 3.9766, + "step": 52210 + }, + { + "epoch": 0.52211, + "grad_norm": 1.0591288618776942, + "learning_rate": 0.003, + "loss": 3.9828, + "step": 52211 + }, + { + "epoch": 0.52212, + "grad_norm": 1.044615962477642, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 52212 + }, + { + "epoch": 0.52213, + "grad_norm": 0.7671313135456987, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 52213 + }, + { + "epoch": 0.52214, + "grad_norm": 0.6699289420192076, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 52214 + }, + { + "epoch": 0.52215, + "grad_norm": 0.7181756544766006, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 52215 + }, + { + "epoch": 0.52216, + "grad_norm": 0.7514435870625349, + "learning_rate": 0.003, + "loss": 3.9752, + "step": 52216 + }, + { + "epoch": 0.52217, + "grad_norm": 0.8682380663073911, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 52217 + }, + { + "epoch": 0.52218, + "grad_norm": 0.8764016598450861, + "learning_rate": 0.003, + "loss": 3.966, + "step": 52218 + }, + { + "epoch": 0.52219, + "grad_norm": 0.8764579940379301, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 52219 + }, + { + "epoch": 0.5222, + "grad_norm": 0.8432900962376805, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 52220 + }, + { + "epoch": 0.52221, + "grad_norm": 0.7142297221624685, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 52221 + }, + { + "epoch": 0.52222, + "grad_norm": 0.7401115219898503, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 52222 + }, + { + "epoch": 0.52223, + "grad_norm": 0.758175438382133, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 52223 + }, + { + "epoch": 0.52224, + "grad_norm": 0.7644809386128751, + "learning_rate": 0.003, + "loss": 3.9924, + "step": 52224 + }, + { + "epoch": 0.52225, + "grad_norm": 0.7353347420765637, + "learning_rate": 0.003, + "loss": 3.974, + "step": 52225 + }, + { + "epoch": 0.52226, + "grad_norm": 0.7881723365158599, + "learning_rate": 0.003, + "loss": 3.987, + "step": 52226 + }, + { + "epoch": 0.52227, + "grad_norm": 0.7529484419473531, + "learning_rate": 0.003, + "loss": 3.9866, + "step": 52227 + }, + { + "epoch": 0.52228, + "grad_norm": 0.7322430749895511, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 52228 + }, + { + "epoch": 0.52229, + "grad_norm": 0.6957318177772608, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 52229 + }, + { + "epoch": 0.5223, + "grad_norm": 0.7019420994585258, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 52230 + }, + { + "epoch": 0.52231, + "grad_norm": 0.6782773888503824, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 52231 + }, + { + "epoch": 0.52232, + "grad_norm": 0.4942578655343354, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 52232 + }, + { + "epoch": 0.52233, + "grad_norm": 0.47856502849284455, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 52233 + }, + { + "epoch": 0.52234, + "grad_norm": 0.47442046106799124, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 52234 + }, + { + "epoch": 0.52235, + "grad_norm": 0.5336403096637523, + "learning_rate": 0.003, + "loss": 4.0114, + "step": 52235 + }, + { + "epoch": 0.52236, + "grad_norm": 0.6336946366561519, + "learning_rate": 0.003, + "loss": 3.9664, + "step": 52236 + }, + { + "epoch": 0.52237, + "grad_norm": 0.7277823146286544, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 52237 + }, + { + "epoch": 0.52238, + "grad_norm": 0.7724533429627617, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 52238 + }, + { + "epoch": 0.52239, + "grad_norm": 0.7000503926794519, + "learning_rate": 0.003, + "loss": 3.9728, + "step": 52239 + }, + { + "epoch": 0.5224, + "grad_norm": 0.6592530800365662, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 52240 + }, + { + "epoch": 0.52241, + "grad_norm": 0.6870053921254524, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 52241 + }, + { + "epoch": 0.52242, + "grad_norm": 0.8260717350935646, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 52242 + }, + { + "epoch": 0.52243, + "grad_norm": 0.9184033213643624, + "learning_rate": 0.003, + "loss": 3.9933, + "step": 52243 + }, + { + "epoch": 0.52244, + "grad_norm": 0.9413680994739213, + "learning_rate": 0.003, + "loss": 3.9824, + "step": 52244 + }, + { + "epoch": 0.52245, + "grad_norm": 1.069545504770155, + "learning_rate": 0.003, + "loss": 4.0121, + "step": 52245 + }, + { + "epoch": 0.52246, + "grad_norm": 0.9682454809922061, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 52246 + }, + { + "epoch": 0.52247, + "grad_norm": 0.840689488004601, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 52247 + }, + { + "epoch": 0.52248, + "grad_norm": 0.907332532863088, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 52248 + }, + { + "epoch": 0.52249, + "grad_norm": 0.7259142933613127, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 52249 + }, + { + "epoch": 0.5225, + "grad_norm": 0.7896009109323708, + "learning_rate": 0.003, + "loss": 3.9914, + "step": 52250 + }, + { + "epoch": 0.52251, + "grad_norm": 0.765417202165686, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 52251 + }, + { + "epoch": 0.52252, + "grad_norm": 0.9267906065540554, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 52252 + }, + { + "epoch": 0.52253, + "grad_norm": 1.0339858420658652, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 52253 + }, + { + "epoch": 0.52254, + "grad_norm": 1.085588474255267, + "learning_rate": 0.003, + "loss": 3.9784, + "step": 52254 + }, + { + "epoch": 0.52255, + "grad_norm": 0.9185991425711042, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 52255 + }, + { + "epoch": 0.52256, + "grad_norm": 0.8997087765189067, + "learning_rate": 0.003, + "loss": 3.9991, + "step": 52256 + }, + { + "epoch": 0.52257, + "grad_norm": 0.8005127636884857, + "learning_rate": 0.003, + "loss": 3.983, + "step": 52257 + }, + { + "epoch": 0.52258, + "grad_norm": 0.7085742816844177, + "learning_rate": 0.003, + "loss": 3.9881, + "step": 52258 + }, + { + "epoch": 0.52259, + "grad_norm": 0.7144241433752274, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 52259 + }, + { + "epoch": 0.5226, + "grad_norm": 0.7364680966508531, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 52260 + }, + { + "epoch": 0.52261, + "grad_norm": 0.8097644855081388, + "learning_rate": 0.003, + "loss": 3.9736, + "step": 52261 + }, + { + "epoch": 0.52262, + "grad_norm": 0.6841100497203786, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 52262 + }, + { + "epoch": 0.52263, + "grad_norm": 0.7372383250009905, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 52263 + }, + { + "epoch": 0.52264, + "grad_norm": 0.6308687286552466, + "learning_rate": 0.003, + "loss": 3.9457, + "step": 52264 + }, + { + "epoch": 0.52265, + "grad_norm": 0.6627440293747998, + "learning_rate": 0.003, + "loss": 3.964, + "step": 52265 + }, + { + "epoch": 0.52266, + "grad_norm": 0.7164914326307051, + "learning_rate": 0.003, + "loss": 3.9549, + "step": 52266 + }, + { + "epoch": 0.52267, + "grad_norm": 0.7426527491938273, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 52267 + }, + { + "epoch": 0.52268, + "grad_norm": 0.7600566780745521, + "learning_rate": 0.003, + "loss": 3.9719, + "step": 52268 + }, + { + "epoch": 0.52269, + "grad_norm": 0.7627282260443946, + "learning_rate": 0.003, + "loss": 3.9792, + "step": 52269 + }, + { + "epoch": 0.5227, + "grad_norm": 0.6683639867628698, + "learning_rate": 0.003, + "loss": 3.9442, + "step": 52270 + }, + { + "epoch": 0.52271, + "grad_norm": 0.6596805571645368, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 52271 + }, + { + "epoch": 0.52272, + "grad_norm": 0.707421340577653, + "learning_rate": 0.003, + "loss": 3.9876, + "step": 52272 + }, + { + "epoch": 0.52273, + "grad_norm": 0.791400225105772, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 52273 + }, + { + "epoch": 0.52274, + "grad_norm": 0.8300172773419022, + "learning_rate": 0.003, + "loss": 3.991, + "step": 52274 + }, + { + "epoch": 0.52275, + "grad_norm": 0.8226101617118229, + "learning_rate": 0.003, + "loss": 3.979, + "step": 52275 + }, + { + "epoch": 0.52276, + "grad_norm": 0.7340371903346149, + "learning_rate": 0.003, + "loss": 3.9536, + "step": 52276 + }, + { + "epoch": 0.52277, + "grad_norm": 0.6809417480208284, + "learning_rate": 0.003, + "loss": 3.9439, + "step": 52277 + }, + { + "epoch": 0.52278, + "grad_norm": 0.5867769212578543, + "learning_rate": 0.003, + "loss": 3.9871, + "step": 52278 + }, + { + "epoch": 0.52279, + "grad_norm": 0.5466842037225674, + "learning_rate": 0.003, + "loss": 3.9487, + "step": 52279 + }, + { + "epoch": 0.5228, + "grad_norm": 0.6734721171525069, + "learning_rate": 0.003, + "loss": 3.9642, + "step": 52280 + }, + { + "epoch": 0.52281, + "grad_norm": 0.6880792408654349, + "learning_rate": 0.003, + "loss": 3.9638, + "step": 52281 + }, + { + "epoch": 0.52282, + "grad_norm": 0.7193667314824124, + "learning_rate": 0.003, + "loss": 3.9602, + "step": 52282 + }, + { + "epoch": 0.52283, + "grad_norm": 0.7927708000741325, + "learning_rate": 0.003, + "loss": 3.9572, + "step": 52283 + }, + { + "epoch": 0.52284, + "grad_norm": 0.8865766138720825, + "learning_rate": 0.003, + "loss": 3.9884, + "step": 52284 + }, + { + "epoch": 0.52285, + "grad_norm": 0.9590878873933184, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 52285 + }, + { + "epoch": 0.52286, + "grad_norm": 0.8610851236166897, + "learning_rate": 0.003, + "loss": 3.9717, + "step": 52286 + }, + { + "epoch": 0.52287, + "grad_norm": 0.739959302232148, + "learning_rate": 0.003, + "loss": 3.9692, + "step": 52287 + }, + { + "epoch": 0.52288, + "grad_norm": 0.7907686266144114, + "learning_rate": 0.003, + "loss": 3.9632, + "step": 52288 + }, + { + "epoch": 0.52289, + "grad_norm": 0.8917148332950889, + "learning_rate": 0.003, + "loss": 3.9775, + "step": 52289 + }, + { + "epoch": 0.5229, + "grad_norm": 0.8977262390662344, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 52290 + }, + { + "epoch": 0.52291, + "grad_norm": 0.9328053140036096, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 52291 + }, + { + "epoch": 0.52292, + "grad_norm": 0.7376419599566089, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 52292 + }, + { + "epoch": 0.52293, + "grad_norm": 0.7710643539143978, + "learning_rate": 0.003, + "loss": 3.9934, + "step": 52293 + }, + { + "epoch": 0.52294, + "grad_norm": 0.7901759935385325, + "learning_rate": 0.003, + "loss": 4.0265, + "step": 52294 + }, + { + "epoch": 0.52295, + "grad_norm": 0.8537334219394428, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 52295 + }, + { + "epoch": 0.52296, + "grad_norm": 0.8469429667790844, + "learning_rate": 0.003, + "loss": 3.988, + "step": 52296 + }, + { + "epoch": 0.52297, + "grad_norm": 0.8551734409280911, + "learning_rate": 0.003, + "loss": 3.9862, + "step": 52297 + }, + { + "epoch": 0.52298, + "grad_norm": 0.6955085830811187, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 52298 + }, + { + "epoch": 0.52299, + "grad_norm": 0.6570827880318949, + "learning_rate": 0.003, + "loss": 3.9492, + "step": 52299 + }, + { + "epoch": 0.523, + "grad_norm": 0.6538229129744411, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 52300 + }, + { + "epoch": 0.52301, + "grad_norm": 0.6962077640198177, + "learning_rate": 0.003, + "loss": 3.9986, + "step": 52301 + }, + { + "epoch": 0.52302, + "grad_norm": 0.6276888364933791, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 52302 + }, + { + "epoch": 0.52303, + "grad_norm": 0.6937288100638863, + "learning_rate": 0.003, + "loss": 3.9618, + "step": 52303 + }, + { + "epoch": 0.52304, + "grad_norm": 0.6170901385477837, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 52304 + }, + { + "epoch": 0.52305, + "grad_norm": 0.5261737311378728, + "learning_rate": 0.003, + "loss": 4.0176, + "step": 52305 + }, + { + "epoch": 0.52306, + "grad_norm": 0.536731257668043, + "learning_rate": 0.003, + "loss": 3.9891, + "step": 52306 + }, + { + "epoch": 0.52307, + "grad_norm": 0.5626217391067821, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 52307 + }, + { + "epoch": 0.52308, + "grad_norm": 0.6453491251611153, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 52308 + }, + { + "epoch": 0.52309, + "grad_norm": 0.7280400375041499, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 52309 + }, + { + "epoch": 0.5231, + "grad_norm": 0.7416750769902253, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 52310 + }, + { + "epoch": 0.52311, + "grad_norm": 0.8276389523879955, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 52311 + }, + { + "epoch": 0.52312, + "grad_norm": 0.9469454770729077, + "learning_rate": 0.003, + "loss": 3.9939, + "step": 52312 + }, + { + "epoch": 0.52313, + "grad_norm": 1.0870439632968205, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 52313 + }, + { + "epoch": 0.52314, + "grad_norm": 0.8336647056783697, + "learning_rate": 0.003, + "loss": 3.9964, + "step": 52314 + }, + { + "epoch": 0.52315, + "grad_norm": 0.8118974214180757, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 52315 + }, + { + "epoch": 0.52316, + "grad_norm": 0.7830038931447811, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 52316 + }, + { + "epoch": 0.52317, + "grad_norm": 0.8058485337519669, + "learning_rate": 0.003, + "loss": 3.9978, + "step": 52317 + }, + { + "epoch": 0.52318, + "grad_norm": 0.7888254553094122, + "learning_rate": 0.003, + "loss": 3.9896, + "step": 52318 + }, + { + "epoch": 0.52319, + "grad_norm": 0.8286358912088918, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 52319 + }, + { + "epoch": 0.5232, + "grad_norm": 0.7808350180239084, + "learning_rate": 0.003, + "loss": 3.9832, + "step": 52320 + }, + { + "epoch": 0.52321, + "grad_norm": 0.8556339762561865, + "learning_rate": 0.003, + "loss": 3.9639, + "step": 52321 + }, + { + "epoch": 0.52322, + "grad_norm": 0.8508454102158259, + "learning_rate": 0.003, + "loss": 3.9751, + "step": 52322 + }, + { + "epoch": 0.52323, + "grad_norm": 0.7635224311981649, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 52323 + }, + { + "epoch": 0.52324, + "grad_norm": 0.828019235262249, + "learning_rate": 0.003, + "loss": 3.9979, + "step": 52324 + }, + { + "epoch": 0.52325, + "grad_norm": 0.9028190806256559, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 52325 + }, + { + "epoch": 0.52326, + "grad_norm": 1.0583874735378531, + "learning_rate": 0.003, + "loss": 4.0029, + "step": 52326 + }, + { + "epoch": 0.52327, + "grad_norm": 0.82034443659247, + "learning_rate": 0.003, + "loss": 4.0162, + "step": 52327 + }, + { + "epoch": 0.52328, + "grad_norm": 0.8060441058495396, + "learning_rate": 0.003, + "loss": 3.9857, + "step": 52328 + }, + { + "epoch": 0.52329, + "grad_norm": 0.9146242808653473, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 52329 + }, + { + "epoch": 0.5233, + "grad_norm": 0.9253641045095262, + "learning_rate": 0.003, + "loss": 4.0209, + "step": 52330 + }, + { + "epoch": 0.52331, + "grad_norm": 1.0194311245204897, + "learning_rate": 0.003, + "loss": 4.0339, + "step": 52331 + }, + { + "epoch": 0.52332, + "grad_norm": 0.9554987150622735, + "learning_rate": 0.003, + "loss": 4.0131, + "step": 52332 + }, + { + "epoch": 0.52333, + "grad_norm": 0.898300726721267, + "learning_rate": 0.003, + "loss": 4.0036, + "step": 52333 + }, + { + "epoch": 0.52334, + "grad_norm": 0.824546520323368, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 52334 + }, + { + "epoch": 0.52335, + "grad_norm": 0.8966716666367525, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 52335 + }, + { + "epoch": 0.52336, + "grad_norm": 0.9220285295966955, + "learning_rate": 0.003, + "loss": 4.0129, + "step": 52336 + }, + { + "epoch": 0.52337, + "grad_norm": 0.9665341059473499, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 52337 + }, + { + "epoch": 0.52338, + "grad_norm": 0.9887060088428383, + "learning_rate": 0.003, + "loss": 3.9961, + "step": 52338 + }, + { + "epoch": 0.52339, + "grad_norm": 0.9198058848603632, + "learning_rate": 0.003, + "loss": 3.9856, + "step": 52339 + }, + { + "epoch": 0.5234, + "grad_norm": 0.8291036492374617, + "learning_rate": 0.003, + "loss": 3.9735, + "step": 52340 + }, + { + "epoch": 0.52341, + "grad_norm": 0.7852012931256086, + "learning_rate": 0.003, + "loss": 3.983, + "step": 52341 + }, + { + "epoch": 0.52342, + "grad_norm": 0.6797937578890951, + "learning_rate": 0.003, + "loss": 3.996, + "step": 52342 + }, + { + "epoch": 0.52343, + "grad_norm": 0.7457963287249918, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 52343 + }, + { + "epoch": 0.52344, + "grad_norm": 0.6842197429168618, + "learning_rate": 0.003, + "loss": 3.9852, + "step": 52344 + }, + { + "epoch": 0.52345, + "grad_norm": 0.6587802620241526, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 52345 + }, + { + "epoch": 0.52346, + "grad_norm": 0.6285247605391844, + "learning_rate": 0.003, + "loss": 3.9455, + "step": 52346 + }, + { + "epoch": 0.52347, + "grad_norm": 0.6106727081179569, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 52347 + }, + { + "epoch": 0.52348, + "grad_norm": 0.6357724229678613, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 52348 + }, + { + "epoch": 0.52349, + "grad_norm": 0.7367226552742204, + "learning_rate": 0.003, + "loss": 3.9849, + "step": 52349 + }, + { + "epoch": 0.5235, + "grad_norm": 0.7751910022151247, + "learning_rate": 0.003, + "loss": 3.9637, + "step": 52350 + }, + { + "epoch": 0.52351, + "grad_norm": 0.750913155660454, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 52351 + }, + { + "epoch": 0.52352, + "grad_norm": 0.7871467457683083, + "learning_rate": 0.003, + "loss": 3.9624, + "step": 52352 + }, + { + "epoch": 0.52353, + "grad_norm": 0.8214048469924692, + "learning_rate": 0.003, + "loss": 3.9831, + "step": 52353 + }, + { + "epoch": 0.52354, + "grad_norm": 0.8411763122406926, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 52354 + }, + { + "epoch": 0.52355, + "grad_norm": 0.8073786168882774, + "learning_rate": 0.003, + "loss": 3.966, + "step": 52355 + }, + { + "epoch": 0.52356, + "grad_norm": 0.7355094585058312, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 52356 + }, + { + "epoch": 0.52357, + "grad_norm": 0.650911438914122, + "learning_rate": 0.003, + "loss": 4.0225, + "step": 52357 + }, + { + "epoch": 0.52358, + "grad_norm": 0.610318280059515, + "learning_rate": 0.003, + "loss": 3.9936, + "step": 52358 + }, + { + "epoch": 0.52359, + "grad_norm": 0.5897539800970019, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 52359 + }, + { + "epoch": 0.5236, + "grad_norm": 0.6229246585170298, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 52360 + }, + { + "epoch": 0.52361, + "grad_norm": 0.6667674751598099, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 52361 + }, + { + "epoch": 0.52362, + "grad_norm": 0.7916266378471538, + "learning_rate": 0.003, + "loss": 3.9385, + "step": 52362 + }, + { + "epoch": 0.52363, + "grad_norm": 0.7778544388321764, + "learning_rate": 0.003, + "loss": 3.9999, + "step": 52363 + }, + { + "epoch": 0.52364, + "grad_norm": 0.6912328864426462, + "learning_rate": 0.003, + "loss": 3.9773, + "step": 52364 + }, + { + "epoch": 0.52365, + "grad_norm": 0.6102793195257177, + "learning_rate": 0.003, + "loss": 4.0006, + "step": 52365 + }, + { + "epoch": 0.52366, + "grad_norm": 0.6741489359309918, + "learning_rate": 0.003, + "loss": 3.9507, + "step": 52366 + }, + { + "epoch": 0.52367, + "grad_norm": 0.6770609884475031, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 52367 + }, + { + "epoch": 0.52368, + "grad_norm": 0.6452725661552073, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 52368 + }, + { + "epoch": 0.52369, + "grad_norm": 0.7133867377561617, + "learning_rate": 0.003, + "loss": 3.9812, + "step": 52369 + }, + { + "epoch": 0.5237, + "grad_norm": 0.75448706396126, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 52370 + }, + { + "epoch": 0.52371, + "grad_norm": 0.7183093447014606, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 52371 + }, + { + "epoch": 0.52372, + "grad_norm": 0.6403270576179153, + "learning_rate": 0.003, + "loss": 3.9691, + "step": 52372 + }, + { + "epoch": 0.52373, + "grad_norm": 0.6339854346401954, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 52373 + }, + { + "epoch": 0.52374, + "grad_norm": 0.6619783348022499, + "learning_rate": 0.003, + "loss": 3.9493, + "step": 52374 + }, + { + "epoch": 0.52375, + "grad_norm": 0.6657416589044135, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 52375 + }, + { + "epoch": 0.52376, + "grad_norm": 0.6762566019643882, + "learning_rate": 0.003, + "loss": 3.9538, + "step": 52376 + }, + { + "epoch": 0.52377, + "grad_norm": 0.7767652913486668, + "learning_rate": 0.003, + "loss": 3.961, + "step": 52377 + }, + { + "epoch": 0.52378, + "grad_norm": 0.8630192150356618, + "learning_rate": 0.003, + "loss": 4.001, + "step": 52378 + }, + { + "epoch": 0.52379, + "grad_norm": 0.8585928770586735, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 52379 + }, + { + "epoch": 0.5238, + "grad_norm": 0.9093332751179625, + "learning_rate": 0.003, + "loss": 3.967, + "step": 52380 + }, + { + "epoch": 0.52381, + "grad_norm": 0.9485170495840596, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 52381 + }, + { + "epoch": 0.52382, + "grad_norm": 0.8160045010022652, + "learning_rate": 0.003, + "loss": 3.9744, + "step": 52382 + }, + { + "epoch": 0.52383, + "grad_norm": 0.6461329110009527, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 52383 + }, + { + "epoch": 0.52384, + "grad_norm": 0.6777265576319738, + "learning_rate": 0.003, + "loss": 3.9701, + "step": 52384 + }, + { + "epoch": 0.52385, + "grad_norm": 0.7143006337699808, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 52385 + }, + { + "epoch": 0.52386, + "grad_norm": 0.7556871444679337, + "learning_rate": 0.003, + "loss": 4.0087, + "step": 52386 + }, + { + "epoch": 0.52387, + "grad_norm": 0.8373476461630203, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 52387 + }, + { + "epoch": 0.52388, + "grad_norm": 0.9029592961609029, + "learning_rate": 0.003, + "loss": 3.9972, + "step": 52388 + }, + { + "epoch": 0.52389, + "grad_norm": 0.9683466914485335, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 52389 + }, + { + "epoch": 0.5239, + "grad_norm": 0.9369906117045026, + "learning_rate": 0.003, + "loss": 3.995, + "step": 52390 + }, + { + "epoch": 0.52391, + "grad_norm": 0.9032188517947554, + "learning_rate": 0.003, + "loss": 4.0047, + "step": 52391 + }, + { + "epoch": 0.52392, + "grad_norm": 0.9549096873445367, + "learning_rate": 0.003, + "loss": 4.0277, + "step": 52392 + }, + { + "epoch": 0.52393, + "grad_norm": 0.9296843503016462, + "learning_rate": 0.003, + "loss": 3.9724, + "step": 52393 + }, + { + "epoch": 0.52394, + "grad_norm": 0.7218329544395403, + "learning_rate": 0.003, + "loss": 3.9941, + "step": 52394 + }, + { + "epoch": 0.52395, + "grad_norm": 0.8799352922046216, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 52395 + }, + { + "epoch": 0.52396, + "grad_norm": 1.0301473626077366, + "learning_rate": 0.003, + "loss": 4.0228, + "step": 52396 + }, + { + "epoch": 0.52397, + "grad_norm": 1.043912354069367, + "learning_rate": 0.003, + "loss": 4.0081, + "step": 52397 + }, + { + "epoch": 0.52398, + "grad_norm": 0.7934420737736398, + "learning_rate": 0.003, + "loss": 3.999, + "step": 52398 + }, + { + "epoch": 0.52399, + "grad_norm": 0.7898222978393095, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 52399 + }, + { + "epoch": 0.524, + "grad_norm": 0.9328194388028664, + "learning_rate": 0.003, + "loss": 3.9836, + "step": 52400 + }, + { + "epoch": 0.52401, + "grad_norm": 1.0069332696941669, + "learning_rate": 0.003, + "loss": 3.9687, + "step": 52401 + }, + { + "epoch": 0.52402, + "grad_norm": 0.9410452983884419, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 52402 + }, + { + "epoch": 0.52403, + "grad_norm": 0.9314854120054926, + "learning_rate": 0.003, + "loss": 4.0017, + "step": 52403 + }, + { + "epoch": 0.52404, + "grad_norm": 0.9608777166776519, + "learning_rate": 0.003, + "loss": 4.0292, + "step": 52404 + }, + { + "epoch": 0.52405, + "grad_norm": 0.9950846178857069, + "learning_rate": 0.003, + "loss": 3.9807, + "step": 52405 + }, + { + "epoch": 0.52406, + "grad_norm": 0.97076208326229, + "learning_rate": 0.003, + "loss": 4.0095, + "step": 52406 + }, + { + "epoch": 0.52407, + "grad_norm": 0.9659767742293452, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 52407 + }, + { + "epoch": 0.52408, + "grad_norm": 0.894767987728884, + "learning_rate": 0.003, + "loss": 4.0196, + "step": 52408 + }, + { + "epoch": 0.52409, + "grad_norm": 0.6735546179627484, + "learning_rate": 0.003, + "loss": 4.0005, + "step": 52409 + }, + { + "epoch": 0.5241, + "grad_norm": 0.6591350558574669, + "learning_rate": 0.003, + "loss": 3.996, + "step": 52410 + }, + { + "epoch": 0.52411, + "grad_norm": 0.6934918050144744, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 52411 + }, + { + "epoch": 0.52412, + "grad_norm": 0.7465257290489464, + "learning_rate": 0.003, + "loss": 3.9835, + "step": 52412 + }, + { + "epoch": 0.52413, + "grad_norm": 0.7785490528251272, + "learning_rate": 0.003, + "loss": 4.0086, + "step": 52413 + }, + { + "epoch": 0.52414, + "grad_norm": 0.7547743753654781, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 52414 + }, + { + "epoch": 0.52415, + "grad_norm": 0.7213361402697858, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 52415 + }, + { + "epoch": 0.52416, + "grad_norm": 0.7395696428548418, + "learning_rate": 0.003, + "loss": 3.9851, + "step": 52416 + }, + { + "epoch": 0.52417, + "grad_norm": 0.7276277133458735, + "learning_rate": 0.003, + "loss": 3.9654, + "step": 52417 + }, + { + "epoch": 0.52418, + "grad_norm": 0.7126379301957574, + "learning_rate": 0.003, + "loss": 3.9667, + "step": 52418 + }, + { + "epoch": 0.52419, + "grad_norm": 0.6711595425892419, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 52419 + }, + { + "epoch": 0.5242, + "grad_norm": 0.6069027770358401, + "learning_rate": 0.003, + "loss": 3.9841, + "step": 52420 + }, + { + "epoch": 0.52421, + "grad_norm": 0.5475959527341789, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 52421 + }, + { + "epoch": 0.52422, + "grad_norm": 0.5426907740978916, + "learning_rate": 0.003, + "loss": 3.9506, + "step": 52422 + }, + { + "epoch": 0.52423, + "grad_norm": 0.5400611544211813, + "learning_rate": 0.003, + "loss": 3.966, + "step": 52423 + }, + { + "epoch": 0.52424, + "grad_norm": 0.5304894386153534, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 52424 + }, + { + "epoch": 0.52425, + "grad_norm": 0.5814199003399341, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 52425 + }, + { + "epoch": 0.52426, + "grad_norm": 0.6239383761468618, + "learning_rate": 0.003, + "loss": 3.9452, + "step": 52426 + }, + { + "epoch": 0.52427, + "grad_norm": 0.7189150179440253, + "learning_rate": 0.003, + "loss": 3.9634, + "step": 52427 + }, + { + "epoch": 0.52428, + "grad_norm": 0.825533609654921, + "learning_rate": 0.003, + "loss": 3.9584, + "step": 52428 + }, + { + "epoch": 0.52429, + "grad_norm": 0.9047820563249681, + "learning_rate": 0.003, + "loss": 3.9518, + "step": 52429 + }, + { + "epoch": 0.5243, + "grad_norm": 0.9115245288313102, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 52430 + }, + { + "epoch": 0.52431, + "grad_norm": 0.7817210030790808, + "learning_rate": 0.003, + "loss": 3.9577, + "step": 52431 + }, + { + "epoch": 0.52432, + "grad_norm": 0.6003841432276847, + "learning_rate": 0.003, + "loss": 3.9822, + "step": 52432 + }, + { + "epoch": 0.52433, + "grad_norm": 0.6120384971352101, + "learning_rate": 0.003, + "loss": 3.9698, + "step": 52433 + }, + { + "epoch": 0.52434, + "grad_norm": 0.5800261405797782, + "learning_rate": 0.003, + "loss": 3.9764, + "step": 52434 + }, + { + "epoch": 0.52435, + "grad_norm": 0.5169066438310583, + "learning_rate": 0.003, + "loss": 3.909, + "step": 52435 + }, + { + "epoch": 0.52436, + "grad_norm": 0.4963580096018927, + "learning_rate": 0.003, + "loss": 3.9517, + "step": 52436 + }, + { + "epoch": 0.52437, + "grad_norm": 0.5172953033648973, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 52437 + }, + { + "epoch": 0.52438, + "grad_norm": 0.5376574256690019, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 52438 + }, + { + "epoch": 0.52439, + "grad_norm": 0.6147281982403515, + "learning_rate": 0.003, + "loss": 3.9626, + "step": 52439 + }, + { + "epoch": 0.5244, + "grad_norm": 0.634937088618743, + "learning_rate": 0.003, + "loss": 3.9531, + "step": 52440 + }, + { + "epoch": 0.52441, + "grad_norm": 0.6424507380758622, + "learning_rate": 0.003, + "loss": 3.9693, + "step": 52441 + }, + { + "epoch": 0.52442, + "grad_norm": 0.7292960780813624, + "learning_rate": 0.003, + "loss": 3.9685, + "step": 52442 + }, + { + "epoch": 0.52443, + "grad_norm": 0.8197922048763575, + "learning_rate": 0.003, + "loss": 3.976, + "step": 52443 + }, + { + "epoch": 0.52444, + "grad_norm": 0.8058902416237554, + "learning_rate": 0.003, + "loss": 3.9679, + "step": 52444 + }, + { + "epoch": 0.52445, + "grad_norm": 0.7410280497924341, + "learning_rate": 0.003, + "loss": 3.9588, + "step": 52445 + }, + { + "epoch": 0.52446, + "grad_norm": 0.734693532833192, + "learning_rate": 0.003, + "loss": 3.9943, + "step": 52446 + }, + { + "epoch": 0.52447, + "grad_norm": 0.613044713979019, + "learning_rate": 0.003, + "loss": 3.9616, + "step": 52447 + }, + { + "epoch": 0.52448, + "grad_norm": 0.7219775335866907, + "learning_rate": 0.003, + "loss": 3.9614, + "step": 52448 + }, + { + "epoch": 0.52449, + "grad_norm": 0.7066681662901284, + "learning_rate": 0.003, + "loss": 3.9516, + "step": 52449 + }, + { + "epoch": 0.5245, + "grad_norm": 0.8319559437906202, + "learning_rate": 0.003, + "loss": 4.0042, + "step": 52450 + }, + { + "epoch": 0.52451, + "grad_norm": 0.9968893567190139, + "learning_rate": 0.003, + "loss": 3.9793, + "step": 52451 + }, + { + "epoch": 0.52452, + "grad_norm": 1.3283804075685746, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 52452 + }, + { + "epoch": 0.52453, + "grad_norm": 0.8124376936688336, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 52453 + }, + { + "epoch": 0.52454, + "grad_norm": 0.7874360348343608, + "learning_rate": 0.003, + "loss": 4.0143, + "step": 52454 + }, + { + "epoch": 0.52455, + "grad_norm": 0.7333501336747404, + "learning_rate": 0.003, + "loss": 3.9633, + "step": 52455 + }, + { + "epoch": 0.52456, + "grad_norm": 0.7901868278823965, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 52456 + }, + { + "epoch": 0.52457, + "grad_norm": 0.8423910256525639, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 52457 + }, + { + "epoch": 0.52458, + "grad_norm": 0.9209087555739339, + "learning_rate": 0.003, + "loss": 4.0061, + "step": 52458 + }, + { + "epoch": 0.52459, + "grad_norm": 0.9408499151112059, + "learning_rate": 0.003, + "loss": 4.0147, + "step": 52459 + }, + { + "epoch": 0.5246, + "grad_norm": 1.0559718101004094, + "learning_rate": 0.003, + "loss": 3.9951, + "step": 52460 + }, + { + "epoch": 0.52461, + "grad_norm": 0.855303235823543, + "learning_rate": 0.003, + "loss": 3.9903, + "step": 52461 + }, + { + "epoch": 0.52462, + "grad_norm": 0.7447278183422813, + "learning_rate": 0.003, + "loss": 3.9496, + "step": 52462 + }, + { + "epoch": 0.52463, + "grad_norm": 0.6645886201021788, + "learning_rate": 0.003, + "loss": 3.9928, + "step": 52463 + }, + { + "epoch": 0.52464, + "grad_norm": 0.6818311483373524, + "learning_rate": 0.003, + "loss": 3.9844, + "step": 52464 + }, + { + "epoch": 0.52465, + "grad_norm": 0.8143455779263395, + "learning_rate": 0.003, + "loss": 3.9635, + "step": 52465 + }, + { + "epoch": 0.52466, + "grad_norm": 0.8539373445344259, + "learning_rate": 0.003, + "loss": 3.966, + "step": 52466 + }, + { + "epoch": 0.52467, + "grad_norm": 0.7302340599987783, + "learning_rate": 0.003, + "loss": 3.995, + "step": 52467 + }, + { + "epoch": 0.52468, + "grad_norm": 0.7120853147142937, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 52468 + }, + { + "epoch": 0.52469, + "grad_norm": 0.7887670557980482, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 52469 + }, + { + "epoch": 0.5247, + "grad_norm": 0.8280877125238435, + "learning_rate": 0.003, + "loss": 3.986, + "step": 52470 + }, + { + "epoch": 0.52471, + "grad_norm": 0.8964724559356619, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 52471 + }, + { + "epoch": 0.52472, + "grad_norm": 0.8203268142652467, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 52472 + }, + { + "epoch": 0.52473, + "grad_norm": 0.6817952616345939, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 52473 + }, + { + "epoch": 0.52474, + "grad_norm": 0.6161807435745541, + "learning_rate": 0.003, + "loss": 3.954, + "step": 52474 + }, + { + "epoch": 0.52475, + "grad_norm": 0.6629177599703477, + "learning_rate": 0.003, + "loss": 3.9553, + "step": 52475 + }, + { + "epoch": 0.52476, + "grad_norm": 0.6475138450972148, + "learning_rate": 0.003, + "loss": 3.9858, + "step": 52476 + }, + { + "epoch": 0.52477, + "grad_norm": 0.7493108529197389, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 52477 + }, + { + "epoch": 0.52478, + "grad_norm": 0.8971809285207132, + "learning_rate": 0.003, + "loss": 3.9901, + "step": 52478 + }, + { + "epoch": 0.52479, + "grad_norm": 0.8482706011720316, + "learning_rate": 0.003, + "loss": 4.0062, + "step": 52479 + }, + { + "epoch": 0.5248, + "grad_norm": 0.8507512542422881, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 52480 + }, + { + "epoch": 0.52481, + "grad_norm": 0.852740140996873, + "learning_rate": 0.003, + "loss": 3.973, + "step": 52481 + }, + { + "epoch": 0.52482, + "grad_norm": 0.7970854847732975, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 52482 + }, + { + "epoch": 0.52483, + "grad_norm": 0.8435598406455986, + "learning_rate": 0.003, + "loss": 3.9765, + "step": 52483 + }, + { + "epoch": 0.52484, + "grad_norm": 0.7979652994011911, + "learning_rate": 0.003, + "loss": 3.959, + "step": 52484 + }, + { + "epoch": 0.52485, + "grad_norm": 0.7180174187931092, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 52485 + }, + { + "epoch": 0.52486, + "grad_norm": 0.7982847091982707, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 52486 + }, + { + "epoch": 0.52487, + "grad_norm": 0.8649972638166771, + "learning_rate": 0.003, + "loss": 3.998, + "step": 52487 + }, + { + "epoch": 0.52488, + "grad_norm": 0.9276280733127225, + "learning_rate": 0.003, + "loss": 3.9855, + "step": 52488 + }, + { + "epoch": 0.52489, + "grad_norm": 0.806590958684507, + "learning_rate": 0.003, + "loss": 3.9867, + "step": 52489 + }, + { + "epoch": 0.5249, + "grad_norm": 0.9123240100525063, + "learning_rate": 0.003, + "loss": 4.0136, + "step": 52490 + }, + { + "epoch": 0.52491, + "grad_norm": 0.887118496869705, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 52491 + }, + { + "epoch": 0.52492, + "grad_norm": 0.8771088804583186, + "learning_rate": 0.003, + "loss": 3.985, + "step": 52492 + }, + { + "epoch": 0.52493, + "grad_norm": 0.7821486587660038, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 52493 + }, + { + "epoch": 0.52494, + "grad_norm": 0.695397239662849, + "learning_rate": 0.003, + "loss": 3.9932, + "step": 52494 + }, + { + "epoch": 0.52495, + "grad_norm": 0.7552756583612686, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 52495 + }, + { + "epoch": 0.52496, + "grad_norm": 0.8009648845144519, + "learning_rate": 0.003, + "loss": 3.9631, + "step": 52496 + }, + { + "epoch": 0.52497, + "grad_norm": 0.8348378367147972, + "learning_rate": 0.003, + "loss": 4.023, + "step": 52497 + }, + { + "epoch": 0.52498, + "grad_norm": 0.9625799046936636, + "learning_rate": 0.003, + "loss": 3.9916, + "step": 52498 + }, + { + "epoch": 0.52499, + "grad_norm": 1.0741364238295996, + "learning_rate": 0.003, + "loss": 3.9818, + "step": 52499 + }, + { + "epoch": 0.525, + "grad_norm": 0.8796786044204461, + "learning_rate": 0.003, + "loss": 3.978, + "step": 52500 + }, + { + "epoch": 0.52501, + "grad_norm": 0.7963349388158524, + "learning_rate": 0.003, + "loss": 3.9734, + "step": 52501 + }, + { + "epoch": 0.52502, + "grad_norm": 0.6838310395814966, + "learning_rate": 0.003, + "loss": 3.9545, + "step": 52502 + }, + { + "epoch": 0.52503, + "grad_norm": 0.6101925771115667, + "learning_rate": 0.003, + "loss": 3.9971, + "step": 52503 + }, + { + "epoch": 0.52504, + "grad_norm": 0.5414877160184792, + "learning_rate": 0.003, + "loss": 3.965, + "step": 52504 + }, + { + "epoch": 0.52505, + "grad_norm": 0.5397126027297398, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 52505 + }, + { + "epoch": 0.52506, + "grad_norm": 0.5356470352108208, + "learning_rate": 0.003, + "loss": 3.9695, + "step": 52506 + }, + { + "epoch": 0.52507, + "grad_norm": 0.5443708617037109, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 52507 + }, + { + "epoch": 0.52508, + "grad_norm": 0.5168719308579192, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 52508 + }, + { + "epoch": 0.52509, + "grad_norm": 0.5843951339107529, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 52509 + }, + { + "epoch": 0.5251, + "grad_norm": 0.6058607879626139, + "learning_rate": 0.003, + "loss": 4.0049, + "step": 52510 + }, + { + "epoch": 0.52511, + "grad_norm": 0.6461409199557474, + "learning_rate": 0.003, + "loss": 3.9464, + "step": 52511 + }, + { + "epoch": 0.52512, + "grad_norm": 0.5294768800709458, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 52512 + }, + { + "epoch": 0.52513, + "grad_norm": 0.46530968890655583, + "learning_rate": 0.003, + "loss": 3.9314, + "step": 52513 + }, + { + "epoch": 0.52514, + "grad_norm": 0.4545175632165498, + "learning_rate": 0.003, + "loss": 3.9449, + "step": 52514 + }, + { + "epoch": 0.52515, + "grad_norm": 0.46094108755007496, + "learning_rate": 0.003, + "loss": 3.9547, + "step": 52515 + }, + { + "epoch": 0.52516, + "grad_norm": 0.5584144065902303, + "learning_rate": 0.003, + "loss": 3.9454, + "step": 52516 + }, + { + "epoch": 0.52517, + "grad_norm": 0.7247996737946771, + "learning_rate": 0.003, + "loss": 3.9771, + "step": 52517 + }, + { + "epoch": 0.52518, + "grad_norm": 0.7756909807746273, + "learning_rate": 0.003, + "loss": 3.9722, + "step": 52518 + }, + { + "epoch": 0.52519, + "grad_norm": 0.7568926006066401, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 52519 + }, + { + "epoch": 0.5252, + "grad_norm": 0.6964098306325962, + "learning_rate": 0.003, + "loss": 3.9596, + "step": 52520 + }, + { + "epoch": 0.52521, + "grad_norm": 0.6222852781439037, + "learning_rate": 0.003, + "loss": 3.9526, + "step": 52521 + }, + { + "epoch": 0.52522, + "grad_norm": 0.6494826998364749, + "learning_rate": 0.003, + "loss": 3.9682, + "step": 52522 + }, + { + "epoch": 0.52523, + "grad_norm": 0.5949769559332249, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 52523 + }, + { + "epoch": 0.52524, + "grad_norm": 0.5897171707481054, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 52524 + }, + { + "epoch": 0.52525, + "grad_norm": 0.6265461908520027, + "learning_rate": 0.003, + "loss": 3.9592, + "step": 52525 + }, + { + "epoch": 0.52526, + "grad_norm": 0.6480776569659872, + "learning_rate": 0.003, + "loss": 3.9794, + "step": 52526 + }, + { + "epoch": 0.52527, + "grad_norm": 0.6385597145409017, + "learning_rate": 0.003, + "loss": 3.9574, + "step": 52527 + }, + { + "epoch": 0.52528, + "grad_norm": 0.6538543537031969, + "learning_rate": 0.003, + "loss": 3.9491, + "step": 52528 + }, + { + "epoch": 0.52529, + "grad_norm": 0.7164561826944149, + "learning_rate": 0.003, + "loss": 3.9702, + "step": 52529 + }, + { + "epoch": 0.5253, + "grad_norm": 0.9249104501944391, + "learning_rate": 0.003, + "loss": 3.9328, + "step": 52530 + }, + { + "epoch": 0.52531, + "grad_norm": 1.2533648047643307, + "learning_rate": 0.003, + "loss": 4.0166, + "step": 52531 + }, + { + "epoch": 0.52532, + "grad_norm": 0.7669665416313498, + "learning_rate": 0.003, + "loss": 3.9563, + "step": 52532 + }, + { + "epoch": 0.52533, + "grad_norm": 0.8048413037163707, + "learning_rate": 0.003, + "loss": 3.9675, + "step": 52533 + }, + { + "epoch": 0.52534, + "grad_norm": 0.8963764427272835, + "learning_rate": 0.003, + "loss": 3.9848, + "step": 52534 + }, + { + "epoch": 0.52535, + "grad_norm": 1.0140596159237276, + "learning_rate": 0.003, + "loss": 3.9619, + "step": 52535 + }, + { + "epoch": 0.52536, + "grad_norm": 1.1771809708752454, + "learning_rate": 0.003, + "loss": 3.9583, + "step": 52536 + }, + { + "epoch": 0.52537, + "grad_norm": 0.7724839157401471, + "learning_rate": 0.003, + "loss": 4.0067, + "step": 52537 + }, + { + "epoch": 0.52538, + "grad_norm": 0.7229883099087815, + "learning_rate": 0.003, + "loss": 4.0097, + "step": 52538 + }, + { + "epoch": 0.52539, + "grad_norm": 0.7965872660683626, + "learning_rate": 0.003, + "loss": 3.9721, + "step": 52539 + }, + { + "epoch": 0.5254, + "grad_norm": 0.8546690984530982, + "learning_rate": 0.003, + "loss": 3.9669, + "step": 52540 + }, + { + "epoch": 0.52541, + "grad_norm": 0.9015149470186724, + "learning_rate": 0.003, + "loss": 3.9564, + "step": 52541 + }, + { + "epoch": 0.52542, + "grad_norm": 0.9258283403034875, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 52542 + }, + { + "epoch": 0.52543, + "grad_norm": 0.8282593606624807, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 52543 + }, + { + "epoch": 0.52544, + "grad_norm": 0.9177879927441818, + "learning_rate": 0.003, + "loss": 3.9869, + "step": 52544 + }, + { + "epoch": 0.52545, + "grad_norm": 0.9760117947013298, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 52545 + }, + { + "epoch": 0.52546, + "grad_norm": 0.8170444650227353, + "learning_rate": 0.003, + "loss": 4.0021, + "step": 52546 + }, + { + "epoch": 0.52547, + "grad_norm": 0.7663260192229285, + "learning_rate": 0.003, + "loss": 4.0164, + "step": 52547 + }, + { + "epoch": 0.52548, + "grad_norm": 0.7168095467633626, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 52548 + }, + { + "epoch": 0.52549, + "grad_norm": 0.8155418932702372, + "learning_rate": 0.003, + "loss": 3.9561, + "step": 52549 + }, + { + "epoch": 0.5255, + "grad_norm": 0.8159228932982941, + "learning_rate": 0.003, + "loss": 3.9854, + "step": 52550 + }, + { + "epoch": 0.52551, + "grad_norm": 0.8533198043212906, + "learning_rate": 0.003, + "loss": 3.9827, + "step": 52551 + }, + { + "epoch": 0.52552, + "grad_norm": 0.8743256020187345, + "learning_rate": 0.003, + "loss": 4.0274, + "step": 52552 + }, + { + "epoch": 0.52553, + "grad_norm": 0.9154012055601369, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 52553 + }, + { + "epoch": 0.52554, + "grad_norm": 0.9661821475404898, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 52554 + }, + { + "epoch": 0.52555, + "grad_norm": 0.9865440874956641, + "learning_rate": 0.003, + "loss": 3.9513, + "step": 52555 + }, + { + "epoch": 0.52556, + "grad_norm": 0.942888310509176, + "learning_rate": 0.003, + "loss": 4.0041, + "step": 52556 + }, + { + "epoch": 0.52557, + "grad_norm": 0.7746490641208224, + "learning_rate": 0.003, + "loss": 3.9842, + "step": 52557 + }, + { + "epoch": 0.52558, + "grad_norm": 0.7693424261450156, + "learning_rate": 0.003, + "loss": 3.9902, + "step": 52558 + }, + { + "epoch": 0.52559, + "grad_norm": 0.8048093839048955, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 52559 + }, + { + "epoch": 0.5256, + "grad_norm": 0.8719916259482035, + "learning_rate": 0.003, + "loss": 4.0002, + "step": 52560 + }, + { + "epoch": 0.52561, + "grad_norm": 0.9925527358016426, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 52561 + }, + { + "epoch": 0.52562, + "grad_norm": 1.1481262854089913, + "learning_rate": 0.003, + "loss": 3.9781, + "step": 52562 + }, + { + "epoch": 0.52563, + "grad_norm": 0.8904645219517935, + "learning_rate": 0.003, + "loss": 4.0046, + "step": 52563 + }, + { + "epoch": 0.52564, + "grad_norm": 0.9162102721678638, + "learning_rate": 0.003, + "loss": 3.9923, + "step": 52564 + }, + { + "epoch": 0.52565, + "grad_norm": 1.0462181804651405, + "learning_rate": 0.003, + "loss": 3.976, + "step": 52565 + }, + { + "epoch": 0.52566, + "grad_norm": 0.8935680933489678, + "learning_rate": 0.003, + "loss": 4.0091, + "step": 52566 + }, + { + "epoch": 0.52567, + "grad_norm": 0.7951732806878727, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 52567 + }, + { + "epoch": 0.52568, + "grad_norm": 0.6875565974841142, + "learning_rate": 0.003, + "loss": 3.9998, + "step": 52568 + }, + { + "epoch": 0.52569, + "grad_norm": 0.6230148444525824, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 52569 + }, + { + "epoch": 0.5257, + "grad_norm": 0.6898394928629497, + "learning_rate": 0.003, + "loss": 3.9965, + "step": 52570 + }, + { + "epoch": 0.52571, + "grad_norm": 0.7043926703851335, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 52571 + }, + { + "epoch": 0.52572, + "grad_norm": 0.6314976244949021, + "learning_rate": 0.003, + "loss": 4.0108, + "step": 52572 + }, + { + "epoch": 0.52573, + "grad_norm": 0.6119646640893073, + "learning_rate": 0.003, + "loss": 3.9502, + "step": 52573 + }, + { + "epoch": 0.52574, + "grad_norm": 0.548237307826896, + "learning_rate": 0.003, + "loss": 3.949, + "step": 52574 + }, + { + "epoch": 0.52575, + "grad_norm": 0.4404794561137054, + "learning_rate": 0.003, + "loss": 3.972, + "step": 52575 + }, + { + "epoch": 0.52576, + "grad_norm": 0.4302278925703599, + "learning_rate": 0.003, + "loss": 3.9762, + "step": 52576 + }, + { + "epoch": 0.52577, + "grad_norm": 0.4602859890855908, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 52577 + }, + { + "epoch": 0.52578, + "grad_norm": 0.4926035905166809, + "learning_rate": 0.003, + "loss": 3.978, + "step": 52578 + }, + { + "epoch": 0.52579, + "grad_norm": 0.5689049523277969, + "learning_rate": 0.003, + "loss": 4.019, + "step": 52579 + }, + { + "epoch": 0.5258, + "grad_norm": 0.6194134242276369, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 52580 + }, + { + "epoch": 0.52581, + "grad_norm": 0.6028264713307541, + "learning_rate": 0.003, + "loss": 3.951, + "step": 52581 + }, + { + "epoch": 0.52582, + "grad_norm": 0.6068912328533884, + "learning_rate": 0.003, + "loss": 3.957, + "step": 52582 + }, + { + "epoch": 0.52583, + "grad_norm": 0.734460342730053, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 52583 + }, + { + "epoch": 0.52584, + "grad_norm": 0.8171154109046076, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 52584 + }, + { + "epoch": 0.52585, + "grad_norm": 0.8378181715553031, + "learning_rate": 0.003, + "loss": 3.984, + "step": 52585 + }, + { + "epoch": 0.52586, + "grad_norm": 0.9516154032105102, + "learning_rate": 0.003, + "loss": 4.0038, + "step": 52586 + }, + { + "epoch": 0.52587, + "grad_norm": 0.7875595663455239, + "learning_rate": 0.003, + "loss": 3.9783, + "step": 52587 + }, + { + "epoch": 0.52588, + "grad_norm": 0.652895070877263, + "learning_rate": 0.003, + "loss": 3.9662, + "step": 52588 + }, + { + "epoch": 0.52589, + "grad_norm": 0.6498471080212308, + "learning_rate": 0.003, + "loss": 3.9565, + "step": 52589 + }, + { + "epoch": 0.5259, + "grad_norm": 0.6232794324452356, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 52590 + }, + { + "epoch": 0.52591, + "grad_norm": 0.6336784010974851, + "learning_rate": 0.003, + "loss": 3.9658, + "step": 52591 + }, + { + "epoch": 0.52592, + "grad_norm": 0.6651521857357856, + "learning_rate": 0.003, + "loss": 3.9374, + "step": 52592 + }, + { + "epoch": 0.52593, + "grad_norm": 0.7337411095355312, + "learning_rate": 0.003, + "loss": 3.9678, + "step": 52593 + }, + { + "epoch": 0.52594, + "grad_norm": 0.7530895340039117, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 52594 + }, + { + "epoch": 0.52595, + "grad_norm": 0.7738441159359636, + "learning_rate": 0.003, + "loss": 3.9859, + "step": 52595 + }, + { + "epoch": 0.52596, + "grad_norm": 0.8387419407865594, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 52596 + }, + { + "epoch": 0.52597, + "grad_norm": 0.9099061928276051, + "learning_rate": 0.003, + "loss": 3.9993, + "step": 52597 + }, + { + "epoch": 0.52598, + "grad_norm": 0.8242942040242632, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 52598 + }, + { + "epoch": 0.52599, + "grad_norm": 0.9223699900891215, + "learning_rate": 0.003, + "loss": 3.9649, + "step": 52599 + }, + { + "epoch": 0.526, + "grad_norm": 1.0302627172311531, + "learning_rate": 0.003, + "loss": 3.991, + "step": 52600 + }, + { + "epoch": 0.52601, + "grad_norm": 0.9490869908042785, + "learning_rate": 0.003, + "loss": 3.985, + "step": 52601 + }, + { + "epoch": 0.52602, + "grad_norm": 0.8418849252872119, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 52602 + }, + { + "epoch": 0.52603, + "grad_norm": 0.8440867762519264, + "learning_rate": 0.003, + "loss": 3.9895, + "step": 52603 + }, + { + "epoch": 0.52604, + "grad_norm": 0.897497153435126, + "learning_rate": 0.003, + "loss": 3.9716, + "step": 52604 + }, + { + "epoch": 0.52605, + "grad_norm": 0.9710443766883652, + "learning_rate": 0.003, + "loss": 3.9882, + "step": 52605 + }, + { + "epoch": 0.52606, + "grad_norm": 1.088169801125963, + "learning_rate": 0.003, + "loss": 3.9847, + "step": 52606 + }, + { + "epoch": 0.52607, + "grad_norm": 0.9002613219191222, + "learning_rate": 0.003, + "loss": 3.9887, + "step": 52607 + }, + { + "epoch": 0.52608, + "grad_norm": 0.7381455377072392, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 52608 + }, + { + "epoch": 0.52609, + "grad_norm": 0.7213984335548429, + "learning_rate": 0.003, + "loss": 3.9754, + "step": 52609 + }, + { + "epoch": 0.5261, + "grad_norm": 0.7534307491648846, + "learning_rate": 0.003, + "loss": 3.9769, + "step": 52610 + }, + { + "epoch": 0.52611, + "grad_norm": 0.8112552915371168, + "learning_rate": 0.003, + "loss": 3.9829, + "step": 52611 + }, + { + "epoch": 0.52612, + "grad_norm": 0.8684292709793846, + "learning_rate": 0.003, + "loss": 3.9985, + "step": 52612 + }, + { + "epoch": 0.52613, + "grad_norm": 0.9301947505779403, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 52613 + }, + { + "epoch": 0.52614, + "grad_norm": 0.8940689551933106, + "learning_rate": 0.003, + "loss": 3.9837, + "step": 52614 + }, + { + "epoch": 0.52615, + "grad_norm": 0.7216613724570933, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 52615 + }, + { + "epoch": 0.52616, + "grad_norm": 0.6349624015998987, + "learning_rate": 0.003, + "loss": 3.96, + "step": 52616 + }, + { + "epoch": 0.52617, + "grad_norm": 0.6623690744998608, + "learning_rate": 0.003, + "loss": 4.0149, + "step": 52617 + }, + { + "epoch": 0.52618, + "grad_norm": 0.665987582702707, + "learning_rate": 0.003, + "loss": 3.9476, + "step": 52618 + }, + { + "epoch": 0.52619, + "grad_norm": 0.6635894350012931, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 52619 + }, + { + "epoch": 0.5262, + "grad_norm": 0.7129559271689773, + "learning_rate": 0.003, + "loss": 3.9661, + "step": 52620 + }, + { + "epoch": 0.52621, + "grad_norm": 0.7121523882031966, + "learning_rate": 0.003, + "loss": 3.9805, + "step": 52621 + }, + { + "epoch": 0.52622, + "grad_norm": 0.597245266487335, + "learning_rate": 0.003, + "loss": 3.9925, + "step": 52622 + }, + { + "epoch": 0.52623, + "grad_norm": 0.538104364651642, + "learning_rate": 0.003, + "loss": 3.9428, + "step": 52623 + }, + { + "epoch": 0.52624, + "grad_norm": 0.5563954488383034, + "learning_rate": 0.003, + "loss": 3.9777, + "step": 52624 + }, + { + "epoch": 0.52625, + "grad_norm": 0.604342133312154, + "learning_rate": 0.003, + "loss": 3.9603, + "step": 52625 + }, + { + "epoch": 0.52626, + "grad_norm": 0.7139397309627125, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 52626 + }, + { + "epoch": 0.52627, + "grad_norm": 0.9305023206161773, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 52627 + }, + { + "epoch": 0.52628, + "grad_norm": 1.1467654577190118, + "learning_rate": 0.003, + "loss": 3.9648, + "step": 52628 + }, + { + "epoch": 0.52629, + "grad_norm": 0.8726914992613298, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 52629 + }, + { + "epoch": 0.5263, + "grad_norm": 0.7603359791890584, + "learning_rate": 0.003, + "loss": 3.9865, + "step": 52630 + }, + { + "epoch": 0.52631, + "grad_norm": 0.7934395585428257, + "learning_rate": 0.003, + "loss": 3.9843, + "step": 52631 + }, + { + "epoch": 0.52632, + "grad_norm": 0.7965592490852328, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 52632 + }, + { + "epoch": 0.52633, + "grad_norm": 0.8456469555988753, + "learning_rate": 0.003, + "loss": 3.9819, + "step": 52633 + }, + { + "epoch": 0.52634, + "grad_norm": 0.8834797822900502, + "learning_rate": 0.003, + "loss": 3.9646, + "step": 52634 + }, + { + "epoch": 0.52635, + "grad_norm": 0.8704111401080308, + "learning_rate": 0.003, + "loss": 3.9853, + "step": 52635 + }, + { + "epoch": 0.52636, + "grad_norm": 0.7899164692844266, + "learning_rate": 0.003, + "loss": 3.9522, + "step": 52636 + }, + { + "epoch": 0.52637, + "grad_norm": 0.7165270383247053, + "learning_rate": 0.003, + "loss": 4.0117, + "step": 52637 + }, + { + "epoch": 0.52638, + "grad_norm": 0.713639041322866, + "learning_rate": 0.003, + "loss": 3.9737, + "step": 52638 + }, + { + "epoch": 0.52639, + "grad_norm": 0.711721184742619, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 52639 + }, + { + "epoch": 0.5264, + "grad_norm": 0.8110960525046952, + "learning_rate": 0.003, + "loss": 3.9885, + "step": 52640 + }, + { + "epoch": 0.52641, + "grad_norm": 0.9899424825619909, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 52641 + }, + { + "epoch": 0.52642, + "grad_norm": 1.0208801708028672, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 52642 + }, + { + "epoch": 0.52643, + "grad_norm": 0.943069261428459, + "learning_rate": 0.003, + "loss": 3.9825, + "step": 52643 + }, + { + "epoch": 0.52644, + "grad_norm": 0.9989929899781129, + "learning_rate": 0.003, + "loss": 3.9907, + "step": 52644 + }, + { + "epoch": 0.52645, + "grad_norm": 0.9448962675426114, + "learning_rate": 0.003, + "loss": 3.9587, + "step": 52645 + }, + { + "epoch": 0.52646, + "grad_norm": 0.7910862304945954, + "learning_rate": 0.003, + "loss": 3.9815, + "step": 52646 + }, + { + "epoch": 0.52647, + "grad_norm": 0.7829109035177996, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 52647 + }, + { + "epoch": 0.52648, + "grad_norm": 0.7154787659778855, + "learning_rate": 0.003, + "loss": 3.9552, + "step": 52648 + }, + { + "epoch": 0.52649, + "grad_norm": 0.6369083927792075, + "learning_rate": 0.003, + "loss": 3.9839, + "step": 52649 + }, + { + "epoch": 0.5265, + "grad_norm": 0.6589028374046242, + "learning_rate": 0.003, + "loss": 3.966, + "step": 52650 + }, + { + "epoch": 0.52651, + "grad_norm": 0.6555282853030615, + "learning_rate": 0.003, + "loss": 3.989, + "step": 52651 + }, + { + "epoch": 0.52652, + "grad_norm": 0.7421237534878847, + "learning_rate": 0.003, + "loss": 4.0057, + "step": 52652 + }, + { + "epoch": 0.52653, + "grad_norm": 0.7113096084306514, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 52653 + }, + { + "epoch": 0.52654, + "grad_norm": 0.8941535318316226, + "learning_rate": 0.003, + "loss": 4.0169, + "step": 52654 + }, + { + "epoch": 0.52655, + "grad_norm": 1.0456130137523294, + "learning_rate": 0.003, + "loss": 4.003, + "step": 52655 + }, + { + "epoch": 0.52656, + "grad_norm": 0.8994361826826581, + "learning_rate": 0.003, + "loss": 3.9888, + "step": 52656 + }, + { + "epoch": 0.52657, + "grad_norm": 0.9300781017438943, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 52657 + }, + { + "epoch": 0.52658, + "grad_norm": 0.9378780032366387, + "learning_rate": 0.003, + "loss": 3.9652, + "step": 52658 + }, + { + "epoch": 0.52659, + "grad_norm": 0.8617338774312111, + "learning_rate": 0.003, + "loss": 3.9892, + "step": 52659 + }, + { + "epoch": 0.5266, + "grad_norm": 0.8075908638930964, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 52660 + }, + { + "epoch": 0.52661, + "grad_norm": 0.7555962180279191, + "learning_rate": 0.003, + "loss": 3.9539, + "step": 52661 + }, + { + "epoch": 0.52662, + "grad_norm": 0.793153435329342, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 52662 + }, + { + "epoch": 0.52663, + "grad_norm": 0.7414282448315858, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 52663 + }, + { + "epoch": 0.52664, + "grad_norm": 0.5959385565183067, + "learning_rate": 0.003, + "loss": 3.9846, + "step": 52664 + }, + { + "epoch": 0.52665, + "grad_norm": 0.5397893545633241, + "learning_rate": 0.003, + "loss": 4.0243, + "step": 52665 + }, + { + "epoch": 0.52666, + "grad_norm": 0.5150512657038689, + "learning_rate": 0.003, + "loss": 4.0024, + "step": 52666 + }, + { + "epoch": 0.52667, + "grad_norm": 0.5124964325820611, + "learning_rate": 0.003, + "loss": 3.9878, + "step": 52667 + }, + { + "epoch": 0.52668, + "grad_norm": 0.48143003582173743, + "learning_rate": 0.003, + "loss": 3.9475, + "step": 52668 + }, + { + "epoch": 0.52669, + "grad_norm": 0.47696122537294466, + "learning_rate": 0.003, + "loss": 3.9647, + "step": 52669 + }, + { + "epoch": 0.5267, + "grad_norm": 0.5184153170728609, + "learning_rate": 0.003, + "loss": 3.9569, + "step": 52670 + }, + { + "epoch": 0.52671, + "grad_norm": 0.586467106905491, + "learning_rate": 0.003, + "loss": 3.9353, + "step": 52671 + }, + { + "epoch": 0.52672, + "grad_norm": 0.6069252926001139, + "learning_rate": 0.003, + "loss": 3.9861, + "step": 52672 + }, + { + "epoch": 0.52673, + "grad_norm": 0.7223022484086221, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 52673 + }, + { + "epoch": 0.52674, + "grad_norm": 0.8107602254460123, + "learning_rate": 0.003, + "loss": 3.9975, + "step": 52674 + }, + { + "epoch": 0.52675, + "grad_norm": 0.7687845755634789, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 52675 + }, + { + "epoch": 0.52676, + "grad_norm": 0.6331981793833797, + "learning_rate": 0.003, + "loss": 3.9731, + "step": 52676 + }, + { + "epoch": 0.52677, + "grad_norm": 0.6893822797678186, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 52677 + }, + { + "epoch": 0.52678, + "grad_norm": 0.6667656814727222, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 52678 + }, + { + "epoch": 0.52679, + "grad_norm": 0.632088384893026, + "learning_rate": 0.003, + "loss": 3.9576, + "step": 52679 + }, + { + "epoch": 0.5268, + "grad_norm": 0.603503022422435, + "learning_rate": 0.003, + "loss": 3.9641, + "step": 52680 + }, + { + "epoch": 0.52681, + "grad_norm": 0.6847615258804165, + "learning_rate": 0.003, + "loss": 3.9519, + "step": 52681 + }, + { + "epoch": 0.52682, + "grad_norm": 0.8188202114765498, + "learning_rate": 0.003, + "loss": 3.9782, + "step": 52682 + }, + { + "epoch": 0.52683, + "grad_norm": 0.8844353198748643, + "learning_rate": 0.003, + "loss": 3.9742, + "step": 52683 + }, + { + "epoch": 0.52684, + "grad_norm": 0.8595146245513373, + "learning_rate": 0.003, + "loss": 3.9743, + "step": 52684 + }, + { + "epoch": 0.52685, + "grad_norm": 0.846349774284694, + "learning_rate": 0.003, + "loss": 3.995, + "step": 52685 + }, + { + "epoch": 0.52686, + "grad_norm": 0.7715956271129659, + "learning_rate": 0.003, + "loss": 3.9689, + "step": 52686 + }, + { + "epoch": 0.52687, + "grad_norm": 0.7293021217185541, + "learning_rate": 0.003, + "loss": 3.9868, + "step": 52687 + }, + { + "epoch": 0.52688, + "grad_norm": 0.6871775158296095, + "learning_rate": 0.003, + "loss": 3.9945, + "step": 52688 + }, + { + "epoch": 0.52689, + "grad_norm": 0.726453755203456, + "learning_rate": 0.003, + "loss": 3.9912, + "step": 52689 + }, + { + "epoch": 0.5269, + "grad_norm": 0.6498673723683493, + "learning_rate": 0.003, + "loss": 3.9611, + "step": 52690 + }, + { + "epoch": 0.52691, + "grad_norm": 0.7628575993988508, + "learning_rate": 0.003, + "loss": 3.9785, + "step": 52691 + }, + { + "epoch": 0.52692, + "grad_norm": 0.9250749034894321, + "learning_rate": 0.003, + "loss": 3.9984, + "step": 52692 + }, + { + "epoch": 0.52693, + "grad_norm": 1.121416231456478, + "learning_rate": 0.003, + "loss": 3.9672, + "step": 52693 + }, + { + "epoch": 0.52694, + "grad_norm": 0.7838503146689596, + "learning_rate": 0.003, + "loss": 3.9889, + "step": 52694 + }, + { + "epoch": 0.52695, + "grad_norm": 0.7870032561675976, + "learning_rate": 0.003, + "loss": 3.9715, + "step": 52695 + }, + { + "epoch": 0.52696, + "grad_norm": 0.7155188136815616, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 52696 + }, + { + "epoch": 0.52697, + "grad_norm": 0.8832205852952483, + "learning_rate": 0.003, + "loss": 3.988, + "step": 52697 + }, + { + "epoch": 0.52698, + "grad_norm": 0.9435148684303516, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 52698 + }, + { + "epoch": 0.52699, + "grad_norm": 0.8794466963752945, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 52699 + }, + { + "epoch": 0.527, + "grad_norm": 0.8398267287634521, + "learning_rate": 0.003, + "loss": 3.9801, + "step": 52700 + }, + { + "epoch": 0.52701, + "grad_norm": 0.8794755032826169, + "learning_rate": 0.003, + "loss": 3.9814, + "step": 52701 + }, + { + "epoch": 0.52702, + "grad_norm": 0.9955036646054011, + "learning_rate": 0.003, + "loss": 3.9905, + "step": 52702 + }, + { + "epoch": 0.52703, + "grad_norm": 0.9955371111097729, + "learning_rate": 0.003, + "loss": 3.9749, + "step": 52703 + }, + { + "epoch": 0.52704, + "grad_norm": 0.9346685734410151, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 52704 + }, + { + "epoch": 0.52705, + "grad_norm": 0.9920378774690249, + "learning_rate": 0.003, + "loss": 4.014, + "step": 52705 + }, + { + "epoch": 0.52706, + "grad_norm": 1.0644159791635772, + "learning_rate": 0.003, + "loss": 4.0103, + "step": 52706 + }, + { + "epoch": 0.52707, + "grad_norm": 0.9696550056807399, + "learning_rate": 0.003, + "loss": 4.0043, + "step": 52707 + }, + { + "epoch": 0.52708, + "grad_norm": 0.9548071771901082, + "learning_rate": 0.003, + "loss": 3.9977, + "step": 52708 + }, + { + "epoch": 0.52709, + "grad_norm": 0.9748777489283714, + "learning_rate": 0.003, + "loss": 4.0094, + "step": 52709 + }, + { + "epoch": 0.5271, + "grad_norm": 0.9193980670943787, + "learning_rate": 0.003, + "loss": 3.9954, + "step": 52710 + }, + { + "epoch": 0.52711, + "grad_norm": 0.7152680739648263, + "learning_rate": 0.003, + "loss": 3.9899, + "step": 52711 + }, + { + "epoch": 0.52712, + "grad_norm": 0.7846219404954466, + "learning_rate": 0.003, + "loss": 3.9726, + "step": 52712 + }, + { + "epoch": 0.52713, + "grad_norm": 0.7500636738555553, + "learning_rate": 0.003, + "loss": 3.9694, + "step": 52713 + }, + { + "epoch": 0.52714, + "grad_norm": 0.7088066511318536, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 52714 + }, + { + "epoch": 0.52715, + "grad_norm": 0.6962959059929522, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 52715 + }, + { + "epoch": 0.52716, + "grad_norm": 0.7010567344486708, + "learning_rate": 0.003, + "loss": 3.9708, + "step": 52716 + }, + { + "epoch": 0.52717, + "grad_norm": 0.7853464852796076, + "learning_rate": 0.003, + "loss": 3.993, + "step": 52717 + }, + { + "epoch": 0.52718, + "grad_norm": 0.7541386849085807, + "learning_rate": 0.003, + "loss": 3.9797, + "step": 52718 + }, + { + "epoch": 0.52719, + "grad_norm": 0.6600130040567489, + "learning_rate": 0.003, + "loss": 3.9817, + "step": 52719 + }, + { + "epoch": 0.5272, + "grad_norm": 0.7088260617572206, + "learning_rate": 0.003, + "loss": 3.988, + "step": 52720 + }, + { + "epoch": 0.52721, + "grad_norm": 0.6534336097029331, + "learning_rate": 0.003, + "loss": 3.9774, + "step": 52721 + }, + { + "epoch": 0.52722, + "grad_norm": 0.7369337821429722, + "learning_rate": 0.003, + "loss": 3.9904, + "step": 52722 + }, + { + "epoch": 0.52723, + "grad_norm": 0.7163239091882103, + "learning_rate": 0.003, + "loss": 3.9653, + "step": 52723 + }, + { + "epoch": 0.52724, + "grad_norm": 0.6034538434614326, + "learning_rate": 0.003, + "loss": 3.981, + "step": 52724 + }, + { + "epoch": 0.52725, + "grad_norm": 0.6236515461221034, + "learning_rate": 0.003, + "loss": 4.0012, + "step": 52725 + }, + { + "epoch": 0.52726, + "grad_norm": 0.6475279118046638, + "learning_rate": 0.003, + "loss": 3.9384, + "step": 52726 + }, + { + "epoch": 0.52727, + "grad_norm": 0.6372932243153523, + "learning_rate": 0.003, + "loss": 3.9875, + "step": 52727 + }, + { + "epoch": 0.52728, + "grad_norm": 0.6761793164521753, + "learning_rate": 0.003, + "loss": 3.983, + "step": 52728 + }, + { + "epoch": 0.52729, + "grad_norm": 0.7855816297002296, + "learning_rate": 0.003, + "loss": 3.9706, + "step": 52729 + }, + { + "epoch": 0.5273, + "grad_norm": 0.808663099179692, + "learning_rate": 0.003, + "loss": 3.9703, + "step": 52730 + }, + { + "epoch": 0.52731, + "grad_norm": 0.8121926100310315, + "learning_rate": 0.003, + "loss": 3.9931, + "step": 52731 + }, + { + "epoch": 0.52732, + "grad_norm": 0.7830178525362819, + "learning_rate": 0.003, + "loss": 3.974, + "step": 52732 + }, + { + "epoch": 0.52733, + "grad_norm": 0.7357502403141337, + "learning_rate": 0.003, + "loss": 3.9379, + "step": 52733 + }, + { + "epoch": 0.52734, + "grad_norm": 0.5995968724786351, + "learning_rate": 0.003, + "loss": 3.9787, + "step": 52734 + }, + { + "epoch": 0.52735, + "grad_norm": 0.5043040082750125, + "learning_rate": 0.003, + "loss": 3.9909, + "step": 52735 + }, + { + "epoch": 0.52736, + "grad_norm": 0.5015253571925367, + "learning_rate": 0.003, + "loss": 3.9389, + "step": 52736 + }, + { + "epoch": 0.52737, + "grad_norm": 0.4869587369817938, + "learning_rate": 0.003, + "loss": 3.941, + "step": 52737 + }, + { + "epoch": 0.52738, + "grad_norm": 0.48424281753656107, + "learning_rate": 0.003, + "loss": 3.9613, + "step": 52738 + }, + { + "epoch": 0.52739, + "grad_norm": 0.5155047849301875, + "learning_rate": 0.003, + "loss": 3.9593, + "step": 52739 + }, + { + "epoch": 0.5274, + "grad_norm": 0.5789810652242999, + "learning_rate": 0.003, + "loss": 3.9556, + "step": 52740 + }, + { + "epoch": 0.52741, + "grad_norm": 0.7274045606287817, + "learning_rate": 0.003, + "loss": 3.974, + "step": 52741 + }, + { + "epoch": 0.52742, + "grad_norm": 0.9151484084687642, + "learning_rate": 0.003, + "loss": 3.9745, + "step": 52742 + }, + { + "epoch": 0.52743, + "grad_norm": 0.9910062977800366, + "learning_rate": 0.003, + "loss": 3.9746, + "step": 52743 + }, + { + "epoch": 0.52744, + "grad_norm": 0.915946491114752, + "learning_rate": 0.003, + "loss": 3.9796, + "step": 52744 + }, + { + "epoch": 0.52745, + "grad_norm": 0.7321477133476577, + "learning_rate": 0.003, + "loss": 3.9879, + "step": 52745 + }, + { + "epoch": 0.52746, + "grad_norm": 0.6382131779341103, + "learning_rate": 0.003, + "loss": 3.9779, + "step": 52746 + }, + { + "epoch": 0.52747, + "grad_norm": 0.7424402734652842, + "learning_rate": 0.003, + "loss": 3.9627, + "step": 52747 + }, + { + "epoch": 0.52748, + "grad_norm": 0.8330653559073504, + "learning_rate": 0.003, + "loss": 3.9956, + "step": 52748 + }, + { + "epoch": 0.52749, + "grad_norm": 1.0277291057951883, + "learning_rate": 0.003, + "loss": 3.9833, + "step": 52749 + }, + { + "epoch": 0.5275, + "grad_norm": 0.9090557819925015, + "learning_rate": 0.003, + "loss": 3.9621, + "step": 52750 + }, + { + "epoch": 0.52751, + "grad_norm": 0.7780762274320868, + "learning_rate": 0.003, + "loss": 4.0102, + "step": 52751 + }, + { + "epoch": 0.52752, + "grad_norm": 0.8525665157270972, + "learning_rate": 0.003, + "loss": 3.9982, + "step": 52752 + }, + { + "epoch": 0.52753, + "grad_norm": 0.9872877458107843, + "learning_rate": 0.003, + "loss": 3.9947, + "step": 52753 + }, + { + "epoch": 0.52754, + "grad_norm": 0.9902909180331366, + "learning_rate": 0.003, + "loss": 4.0153, + "step": 52754 + }, + { + "epoch": 0.52755, + "grad_norm": 0.9548992088849255, + "learning_rate": 0.003, + "loss": 3.9942, + "step": 52755 + }, + { + "epoch": 0.52756, + "grad_norm": 1.0117746074170435, + "learning_rate": 0.003, + "loss": 4.0033, + "step": 52756 + }, + { + "epoch": 0.52757, + "grad_norm": 0.9481521126883569, + "learning_rate": 0.003, + "loss": 4.0085, + "step": 52757 + }, + { + "epoch": 0.52758, + "grad_norm": 0.9978412462299636, + "learning_rate": 0.003, + "loss": 4.0053, + "step": 52758 + }, + { + "epoch": 0.52759, + "grad_norm": 0.9676554383966641, + "learning_rate": 0.003, + "loss": 3.9957, + "step": 52759 + }, + { + "epoch": 0.5276, + "grad_norm": 0.9715719929817154, + "learning_rate": 0.003, + "loss": 3.9748, + "step": 52760 + }, + { + "epoch": 0.52761, + "grad_norm": 0.8697803741452906, + "learning_rate": 0.003, + "loss": 4.0316, + "step": 52761 + }, + { + "epoch": 0.52762, + "grad_norm": 0.7366112309664938, + "learning_rate": 0.003, + "loss": 3.9826, + "step": 52762 + }, + { + "epoch": 0.52763, + "grad_norm": 0.6490547031197411, + "learning_rate": 0.003, + "loss": 3.9757, + "step": 52763 + }, + { + "epoch": 0.52764, + "grad_norm": 0.598854270185046, + "learning_rate": 0.003, + "loss": 3.9959, + "step": 52764 + }, + { + "epoch": 0.52765, + "grad_norm": 0.6207283899529249, + "learning_rate": 0.003, + "loss": 4.0142, + "step": 52765 + }, + { + "epoch": 0.52766, + "grad_norm": 0.6375404543943657, + "learning_rate": 0.003, + "loss": 3.9963, + "step": 52766 + }, + { + "epoch": 0.52767, + "grad_norm": 0.6530975295649295, + "learning_rate": 0.003, + "loss": 3.952, + "step": 52767 + }, + { + "epoch": 0.52768, + "grad_norm": 0.6641131434632155, + "learning_rate": 0.003, + "loss": 3.961, + "step": 52768 + }, + { + "epoch": 0.52769, + "grad_norm": 0.6953532307062301, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 52769 + }, + { + "epoch": 0.5277, + "grad_norm": 0.7181255515806186, + "learning_rate": 0.003, + "loss": 3.9688, + "step": 52770 + }, + { + "epoch": 0.52771, + "grad_norm": 0.7219776294418363, + "learning_rate": 0.003, + "loss": 4.0056, + "step": 52771 + }, + { + "epoch": 0.52772, + "grad_norm": 0.6353833212874285, + "learning_rate": 0.003, + "loss": 3.9948, + "step": 52772 + }, + { + "epoch": 0.52773, + "grad_norm": 0.6020636427522913, + "learning_rate": 0.003, + "loss": 3.9739, + "step": 52773 + }, + { + "epoch": 0.52774, + "grad_norm": 0.5189923387991365, + "learning_rate": 0.003, + "loss": 4.0206, + "step": 52774 + }, + { + "epoch": 0.52775, + "grad_norm": 0.53647531446728, + "learning_rate": 0.003, + "loss": 3.9973, + "step": 52775 + }, + { + "epoch": 0.52776, + "grad_norm": 0.5619004146166813, + "learning_rate": 0.003, + "loss": 3.9381, + "step": 52776 + }, + { + "epoch": 0.52777, + "grad_norm": 0.7137995096599765, + "learning_rate": 0.003, + "loss": 3.9659, + "step": 52777 + }, + { + "epoch": 0.52778, + "grad_norm": 0.8759059975866078, + "learning_rate": 0.003, + "loss": 4.0055, + "step": 52778 + }, + { + "epoch": 0.52779, + "grad_norm": 1.0344163149001484, + "learning_rate": 0.003, + "loss": 3.9651, + "step": 52779 + }, + { + "epoch": 0.5278, + "grad_norm": 0.9546320413271778, + "learning_rate": 0.003, + "loss": 4.0008, + "step": 52780 + }, + { + "epoch": 0.52781, + "grad_norm": 0.9421954798235845, + "learning_rate": 0.003, + "loss": 3.9834, + "step": 52781 + }, + { + "epoch": 0.52782, + "grad_norm": 1.038654923217241, + "learning_rate": 0.003, + "loss": 3.9573, + "step": 52782 + }, + { + "epoch": 0.52783, + "grad_norm": 0.8633102869848194, + "learning_rate": 0.003, + "loss": 3.9922, + "step": 52783 + }, + { + "epoch": 0.52784, + "grad_norm": 0.8555769808191505, + "learning_rate": 0.003, + "loss": 4.0009, + "step": 52784 + }, + { + "epoch": 0.52785, + "grad_norm": 0.9070334576965914, + "learning_rate": 0.003, + "loss": 3.9949, + "step": 52785 + }, + { + "epoch": 0.52786, + "grad_norm": 0.842224704808037, + "learning_rate": 0.003, + "loss": 4.0032, + "step": 52786 + }, + { + "epoch": 0.52787, + "grad_norm": 0.7676758339729327, + "learning_rate": 0.003, + "loss": 3.9799, + "step": 52787 + }, + { + "epoch": 0.52788, + "grad_norm": 0.7510201639063429, + "learning_rate": 0.003, + "loss": 4.0037, + "step": 52788 + }, + { + "epoch": 0.52789, + "grad_norm": 0.6405181598589965, + "learning_rate": 0.003, + "loss": 3.9798, + "step": 52789 + }, + { + "epoch": 0.5279, + "grad_norm": 0.7705493340232603, + "learning_rate": 0.003, + "loss": 3.9806, + "step": 52790 + }, + { + "epoch": 0.52791, + "grad_norm": 0.7979172443339704, + "learning_rate": 0.003, + "loss": 3.9768, + "step": 52791 + }, + { + "epoch": 0.52792, + "grad_norm": 0.7822799447787717, + "learning_rate": 0.003, + "loss": 4.0015, + "step": 52792 + }, + { + "epoch": 0.52793, + "grad_norm": 0.8048925703174461, + "learning_rate": 0.003, + "loss": 3.9711, + "step": 52793 + }, + { + "epoch": 0.52794, + "grad_norm": 0.8219317048431697, + "learning_rate": 0.003, + "loss": 3.9368, + "step": 52794 + }, + { + "epoch": 0.52795, + "grad_norm": 0.7684223610513445, + "learning_rate": 0.003, + "loss": 3.9786, + "step": 52795 + }, + { + "epoch": 0.52796, + "grad_norm": 0.6587978374969795, + "learning_rate": 0.003, + "loss": 3.9894, + "step": 52796 + }, + { + "epoch": 0.52797, + "grad_norm": 0.6465268701935916, + "learning_rate": 0.003, + "loss": 3.9699, + "step": 52797 + }, + { + "epoch": 0.52798, + "grad_norm": 0.6243137614515035, + "learning_rate": 0.003, + "loss": 4.0023, + "step": 52798 + }, + { + "epoch": 0.52799, + "grad_norm": 0.6324595238250527, + "learning_rate": 0.003, + "loss": 3.9643, + "step": 52799 + }, + { + "epoch": 0.528, + "grad_norm": 0.7306666689334204, + "learning_rate": 0.003, + "loss": 3.9763, + "step": 52800 + } + ], + "logging_steps": 1, + "max_steps": 100000, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 25, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 2.0931843427663872e+18, + "train_batch_size": 512, + "trial_name": null, + "trial_params": null +}